diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,70193 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8776958290248525,
+  "eval_steps": 500,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.776958290248525e-05,
+      "grad_norm": 0.40625,
+      "learning_rate": 3e-06,
+      "loss": 1.7689,
+      "step": 1
+    },
+    {
+      "epoch": 0.0001755391658049705,
+      "grad_norm": 0.38671875,
+      "learning_rate": 6e-06,
+      "loss": 1.7451,
+      "step": 2
+    },
+    {
+      "epoch": 0.00026330874870745575,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9e-06,
+      "loss": 1.7571,
+      "step": 3
+    },
+    {
+      "epoch": 0.000351078331609941,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2e-05,
+      "loss": 1.7319,
+      "step": 4
+    },
+    {
+      "epoch": 0.00043884791451242626,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5e-05,
+      "loss": 1.7674,
+      "step": 5
+    },
+    {
+      "epoch": 0.0005266174974149115,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8e-05,
+      "loss": 1.7341,
+      "step": 6
+    },
+    {
+      "epoch": 0.0006143870803173967,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.1000000000000002e-05,
+      "loss": 1.7525,
+      "step": 7
+    },
+    {
+      "epoch": 0.000702156663219882,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.4e-05,
+      "loss": 1.7367,
+      "step": 8
+    },
+    {
+      "epoch": 0.0007899262461223672,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.7e-05,
+      "loss": 1.7394,
+      "step": 9
+    },
+    {
+      "epoch": 0.0008776958290248525,
+      "grad_norm": 0.328125,
+      "learning_rate": 3e-05,
+      "loss": 1.7194,
+      "step": 10
+    },
+    {
+      "epoch": 0.0009654654119273377,
+      "grad_norm": 0.31640625,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 1.7542,
+      "step": 11
+    },
+    {
+      "epoch": 0.001053234994829823,
+      "grad_norm": 0.310546875,
+      "learning_rate": 3.6e-05,
+      "loss": 1.7498,
+      "step": 12
+    },
+    {
+      "epoch": 0.0011410045777323083,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.9e-05,
+      "loss": 1.7251,
+      "step": 13
+    },
+    {
+      "epoch": 0.0012287741606347934,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.2000000000000004e-05,
+      "loss": 1.7456,
+      "step": 14
+    },
+    {
+      "epoch": 0.0013165437435372787,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 1.7261,
+      "step": 15
+    },
+    {
+      "epoch": 0.001404313326439764,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.8e-05,
+      "loss": 1.7426,
+      "step": 16
+    },
+    {
+      "epoch": 0.0014920829093422493,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 5.1000000000000006e-05,
+      "loss": 1.7373,
+      "step": 17
+    },
+    {
+      "epoch": 0.0015798524922447344,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.4e-05,
+      "loss": 1.7427,
+      "step": 18
+    },
+    {
+      "epoch": 0.0016676220751472197,
+      "grad_norm": 0.244140625,
+      "learning_rate": 5.7e-05,
+      "loss": 1.7142,
+      "step": 19
+    },
+    {
+      "epoch": 0.001755391658049705,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6e-05,
+      "loss": 1.7586,
+      "step": 20
+    },
+    {
+      "epoch": 0.0018431612409521903,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.3e-05,
+      "loss": 1.7348,
+      "step": 21
+    },
+    {
+      "epoch": 0.0019309308238546754,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 1.7164,
+      "step": 22
+    },
+    {
+      "epoch": 0.002018700406757161,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.9e-05,
+      "loss": 1.7185,
+      "step": 23
+    },
+    {
+      "epoch": 0.002106469989659646,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.2e-05,
+      "loss": 1.7122,
+      "step": 24
+    },
+    {
+      "epoch": 0.002194239572562131,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.7534,
+      "step": 25
+    },
+    {
+      "epoch": 0.0022820091554646166,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.8e-05,
+      "loss": 1.7097,
+      "step": 26
+    },
+    {
+      "epoch": 0.0023697787383671017,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.1e-05,
+      "loss": 1.7264,
+      "step": 27
+    },
+    {
+      "epoch": 0.002457548321269587,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.400000000000001e-05,
+      "loss": 1.7221,
+      "step": 28
+    },
+    {
+      "epoch": 0.0025453179041720723,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.7e-05,
+      "loss": 1.7413,
+      "step": 29
+    },
+    {
+      "epoch": 0.0026330874870745574,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 1.7129,
+      "step": 30
+    },
+    {
+      "epoch": 0.002720857069977043,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.3e-05,
+      "loss": 1.6926,
+      "step": 31
+    },
+    {
+      "epoch": 0.002808626652879528,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.6e-05,
+      "loss": 1.7416,
+      "step": 32
+    },
+    {
+      "epoch": 0.002896396235782013,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 1.7328,
+      "step": 33
+    },
+    {
+      "epoch": 0.0029841658186844987,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00010200000000000001,
+      "loss": 1.7141,
+      "step": 34
+    },
+    {
+      "epoch": 0.0030719354015869837,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010500000000000002,
+      "loss": 1.7105,
+      "step": 35
+    },
+    {
+      "epoch": 0.003159704984489469,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000108,
+      "loss": 1.7226,
+      "step": 36
+    },
+    {
+      "epoch": 0.0032474745673919544,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000111,
+      "loss": 1.7251,
+      "step": 37
+    },
+    {
+      "epoch": 0.0033352441502944394,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.000114,
+      "loss": 1.7446,
+      "step": 38
+    },
+    {
+      "epoch": 0.003423013733196925,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.000117,
+      "loss": 1.741,
+      "step": 39
+    },
+    {
+      "epoch": 0.00351078331609941,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012,
+      "loss": 1.6584,
+      "step": 40
+    },
+    {
+      "epoch": 0.003598552899001895,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.000123,
+      "loss": 1.6891,
+      "step": 41
+    },
+    {
+      "epoch": 0.0036863224819043807,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.000126,
+      "loss": 1.7028,
+      "step": 42
+    },
+    {
+      "epoch": 0.0037740920648068658,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000129,
+      "loss": 1.6706,
+      "step": 43
+    },
+    {
+      "epoch": 0.003861861647709351,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 1.7151,
+      "step": 44
+    },
+    {
+      "epoch": 0.003949631230611836,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000135,
+      "loss": 1.7094,
+      "step": 45
+    },
+    {
+      "epoch": 0.004037400813514322,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000138,
+      "loss": 1.7015,
+      "step": 46
+    },
+    {
+      "epoch": 0.0041251703964168066,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.000141,
+      "loss": 1.6753,
+      "step": 47
+    },
+    {
+      "epoch": 0.004212939979319292,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000144,
+      "loss": 1.6983,
+      "step": 48
+    },
+    {
+      "epoch": 0.004300709562221778,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000147,
+      "loss": 1.6713,
+      "step": 49
+    },
+    {
+      "epoch": 0.004388479145124262,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.7099,
+      "step": 50
+    },
+    {
+      "epoch": 0.004476248728026748,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.000153,
+      "loss": 1.698,
+      "step": 51
+    },
+    {
+      "epoch": 0.004564018310929233,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.000156,
+      "loss": 1.6855,
+      "step": 52
+    },
+    {
+      "epoch": 0.004651787893831718,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.000159,
+      "loss": 1.7034,
+      "step": 53
+    },
+    {
+      "epoch": 0.0047395574767342035,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000162,
+      "loss": 1.7015,
+      "step": 54
+    },
+    {
+      "epoch": 0.004827327059636689,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.000165,
+      "loss": 1.6715,
+      "step": 55
+    },
+    {
+      "epoch": 0.004915096642539174,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00016800000000000002,
+      "loss": 1.6808,
+      "step": 56
+    },
+    {
+      "epoch": 0.005002866225441659,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.000171,
+      "loss": 1.6848,
+      "step": 57
+    },
+    {
+      "epoch": 0.005090635808344145,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.000174,
+      "loss": 1.6683,
+      "step": 58
+    },
+    {
+      "epoch": 0.005178405391246629,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.000177,
+      "loss": 1.6896,
+      "step": 59
+    },
+    {
+      "epoch": 0.005266174974149115,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 1.7075,
+      "step": 60
+    },
+    {
+      "epoch": 0.0053539445570516,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.000183,
+      "loss": 1.6605,
+      "step": 61
+    },
+    {
+      "epoch": 0.005441714139954086,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000186,
+      "loss": 1.6536,
+      "step": 62
+    },
+    {
+      "epoch": 0.005529483722856571,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.000189,
+      "loss": 1.6499,
+      "step": 63
+    },
+    {
+      "epoch": 0.005617253305759056,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000192,
+      "loss": 1.6682,
+      "step": 64
+    },
+    {
+      "epoch": 0.005705022888661542,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019500000000000002,
+      "loss": 1.6727,
+      "step": 65
+    },
+    {
+      "epoch": 0.005792792471564026,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 1.6554,
+      "step": 66
+    },
+    {
+      "epoch": 0.005880562054466512,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.000201,
+      "loss": 1.651,
+      "step": 67
+    },
+    {
+      "epoch": 0.005968331637368997,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00020400000000000003,
+      "loss": 1.6457,
+      "step": 68
+    },
+    {
+      "epoch": 0.006056101220271482,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00020700000000000002,
+      "loss": 1.6536,
+      "step": 69
+    },
+    {
+      "epoch": 0.0061438708031739675,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00021000000000000004,
+      "loss": 1.6843,
+      "step": 70
+    },
+    {
+      "epoch": 0.006231640386076453,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00021299999999999997,
+      "loss": 1.6222,
+      "step": 71
+    },
+    {
+      "epoch": 0.006319409968978938,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000216,
+      "loss": 1.6791,
+      "step": 72
+    },
+    {
+      "epoch": 0.006407179551881423,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00021899999999999998,
+      "loss": 1.6917,
+      "step": 73
+    },
+    {
+      "epoch": 0.006494949134783909,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.000222,
+      "loss": 1.6895,
+      "step": 74
+    },
+    {
+      "epoch": 0.006582718717686393,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.000225,
+      "loss": 1.7024,
+      "step": 75
+    },
+    {
+      "epoch": 0.006670488300588879,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.000228,
+      "loss": 1.6649,
+      "step": 76
+    },
+    {
+      "epoch": 0.006758257883491364,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000231,
+      "loss": 1.6537,
+      "step": 77
+    },
+    {
+      "epoch": 0.00684602746639385,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.000234,
+      "loss": 1.6345,
+      "step": 78
+    },
+    {
+      "epoch": 0.006933797049296335,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00023700000000000001,
+      "loss": 1.6799,
+      "step": 79
+    },
+    {
+      "epoch": 0.00702156663219882,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00024,
+      "loss": 1.6535,
+      "step": 80
+    },
+    {
+      "epoch": 0.007109336215101306,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00024300000000000002,
+      "loss": 1.6421,
+      "step": 81
+    },
+    {
+      "epoch": 0.00719710579800379,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.000246,
+      "loss": 1.6371,
+      "step": 82
+    },
+    {
+      "epoch": 0.007284875380906276,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00024900000000000004,
+      "loss": 1.6963,
+      "step": 83
+    },
+    {
+      "epoch": 0.007372644963808761,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.000252,
+      "loss": 1.6568,
+      "step": 84
+    },
+    {
+      "epoch": 0.007460414546711246,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.000255,
+      "loss": 1.6686,
+      "step": 85
+    },
+    {
+      "epoch": 0.0075481841296137315,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000258,
+      "loss": 1.6236,
+      "step": 86
+    },
+    {
+      "epoch": 0.007635953712516217,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.000261,
+      "loss": 1.6198,
+      "step": 87
+    },
+    {
+      "epoch": 0.007723723295418702,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00026399999999999997,
+      "loss": 1.6159,
+      "step": 88
+    },
+    {
+      "epoch": 0.007811492878321187,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.000267,
+      "loss": 1.6369,
+      "step": 89
+    },
+    {
+      "epoch": 0.007899262461223673,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00027,
+      "loss": 1.5979,
+      "step": 90
+    },
+    {
+      "epoch": 0.007987032044126157,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000273,
+      "loss": 1.6274,
+      "step": 91
+    },
+    {
+      "epoch": 0.008074801627028644,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000276,
+      "loss": 1.6396,
+      "step": 92
+    },
+    {
+      "epoch": 0.008162571209931128,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000279,
+      "loss": 1.6311,
+      "step": 93
+    },
+    {
+      "epoch": 0.008250340792833613,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.000282,
+      "loss": 1.6497,
+      "step": 94
+    },
+    {
+      "epoch": 0.0083381103757361,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.000285,
+      "loss": 1.6387,
+      "step": 95
+    },
+    {
+      "epoch": 0.008425879958638584,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.000288,
+      "loss": 1.6348,
+      "step": 96
+    },
+    {
+      "epoch": 0.008513649541541069,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.000291,
+      "loss": 1.6238,
+      "step": 97
+    },
+    {
+      "epoch": 0.008601419124443555,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.000294,
+      "loss": 1.6424,
+      "step": 98
+    },
+    {
+      "epoch": 0.00868918870734604,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.000297,
+      "loss": 1.6115,
+      "step": 99
+    },
+    {
+      "epoch": 0.008776958290248525,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00030000000000000003,
+      "loss": 1.6163,
+      "step": 100
+    },
+    {
+      "epoch": 0.008864727873151011,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00030300000000000005,
+      "loss": 1.6289,
+      "step": 101
+    },
+    {
+      "epoch": 0.008952497456053496,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.000306,
+      "loss": 1.6104,
+      "step": 102
+    },
+    {
+      "epoch": 0.00904026703895598,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000309,
+      "loss": 1.5959,
+      "step": 103
+    },
+    {
+      "epoch": 0.009128036621858467,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.000312,
+      "loss": 1.6469,
+      "step": 104
+    },
+    {
+      "epoch": 0.009215806204760951,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.000315,
+      "loss": 1.6033,
+      "step": 105
+    },
+    {
+      "epoch": 0.009303575787663436,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.000318,
+      "loss": 1.628,
+      "step": 106
+    },
+    {
+      "epoch": 0.009391345370565922,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.000321,
+      "loss": 1.6601,
+      "step": 107
+    },
+    {
+      "epoch": 0.009479114953468407,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.000324,
+      "loss": 1.6207,
+      "step": 108
+    },
+    {
+      "epoch": 0.009566884536370892,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.000327,
+      "loss": 1.5845,
+      "step": 109
+    },
+    {
+      "epoch": 0.009654654119273378,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00033,
+      "loss": 1.6281,
+      "step": 110
+    },
+    {
+      "epoch": 0.009742423702175863,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.000333,
+      "loss": 1.6241,
+      "step": 111
+    },
+    {
+      "epoch": 0.009830193285078347,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00033600000000000004,
+      "loss": 1.6173,
+      "step": 112
+    },
+    {
+      "epoch": 0.009917962867980834,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.000339,
+      "loss": 1.6026,
+      "step": 113
+    },
+    {
+      "epoch": 0.010005732450883318,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.000342,
+      "loss": 1.6295,
+      "step": 114
+    },
+    {
+      "epoch": 0.010093502033785803,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00034500000000000004,
+      "loss": 1.6142,
+      "step": 115
+    },
+    {
+      "epoch": 0.01018127161668829,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000348,
+      "loss": 1.6091,
+      "step": 116
+    },
+    {
+      "epoch": 0.010269041199590774,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.000351,
+      "loss": 1.5936,
+      "step": 117
+    },
+    {
+      "epoch": 0.010356810782493259,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.000354,
+      "loss": 1.6154,
+      "step": 118
+    },
+    {
+      "epoch": 0.010444580365395745,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.000357,
+      "loss": 1.5782,
+      "step": 119
+    },
+    {
+      "epoch": 0.01053234994829823,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00035999999999999997,
+      "loss": 1.6065,
+      "step": 120
+    },
+    {
+      "epoch": 0.010620119531200716,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.000363,
+      "loss": 1.5985,
+      "step": 121
+    },
+    {
+      "epoch": 0.0107078891141032,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.000366,
+      "loss": 1.5969,
+      "step": 122
+    },
+    {
+      "epoch": 0.010795658697005685,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.000369,
+      "loss": 1.6173,
+      "step": 123
+    },
+    {
+      "epoch": 0.010883428279908172,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.000372,
+      "loss": 1.5828,
+      "step": 124
+    },
+    {
+      "epoch": 0.010971197862810656,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000375,
+      "loss": 1.5594,
+      "step": 125
+    },
+    {
+      "epoch": 0.011058967445713141,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000378,
+      "loss": 1.6131,
+      "step": 126
+    },
+    {
+      "epoch": 0.011146737028615628,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.000381,
+      "loss": 1.6203,
+      "step": 127
+    },
+    {
+      "epoch": 0.011234506611518112,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.000384,
+      "loss": 1.598,
+      "step": 128
+    },
+    {
+      "epoch": 0.011322276194420597,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00038700000000000003,
+      "loss": 1.5817,
+      "step": 129
+    },
+    {
+      "epoch": 0.011410045777323083,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00039000000000000005,
+      "loss": 1.5968,
+      "step": 130
+    },
+    {
+      "epoch": 0.011497815360225568,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000393,
+      "loss": 1.6066,
+      "step": 131
+    },
+    {
+      "epoch": 0.011585584943128053,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00039600000000000003,
+      "loss": 1.6208,
+      "step": 132
+    },
+    {
+      "epoch": 0.011673354526030539,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00039900000000000005,
+      "loss": 1.5973,
+      "step": 133
+    },
+    {
+      "epoch": 0.011761124108933024,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000402,
+      "loss": 1.588,
+      "step": 134
+    },
+    {
+      "epoch": 0.011848893691835508,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00040500000000000003,
+      "loss": 1.6202,
+      "step": 135
+    },
+    {
+      "epoch": 0.011936663274737995,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00040800000000000005,
+      "loss": 1.5611,
+      "step": 136
+    },
+    {
+      "epoch": 0.01202443285764048,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000411,
+      "loss": 1.5582,
+      "step": 137
+    },
+    {
+      "epoch": 0.012112202440542964,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00041400000000000003,
+      "loss": 1.5798,
+      "step": 138
+    },
+    {
+      "epoch": 0.01219997202344545,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00041700000000000005,
+      "loss": 1.5895,
+      "step": 139
+    },
+    {
+      "epoch": 0.012287741606347935,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00042000000000000007,
+      "loss": 1.605,
+      "step": 140
+    },
+    {
+      "epoch": 0.01237551118925042,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.000423,
+      "loss": 1.5972,
+      "step": 141
+    },
+    {
+      "epoch": 0.012463280772152906,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00042599999999999995,
+      "loss": 1.5651,
+      "step": 142
+    },
+    {
+      "epoch": 0.01255105035505539,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00042899999999999997,
+      "loss": 1.5516,
+      "step": 143
+    },
+    {
+      "epoch": 0.012638819937957875,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.000432,
+      "loss": 1.5887,
+      "step": 144
+    },
+    {
+      "epoch": 0.012726589520860362,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.000435,
+      "loss": 1.5829,
+      "step": 145
+    },
+    {
+      "epoch": 0.012814359103762846,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00043799999999999997,
+      "loss": 1.5895,
+      "step": 146
+    },
+    {
+      "epoch": 0.012902128686665331,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.000441,
+      "loss": 1.6111,
+      "step": 147
+    },
+    {
+      "epoch": 0.012989898269567817,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.000444,
+      "loss": 1.5881,
+      "step": 148
+    },
+    {
+      "epoch": 0.013077667852470302,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00044699999999999997,
+      "loss": 1.5744,
+      "step": 149
+    },
+    {
+      "epoch": 0.013165437435372787,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00045,
+      "loss": 1.5884,
+      "step": 150
+    },
+    {
+      "epoch": 0.013253207018275273,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.000453,
+      "loss": 1.5649,
+      "step": 151
+    },
+    {
+      "epoch": 0.013340976601177758,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.000456,
+      "loss": 1.5982,
+      "step": 152
+    },
+    {
+      "epoch": 0.013428746184080244,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000459,
+      "loss": 1.5878,
+      "step": 153
+    },
+    {
+      "epoch": 0.013516515766982729,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000462,
+      "loss": 1.5519,
+      "step": 154
+    },
+    {
+      "epoch": 0.013604285349885213,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000465,
+      "loss": 1.5693,
+      "step": 155
+    },
+    {
+      "epoch": 0.0136920549327877,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.000468,
+      "loss": 1.5546,
+      "step": 156
+    },
+    {
+      "epoch": 0.013779824515690185,
+      "grad_norm": 0.375,
+      "learning_rate": 0.000471,
+      "loss": 1.5526,
+      "step": 157
+    },
+    {
+      "epoch": 0.01386759409859267,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00047400000000000003,
+      "loss": 1.5593,
+      "step": 158
+    },
+    {
+      "epoch": 0.013955363681495156,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000477,
+      "loss": 1.5608,
+      "step": 159
+    },
+    {
+      "epoch": 0.01404313326439764,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048,
+      "loss": 1.5527,
+      "step": 160
+    },
+    {
+      "epoch": 0.014130902847300125,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00048300000000000003,
+      "loss": 1.5818,
+      "step": 161
+    },
+    {
+      "epoch": 0.014218672430202611,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00048600000000000005,
+      "loss": 1.5687,
+      "step": 162
+    },
+    {
+      "epoch": 0.014306442013105096,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004890000000000001,
+      "loss": 1.5944,
+      "step": 163
+    },
+    {
+      "epoch": 0.01439421159600758,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.000492,
+      "loss": 1.6024,
+      "step": 164
+    },
+    {
+      "epoch": 0.014481981178910067,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000495,
+      "loss": 1.561,
+      "step": 165
+    },
+    {
+      "epoch": 0.014569750761812552,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004980000000000001,
+      "loss": 1.576,
+      "step": 166
+    },
+    {
+      "epoch": 0.014657520344715036,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.000501,
+      "loss": 1.536,
+      "step": 167
+    },
+    {
+      "epoch": 0.014745289927617523,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000504,
+      "loss": 1.5691,
+      "step": 168
+    },
+    {
+      "epoch": 0.014833059510520007,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0005070000000000001,
+      "loss": 1.5502,
+      "step": 169
+    },
+    {
+      "epoch": 0.014920829093422492,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00051,
+      "loss": 1.5524,
+      "step": 170
+    },
+    {
+      "epoch": 0.015008598676324978,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.000513,
+      "loss": 1.582,
+      "step": 171
+    },
+    {
+      "epoch": 0.015096368259227463,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000516,
+      "loss": 1.5702,
+      "step": 172
+    },
+    {
+      "epoch": 0.015184137842129948,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005189999999999999,
+      "loss": 1.578,
+      "step": 173
+    },
+    {
+      "epoch": 0.015271907425032434,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000522,
+      "loss": 1.534,
+      "step": 174
+    },
+    {
+      "epoch": 0.015359677007934919,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.000525,
+      "loss": 1.5624,
+      "step": 175
+    },
+    {
+      "epoch": 0.015447446590837403,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0005279999999999999,
+      "loss": 1.5523,
+      "step": 176
+    },
+    {
+      "epoch": 0.01553521617373989,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.000531,
+      "loss": 1.5955,
+      "step": 177
+    },
+    {
+      "epoch": 0.015622985756642374,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.000534,
+      "loss": 1.5625,
+      "step": 178
+    },
+    {
+      "epoch": 0.01571075533954486,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.000537,
+      "loss": 1.553,
+      "step": 179
+    },
+    {
+      "epoch": 0.015798524922447345,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00054,
+      "loss": 1.5868,
+      "step": 180
+    },
+    {
+      "epoch": 0.01588629450534983,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.000543,
+      "loss": 1.5586,
+      "step": 181
+    },
+    {
+      "epoch": 0.015974064088252315,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000546,
+      "loss": 1.5364,
+      "step": 182
+    },
+    {
+      "epoch": 0.0160618336711548,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.000549,
+      "loss": 1.5587,
+      "step": 183
+    },
+    {
+      "epoch": 0.016149603254057288,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.000552,
+      "loss": 1.5493,
+      "step": 184
+    },
+    {
+      "epoch": 0.016237372836959772,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.000555,
+      "loss": 1.5635,
+      "step": 185
+    },
+    {
+      "epoch": 0.016325142419862257,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000558,
+      "loss": 1.5035,
+      "step": 186
+    },
+    {
+      "epoch": 0.01641291200276474,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000561,
+      "loss": 1.5546,
+      "step": 187
+    },
+    {
+      "epoch": 0.016500681585667226,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000564,
+      "loss": 1.5222,
+      "step": 188
+    },
+    {
+      "epoch": 0.01658845116856971,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.000567,
+      "loss": 1.5477,
+      "step": 189
+    },
+    {
+      "epoch": 0.0166762207514722,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00057,
+      "loss": 1.5579,
+      "step": 190
+    },
+    {
+      "epoch": 0.016763990334374684,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.000573,
+      "loss": 1.5225,
+      "step": 191
+    },
+    {
+      "epoch": 0.01685175991727717,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.000576,
+      "loss": 1.5301,
+      "step": 192
+    },
+    {
+      "epoch": 0.016939529500179653,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.000579,
+      "loss": 1.5545,
+      "step": 193
+    },
+    {
+      "epoch": 0.017027299083082138,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.000582,
+      "loss": 1.5481,
+      "step": 194
+    },
+    {
+      "epoch": 0.017115068665984622,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000585,
+      "loss": 1.5556,
+      "step": 195
+    },
+    {
+      "epoch": 0.01720283824888711,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.000588,
+      "loss": 1.5607,
+      "step": 196
+    },
+    {
+      "epoch": 0.017290607831789595,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000591,
+      "loss": 1.5166,
+      "step": 197
+    },
+    {
+      "epoch": 0.01737837741469208,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.000594,
+      "loss": 1.5709,
+      "step": 198
+    },
+    {
+      "epoch": 0.017466146997594564,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005970000000000001,
+      "loss": 1.5871,
+      "step": 199
+    },
+    {
+      "epoch": 0.01755391658049705,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0006000000000000001,
+      "loss": 1.5793,
+      "step": 200
+    },
+    {
+      "epoch": 0.017641686163399534,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.000603,
+      "loss": 1.5614,
+      "step": 201
+    },
+    {
+      "epoch": 0.017729455746302022,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0006060000000000001,
+      "loss": 1.566,
+      "step": 202
+    },
+    {
+      "epoch": 0.017817225329204506,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0006090000000000001,
+      "loss": 1.5251,
+      "step": 203
+    },
+    {
+      "epoch": 0.01790499491210699,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.000612,
+      "loss": 1.5759,
+      "step": 204
+    },
+    {
+      "epoch": 0.017992764495009476,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000615,
+      "loss": 1.568,
+      "step": 205
+    },
+    {
+      "epoch": 0.01808053407791196,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.000618,
+      "loss": 1.5376,
+      "step": 206
+    },
+    {
+      "epoch": 0.01816830366081445,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.000621,
+      "loss": 1.529,
+      "step": 207
+    },
+    {
+      "epoch": 0.018256073243716933,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.000624,
+      "loss": 1.5411,
+      "step": 208
+    },
+    {
+      "epoch": 0.018343842826619418,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.000627,
+      "loss": 1.5323,
+      "step": 209
+    },
+    {
+      "epoch": 0.018431612409521902,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00063,
+      "loss": 1.5671,
+      "step": 210
+    },
+    {
+      "epoch": 0.018519381992424387,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.000633,
+      "loss": 1.528,
+      "step": 211
+    },
+    {
+      "epoch": 0.018607151575326872,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000636,
+      "loss": 1.5338,
+      "step": 212
+    },
+    {
+      "epoch": 0.01869492115822936,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000639,
+      "loss": 1.4899,
+      "step": 213
+    },
+    {
+      "epoch": 0.018782690741131845,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.000642,
+      "loss": 1.5698,
+      "step": 214
+    },
+    {
+      "epoch": 0.01887046032403433,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.000645,
+      "loss": 1.5581,
+      "step": 215
+    },
+    {
+      "epoch": 0.018958229906936814,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000648,
+      "loss": 1.5541,
+      "step": 216
+    },
+    {
+      "epoch": 0.0190459994898393,
+      "grad_norm": 0.25,
+      "learning_rate": 0.000651,
+      "loss": 1.5394,
+      "step": 217
+    },
+    {
+      "epoch": 0.019133769072741783,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.000654,
+      "loss": 1.5317,
+      "step": 218
+    },
+    {
+      "epoch": 0.01922153865564427,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.000657,
+      "loss": 1.4966,
+      "step": 219
+    },
+    {
+      "epoch": 0.019309308238546756,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00066,
+      "loss": 1.54,
+      "step": 220
+    },
+    {
+      "epoch": 0.01939707782144924,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0006630000000000001,
+      "loss": 1.5467,
+      "step": 221
+    },
+    {
+      "epoch": 0.019484847404351725,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000666,
+      "loss": 1.5047,
+      "step": 222
+    },
+    {
+      "epoch": 0.01957261698725421,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000669,
+      "loss": 1.5274,
+      "step": 223
+    },
+    {
+      "epoch": 0.019660386570156695,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0006720000000000001,
+      "loss": 1.5205,
+      "step": 224
+    },
+    {
+      "epoch": 0.019748156153059183,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.000675,
+      "loss": 1.528,
+      "step": 225
+    },
+    {
+      "epoch": 0.019835925735961667,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.000678,
+      "loss": 1.5186,
+      "step": 226
+    },
+    {
+      "epoch": 0.019923695318864152,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0006810000000000001,
+      "loss": 1.5292,
+      "step": 227
+    },
+    {
+      "epoch": 0.020011464901766637,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.000684,
+      "loss": 1.5265,
+      "step": 228
+    },
+    {
+      "epoch": 0.02009923448466912,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.000687,
+      "loss": 1.4965,
+      "step": 229
+    },
+    {
+      "epoch": 0.020187004067571606,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0006900000000000001,
+      "loss": 1.4951,
+      "step": 230
+    },
+    {
+      "epoch": 0.020274773650474094,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.000693,
+      "loss": 1.5387,
+      "step": 231
+    },
+    {
+      "epoch": 0.02036254323337658,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.000696,
+      "loss": 1.5147,
+      "step": 232
+    },
+    {
+      "epoch": 0.020450312816279063,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0006990000000000001,
+      "loss": 1.4739,
+      "step": 233
+    },
+    {
+      "epoch": 0.020538082399181548,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.000702,
+      "loss": 1.5632,
+      "step": 234
+    },
+    {
+      "epoch": 0.020625851982084033,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000705,
+      "loss": 1.5118,
+      "step": 235
+    },
+    {
+      "epoch": 0.020713621564986517,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000708,
+      "loss": 1.5593,
+      "step": 236
+    },
+    {
+      "epoch": 0.020801391147889006,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0007109999999999999,
+      "loss": 1.5134,
+      "step": 237
+    },
+    {
+      "epoch": 0.02088916073079149,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.000714,
+      "loss": 1.4698,
+      "step": 238
+    },
+    {
+      "epoch": 0.020976930313693975,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000717,
+      "loss": 1.5551,
+      "step": 239
+    },
+    {
+      "epoch": 0.02106469989659646,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0007199999999999999,
+      "loss": 1.5045,
+      "step": 240
+    },
+    {
+      "epoch": 0.021152469479498944,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.000723,
+      "loss": 1.5291,
+      "step": 241
+    },
+    {
+      "epoch": 0.021240239062401432,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000726,
+      "loss": 1.5537,
+      "step": 242
+    },
+    {
+      "epoch": 0.021328008645303917,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.000729,
+      "loss": 1.5157,
+      "step": 243
+    },
+    {
+      "epoch": 0.0214157782282064,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.000732,
+      "loss": 1.4763,
+      "step": 244
+    },
+    {
+      "epoch": 0.021503547811108886,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.000735,
+      "loss": 1.5265,
+      "step": 245
+    },
+    {
+      "epoch": 0.02159131739401137,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000738,
+      "loss": 1.5019,
+      "step": 246
+    },
+    {
+      "epoch": 0.021679086976913856,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.000741,
+      "loss": 1.5275,
+      "step": 247
+    },
+    {
+      "epoch": 0.021766856559816344,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000744,
+      "loss": 1.4965,
+      "step": 248
+    },
+    {
+      "epoch": 0.02185462614271883,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.000747,
+      "loss": 1.5335,
+      "step": 249
+    },
+    {
+      "epoch": 0.021942395725621313,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00075,
+      "loss": 1.4928,
+      "step": 250
+    },
+    {
+      "epoch": 0.022030165308523798,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.000753,
+      "loss": 1.5167,
+      "step": 251
+    },
+    {
+      "epoch": 0.022117934891426282,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.000756,
+      "loss": 1.4861,
+      "step": 252
+    },
+    {
+      "epoch": 0.022205704474328767,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000759,
+      "loss": 1.4847,
+      "step": 253
+    },
+    {
+      "epoch": 0.022293474057231255,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.000762,
+      "loss": 1.5386,
+      "step": 254
+    },
+    {
+      "epoch": 0.02238124364013374,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0007650000000000001,
+      "loss": 1.5287,
+      "step": 255
+    },
+    {
+      "epoch": 0.022469013223036224,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000768,
+      "loss": 1.5219,
+      "step": 256
+    },
+    {
+      "epoch": 0.02255678280593871,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.000771,
+      "loss": 1.5138,
+      "step": 257
+    },
+    {
+      "epoch": 0.022644552388841194,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0007740000000000001,
+      "loss": 1.5172,
+      "step": 258
+    },
+    {
+      "epoch": 0.02273232197174368,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.000777,
+      "loss": 1.4874,
+      "step": 259
+    },
+    {
+      "epoch": 0.022820091554646166,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0007800000000000001,
+      "loss": 1.5269,
+      "step": 260
+    },
+    {
+      "epoch": 0.02290786113754865,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0007830000000000001,
+      "loss": 1.5049,
+      "step": 261
+    },
+    {
+      "epoch": 0.022995630720451136,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000786,
+      "loss": 1.5297,
+      "step": 262
+    },
+    {
+      "epoch": 0.02308340030335362,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0007890000000000001,
+      "loss": 1.516,
+      "step": 263
+    },
+    {
+      "epoch": 0.023171169886256105,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0007920000000000001,
+      "loss": 1.5185,
+      "step": 264
+    },
+    {
+      "epoch": 0.02325893946915859,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.000795,
+      "loss": 1.4899,
+      "step": 265
+    },
+    {
+      "epoch": 0.023346709052061078,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0007980000000000001,
+      "loss": 1.5057,
+      "step": 266
+    },
+    {
+      "epoch": 0.023434478634963563,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0008010000000000001,
+      "loss": 1.518,
+      "step": 267
+    },
+    {
+      "epoch": 0.023522248217866047,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000804,
+      "loss": 1.4779,
+      "step": 268
+    },
+    {
+      "epoch": 0.023610017800768532,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0008070000000000001,
+      "loss": 1.483,
+      "step": 269
+    },
+    {
+      "epoch": 0.023697787383671017,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0008100000000000001,
+      "loss": 1.5232,
+      "step": 270
+    },
+    {
+      "epoch": 0.023785556966573505,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.000813,
+      "loss": 1.5,
+      "step": 271
+    },
+    {
+      "epoch": 0.02387332654947599,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0008160000000000001,
+      "loss": 1.4608,
+      "step": 272
+    },
+    {
+      "epoch": 0.023961096132378474,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0008190000000000001,
+      "loss": 1.4699,
+      "step": 273
+    },
+    {
+      "epoch": 0.02404886571528096,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000822,
+      "loss": 1.493,
+      "step": 274
+    },
+    {
+      "epoch": 0.024136635298183443,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0008250000000000001,
+      "loss": 1.5142,
+      "step": 275
+    },
+    {
+      "epoch": 0.024224404881085928,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0008280000000000001,
+      "loss": 1.5442,
+      "step": 276
+    },
+    {
+      "epoch": 0.024312174463988416,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0008310000000000001,
+      "loss": 1.5254,
+      "step": 277
+    },
+    {
+      "epoch": 0.0243999440468909,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0008340000000000001,
+      "loss": 1.4906,
+      "step": 278
+    },
+    {
+      "epoch": 0.024487713629793385,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0008370000000000001,
+      "loss": 1.4933,
+      "step": 279
+    },
+    {
+      "epoch": 0.02457548321269587,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0008400000000000001,
+      "loss": 1.5112,
+      "step": 280
+    },
+    {
+      "epoch": 0.024663252795598355,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0008430000000000001,
+      "loss": 1.4743,
+      "step": 281
+    },
+    {
+      "epoch": 0.02475102237850084,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000846,
+      "loss": 1.5153,
+      "step": 282
+    },
+    {
+      "epoch": 0.024838791961403327,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0008489999999999999,
+      "loss": 1.4923,
+      "step": 283
+    },
+    {
+      "epoch": 0.024926561544305812,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0008519999999999999,
+      "loss": 1.4636,
+      "step": 284
+    },
+    {
+      "epoch": 0.025014331127208297,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000855,
+      "loss": 1.5013,
+      "step": 285
+    },
+    {
+      "epoch": 0.02510210071011078,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0008579999999999999,
+      "loss": 1.4789,
+      "step": 286
+    },
+    {
+      "epoch": 0.025189870293013266,
+      "grad_norm": 0.25,
+      "learning_rate": 0.000861,
+      "loss": 1.4871,
+      "step": 287
+    },
+    {
+      "epoch": 0.02527763987591575,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.000864,
+      "loss": 1.4758,
+      "step": 288
+    },
+    {
+      "epoch": 0.02536540945881824,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0008669999999999999,
+      "loss": 1.4956,
+      "step": 289
+    },
+    {
+      "epoch": 0.025453179041720723,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00087,
+      "loss": 1.5027,
+      "step": 290
+    },
+    {
+      "epoch": 0.025540948624623208,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.000873,
+      "loss": 1.503,
+      "step": 291
+    },
+    {
+      "epoch": 0.025628718207525693,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0008759999999999999,
+      "loss": 1.4981,
+      "step": 292
+    },
+    {
+      "epoch": 0.025716487790428177,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.000879,
+      "loss": 1.5098,
+      "step": 293
+    },
+    {
+      "epoch": 0.025804257373330662,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000882,
+      "loss": 1.4731,
+      "step": 294
+    },
+    {
+      "epoch": 0.02589202695623315,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0008849999999999999,
+      "loss": 1.4688,
+      "step": 295
+    },
+    {
+      "epoch": 0.025979796539135635,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.000888,
+      "loss": 1.4874,
+      "step": 296
+    },
+    {
+      "epoch": 0.02606756612203812,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.000891,
+      "loss": 1.4911,
+      "step": 297
+    },
+    {
+      "epoch": 0.026155335704940604,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0008939999999999999,
+      "loss": 1.51,
+      "step": 298
+    },
+    {
+      "epoch": 0.02624310528784309,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.000897,
+      "loss": 1.509,
+      "step": 299
+    },
+    {
+      "epoch": 0.026330874870745574,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0009,
+      "loss": 1.5112,
+      "step": 300
+    },
+    {
+      "epoch": 0.02641864445364806,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0009029999999999999,
+      "loss": 1.5105,
+      "step": 301
+    },
+    {
+      "epoch": 0.026506414036550546,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000906,
+      "loss": 1.5134,
+      "step": 302
+    },
+    {
+      "epoch": 0.02659418361945303,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.000909,
+      "loss": 1.4693,
+      "step": 303
+    },
+    {
+      "epoch": 0.026681953202355516,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.000912,
+      "loss": 1.4779,
+      "step": 304
+    },
+    {
+      "epoch": 0.026769722785258,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.000915,
+      "loss": 1.4689,
+      "step": 305
+    },
+    {
+      "epoch": 0.02685749236816049,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.000918,
+      "loss": 1.4608,
+      "step": 306
+    },
+    {
+      "epoch": 0.026945261951062973,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.000921,
+      "loss": 1.5206,
+      "step": 307
+    },
+    {
+      "epoch": 0.027033031533965458,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000924,
+      "loss": 1.4968,
+      "step": 308
+    },
+    {
+      "epoch": 0.027120801116867942,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000927,
+      "loss": 1.4968,
+      "step": 309
+    },
+    {
+      "epoch": 0.027208570699770427,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00093,
+      "loss": 1.4736,
+      "step": 310
+    },
+    {
+      "epoch": 0.02729634028267291,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.000933,
+      "loss": 1.4864,
+      "step": 311
+    },
+    {
+      "epoch": 0.0273841098655754,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000936,
+      "loss": 1.4975,
+      "step": 312
+    },
+    {
+      "epoch": 0.027471879448477884,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0009390000000000001,
+      "loss": 1.4502,
+      "step": 313
+    },
+    {
+      "epoch": 0.02755964903138037,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.000942,
+      "loss": 1.4725,
+      "step": 314
+    },
+    {
+      "epoch": 0.027647418614282854,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000945,
+      "loss": 1.505,
+      "step": 315
+    },
+    {
+      "epoch": 0.02773518819718534,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0009480000000000001,
+      "loss": 1.4774,
+      "step": 316
+    },
+    {
+      "epoch": 0.027822957780087823,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.000951,
+      "loss": 1.5004,
+      "step": 317
+    },
+    {
+      "epoch": 0.02791072736299031,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.000954,
+      "loss": 1.4844,
+      "step": 318
+    },
+    {
+      "epoch": 0.027998496945892796,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0009570000000000001,
+      "loss": 1.4924,
+      "step": 319
+    },
+    {
+      "epoch": 0.02808626652879528,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00096,
+      "loss": 1.4896,
+      "step": 320
+    },
+    {
+      "epoch": 0.028174036111697765,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.000963,
+      "loss": 1.5079,
+      "step": 321
+    },
+    {
+      "epoch": 0.02826180569460025,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0009660000000000001,
+      "loss": 1.4842,
+      "step": 322
+    },
+    {
+      "epoch": 0.028349575277502734,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000969,
+      "loss": 1.4825,
+      "step": 323
+    },
+    {
+      "epoch": 0.028437344860405223,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0009720000000000001,
+      "loss": 1.502,
+      "step": 324
+    },
+    {
+      "epoch": 0.028525114443307707,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0009750000000000001,
+      "loss": 1.4996,
+      "step": 325
+    },
+    {
+      "epoch": 0.028612884026210192,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0009780000000000001,
+      "loss": 1.4486,
+      "step": 326
+    },
+    {
+      "epoch": 0.028700653609112677,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000981,
+      "loss": 1.483,
+      "step": 327
+    },
+    {
+      "epoch": 0.02878842319201516,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000984,
+      "loss": 1.465,
+      "step": 328
+    },
+    {
+      "epoch": 0.028876192774917646,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.000987,
+      "loss": 1.4744,
+      "step": 329
+    },
+    {
+      "epoch": 0.028963962357820134,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00099,
+      "loss": 1.4533,
+      "step": 330
+    },
+    {
+      "epoch": 0.02905173194072262,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0009930000000000002,
+      "loss": 1.4598,
+      "step": 331
+    },
+    {
+      "epoch": 0.029139501523625103,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0009960000000000001,
+      "loss": 1.4901,
+      "step": 332
+    },
+    {
+      "epoch": 0.029227271106527588,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.000999,
+      "loss": 1.5001,
+      "step": 333
+    },
+    {
+      "epoch": 0.029315040689430073,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.001002,
+      "loss": 1.4961,
+      "step": 334
+    },
+    {
+      "epoch": 0.029402810272332557,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.001005,
+      "loss": 1.4744,
+      "step": 335
+    },
+    {
+      "epoch": 0.029490579855235045,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.001008,
+      "loss": 1.4319,
+      "step": 336
+    },
+    {
+      "epoch": 0.02957834943813753,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0010110000000000002,
+      "loss": 1.4544,
+      "step": 337
+    },
+    {
+      "epoch": 0.029666119021040015,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0010140000000000001,
+      "loss": 1.4933,
+      "step": 338
+    },
+    {
+      "epoch": 0.0297538886039425,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0010170000000000001,
+      "loss": 1.4171,
+      "step": 339
+    },
+    {
+      "epoch": 0.029841658186844984,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00102,
+      "loss": 1.4674,
+      "step": 340
+    },
+    {
+      "epoch": 0.029929427769747472,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.001023,
+      "loss": 1.4816,
+      "step": 341
+    },
+    {
+      "epoch": 0.030017197352649957,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001026,
+      "loss": 1.5247,
+      "step": 342
+    },
+    {
+      "epoch": 0.03010496693555244,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0010290000000000002,
+      "loss": 1.5058,
+      "step": 343
+    },
+    {
+      "epoch": 0.030192736518454926,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001032,
+      "loss": 1.4926,
+      "step": 344
+    },
+    {
+      "epoch": 0.03028050610135741,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.001035,
+      "loss": 1.4171,
+      "step": 345
+    },
+    {
+      "epoch": 0.030368275684259895,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0010379999999999999,
+      "loss": 1.4589,
+      "step": 346
+    },
+    {
+      "epoch": 0.030456045267162384,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.001041,
+      "loss": 1.4681,
+      "step": 347
+    },
+    {
+      "epoch": 0.030543814850064868,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.001044,
+      "loss": 1.4715,
+      "step": 348
+    },
+    {
+      "epoch": 0.030631584432967353,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.001047,
+      "loss": 1.4296,
+      "step": 349
+    },
+    {
+      "epoch": 0.030719354015869837,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00105,
+      "loss": 1.4862,
+      "step": 350
+    },
+    {
+      "epoch": 0.030807123598772322,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.001053,
+      "loss": 1.4369,
+      "step": 351
+    },
+    {
+      "epoch": 0.030894893181674807,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0010559999999999999,
+      "loss": 1.506,
+      "step": 352
+    },
+    {
+      "epoch": 0.030982662764577295,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001059,
+      "loss": 1.4843,
+      "step": 353
+    },
+    {
+      "epoch": 0.03107043234747978,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.001062,
+      "loss": 1.4721,
+      "step": 354
+    },
+    {
+      "epoch": 0.031158201930382264,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.001065,
+      "loss": 1.4635,
+      "step": 355
+    },
+    {
+      "epoch": 0.03124597151328475,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.001068,
+      "loss": 1.5184,
+      "step": 356
+    },
+    {
+      "epoch": 0.03133374109618724,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001071,
+      "loss": 1.4645,
+      "step": 357
+    },
+    {
+      "epoch": 0.03142151067908972,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.001074,
+      "loss": 1.4442,
+      "step": 358
+    },
+    {
+      "epoch": 0.031509280261992206,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.001077,
+      "loss": 1.4393,
+      "step": 359
+    },
+    {
+      "epoch": 0.03159704984489469,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00108,
+      "loss": 1.495,
+      "step": 360
+    },
+    {
+      "epoch": 0.031684819427797176,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001083,
+      "loss": 1.4464,
+      "step": 361
+    },
+    {
+      "epoch": 0.03177258901069966,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.001086,
+      "loss": 1.4554,
+      "step": 362
+    },
+    {
+      "epoch": 0.031860358593602145,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.001089,
+      "loss": 1.4758,
+      "step": 363
+    },
+    {
+      "epoch": 0.03194812817650463,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.001092,
+      "loss": 1.4644,
+      "step": 364
+    },
+    {
+      "epoch": 0.032035897759407114,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.001095,
+      "loss": 1.4704,
+      "step": 365
+    },
+    {
+      "epoch": 0.0321236673423096,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.001098,
+      "loss": 1.4682,
+      "step": 366
+    },
+    {
+      "epoch": 0.032211436925212084,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.001101,
+      "loss": 1.4806,
+      "step": 367
+    },
+    {
+      "epoch": 0.032299206508114575,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.001104,
+      "loss": 1.4375,
+      "step": 368
+    },
+    {
+      "epoch": 0.03238697609101706,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.001107,
+      "loss": 1.4962,
+      "step": 369
+    },
+    {
+      "epoch": 0.032474745673919544,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00111,
+      "loss": 1.4409,
+      "step": 370
+    },
+    {
+      "epoch": 0.03256251525682203,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.001113,
+      "loss": 1.5093,
+      "step": 371
+    },
+    {
+      "epoch": 0.032650284839724514,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.001116,
+      "loss": 1.4576,
+      "step": 372
+    },
+    {
+      "epoch": 0.032738054422627,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.001119,
+      "loss": 1.4691,
+      "step": 373
+    },
+    {
+      "epoch": 0.03282582400552948,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.001122,
+      "loss": 1.4446,
+      "step": 374
+    },
+    {
+      "epoch": 0.03291359358843197,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0011250000000000001,
+      "loss": 1.4638,
+      "step": 375
+    },
+    {
+      "epoch": 0.03300136317133445,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.001128,
+      "loss": 1.4647,
+      "step": 376
+    },
+    {
+      "epoch": 0.03308913275423694,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.001131,
+      "loss": 1.4811,
+      "step": 377
+    },
+    {
+      "epoch": 0.03317690233713942,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.001134,
+      "loss": 1.4713,
+      "step": 378
+    },
+    {
+      "epoch": 0.03326467192004191,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.001137,
+      "loss": 1.455,
+      "step": 379
+    },
+    {
+      "epoch": 0.0333524415029444,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00114,
+      "loss": 1.4806,
+      "step": 380
+    },
+    {
+      "epoch": 0.03344021108584688,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0011430000000000001,
+      "loss": 1.4486,
+      "step": 381
+    },
+    {
+      "epoch": 0.03352798066874937,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.001146,
+      "loss": 1.5052,
+      "step": 382
+    },
+    {
+      "epoch": 0.03361575025165185,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.001149,
+      "loss": 1.4378,
+      "step": 383
+    },
+    {
+      "epoch": 0.03370351983455434,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.001152,
+      "loss": 1.4448,
+      "step": 384
+    },
+    {
+      "epoch": 0.03379128941745682,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.001155,
+      "loss": 1.4563,
+      "step": 385
+    },
+    {
+      "epoch": 0.033879059000359306,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.001158,
+      "loss": 1.4746,
+      "step": 386
+    },
+    {
+      "epoch": 0.03396682858326179,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0011610000000000001,
+      "loss": 1.4195,
+      "step": 387
+    },
+    {
+      "epoch": 0.034054598166164275,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.001164,
+      "loss": 1.4748,
+      "step": 388
+    },
+    {
+      "epoch": 0.03414236774906676,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.001167,
+      "loss": 1.4435,
+      "step": 389
+    },
+    {
+      "epoch": 0.034230137331969245,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00117,
+      "loss": 1.5054,
+      "step": 390
+    },
+    {
+      "epoch": 0.034317906914871736,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001173,
+      "loss": 1.4637,
+      "step": 391
+    },
+    {
+      "epoch": 0.03440567649777422,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.001176,
+      "loss": 1.4625,
+      "step": 392
+    },
+    {
+      "epoch": 0.034493446080676705,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0011790000000000001,
+      "loss": 1.4551,
+      "step": 393
+    },
+    {
+      "epoch": 0.03458121566357919,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.001182,
+      "loss": 1.4235,
+      "step": 394
+    },
+    {
+      "epoch": 0.034668985246481675,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.001185,
+      "loss": 1.4662,
+      "step": 395
+    },
+    {
+      "epoch": 0.03475675482938416,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.001188,
+      "loss": 1.4661,
+      "step": 396
+    },
+    {
+      "epoch": 0.034844524412286644,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.001191,
+      "loss": 1.4496,
+      "step": 397
+    },
+    {
+      "epoch": 0.03493229399518913,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0011940000000000002,
+      "loss": 1.4556,
+      "step": 398
+    },
+    {
+      "epoch": 0.03502006357809161,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0011970000000000001,
+      "loss": 1.4646,
+      "step": 399
+    },
+    {
+      "epoch": 0.0351078331609941,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0012000000000000001,
+      "loss": 1.4353,
+      "step": 400
+    },
+    {
+      "epoch": 0.03519560274389658,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.001203,
+      "loss": 1.4552,
+      "step": 401
+    },
+    {
+      "epoch": 0.03528337232679907,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.001206,
+      "loss": 1.4238,
+      "step": 402
+    },
+    {
+      "epoch": 0.03537114190970156,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.001209,
+      "loss": 1.467,
+      "step": 403
+    },
+    {
+      "epoch": 0.035458911492604044,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0012120000000000002,
+      "loss": 1.488,
+      "step": 404
+    },
+    {
+      "epoch": 0.03554668107550653,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0012150000000000002,
+      "loss": 1.505,
+      "step": 405
+    },
+    {
+      "epoch": 0.03563445065840901,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0012180000000000001,
+      "loss": 1.4489,
+      "step": 406
+    },
+    {
+      "epoch": 0.0357222202413115,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0012209999999999999,
+      "loss": 1.4478,
+      "step": 407
+    },
+    {
+      "epoch": 0.03580998982421398,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.001224,
+      "loss": 1.5016,
+      "step": 408
+    },
+    {
+      "epoch": 0.03589775940711647,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.001227,
+      "loss": 1.4721,
+      "step": 409
+    },
+    {
+      "epoch": 0.03598552899001895,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00123,
+      "loss": 1.4401,
+      "step": 410
+    },
+    {
+      "epoch": 0.036073298572921436,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.001233,
+      "loss": 1.4316,
+      "step": 411
+    },
+    {
+      "epoch": 0.03616106815582392,
+      "grad_norm": 0.375,
+      "learning_rate": 0.001236,
+      "loss": 1.4899,
+      "step": 412
+    },
+    {
+      "epoch": 0.036248837738726405,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0012389999999999999,
+      "loss": 1.459,
+      "step": 413
+    },
+    {
+      "epoch": 0.0363366073216289,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.001242,
+      "loss": 1.4614,
+      "step": 414
+    },
+    {
+      "epoch": 0.03642437690453138,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001245,
+      "loss": 1.4686,
+      "step": 415
+    },
+    {
+      "epoch": 0.036512146487433866,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001248,
+      "loss": 1.4314,
+      "step": 416
+    },
+    {
+      "epoch": 0.03659991607033635,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.001251,
+      "loss": 1.4779,
+      "step": 417
+    },
+    {
+      "epoch": 0.036687685653238836,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.001254,
+      "loss": 1.4588,
+      "step": 418
+    },
+    {
+      "epoch": 0.03677545523614132,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0012569999999999999,
+      "loss": 1.4517,
+      "step": 419
+    },
+    {
+      "epoch": 0.036863224819043805,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00126,
+      "loss": 1.4708,
+      "step": 420
+    },
+    {
+      "epoch": 0.03695099440194629,
+      "grad_norm": 0.25,
+      "learning_rate": 0.001263,
+      "loss": 1.3873,
+      "step": 421
+    },
+    {
+      "epoch": 0.037038763984848774,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.001266,
+      "loss": 1.4315,
+      "step": 422
+    },
+    {
+      "epoch": 0.03712653356775126,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.001269,
+      "loss": 1.4852,
+      "step": 423
+    },
+    {
+      "epoch": 0.037214303150653744,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.001272,
+      "loss": 1.4162,
+      "step": 424
+    },
+    {
+      "epoch": 0.03730207273355623,
+      "grad_norm": 0.375,
+      "learning_rate": 0.001275,
+      "loss": 1.4792,
+      "step": 425
+    },
+    {
+      "epoch": 0.03738984231645872,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001278,
+      "loss": 1.4004,
+      "step": 426
+    },
+    {
+      "epoch": 0.037477611899361205,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001281,
+      "loss": 1.4025,
+      "step": 427
+    },
+    {
+      "epoch": 0.03756538148226369,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.001284,
+      "loss": 1.4255,
+      "step": 428
+    },
+    {
+      "epoch": 0.037653151065166174,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.001287,
+      "loss": 1.4704,
+      "step": 429
+    },
+    {
+      "epoch": 0.03774092064806866,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00129,
+      "loss": 1.3953,
+      "step": 430
+    },
+    {
+      "epoch": 0.03782869023097114,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.001293,
+      "loss": 1.4279,
+      "step": 431
+    },
+    {
+      "epoch": 0.03791645981387363,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.001296,
+      "loss": 1.4645,
+      "step": 432
+    },
+    {
+      "epoch": 0.03800422939677611,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.001299,
+      "loss": 1.4312,
+      "step": 433
+    },
+    {
+      "epoch": 0.0380919989796786,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.001302,
+      "loss": 1.4486,
+      "step": 434
+    },
+    {
+      "epoch": 0.03817976856258108,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.001305,
+      "loss": 1.4482,
+      "step": 435
+    },
+    {
+      "epoch": 0.038267538145483566,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.001308,
+      "loss": 1.4681,
+      "step": 436
+    },
+    {
+      "epoch": 0.03835530772838605,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.001311,
+      "loss": 1.4597,
+      "step": 437
+    },
+    {
+      "epoch": 0.03844307731128854,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.001314,
+      "loss": 1.4637,
+      "step": 438
+    },
+    {
+      "epoch": 0.03853084689419103,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.001317,
+      "loss": 1.4723,
+      "step": 439
+    },
+    {
+      "epoch": 0.03861861647709351,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00132,
+      "loss": 1.4858,
+      "step": 440
+    },
+    {
+      "epoch": 0.038706386059996,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.001323,
+      "loss": 1.4264,
+      "step": 441
+    },
+    {
+      "epoch": 0.03879415564289848,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0013260000000000001,
+      "loss": 1.4633,
+      "step": 442
+    },
+    {
+      "epoch": 0.038881925225800966,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001329,
+      "loss": 1.4634,
+      "step": 443
+    },
+    {
+      "epoch": 0.03896969480870345,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.001332,
+      "loss": 1.4455,
+      "step": 444
+    },
+    {
+      "epoch": 0.039057464391605935,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.001335,
+      "loss": 1.4572,
+      "step": 445
+    },
+    {
+      "epoch": 0.03914523397450842,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.001338,
+      "loss": 1.4889,
+      "step": 446
+    },
+    {
+      "epoch": 0.039233003557410905,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.001341,
+      "loss": 1.4454,
+      "step": 447
+    },
+    {
+      "epoch": 0.03932077314031339,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0013440000000000001,
+      "loss": 1.4824,
+      "step": 448
+    },
+    {
+      "epoch": 0.03940854272321588,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.001347,
+      "loss": 1.4214,
+      "step": 449
+    },
+    {
+      "epoch": 0.039496312306118365,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00135,
+      "loss": 1.4663,
+      "step": 450
+    },
+    {
+      "epoch": 0.03958408188902085,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001353,
+      "loss": 1.4223,
+      "step": 451
+    },
+    {
+      "epoch": 0.039671851471923335,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.001356,
+      "loss": 1.4393,
+      "step": 452
+    },
+    {
+      "epoch": 0.03975962105482582,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.001359,
+      "loss": 1.4849,
+      "step": 453
+    },
+    {
+      "epoch": 0.039847390637728304,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0013620000000000001,
+      "loss": 1.4257,
+      "step": 454
+    },
+    {
+      "epoch": 0.03993516022063079,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0013650000000000001,
+      "loss": 1.451,
+      "step": 455
+    },
+    {
+      "epoch": 0.04002292980353327,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.001368,
+      "loss": 1.466,
+      "step": 456
+    },
+    {
+      "epoch": 0.04011069938643576,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.001371,
+      "loss": 1.4471,
+      "step": 457
+    },
+    {
+      "epoch": 0.04019846896933824,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.001374,
+      "loss": 1.3836,
+      "step": 458
+    },
+    {
+      "epoch": 0.04028623855224073,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0013770000000000002,
+      "loss": 1.4219,
+      "step": 459
+    },
+    {
+      "epoch": 0.04037400813514321,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0013800000000000002,
+      "loss": 1.4534,
+      "step": 460
+    },
+    {
+      "epoch": 0.040461777718045704,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0013830000000000001,
+      "loss": 1.4593,
+      "step": 461
+    },
+    {
+      "epoch": 0.04054954730094819,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.001386,
+      "loss": 1.4534,
+      "step": 462
+    },
+    {
+      "epoch": 0.04063731688385067,
+      "grad_norm": 1.625,
+      "learning_rate": 0.001389,
+      "loss": 1.4809,
+      "step": 463
+    },
+    {
+      "epoch": 0.04072508646675316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.001392,
+      "loss": 1.452,
+      "step": 464
+    },
+    {
+      "epoch": 0.04081285604965564,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0013950000000000002,
+      "loss": 1.4777,
+      "step": 465
+    },
+    {
+      "epoch": 0.04090062563255813,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0013980000000000002,
+      "loss": 1.4226,
+      "step": 466
+    },
+    {
+      "epoch": 0.04098839521546061,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0014010000000000001,
+      "loss": 1.4141,
+      "step": 467
+    },
+    {
+      "epoch": 0.041076164798363096,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.001404,
+      "loss": 1.4417,
+      "step": 468
+    },
+    {
+      "epoch": 0.04116393438126558,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.001407,
+      "loss": 1.4534,
+      "step": 469
+    },
+    {
+      "epoch": 0.041251703964168066,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00141,
+      "loss": 1.4575,
+      "step": 470
+    },
+    {
+      "epoch": 0.04133947354707055,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.001413,
+      "loss": 1.4112,
+      "step": 471
+    },
+    {
+      "epoch": 0.041427243129973035,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001416,
+      "loss": 1.4343,
+      "step": 472
+    },
+    {
+      "epoch": 0.041515012712875526,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.001419,
+      "loss": 1.4372,
+      "step": 473
+    },
+    {
+      "epoch": 0.04160278229577801,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0014219999999999999,
+      "loss": 1.4323,
+      "step": 474
+    },
+    {
+      "epoch": 0.041690551878680496,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.001425,
+      "loss": 1.4271,
+      "step": 475
+    },
+    {
+      "epoch": 0.04177832146158298,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001428,
+      "loss": 1.4238,
+      "step": 476
+    },
+    {
+      "epoch": 0.041866091044485465,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001431,
+      "loss": 1.4322,
+      "step": 477
+    },
+    {
+      "epoch": 0.04195386062738795,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.001434,
+      "loss": 1.4554,
+      "step": 478
+    },
+    {
+      "epoch": 0.042041630210290434,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.001437,
+      "loss": 1.4249,
+      "step": 479
+    },
+    {
+      "epoch": 0.04212939979319292,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0014399999999999999,
+      "loss": 1.4723,
+      "step": 480
+    },
+    {
+      "epoch": 0.042217169376095404,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.001443,
+      "loss": 1.4306,
+      "step": 481
+    },
+    {
+      "epoch": 0.04230493895899789,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001446,
+      "loss": 1.454,
+      "step": 482
+    },
+    {
+      "epoch": 0.04239270854190037,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.001449,
+      "loss": 1.4104,
+      "step": 483
+    },
+    {
+      "epoch": 0.042480478124802865,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.001452,
+      "loss": 1.4182,
+      "step": 484
+    },
+    {
+      "epoch": 0.04256824770770535,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.001455,
+      "loss": 1.456,
+      "step": 485
+    },
+    {
+      "epoch": 0.042656017290607834,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.001458,
+      "loss": 1.4492,
+      "step": 486
+    },
+    {
+      "epoch": 0.04274378687351032,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.001461,
+      "loss": 1.4531,
+      "step": 487
+    },
+    {
+      "epoch": 0.0428315564564128,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.001464,
+      "loss": 1.4238,
+      "step": 488
+    },
+    {
+      "epoch": 0.04291932603931529,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.001467,
+      "loss": 1.3839,
+      "step": 489
+    },
+    {
+      "epoch": 0.04300709562221777,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00147,
+      "loss": 1.4123,
+      "step": 490
+    },
+    {
+      "epoch": 0.04309486520512026,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.001473,
+      "loss": 1.499,
+      "step": 491
+    },
+    {
+      "epoch": 0.04318263478802274,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001476,
+      "loss": 1.4202,
+      "step": 492
+    },
+    {
+      "epoch": 0.043270404370925226,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.001479,
+      "loss": 1.4173,
+      "step": 493
+    },
+    {
+      "epoch": 0.04335817395382771,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.001482,
+      "loss": 1.4403,
+      "step": 494
+    },
+    {
+      "epoch": 0.043445943536730196,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.001485,
+      "loss": 1.4078,
+      "step": 495
+    },
+    {
+      "epoch": 0.04353371311963269,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.001488,
+      "loss": 1.4551,
+      "step": 496
+    },
+    {
+      "epoch": 0.04362148270253517,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.001491,
+      "loss": 1.4303,
+      "step": 497
+    },
+    {
+      "epoch": 0.04370925228543766,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.001494,
+      "loss": 1.3937,
+      "step": 498
+    },
+    {
+      "epoch": 0.04379702186834014,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.001497,
+      "loss": 1.464,
+      "step": 499
+    },
+    {
+      "epoch": 0.043884791451242626,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0015,
+      "loss": 1.4162,
+      "step": 500
+    },
+    {
+      "epoch": 0.043884791451242626,
+      "eval_loss": 1.4264767169952393,
+      "eval_runtime": 436.4842,
+      "eval_samples_per_second": 33.77,
+      "eval_steps_per_second": 8.442,
+      "step": 500
+    },
+    {
+      "epoch": 0.04397256103414511,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.001503,
+      "loss": 1.4449,
+      "step": 501
+    },
+    {
+      "epoch": 0.044060330617047595,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.001506,
+      "loss": 1.4514,
+      "step": 502
+    },
+    {
+      "epoch": 0.04414810019995008,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0015090000000000001,
+      "loss": 1.394,
+      "step": 503
+    },
+    {
+      "epoch": 0.044235869782852565,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.001512,
+      "loss": 1.4512,
+      "step": 504
+    },
+    {
+      "epoch": 0.04432363936575505,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.001515,
+      "loss": 1.4325,
+      "step": 505
+    },
+    {
+      "epoch": 0.044411408948657534,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.001518,
+      "loss": 1.4595,
+      "step": 506
+    },
+    {
+      "epoch": 0.044499178531560026,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.001521,
+      "loss": 1.4304,
+      "step": 507
+    },
+    {
+      "epoch": 0.04458694811446251,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001524,
+      "loss": 1.4345,
+      "step": 508
+    },
+    {
+      "epoch": 0.044674717697364995,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0015270000000000001,
+      "loss": 1.4321,
+      "step": 509
+    },
+    {
+      "epoch": 0.04476248728026748,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0015300000000000001,
+      "loss": 1.4396,
+      "step": 510
+    },
+    {
+      "epoch": 0.044850256863169964,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.001533,
+      "loss": 1.4269,
+      "step": 511
+    },
+    {
+      "epoch": 0.04493802644607245,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.001536,
+      "loss": 1.424,
+      "step": 512
+    },
+    {
+      "epoch": 0.04502579602897493,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.001539,
+      "loss": 1.4528,
+      "step": 513
+    },
+    {
+      "epoch": 0.04511356561187742,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.001542,
+      "loss": 1.4619,
+      "step": 514
+    },
+    {
+      "epoch": 0.0452013351947799,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0015450000000000001,
+      "loss": 1.3829,
+      "step": 515
+    },
+    {
+      "epoch": 0.04528910477768239,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0015480000000000001,
+      "loss": 1.3826,
+      "step": 516
+    },
+    {
+      "epoch": 0.04537687436058487,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001551,
+      "loss": 1.4539,
+      "step": 517
+    },
+    {
+      "epoch": 0.04546464394348736,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001554,
+      "loss": 1.4542,
+      "step": 518
+    },
+    {
+      "epoch": 0.04555241352638985,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.001557,
+      "loss": 1.3974,
+      "step": 519
+    },
+    {
+      "epoch": 0.04564018310929233,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0015600000000000002,
+      "loss": 1.3831,
+      "step": 520
+    },
+    {
+      "epoch": 0.04572795269219482,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0015630000000000002,
+      "loss": 1.4389,
+      "step": 521
+    },
+    {
+      "epoch": 0.0458157222750973,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0015660000000000001,
+      "loss": 1.4368,
+      "step": 522
+    },
+    {
+      "epoch": 0.04590349185799979,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.001569,
+      "loss": 1.4652,
+      "step": 523
+    },
+    {
+      "epoch": 0.04599126144090227,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.001572,
+      "loss": 1.4344,
+      "step": 524
+    },
+    {
+      "epoch": 0.046079031023804756,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.001575,
+      "loss": 1.413,
+      "step": 525
+    },
+    {
+      "epoch": 0.04616680060670724,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0015780000000000002,
+      "loss": 1.4295,
+      "step": 526
+    },
+    {
+      "epoch": 0.046254570189609726,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0015810000000000002,
+      "loss": 1.4291,
+      "step": 527
+    },
+    {
+      "epoch": 0.04634233977251221,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0015840000000000001,
+      "loss": 1.4636,
+      "step": 528
+    },
+    {
+      "epoch": 0.046430109355414695,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.001587,
+      "loss": 1.4616,
+      "step": 529
+    },
+    {
+      "epoch": 0.04651787893831718,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00159,
+      "loss": 1.4599,
+      "step": 530
+    },
+    {
+      "epoch": 0.04660564852121967,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001593,
+      "loss": 1.4816,
+      "step": 531
+    },
+    {
+      "epoch": 0.046693418104122156,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0015960000000000002,
+      "loss": 1.4006,
+      "step": 532
+    },
+    {
+      "epoch": 0.04678118768702464,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0015990000000000002,
+      "loss": 1.4182,
+      "step": 533
+    },
+    {
+      "epoch": 0.046868957269927125,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0016020000000000001,
+      "loss": 1.4418,
+      "step": 534
+    },
+    {
+      "epoch": 0.04695672685282961,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.001605,
+      "loss": 1.3954,
+      "step": 535
+    },
+    {
+      "epoch": 0.047044496435732094,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.001608,
+      "loss": 1.4086,
+      "step": 536
+    },
+    {
+      "epoch": 0.04713226601863458,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0016110000000000002,
+      "loss": 1.4538,
+      "step": 537
+    },
+    {
+      "epoch": 0.047220035601537064,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0016140000000000002,
+      "loss": 1.4082,
+      "step": 538
+    },
+    {
+      "epoch": 0.04730780518443955,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0016170000000000002,
+      "loss": 1.4373,
+      "step": 539
+    },
+    {
+      "epoch": 0.04739557476734203,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0016200000000000001,
+      "loss": 1.4836,
+      "step": 540
+    },
+    {
+      "epoch": 0.04748334435024452,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.001623,
+      "loss": 1.4136,
+      "step": 541
+    },
+    {
+      "epoch": 0.04757111393314701,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.001626,
+      "loss": 1.4191,
+      "step": 542
+    },
+    {
+      "epoch": 0.047658883516049494,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0016290000000000002,
+      "loss": 1.4583,
+      "step": 543
+    },
+    {
+      "epoch": 0.04774665309895198,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0016320000000000002,
+      "loss": 1.4471,
+      "step": 544
+    },
+    {
+      "epoch": 0.04783442268185446,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0016350000000000002,
+      "loss": 1.5006,
+      "step": 545
+    },
+    {
+      "epoch": 0.04792219226475695,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0016380000000000001,
+      "loss": 1.4461,
+      "step": 546
+    },
+    {
+      "epoch": 0.04800996184765943,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.001641,
+      "loss": 1.4521,
+      "step": 547
+    },
+    {
+      "epoch": 0.04809773143056192,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.001644,
+      "loss": 1.4113,
+      "step": 548
+    },
+    {
+      "epoch": 0.0481855010134644,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0016470000000000002,
+      "loss": 1.4362,
+      "step": 549
+    },
+    {
+      "epoch": 0.048273270596366886,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0016500000000000002,
+      "loss": 1.3773,
+      "step": 550
+    },
+    {
+      "epoch": 0.04836104017926937,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0016530000000000002,
+      "loss": 1.4137,
+      "step": 551
+    },
+    {
+      "epoch": 0.048448809762171856,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0016560000000000001,
+      "loss": 1.3973,
+      "step": 552
+    },
+    {
+      "epoch": 0.04853657934507434,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.001659,
+      "loss": 1.4353,
+      "step": 553
+    },
+    {
+      "epoch": 0.04862434892797683,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0016620000000000003,
+      "loss": 1.3787,
+      "step": 554
+    },
+    {
+      "epoch": 0.04871211851087932,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0016650000000000002,
+      "loss": 1.4034,
+      "step": 555
+    },
+    {
+      "epoch": 0.0487998880937818,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0016680000000000002,
+      "loss": 1.4361,
+      "step": 556
+    },
+    {
+      "epoch": 0.048887657676684286,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0016710000000000002,
+      "loss": 1.4113,
+      "step": 557
+    },
+    {
+      "epoch": 0.04897542725958677,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0016740000000000001,
+      "loss": 1.4935,
+      "step": 558
+    },
+    {
+      "epoch": 0.049063196842489255,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001677,
+      "loss": 1.4107,
+      "step": 559
+    },
+    {
+      "epoch": 0.04915096642539174,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0016800000000000003,
+      "loss": 1.4258,
+      "step": 560
+    },
+    {
+      "epoch": 0.049238736008294225,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0016830000000000003,
+      "loss": 1.4335,
+      "step": 561
+    },
+    {
+      "epoch": 0.04932650559119671,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0016860000000000002,
+      "loss": 1.432,
+      "step": 562
+    },
+    {
+      "epoch": 0.049414275174099194,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.001689,
+      "loss": 1.4373,
+      "step": 563
+    },
+    {
+      "epoch": 0.04950204475700168,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.001692,
+      "loss": 1.3948,
+      "step": 564
+    },
+    {
+      "epoch": 0.04958981433990416,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.001695,
+      "loss": 1.4371,
+      "step": 565
+    },
+    {
+      "epoch": 0.049677583922806655,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0016979999999999999,
+      "loss": 1.3975,
+      "step": 566
+    },
+    {
+      "epoch": 0.04976535350570914,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0017009999999999998,
+      "loss": 1.4313,
+      "step": 567
+    },
+    {
+      "epoch": 0.049853123088611624,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0017039999999999998,
+      "loss": 1.4414,
+      "step": 568
+    },
+    {
+      "epoch": 0.04994089267151411,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.001707,
+      "loss": 1.4192,
+      "step": 569
+    },
+    {
+      "epoch": 0.050028662254416593,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00171,
+      "loss": 1.425,
+      "step": 570
+    },
+    {
+      "epoch": 0.05011643183731908,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.001713,
+      "loss": 1.421,
+      "step": 571
+    },
+    {
+      "epoch": 0.05020420142022156,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0017159999999999999,
+      "loss": 1.4101,
+      "step": 572
+    },
+    {
+      "epoch": 0.05029197100312405,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0017189999999999998,
+      "loss": 1.4414,
+      "step": 573
+    },
+    {
+      "epoch": 0.05037974058602653,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.001722,
+      "loss": 1.4248,
+      "step": 574
+    },
+    {
+      "epoch": 0.05046751016892902,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.001725,
+      "loss": 1.4058,
+      "step": 575
+    },
+    {
+      "epoch": 0.0505552797518315,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.001728,
+      "loss": 1.4156,
+      "step": 576
+    },
+    {
+      "epoch": 0.05064304933473399,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.001731,
+      "loss": 1.4011,
+      "step": 577
+    },
+    {
+      "epoch": 0.05073081891763648,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0017339999999999999,
+      "loss": 1.4422,
+      "step": 578
+    },
+    {
+      "epoch": 0.05081858850053896,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0017369999999999998,
+      "loss": 1.4312,
+      "step": 579
+    },
+    {
+      "epoch": 0.05090635808344145,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00174,
+      "loss": 1.4641,
+      "step": 580
+    },
+    {
+      "epoch": 0.05099412766634393,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.001743,
+      "loss": 1.4104,
+      "step": 581
+    },
+    {
+      "epoch": 0.051081897249246416,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.001746,
+      "loss": 1.3759,
+      "step": 582
+    },
+    {
+      "epoch": 0.0511696668321489,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.001749,
+      "loss": 1.4461,
+      "step": 583
+    },
+    {
+      "epoch": 0.051257436415051386,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0017519999999999999,
+      "loss": 1.4052,
+      "step": 584
+    },
+    {
+      "epoch": 0.05134520599795387,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0017549999999999998,
+      "loss": 1.4361,
+      "step": 585
+    },
+    {
+      "epoch": 0.051432975580856355,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.001758,
+      "loss": 1.4081,
+      "step": 586
+    },
+    {
+      "epoch": 0.05152074516375884,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.001761,
+      "loss": 1.4229,
+      "step": 587
+    },
+    {
+      "epoch": 0.051608514746661324,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.001764,
+      "loss": 1.4724,
+      "step": 588
+    },
+    {
+      "epoch": 0.051696284329563816,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.001767,
+      "loss": 1.4288,
+      "step": 589
+    },
+    {
+      "epoch": 0.0517840539124663,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0017699999999999999,
+      "loss": 1.5109,
+      "step": 590
+    },
+    {
+      "epoch": 0.051871823495368785,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.001773,
+      "loss": 1.4384,
+      "step": 591
+    },
+    {
+      "epoch": 0.05195959307827127,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.001776,
+      "loss": 1.3817,
+      "step": 592
+    },
+    {
+      "epoch": 0.052047362661173754,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.001779,
+      "loss": 1.4257,
+      "step": 593
+    },
+    {
+      "epoch": 0.05213513224407624,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.001782,
+      "loss": 1.4052,
+      "step": 594
+    },
+    {
+      "epoch": 0.052222901826978724,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.001785,
+      "loss": 1.4482,
+      "step": 595
+    },
+    {
+      "epoch": 0.05231067140988121,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0017879999999999999,
+      "loss": 1.4112,
+      "step": 596
+    },
+    {
+      "epoch": 0.05239844099278369,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001791,
+      "loss": 1.4303,
+      "step": 597
+    },
+    {
+      "epoch": 0.05248621057568618,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001794,
+      "loss": 1.4169,
+      "step": 598
+    },
+    {
+      "epoch": 0.05257398015858866,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.001797,
+      "loss": 1.408,
+      "step": 599
+    },
+    {
+      "epoch": 0.05266174974149115,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0018,
+      "loss": 1.4476,
+      "step": 600
+    },
+    {
+      "epoch": 0.05274951932439364,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.001803,
+      "loss": 1.4048,
+      "step": 601
+    },
+    {
+      "epoch": 0.05283728890729612,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0018059999999999999,
+      "loss": 1.3852,
+      "step": 602
+    },
+    {
+      "epoch": 0.05292505849019861,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.001809,
+      "loss": 1.4454,
+      "step": 603
+    },
+    {
+      "epoch": 0.05301282807310109,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.001812,
+      "loss": 1.4336,
+      "step": 604
+    },
+    {
+      "epoch": 0.05310059765600358,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.001815,
+      "loss": 1.4297,
+      "step": 605
+    },
+    {
+      "epoch": 0.05318836723890606,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.001818,
+      "loss": 1.4429,
+      "step": 606
+    },
+    {
+      "epoch": 0.05327613682180855,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.001821,
+      "loss": 1.4271,
+      "step": 607
+    },
+    {
+      "epoch": 0.05336390640471103,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.001824,
+      "loss": 1.4233,
+      "step": 608
+    },
+    {
+      "epoch": 0.053451675987613516,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.001827,
+      "loss": 1.3934,
+      "step": 609
+    },
+    {
+      "epoch": 0.053539445570516,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00183,
+      "loss": 1.4198,
+      "step": 610
+    },
+    {
+      "epoch": 0.053627215153418485,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.001833,
+      "loss": 1.3836,
+      "step": 611
+    },
+    {
+      "epoch": 0.05371498473632098,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.001836,
+      "loss": 1.4215,
+      "step": 612
+    },
+    {
+      "epoch": 0.05380275431922346,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001839,
+      "loss": 1.4335,
+      "step": 613
+    },
+    {
+      "epoch": 0.053890523902125946,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.001842,
+      "loss": 1.4146,
+      "step": 614
+    },
+    {
+      "epoch": 0.05397829348502843,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.001845,
+      "loss": 1.3886,
+      "step": 615
+    },
+    {
+      "epoch": 0.054066063067930915,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.001848,
+      "loss": 1.4687,
+      "step": 616
+    },
+    {
+      "epoch": 0.0541538326508334,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.001851,
+      "loss": 1.4417,
+      "step": 617
+    },
+    {
+      "epoch": 0.054241602233735885,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001854,
+      "loss": 1.4007,
+      "step": 618
+    },
+    {
+      "epoch": 0.05432937181663837,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001857,
+      "loss": 1.3945,
+      "step": 619
+    },
+    {
+      "epoch": 0.054417141399540854,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00186,
+      "loss": 1.4588,
+      "step": 620
+    },
+    {
+      "epoch": 0.05450491098244334,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.001863,
+      "loss": 1.3761,
+      "step": 621
+    },
+    {
+      "epoch": 0.05459268056534582,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.001866,
+      "loss": 1.4146,
+      "step": 622
+    },
+    {
+      "epoch": 0.05468045014824831,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.001869,
+      "loss": 1.4337,
+      "step": 623
+    },
+    {
+      "epoch": 0.0547682197311508,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.001872,
+      "loss": 1.4874,
+      "step": 624
+    },
+    {
+      "epoch": 0.054855989314053284,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.001875,
+      "loss": 1.4406,
+      "step": 625
+    },
+    {
+      "epoch": 0.05494375889695577,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0018780000000000001,
+      "loss": 1.5081,
+      "step": 626
+    },
+    {
+      "epoch": 0.055031528479858254,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.001881,
+      "loss": 1.4169,
+      "step": 627
+    },
+    {
+      "epoch": 0.05511929806276074,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.001884,
+      "loss": 1.4722,
+      "step": 628
+    },
+    {
+      "epoch": 0.05520706764566322,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.001887,
+      "loss": 1.4118,
+      "step": 629
+    },
+    {
+      "epoch": 0.05529483722856571,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00189,
+      "loss": 1.4665,
+      "step": 630
+    },
+    {
+      "epoch": 0.05538260681146819,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0018930000000000002,
+      "loss": 1.4355,
+      "step": 631
+    },
+    {
+      "epoch": 0.05547037639437068,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0018960000000000001,
+      "loss": 1.43,
+      "step": 632
+    },
+    {
+      "epoch": 0.05555814597727316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.001899,
+      "loss": 1.3824,
+      "step": 633
+    },
+    {
+      "epoch": 0.055645915560175646,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.001902,
+      "loss": 1.4145,
+      "step": 634
+    },
+    {
+      "epoch": 0.05573368514307813,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.001905,
+      "loss": 1.4662,
+      "step": 635
+    },
+    {
+      "epoch": 0.05582145472598062,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.001908,
+      "loss": 1.4029,
+      "step": 636
+    },
+    {
+      "epoch": 0.05590922430888311,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0019110000000000002,
+      "loss": 1.385,
+      "step": 637
+    },
+    {
+      "epoch": 0.05599699389178559,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0019140000000000001,
+      "loss": 1.4255,
+      "step": 638
+    },
+    {
+      "epoch": 0.056084763474688076,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.001917,
+      "loss": 1.4397,
+      "step": 639
+    },
+    {
+      "epoch": 0.05617253305759056,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00192,
+      "loss": 1.4382,
+      "step": 640
+    },
+    {
+      "epoch": 0.056260302640493046,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.001923,
+      "loss": 1.4125,
+      "step": 641
+    },
+    {
+      "epoch": 0.05634807222339553,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.001926,
+      "loss": 1.4425,
+      "step": 642
+    },
+    {
+      "epoch": 0.056435841806298015,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0019290000000000002,
+      "loss": 1.4217,
+      "step": 643
+    },
+    {
+      "epoch": 0.0565236113892005,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0019320000000000001,
+      "loss": 1.3992,
+      "step": 644
+    },
+    {
+      "epoch": 0.056611380972102984,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.001935,
+      "loss": 1.3734,
+      "step": 645
+    },
+    {
+      "epoch": 0.05669915055500547,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.001938,
+      "loss": 1.4064,
+      "step": 646
+    },
+    {
+      "epoch": 0.05678692013790796,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.001941,
+      "loss": 1.4006,
+      "step": 647
+    },
+    {
+      "epoch": 0.056874689720810445,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0019440000000000002,
+      "loss": 1.3646,
+      "step": 648
+    },
+    {
+      "epoch": 0.05696245930371293,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0019470000000000002,
+      "loss": 1.4171,
+      "step": 649
+    },
+    {
+      "epoch": 0.057050228886615414,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0019500000000000001,
+      "loss": 1.4012,
+      "step": 650
+    },
+    {
+      "epoch": 0.0571379984695179,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.001953,
+      "loss": 1.434,
+      "step": 651
+    },
+    {
+      "epoch": 0.057225768052420384,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0019560000000000003,
+      "loss": 1.4133,
+      "step": 652
+    },
+    {
+      "epoch": 0.05731353763532287,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0019590000000000002,
+      "loss": 1.3874,
+      "step": 653
+    },
+    {
+      "epoch": 0.05740130721822535,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.001962,
+      "loss": 1.4022,
+      "step": 654
+    },
+    {
+      "epoch": 0.05748907680112784,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.001965,
+      "loss": 1.4111,
+      "step": 655
+    },
+    {
+      "epoch": 0.05757684638403032,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.001968,
+      "loss": 1.454,
+      "step": 656
+    },
+    {
+      "epoch": 0.05766461596693281,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.001971,
+      "loss": 1.4611,
+      "step": 657
+    },
+    {
+      "epoch": 0.05775238554983529,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.001974,
+      "loss": 1.4277,
+      "step": 658
+    },
+    {
+      "epoch": 0.05784015513273778,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.001977,
+      "loss": 1.4443,
+      "step": 659
+    },
+    {
+      "epoch": 0.05792792471564027,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00198,
+      "loss": 1.3763,
+      "step": 660
+    },
+    {
+      "epoch": 0.05801569429854275,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.001983,
+      "loss": 1.3979,
+      "step": 661
+    },
+    {
+      "epoch": 0.05810346388144524,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0019860000000000004,
+      "loss": 1.4839,
+      "step": 662
+    },
+    {
+      "epoch": 0.05819123346434772,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0019890000000000003,
+      "loss": 1.3999,
+      "step": 663
+    },
+    {
+      "epoch": 0.05827900304725021,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0019920000000000003,
+      "loss": 1.3618,
+      "step": 664
+    },
+    {
+      "epoch": 0.05836677263015269,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0019950000000000002,
+      "loss": 1.4032,
+      "step": 665
+    },
+    {
+      "epoch": 0.058454542213055176,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.001998,
+      "loss": 1.4011,
+      "step": 666
+    },
+    {
+      "epoch": 0.05854231179595766,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002001,
+      "loss": 1.4109,
+      "step": 667
+    },
+    {
+      "epoch": 0.058630081378860145,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002004,
+      "loss": 1.4714,
+      "step": 668
+    },
+    {
+      "epoch": 0.05871785096176263,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002007,
+      "loss": 1.3933,
+      "step": 669
+    },
+    {
+      "epoch": 0.058805620544665115,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00201,
+      "loss": 1.3808,
+      "step": 670
+    },
+    {
+      "epoch": 0.058893390127567606,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002013,
+      "loss": 1.3896,
+      "step": 671
+    },
+    {
+      "epoch": 0.05898115971047009,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.002016,
+      "loss": 1.391,
+      "step": 672
+    },
+    {
+      "epoch": 0.059068929293372575,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002019,
+      "loss": 1.378,
+      "step": 673
+    },
+    {
+      "epoch": 0.05915669887627506,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0020220000000000004,
+      "loss": 1.4274,
+      "step": 674
+    },
+    {
+      "epoch": 0.059244468459177545,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0020250000000000003,
+      "loss": 1.4078,
+      "step": 675
+    },
+    {
+      "epoch": 0.05933223804208003,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0020280000000000003,
+      "loss": 1.4316,
+      "step": 676
+    },
+    {
+      "epoch": 0.059420007624982514,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0020310000000000003,
+      "loss": 1.4221,
+      "step": 677
+    },
+    {
+      "epoch": 0.059507777207885,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0020340000000000002,
+      "loss": 1.4176,
+      "step": 678
+    },
+    {
+      "epoch": 0.05959554679078748,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.002037,
+      "loss": 1.4378,
+      "step": 679
+    },
+    {
+      "epoch": 0.05968331637368997,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00204,
+      "loss": 1.4158,
+      "step": 680
+    },
+    {
+      "epoch": 0.05977108595659245,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.002043,
+      "loss": 1.4449,
+      "step": 681
+    },
+    {
+      "epoch": 0.059858855539494944,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.002046,
+      "loss": 1.4289,
+      "step": 682
+    },
+    {
+      "epoch": 0.05994662512239743,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002049,
+      "loss": 1.4242,
+      "step": 683
+    },
+    {
+      "epoch": 0.060034394705299914,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002052,
+      "loss": 1.3802,
+      "step": 684
+    },
+    {
+      "epoch": 0.0601221642882024,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0020550000000000004,
+      "loss": 1.4551,
+      "step": 685
+    },
+    {
+      "epoch": 0.06020993387110488,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.0020580000000000004,
+      "loss": 1.9919,
+      "step": 686
+    },
+    {
+      "epoch": 0.06029770345400737,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0020610000000000003,
+      "loss": 1.469,
+      "step": 687
+    },
+    {
+      "epoch": 0.06038547303690985,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.002064,
+      "loss": 1.4125,
+      "step": 688
+    },
+    {
+      "epoch": 0.06047324261981234,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002067,
+      "loss": 1.4094,
+      "step": 689
+    },
+    {
+      "epoch": 0.06056101220271482,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00207,
+      "loss": 1.3779,
+      "step": 690
+    },
+    {
+      "epoch": 0.060648781785617306,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0020729999999999998,
+      "loss": 1.4507,
+      "step": 691
+    },
+    {
+      "epoch": 0.06073655136851979,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0020759999999999997,
+      "loss": 1.4563,
+      "step": 692
+    },
+    {
+      "epoch": 0.060824320951422275,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0020789999999999997,
+      "loss": 1.4445,
+      "step": 693
+    },
+    {
+      "epoch": 0.06091209053432477,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002082,
+      "loss": 1.432,
+      "step": 694
+    },
+    {
+      "epoch": 0.06099986011722725,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002085,
+      "loss": 1.3819,
+      "step": 695
+    },
+    {
+      "epoch": 0.061087629700129736,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002088,
+      "loss": 1.4557,
+      "step": 696
+    },
+    {
+      "epoch": 0.06117539928303222,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002091,
+      "loss": 1.442,
+      "step": 697
+    },
+    {
+      "epoch": 0.061263168865934706,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002094,
+      "loss": 1.4193,
+      "step": 698
+    },
+    {
+      "epoch": 0.06135093844883719,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.002097,
+      "loss": 1.4591,
+      "step": 699
+    },
+    {
+      "epoch": 0.061438708031739675,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0021,
+      "loss": 1.4352,
+      "step": 700
+    },
+    {
+      "epoch": 0.06152647761464216,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.002103,
+      "loss": 1.4033,
+      "step": 701
+    },
+    {
+      "epoch": 0.061614247197544644,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.002106,
+      "loss": 1.4272,
+      "step": 702
+    },
+    {
+      "epoch": 0.06170201678044713,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0021089999999999998,
+      "loss": 1.39,
+      "step": 703
+    },
+    {
+      "epoch": 0.061789786363349614,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0021119999999999997,
+      "loss": 1.3744,
+      "step": 704
+    },
+    {
+      "epoch": 0.0618775559462521,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.002115,
+      "loss": 1.4513,
+      "step": 705
+    },
+    {
+      "epoch": 0.06196532552915459,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002118,
+      "loss": 1.428,
+      "step": 706
+    },
+    {
+      "epoch": 0.062053095112057075,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002121,
+      "loss": 1.4149,
+      "step": 707
+    },
+    {
+      "epoch": 0.06214086469495956,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002124,
+      "loss": 1.4335,
+      "step": 708
+    },
+    {
+      "epoch": 0.062228634277862044,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.002127,
+      "loss": 1.4268,
+      "step": 709
+    },
+    {
+      "epoch": 0.06231640386076453,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00213,
+      "loss": 1.4299,
+      "step": 710
+    },
+    {
+      "epoch": 0.06240417344366701,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002133,
+      "loss": 1.428,
+      "step": 711
+    },
+    {
+      "epoch": 0.0624919430265695,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002136,
+      "loss": 1.377,
+      "step": 712
+    },
+    {
+      "epoch": 0.06257971260947198,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002139,
+      "loss": 1.3745,
+      "step": 713
+    },
+    {
+      "epoch": 0.06266748219237447,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.002142,
+      "loss": 1.4129,
+      "step": 714
+    },
+    {
+      "epoch": 0.06275525177527695,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0021449999999999998,
+      "loss": 1.4237,
+      "step": 715
+    },
+    {
+      "epoch": 0.06284302135817944,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.002148,
+      "loss": 1.4042,
+      "step": 716
+    },
+    {
+      "epoch": 0.06293079094108192,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.002151,
+      "loss": 1.401,
+      "step": 717
+    },
+    {
+      "epoch": 0.06301856052398441,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002154,
+      "loss": 1.3501,
+      "step": 718
+    },
+    {
+      "epoch": 0.06310633010688689,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002157,
+      "loss": 1.4434,
+      "step": 719
+    },
+    {
+      "epoch": 0.06319409968978938,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00216,
+      "loss": 1.4139,
+      "step": 720
+    },
+    {
+      "epoch": 0.06328186927269186,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.002163,
+      "loss": 1.3906,
+      "step": 721
+    },
+    {
+      "epoch": 0.06336963885559435,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002166,
+      "loss": 1.3987,
+      "step": 722
+    },
+    {
+      "epoch": 0.06345740843849684,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.002169,
+      "loss": 1.4363,
+      "step": 723
+    },
+    {
+      "epoch": 0.06354517802139932,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.002172,
+      "loss": 1.3764,
+      "step": 724
+    },
+    {
+      "epoch": 0.06363294760430181,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002175,
+      "loss": 1.4358,
+      "step": 725
+    },
+    {
+      "epoch": 0.06372071718720429,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002178,
+      "loss": 1.3627,
+      "step": 726
+    },
+    {
+      "epoch": 0.06380848677010678,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0021809999999999998,
+      "loss": 1.4233,
+      "step": 727
+    },
+    {
+      "epoch": 0.06389625635300926,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.002184,
+      "loss": 1.3925,
+      "step": 728
+    },
+    {
+      "epoch": 0.06398402593591175,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002187,
+      "loss": 1.4323,
+      "step": 729
+    },
+    {
+      "epoch": 0.06407179551881423,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00219,
+      "loss": 1.4255,
+      "step": 730
+    },
+    {
+      "epoch": 0.06415956510171672,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.002193,
+      "loss": 1.4131,
+      "step": 731
+    },
+    {
+      "epoch": 0.0642473346846192,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.002196,
+      "loss": 1.415,
+      "step": 732
+    },
+    {
+      "epoch": 0.06433510426752169,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002199,
+      "loss": 1.4095,
+      "step": 733
+    },
+    {
+      "epoch": 0.06442287385042417,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.002202,
+      "loss": 1.4265,
+      "step": 734
+    },
+    {
+      "epoch": 0.06451064343332666,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.002205,
+      "loss": 1.3979,
+      "step": 735
+    },
+    {
+      "epoch": 0.06459841301622915,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002208,
+      "loss": 1.3926,
+      "step": 736
+    },
+    {
+      "epoch": 0.06468618259913163,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002211,
+      "loss": 1.3872,
+      "step": 737
+    },
+    {
+      "epoch": 0.06477395218203412,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002214,
+      "loss": 1.4486,
+      "step": 738
+    },
+    {
+      "epoch": 0.0648617217649366,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0022170000000000002,
+      "loss": 1.3547,
+      "step": 739
+    },
+    {
+      "epoch": 0.06494949134783909,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00222,
+      "loss": 1.4189,
+      "step": 740
+    },
+    {
+      "epoch": 0.06503726093074157,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002223,
+      "loss": 1.4166,
+      "step": 741
+    },
+    {
+      "epoch": 0.06512503051364406,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.002226,
+      "loss": 1.3736,
+      "step": 742
+    },
+    {
+      "epoch": 0.06521280009654654,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.002229,
+      "loss": 1.4195,
+      "step": 743
+    },
+    {
+      "epoch": 0.06530056967944903,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002232,
+      "loss": 1.4215,
+      "step": 744
+    },
+    {
+      "epoch": 0.0653883392623515,
+      "grad_norm": 0.75,
+      "learning_rate": 0.002235,
+      "loss": 1.4432,
+      "step": 745
+    },
+    {
+      "epoch": 0.065476108845254,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002238,
+      "loss": 1.372,
+      "step": 746
+    },
+    {
+      "epoch": 0.06556387842815649,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002241,
+      "loss": 1.4103,
+      "step": 747
+    },
+    {
+      "epoch": 0.06565164801105897,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002244,
+      "loss": 1.4322,
+      "step": 748
+    },
+    {
+      "epoch": 0.06573941759396146,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002247,
+      "loss": 1.3891,
+      "step": 749
+    },
+    {
+      "epoch": 0.06582718717686394,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0022500000000000003,
+      "loss": 1.4126,
+      "step": 750
+    },
+    {
+      "epoch": 0.06591495675976643,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0022530000000000002,
+      "loss": 1.3994,
+      "step": 751
+    },
+    {
+      "epoch": 0.0660027263426689,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.002256,
+      "loss": 1.3941,
+      "step": 752
+    },
+    {
+      "epoch": 0.0660904959255714,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.002259,
+      "loss": 1.3345,
+      "step": 753
+    },
+    {
+      "epoch": 0.06617826550847387,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.002262,
+      "loss": 1.4136,
+      "step": 754
+    },
+    {
+      "epoch": 0.06626603509137637,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.002265,
+      "loss": 1.3705,
+      "step": 755
+    },
+    {
+      "epoch": 0.06635380467427884,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002268,
+      "loss": 1.4587,
+      "step": 756
+    },
+    {
+      "epoch": 0.06644157425718134,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.002271,
+      "loss": 1.432,
+      "step": 757
+    },
+    {
+      "epoch": 0.06652934384008383,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.002274,
+      "loss": 1.3662,
+      "step": 758
+    },
+    {
+      "epoch": 0.0666171134229863,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.002277,
+      "loss": 1.387,
+      "step": 759
+    },
+    {
+      "epoch": 0.0667048830058888,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00228,
+      "loss": 1.3745,
+      "step": 760
+    },
+    {
+      "epoch": 0.06679265258879127,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002283,
+      "loss": 1.4119,
+      "step": 761
+    },
+    {
+      "epoch": 0.06688042217169377,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0022860000000000003,
+      "loss": 1.3951,
+      "step": 762
+    },
+    {
+      "epoch": 0.06696819175459624,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0022890000000000002,
+      "loss": 1.4563,
+      "step": 763
+    },
+    {
+      "epoch": 0.06705596133749873,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.002292,
+      "loss": 1.3564,
+      "step": 764
+    },
+    {
+      "epoch": 0.06714373092040121,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002295,
+      "loss": 1.4115,
+      "step": 765
+    },
+    {
+      "epoch": 0.0672315005033037,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002298,
+      "loss": 1.4028,
+      "step": 766
+    },
+    {
+      "epoch": 0.06731927008620618,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.002301,
+      "loss": 1.3478,
+      "step": 767
+    },
+    {
+      "epoch": 0.06740703966910867,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002304,
+      "loss": 1.3808,
+      "step": 768
+    },
+    {
+      "epoch": 0.06749480925201115,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.002307,
+      "loss": 1.3547,
+      "step": 769
+    },
+    {
+      "epoch": 0.06758257883491364,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00231,
+      "loss": 1.4304,
+      "step": 770
+    },
+    {
+      "epoch": 0.06767034841781613,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.002313,
+      "loss": 1.4404,
+      "step": 771
+    },
+    {
+      "epoch": 0.06775811800071861,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002316,
+      "loss": 1.4134,
+      "step": 772
+    },
+    {
+      "epoch": 0.0678458875836211,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0023190000000000003,
+      "loss": 1.4424,
+      "step": 773
+    },
+    {
+      "epoch": 0.06793365716652358,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0023220000000000003,
+      "loss": 1.4063,
+      "step": 774
+    },
+    {
+      "epoch": 0.06802142674942607,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0023250000000000002,
+      "loss": 1.444,
+      "step": 775
+    },
+    {
+      "epoch": 0.06810919633232855,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002328,
+      "loss": 1.4046,
+      "step": 776
+    },
+    {
+      "epoch": 0.06819696591523104,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002331,
+      "loss": 1.3615,
+      "step": 777
+    },
+    {
+      "epoch": 0.06828473549813352,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.002334,
+      "loss": 1.3941,
+      "step": 778
+    },
+    {
+      "epoch": 0.06837250508103601,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002337,
+      "loss": 1.4284,
+      "step": 779
+    },
+    {
+      "epoch": 0.06846027466393849,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00234,
+      "loss": 1.4889,
+      "step": 780
+    },
+    {
+      "epoch": 0.06854804424684098,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002343,
+      "loss": 1.4139,
+      "step": 781
+    },
+    {
+      "epoch": 0.06863581382974347,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.002346,
+      "loss": 1.3882,
+      "step": 782
+    },
+    {
+      "epoch": 0.06872358341264595,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.002349,
+      "loss": 1.3539,
+      "step": 783
+    },
+    {
+      "epoch": 0.06881135299554844,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002352,
+      "loss": 1.3903,
+      "step": 784
+    },
+    {
+      "epoch": 0.06889912257845092,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0023550000000000003,
+      "loss": 1.4078,
+      "step": 785
+    },
+    {
+      "epoch": 0.06898689216135341,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0023580000000000003,
+      "loss": 1.4124,
+      "step": 786
+    },
+    {
+      "epoch": 0.06907466174425589,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0023610000000000003,
+      "loss": 1.404,
+      "step": 787
+    },
+    {
+      "epoch": 0.06916243132715838,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.002364,
+      "loss": 1.4122,
+      "step": 788
+    },
+    {
+      "epoch": 0.06925020091006086,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.002367,
+      "loss": 1.3725,
+      "step": 789
+    },
+    {
+      "epoch": 0.06933797049296335,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00237,
+      "loss": 1.379,
+      "step": 790
+    },
+    {
+      "epoch": 0.06942574007586583,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.002373,
+      "loss": 1.3556,
+      "step": 791
+    },
+    {
+      "epoch": 0.06951350965876832,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002376,
+      "loss": 1.361,
+      "step": 792
+    },
+    {
+      "epoch": 0.06960127924167081,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.002379,
+      "loss": 1.4277,
+      "step": 793
+    },
+    {
+      "epoch": 0.06968904882457329,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002382,
+      "loss": 1.3974,
+      "step": 794
+    },
+    {
+      "epoch": 0.06977681840747578,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002385,
+      "loss": 1.4055,
+      "step": 795
+    },
+    {
+      "epoch": 0.06986458799037826,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0023880000000000004,
+      "loss": 1.3921,
+      "step": 796
+    },
+    {
+      "epoch": 0.06995235757328075,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0023910000000000003,
+      "loss": 1.4157,
+      "step": 797
+    },
+    {
+      "epoch": 0.07004012715618323,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0023940000000000003,
+      "loss": 1.3959,
+      "step": 798
+    },
+    {
+      "epoch": 0.07012789673908572,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0023970000000000003,
+      "loss": 1.3814,
+      "step": 799
+    },
+    {
+      "epoch": 0.0702156663219882,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0024000000000000002,
+      "loss": 1.3708,
+      "step": 800
+    },
+    {
+      "epoch": 0.07030343590489069,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002403,
+      "loss": 1.4212,
+      "step": 801
+    },
+    {
+      "epoch": 0.07039120548779317,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002406,
+      "loss": 1.3823,
+      "step": 802
+    },
+    {
+      "epoch": 0.07047897507069566,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002409,
+      "loss": 1.4311,
+      "step": 803
+    },
+    {
+      "epoch": 0.07056674465359813,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002412,
+      "loss": 1.3919,
+      "step": 804
+    },
+    {
+      "epoch": 0.07065451423650063,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002415,
+      "loss": 1.4003,
+      "step": 805
+    },
+    {
+      "epoch": 0.07074228381940312,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002418,
+      "loss": 1.4254,
+      "step": 806
+    },
+    {
+      "epoch": 0.0708300534023056,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0024210000000000004,
+      "loss": 1.3777,
+      "step": 807
+    },
+    {
+      "epoch": 0.07091782298520809,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0024240000000000004,
+      "loss": 1.4008,
+      "step": 808
+    },
+    {
+      "epoch": 0.07100559256811056,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0024270000000000003,
+      "loss": 1.3804,
+      "step": 809
+    },
+    {
+      "epoch": 0.07109336215101306,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0024300000000000003,
+      "loss": 1.4175,
+      "step": 810
+    },
+    {
+      "epoch": 0.07118113173391553,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0024330000000000003,
+      "loss": 1.398,
+      "step": 811
+    },
+    {
+      "epoch": 0.07126890131681803,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0024360000000000002,
+      "loss": 1.3582,
+      "step": 812
+    },
+    {
+      "epoch": 0.0713566708997205,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0024389999999999998,
+      "loss": 1.4143,
+      "step": 813
+    },
+    {
+      "epoch": 0.071444440482623,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0024419999999999997,
+      "loss": 1.4149,
+      "step": 814
+    },
+    {
+      "epoch": 0.07153221006552547,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0024449999999999997,
+      "loss": 1.3686,
+      "step": 815
+    },
+    {
+      "epoch": 0.07161997964842796,
+      "grad_norm": 3.828125,
+      "learning_rate": 0.002448,
+      "loss": 1.4106,
+      "step": 816
+    },
+    {
+      "epoch": 0.07170774923133046,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002451,
+      "loss": 1.4604,
+      "step": 817
+    },
+    {
+      "epoch": 0.07179551881423293,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.002454,
+      "loss": 1.4668,
+      "step": 818
+    },
+    {
+      "epoch": 0.07188328839713543,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.002457,
+      "loss": 1.3775,
+      "step": 819
+    },
+    {
+      "epoch": 0.0719710579800379,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00246,
+      "loss": 1.3806,
+      "step": 820
+    },
+    {
+      "epoch": 0.0720588275629404,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.002463,
+      "loss": 1.4512,
+      "step": 821
+    },
+    {
+      "epoch": 0.07214659714584287,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002466,
+      "loss": 1.4076,
+      "step": 822
+    },
+    {
+      "epoch": 0.07223436672874536,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002469,
+      "loss": 1.3972,
+      "step": 823
+    },
+    {
+      "epoch": 0.07232213631164784,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002472,
+      "loss": 1.4491,
+      "step": 824
+    },
+    {
+      "epoch": 0.07240990589455033,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0024749999999999998,
+      "loss": 1.4254,
+      "step": 825
+    },
+    {
+      "epoch": 0.07249767547745281,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0024779999999999997,
+      "loss": 1.3944,
+      "step": 826
+    },
+    {
+      "epoch": 0.0725854450603553,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.002481,
+      "loss": 1.4159,
+      "step": 827
+    },
+    {
+      "epoch": 0.0726732146432578,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002484,
+      "loss": 1.3851,
+      "step": 828
+    },
+    {
+      "epoch": 0.07276098422616027,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002487,
+      "loss": 1.3845,
+      "step": 829
+    },
+    {
+      "epoch": 0.07284875380906276,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00249,
+      "loss": 1.4024,
+      "step": 830
+    },
+    {
+      "epoch": 0.07293652339196524,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.002493,
+      "loss": 1.3894,
+      "step": 831
+    },
+    {
+      "epoch": 0.07302429297486773,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.002496,
+      "loss": 1.4546,
+      "step": 832
+    },
+    {
+      "epoch": 0.07311206255777021,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002499,
+      "loss": 1.4338,
+      "step": 833
+    },
+    {
+      "epoch": 0.0731998321406727,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.002502,
+      "loss": 1.4251,
+      "step": 834
+    },
+    {
+      "epoch": 0.07328760172357518,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002505,
+      "loss": 1.4107,
+      "step": 835
+    },
+    {
+      "epoch": 0.07337537130647767,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.002508,
+      "loss": 1.3467,
+      "step": 836
+    },
+    {
+      "epoch": 0.07346314088938015,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0025109999999999998,
+      "loss": 1.3859,
+      "step": 837
+    },
+    {
+      "epoch": 0.07355091047228264,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0025139999999999997,
+      "loss": 1.3957,
+      "step": 838
+    },
+    {
+      "epoch": 0.07363868005518512,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.002517,
+      "loss": 1.3665,
+      "step": 839
+    },
+    {
+      "epoch": 0.07372644963808761,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00252,
+      "loss": 1.3961,
+      "step": 840
+    },
+    {
+      "epoch": 0.0738142192209901,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002523,
+      "loss": 1.3849,
+      "step": 841
+    },
+    {
+      "epoch": 0.07390198880389258,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002526,
+      "loss": 1.352,
+      "step": 842
+    },
+    {
+      "epoch": 0.07398975838679507,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.002529,
+      "loss": 1.4165,
+      "step": 843
+    },
+    {
+      "epoch": 0.07407752796969755,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.002532,
+      "loss": 1.4083,
+      "step": 844
+    },
+    {
+      "epoch": 0.07416529755260004,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.002535,
+      "loss": 1.4197,
+      "step": 845
+    },
+    {
+      "epoch": 0.07425306713550252,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.002538,
+      "loss": 1.4065,
+      "step": 846
+    },
+    {
+      "epoch": 0.07434083671840501,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.002541,
+      "loss": 1.3912,
+      "step": 847
+    },
+    {
+      "epoch": 0.07442860630130749,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002544,
+      "loss": 1.4331,
+      "step": 848
+    },
+    {
+      "epoch": 0.07451637588420998,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002547,
+      "loss": 1.3746,
+      "step": 849
+    },
+    {
+      "epoch": 0.07460414546711246,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00255,
+      "loss": 1.3978,
+      "step": 850
+    },
+    {
+      "epoch": 0.07469191505001495,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.002553,
+      "loss": 1.4342,
+      "step": 851
+    },
+    {
+      "epoch": 0.07477968463291744,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002556,
+      "loss": 1.369,
+      "step": 852
+    },
+    {
+      "epoch": 0.07486745421581992,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002559,
+      "loss": 1.4062,
+      "step": 853
+    },
+    {
+      "epoch": 0.07495522379872241,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002562,
+      "loss": 1.3848,
+      "step": 854
+    },
+    {
+      "epoch": 0.07504299338162489,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002565,
+      "loss": 1.3805,
+      "step": 855
+    },
+    {
+      "epoch": 0.07513076296452738,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.002568,
+      "loss": 1.3979,
+      "step": 856
+    },
+    {
+      "epoch": 0.07521853254742986,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002571,
+      "loss": 1.3539,
+      "step": 857
+    },
+    {
+      "epoch": 0.07530630213033235,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.002574,
+      "loss": 1.3399,
+      "step": 858
+    },
+    {
+      "epoch": 0.07539407171323483,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.002577,
+      "loss": 1.4018,
+      "step": 859
+    },
+    {
+      "epoch": 0.07548184129613732,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00258,
+      "loss": 1.4543,
+      "step": 860
+    },
+    {
+      "epoch": 0.0755696108790398,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0025830000000000002,
+      "loss": 1.4085,
+      "step": 861
+    },
+    {
+      "epoch": 0.07565738046194229,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002586,
+      "loss": 1.4448,
+      "step": 862
+    },
+    {
+      "epoch": 0.07574515004484478,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.002589,
+      "loss": 1.4506,
+      "step": 863
+    },
+    {
+      "epoch": 0.07583291962774726,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002592,
+      "loss": 1.4231,
+      "step": 864
+    },
+    {
+      "epoch": 0.07592068921064975,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002595,
+      "loss": 1.3962,
+      "step": 865
+    },
+    {
+      "epoch": 0.07600845879355222,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.002598,
+      "loss": 1.3933,
+      "step": 866
+    },
+    {
+      "epoch": 0.07609622837645472,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.002601,
+      "loss": 1.4323,
+      "step": 867
+    },
+    {
+      "epoch": 0.0761839979593572,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.002604,
+      "loss": 1.4321,
+      "step": 868
+    },
+    {
+      "epoch": 0.07627176754225969,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002607,
+      "loss": 1.403,
+      "step": 869
+    },
+    {
+      "epoch": 0.07635953712516216,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00261,
+      "loss": 1.3849,
+      "step": 870
+    },
+    {
+      "epoch": 0.07644730670806466,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002613,
+      "loss": 1.4362,
+      "step": 871
+    },
+    {
+      "epoch": 0.07653507629096713,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002616,
+      "loss": 1.3857,
+      "step": 872
+    },
+    {
+      "epoch": 0.07662284587386962,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0026190000000000002,
+      "loss": 1.424,
+      "step": 873
+    },
+    {
+      "epoch": 0.0767106154567721,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.002622,
+      "loss": 1.4226,
+      "step": 874
+    },
+    {
+      "epoch": 0.0767983850396746,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002625,
+      "loss": 1.4058,
+      "step": 875
+    },
+    {
+      "epoch": 0.07688615462257709,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002628,
+      "loss": 1.3957,
+      "step": 876
+    },
+    {
+      "epoch": 0.07697392420547956,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002631,
+      "loss": 1.4533,
+      "step": 877
+    },
+    {
+      "epoch": 0.07706169378838205,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002634,
+      "loss": 1.3528,
+      "step": 878
+    },
+    {
+      "epoch": 0.07714946337128453,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002637,
+      "loss": 1.3916,
+      "step": 879
+    },
+    {
+      "epoch": 0.07723723295418702,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00264,
+      "loss": 1.449,
+      "step": 880
+    },
+    {
+      "epoch": 0.0773250025370895,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002643,
+      "loss": 1.4141,
+      "step": 881
+    },
+    {
+      "epoch": 0.077412772119992,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002646,
+      "loss": 1.4228,
+      "step": 882
+    },
+    {
+      "epoch": 0.07750054170289447,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.002649,
+      "loss": 1.3928,
+      "step": 883
+    },
+    {
+      "epoch": 0.07758831128579696,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0026520000000000003,
+      "loss": 1.3593,
+      "step": 884
+    },
+    {
+      "epoch": 0.07767608086869944,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0026550000000000002,
+      "loss": 1.3993,
+      "step": 885
+    },
+    {
+      "epoch": 0.07776385045160193,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.002658,
+      "loss": 1.3602,
+      "step": 886
+    },
+    {
+      "epoch": 0.07785162003450442,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002661,
+      "loss": 1.4019,
+      "step": 887
+    },
+    {
+      "epoch": 0.0779393896174069,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002664,
+      "loss": 1.3701,
+      "step": 888
+    },
+    {
+      "epoch": 0.07802715920030939,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.002667,
+      "loss": 1.4444,
+      "step": 889
+    },
+    {
+      "epoch": 0.07811492878321187,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00267,
+      "loss": 1.401,
+      "step": 890
+    },
+    {
+      "epoch": 0.07820269836611436,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002673,
+      "loss": 1.409,
+      "step": 891
+    },
+    {
+      "epoch": 0.07829046794901684,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.002676,
+      "loss": 1.3678,
+      "step": 892
+    },
+    {
+      "epoch": 0.07837823753191933,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.002679,
+      "loss": 1.4032,
+      "step": 893
+    },
+    {
+      "epoch": 0.07846600711482181,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.002682,
+      "loss": 1.3723,
+      "step": 894
+    },
+    {
+      "epoch": 0.0785537766977243,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0026850000000000003,
+      "loss": 1.3817,
+      "step": 895
+    },
+    {
+      "epoch": 0.07864154628062678,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0026880000000000003,
+      "loss": 1.4086,
+      "step": 896
+    },
+    {
+      "epoch": 0.07872931586352927,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0026910000000000002,
+      "loss": 1.3719,
+      "step": 897
+    },
+    {
+      "epoch": 0.07881708544643176,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.002694,
+      "loss": 1.3896,
+      "step": 898
+    },
+    {
+      "epoch": 0.07890485502933424,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.002697,
+      "loss": 1.4354,
+      "step": 899
+    },
+    {
+      "epoch": 0.07899262461223673,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0027,
+      "loss": 1.4313,
+      "step": 900
+    },
+    {
+      "epoch": 0.07908039419513921,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.002703,
+      "loss": 1.4053,
+      "step": 901
+    },
+    {
+      "epoch": 0.0791681637780417,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002706,
+      "loss": 1.4214,
+      "step": 902
+    },
+    {
+      "epoch": 0.07925593336094418,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002709,
+      "loss": 1.3898,
+      "step": 903
+    },
+    {
+      "epoch": 0.07934370294384667,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.002712,
+      "loss": 1.4251,
+      "step": 904
+    },
+    {
+      "epoch": 0.07943147252674915,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.002715,
+      "loss": 1.3688,
+      "step": 905
+    },
+    {
+      "epoch": 0.07951924210965164,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002718,
+      "loss": 1.377,
+      "step": 906
+    },
+    {
+      "epoch": 0.07960701169255412,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0027210000000000003,
+      "loss": 1.4229,
+      "step": 907
+    },
+    {
+      "epoch": 0.07969478127545661,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0027240000000000003,
+      "loss": 1.4327,
+      "step": 908
+    },
+    {
+      "epoch": 0.07978255085835909,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0027270000000000003,
+      "loss": 1.3982,
+      "step": 909
+    },
+    {
+      "epoch": 0.07987032044126158,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0027300000000000002,
+      "loss": 1.392,
+      "step": 910
+    },
+    {
+      "epoch": 0.07995809002416407,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002733,
+      "loss": 1.4287,
+      "step": 911
+    },
+    {
+      "epoch": 0.08004585960706655,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.002736,
+      "loss": 1.3982,
+      "step": 912
+    },
+    {
+      "epoch": 0.08013362918996904,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.002739,
+      "loss": 1.3988,
+      "step": 913
+    },
+    {
+      "epoch": 0.08022139877287152,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002742,
+      "loss": 1.3755,
+      "step": 914
+    },
+    {
+      "epoch": 0.08030916835577401,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002745,
+      "loss": 1.4671,
+      "step": 915
+    },
+    {
+      "epoch": 0.08039693793867649,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002748,
+      "loss": 1.4094,
+      "step": 916
+    },
+    {
+      "epoch": 0.08048470752157898,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002751,
+      "loss": 1.4038,
+      "step": 917
+    },
+    {
+      "epoch": 0.08057247710448145,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0027540000000000004,
+      "loss": 1.4396,
+      "step": 918
+    },
+    {
+      "epoch": 0.08066024668738395,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0027570000000000003,
+      "loss": 1.3817,
+      "step": 919
+    },
+    {
+      "epoch": 0.08074801627028642,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0027600000000000003,
+      "loss": 1.382,
+      "step": 920
+    },
+    {
+      "epoch": 0.08083578585318892,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0027630000000000003,
+      "loss": 1.4015,
+      "step": 921
+    },
+    {
+      "epoch": 0.08092355543609141,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0027660000000000002,
+      "loss": 1.4056,
+      "step": 922
+    },
+    {
+      "epoch": 0.08101132501899388,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.002769,
+      "loss": 1.4484,
+      "step": 923
+    },
+    {
+      "epoch": 0.08109909460189638,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.002772,
+      "loss": 1.4534,
+      "step": 924
+    },
+    {
+      "epoch": 0.08118686418479885,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.002775,
+      "loss": 1.3962,
+      "step": 925
+    },
+    {
+      "epoch": 0.08127463376770135,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.002778,
+      "loss": 1.3683,
+      "step": 926
+    },
+    {
+      "epoch": 0.08136240335060382,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002781,
+      "loss": 1.3443,
+      "step": 927
+    },
+    {
+      "epoch": 0.08145017293350632,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.002784,
+      "loss": 1.3491,
+      "step": 928
+    },
+    {
+      "epoch": 0.08153794251640879,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0027870000000000004,
+      "loss": 1.3446,
+      "step": 929
+    },
+    {
+      "epoch": 0.08162571209931128,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0027900000000000004,
+      "loss": 1.3753,
+      "step": 930
+    },
+    {
+      "epoch": 0.08171348168221376,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0027930000000000003,
+      "loss": 1.4317,
+      "step": 931
+    },
+    {
+      "epoch": 0.08180125126511625,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0027960000000000003,
+      "loss": 1.4003,
+      "step": 932
+    },
+    {
+      "epoch": 0.08188902084801875,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0027990000000000003,
+      "loss": 1.3705,
+      "step": 933
+    },
+    {
+      "epoch": 0.08197679043092122,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0028020000000000002,
+      "loss": 1.4637,
+      "step": 934
+    },
+    {
+      "epoch": 0.08206456001382371,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002805,
+      "loss": 1.3788,
+      "step": 935
+    },
+    {
+      "epoch": 0.08215232959672619,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.002808,
+      "loss": 1.4177,
+      "step": 936
+    },
+    {
+      "epoch": 0.08224009917962868,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.002811,
+      "loss": 1.365,
+      "step": 937
+    },
+    {
+      "epoch": 0.08232786876253116,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.002814,
+      "loss": 1.403,
+      "step": 938
+    },
+    {
+      "epoch": 0.08241563834543365,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.002817,
+      "loss": 1.4323,
+      "step": 939
+    },
+    {
+      "epoch": 0.08250340792833613,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00282,
+      "loss": 1.3889,
+      "step": 940
+    },
+    {
+      "epoch": 0.08259117751123862,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.002823,
+      "loss": 1.3413,
+      "step": 941
+    },
+    {
+      "epoch": 0.0826789470941411,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002826,
+      "loss": 1.4498,
+      "step": 942
+    },
+    {
+      "epoch": 0.08276671667704359,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.002829,
+      "loss": 1.4469,
+      "step": 943
+    },
+    {
+      "epoch": 0.08285448625994607,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.002832,
+      "loss": 1.3705,
+      "step": 944
+    },
+    {
+      "epoch": 0.08294225584284856,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002835,
+      "loss": 1.3415,
+      "step": 945
+    },
+    {
+      "epoch": 0.08303002542575105,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002838,
+      "loss": 1.4412,
+      "step": 946
+    },
+    {
+      "epoch": 0.08311779500865353,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0028409999999999998,
+      "loss": 1.4073,
+      "step": 947
+    },
+    {
+      "epoch": 0.08320556459155602,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0028439999999999997,
+      "loss": 1.4343,
+      "step": 948
+    },
+    {
+      "epoch": 0.0832933341744585,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.002847,
+      "loss": 1.427,
+      "step": 949
+    },
+    {
+      "epoch": 0.08338110375736099,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00285,
+      "loss": 1.3348,
+      "step": 950
+    },
+    {
+      "epoch": 0.08346887334026347,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.002853,
+      "loss": 1.3768,
+      "step": 951
+    },
+    {
+      "epoch": 0.08355664292316596,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002856,
+      "loss": 1.4057,
+      "step": 952
+    },
+    {
+      "epoch": 0.08364441250606844,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002859,
+      "loss": 1.4168,
+      "step": 953
+    },
+    {
+      "epoch": 0.08373218208897093,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002862,
+      "loss": 1.3865,
+      "step": 954
+    },
+    {
+      "epoch": 0.08381995167187341,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002865,
+      "loss": 1.447,
+      "step": 955
+    },
+    {
+      "epoch": 0.0839077212547759,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.002868,
+      "loss": 1.3856,
+      "step": 956
+    },
+    {
+      "epoch": 0.08399549083767839,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002871,
+      "loss": 1.4196,
+      "step": 957
+    },
+    {
+      "epoch": 0.08408326042058087,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.002874,
+      "loss": 1.3786,
+      "step": 958
+    },
+    {
+      "epoch": 0.08417103000348336,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002877,
+      "loss": 1.3898,
+      "step": 959
+    },
+    {
+      "epoch": 0.08425879958638584,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0028799999999999997,
+      "loss": 1.4289,
+      "step": 960
+    },
+    {
+      "epoch": 0.08434656916928833,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.002883,
+      "loss": 1.3777,
+      "step": 961
+    },
+    {
+      "epoch": 0.08443433875219081,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002886,
+      "loss": 1.4581,
+      "step": 962
+    },
+    {
+      "epoch": 0.0845221083350933,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.002889,
+      "loss": 1.3526,
+      "step": 963
+    },
+    {
+      "epoch": 0.08460987791799578,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002892,
+      "loss": 1.404,
+      "step": 964
+    },
+    {
+      "epoch": 0.08469764750089827,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002895,
+      "loss": 1.3726,
+      "step": 965
+    },
+    {
+      "epoch": 0.08478541708380075,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.002898,
+      "loss": 1.423,
+      "step": 966
+    },
+    {
+      "epoch": 0.08487318666670324,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002901,
+      "loss": 1.4026,
+      "step": 967
+    },
+    {
+      "epoch": 0.08496095624960573,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002904,
+      "loss": 1.4215,
+      "step": 968
+    },
+    {
+      "epoch": 0.0850487258325082,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002907,
+      "loss": 1.3334,
+      "step": 969
+    },
+    {
+      "epoch": 0.0851364954154107,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00291,
+      "loss": 1.4332,
+      "step": 970
+    },
+    {
+      "epoch": 0.08522426499831318,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002913,
+      "loss": 1.3709,
+      "step": 971
+    },
+    {
+      "epoch": 0.08531203458121567,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002916,
+      "loss": 1.378,
+      "step": 972
+    },
+    {
+      "epoch": 0.08539980416411815,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.002919,
+      "loss": 1.3483,
+      "step": 973
+    },
+    {
+      "epoch": 0.08548757374702064,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.002922,
+      "loss": 1.3419,
+      "step": 974
+    },
+    {
+      "epoch": 0.08557534332992311,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.002925,
+      "loss": 1.4303,
+      "step": 975
+    },
+    {
+      "epoch": 0.0856631129128256,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.002928,
+      "loss": 1.3703,
+      "step": 976
+    },
+    {
+      "epoch": 0.08575088249572808,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002931,
+      "loss": 1.3757,
+      "step": 977
+    },
+    {
+      "epoch": 0.08583865207863058,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.002934,
+      "loss": 1.4041,
+      "step": 978
+    },
+    {
+      "epoch": 0.08592642166153307,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002937,
+      "loss": 1.3838,
+      "step": 979
+    },
+    {
+      "epoch": 0.08601419124443554,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00294,
+      "loss": 1.3663,
+      "step": 980
+    },
+    {
+      "epoch": 0.08610196082733804,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002943,
+      "loss": 1.3884,
+      "step": 981
+    },
+    {
+      "epoch": 0.08618973041024051,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.002946,
+      "loss": 1.4175,
+      "step": 982
+    },
+    {
+      "epoch": 0.086277499993143,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0029490000000000002,
+      "loss": 1.3364,
+      "step": 983
+    },
+    {
+      "epoch": 0.08636526957604548,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002952,
+      "loss": 1.3714,
+      "step": 984
+    },
+    {
+      "epoch": 0.08645303915894798,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.002955,
+      "loss": 1.3786,
+      "step": 985
+    },
+    {
+      "epoch": 0.08654080874185045,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002958,
+      "loss": 1.4111,
+      "step": 986
+    },
+    {
+      "epoch": 0.08662857832475294,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.002961,
+      "loss": 1.4381,
+      "step": 987
+    },
+    {
+      "epoch": 0.08671634790765542,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002964,
+      "loss": 1.3721,
+      "step": 988
+    },
+    {
+      "epoch": 0.08680411749055791,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002967,
+      "loss": 1.4126,
+      "step": 989
+    },
+    {
+      "epoch": 0.08689188707346039,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00297,
+      "loss": 1.4663,
+      "step": 990
+    },
+    {
+      "epoch": 0.08697965665636288,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002973,
+      "loss": 1.382,
+      "step": 991
+    },
+    {
+      "epoch": 0.08706742623926537,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002976,
+      "loss": 1.4019,
+      "step": 992
+    },
+    {
+      "epoch": 0.08715519582216785,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002979,
+      "loss": 1.462,
+      "step": 993
+    },
+    {
+      "epoch": 0.08724296540507034,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002982,
+      "loss": 1.421,
+      "step": 994
+    },
+    {
+      "epoch": 0.08733073498797282,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0029850000000000002,
+      "loss": 1.3908,
+      "step": 995
+    },
+    {
+      "epoch": 0.08741850457087531,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.002988,
+      "loss": 1.3896,
+      "step": 996
+    },
+    {
+      "epoch": 0.08750627415377779,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002991,
+      "loss": 1.4022,
+      "step": 997
+    },
+    {
+      "epoch": 0.08759404373668028,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.002994,
+      "loss": 1.3985,
+      "step": 998
+    },
+    {
+      "epoch": 0.08768181331958276,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.002997,
+      "loss": 1.3715,
+      "step": 999
+    },
+    {
+      "epoch": 0.08776958290248525,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.003,
+      "loss": 1.3632,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08776958290248525,
+      "eval_loss": 1.3825047016143799,
+      "eval_runtime": 437.7802,
+      "eval_samples_per_second": 33.67,
+      "eval_steps_per_second": 8.417,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08785735248538773,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0029999999333801706,
+      "loss": 1.4102,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08794512206829022,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0029999997335206897,
+      "loss": 1.4553,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08803289165119271,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0029999994004215767,
+      "loss": 1.3328,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08812066123409519,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002999998934082865,
+      "loss": 1.4013,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08820843081699768,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0029999983345045996,
+      "loss": 1.3431,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08829620039990016,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0029999976016868404,
+      "loss": 1.3888,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08838396998280265,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0029999967356296596,
+      "loss": 1.4219,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08847173956570513,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002999995736333143,
+      "loss": 1.4155,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08855950914860762,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0029999946037973886,
+      "loss": 1.3674,
+      "step": 1009
+    },
+    {
+      "epoch": 0.0886472787315101,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0029999933380225087,
+      "loss": 1.3753,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08873504831441259,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0029999919390086274,
+      "loss": 1.3757,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08882281789731507,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0029999904067558834,
+      "loss": 1.3702,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08891058748021756,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002999988741264429,
+      "loss": 1.3923,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08899835706312005,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0029999869425344264,
+      "loss": 1.37,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08908612664602253,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0029999850105660547,
+      "loss": 1.4014,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08917389622892502,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0029999829453595038,
+      "loss": 1.3861,
+      "step": 1016
+    },
+    {
+      "epoch": 0.0892616658118275,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002999980746914978,
+      "loss": 1.4112,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08934943539472999,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002999978415232694,
+      "loss": 1.3873,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08943720497763247,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002999975950312882,
+      "loss": 1.404,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08952497456053496,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002999973352155785,
+      "loss": 1.4123,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08961274414343744,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00299997062076166,
+      "loss": 1.4009,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08970051372633993,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0029999677561307766,
+      "loss": 1.352,
+      "step": 1022
+    },
+    {
+      "epoch": 0.0897882833092424,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002999964758263417,
+      "loss": 1.4426,
+      "step": 1023
+    },
+    {
+      "epoch": 0.0898760528921449,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.002999961627159877,
+      "loss": 1.4257,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08996382247504738,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.002999958362820466,
+      "loss": 1.4002,
+      "step": 1025
+    },
+    {
+      "epoch": 0.09005159205794987,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0029999549652455063,
+      "loss": 1.413,
+      "step": 1026
+    },
+    {
+      "epoch": 0.09013936164085236,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.002999951434435333,
+      "loss": 1.3596,
+      "step": 1027
+    },
+    {
+      "epoch": 0.09022713122375484,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0029999477703902947,
+      "loss": 1.4353,
+      "step": 1028
+    },
+    {
+      "epoch": 0.09031490080665733,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0029999439731107527,
+      "loss": 1.3752,
+      "step": 1029
+    },
+    {
+      "epoch": 0.0904026703895598,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.002999940042597082,
+      "loss": 1.386,
+      "step": 1030
+    },
+    {
+      "epoch": 0.0904904399724623,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0029999359788496715,
+      "loss": 1.3796,
+      "step": 1031
+    },
+    {
+      "epoch": 0.09057820955536477,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00299993178186892,
+      "loss": 1.3642,
+      "step": 1032
+    },
+    {
+      "epoch": 0.09066597913826727,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0029999274516552435,
+      "loss": 1.3919,
+      "step": 1033
+    },
+    {
+      "epoch": 0.09075374872116974,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002999922988209069,
+      "loss": 1.3464,
+      "step": 1034
+    },
+    {
+      "epoch": 0.09084151830407224,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002999918391530837,
+      "loss": 1.3758,
+      "step": 1035
+    },
+    {
+      "epoch": 0.09092928788697471,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002999913661621001,
+      "loss": 1.408,
+      "step": 1036
+    },
+    {
+      "epoch": 0.0910170574698772,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0029999087984800277,
+      "loss": 1.3728,
+      "step": 1037
+    },
+    {
+      "epoch": 0.0911048270527797,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.002999903802108398,
+      "loss": 1.3839,
+      "step": 1038
+    },
+    {
+      "epoch": 0.09119259663568217,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029998986725066033,
+      "loss": 1.3796,
+      "step": 1039
+    },
+    {
+      "epoch": 0.09128036621858467,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002999893409675152,
+      "loss": 1.3723,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09136813580148714,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0029998880136145613,
+      "loss": 1.3627,
+      "step": 1041
+    },
+    {
+      "epoch": 0.09145590538438964,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002999882484325366,
+      "loss": 1.4139,
+      "step": 1042
+    },
+    {
+      "epoch": 0.09154367496729211,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00299987682180811,
+      "loss": 1.4144,
+      "step": 1043
+    },
+    {
+      "epoch": 0.0916314445501946,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002999871026063353,
+      "loss": 1.335,
+      "step": 1044
+    },
+    {
+      "epoch": 0.09171921413309708,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.002999865097091667,
+      "loss": 1.36,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09180698371599957,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0029998590348936376,
+      "loss": 1.3949,
+      "step": 1046
+    },
+    {
+      "epoch": 0.09189475329890205,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0029998528394698616,
+      "loss": 1.3764,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09198252288180454,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002999846510820952,
+      "loss": 1.3687,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09207029246470703,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0029998400489475332,
+      "loss": 1.3591,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09215806204760951,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0029998334538502416,
+      "loss": 1.3819,
+      "step": 1050
+    },
+    {
+      "epoch": 0.092245831630512,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0029998267255297298,
+      "loss": 1.3957,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09233360121341448,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0029998198639866612,
+      "loss": 1.3797,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09242137079631697,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0029998128692217125,
+      "loss": 1.4328,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09250914037921945,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0029998057412355754,
+      "loss": 1.425,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09259690996212194,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.002999798480028952,
+      "loss": 1.4552,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09268467954502442,
+      "grad_norm": 4.5625,
+      "learning_rate": 0.0029997910856025597,
+      "loss": 1.4378,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09277244912792691,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0029997835579571278,
+      "loss": 1.397,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09286021871082939,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0029997758970933997,
+      "loss": 1.3356,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09294798829373188,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0029997681030121313,
+      "loss": 1.422,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09303575787663436,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002999760175714092,
+      "loss": 1.4501,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09312352745953685,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0029997521152000644,
+      "loss": 1.4169,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09321129704243934,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0029997439214708432,
+      "loss": 1.3878,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09329906662534182,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.002999735594527238,
+      "loss": 1.406,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09338683620824431,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00299972713437007,
+      "loss": 1.3747,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09347460579114679,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0029997185410001745,
+      "loss": 1.3693,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09356237537404928,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0029997098144183996,
+      "loss": 1.4782,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09365014495695176,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0029997009546256065,
+      "loss": 1.3834,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09373791453985425,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00299969196162267,
+      "loss": 1.3232,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09382568412275673,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0029996828354104765,
+      "loss": 1.4099,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09391345370565922,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002999673575989929,
+      "loss": 1.3691,
+      "step": 1070
+    },
+    {
+      "epoch": 0.0940012232885617,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002999664183361939,
+      "loss": 1.4001,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09408899287146419,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002999654657527434,
+      "loss": 1.3827,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09417676245436668,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002999644998487355,
+      "loss": 1.3673,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09426453203726916,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0029996352062426554,
+      "loss": 1.3937,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09435230162017165,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0029996252807943007,
+      "loss": 1.359,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09444007120307413,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0029996152221432714,
+      "loss": 1.4212,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09452784078597662,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0029996050302905594,
+      "loss": 1.3969,
+      "step": 1077
+    },
+    {
+      "epoch": 0.0946156103688791,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002999594705237171,
+      "loss": 1.4725,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09470337995178159,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002999584246984126,
+      "loss": 1.3726,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09479114953468407,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.002999573655532455,
+      "loss": 1.3471,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09487891911758656,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0029995629308832046,
+      "loss": 1.3358,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09496668870048904,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.002999552073037433,
+      "loss": 1.3811,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09505445828339153,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002999541081996211,
+      "loss": 1.4059,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09514222786629402,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0029995299577606246,
+      "loss": 1.411,
+      "step": 1084
+    },
+    {
+      "epoch": 0.0952299974491965,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0029995187003317714,
+      "loss": 1.344,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09531776703209899,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0029995073097107616,
+      "loss": 1.4106,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09540553661500147,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0029994957858987207,
+      "loss": 1.41,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09549330619790396,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0029994841288967854,
+      "loss": 1.3824,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09558107578080643,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.002999472338706106,
+      "loss": 1.3149,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09566884536370893,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.002999460415327846,
+      "loss": 1.4327,
+      "step": 1090
+    },
+    {
+      "epoch": 0.0957566149466114,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0029994483587631833,
+      "loss": 1.3614,
+      "step": 1091
+    },
+    {
+      "epoch": 0.0958443845295139,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0029994361690133067,
+      "loss": 1.4151,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09593215411241637,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0029994238460794194,
+      "loss": 1.4326,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09601992369531887,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0029994113899627383,
+      "loss": 1.4586,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09610769327822134,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0029993988006644925,
+      "loss": 1.4072,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09619546286112383,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002999386078185924,
+      "loss": 1.3702,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09628323244402633,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002999373222528289,
+      "loss": 1.3838,
+      "step": 1097
+    },
+    {
+      "epoch": 0.0963710020269288,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0029993602336928563,
+      "loss": 1.3674,
+      "step": 1098
+    },
+    {
+      "epoch": 0.0964587716098313,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029993471116809074,
+      "loss": 1.4215,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09654654119273377,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0029993338564937378,
+      "loss": 1.3732,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09663431077563626,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0029993204681326557,
+      "loss": 1.3703,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09672208035853874,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002999306946598982,
+      "loss": 1.3311,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09680984994144123,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.002999293291894052,
+      "loss": 1.4013,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09689761952434371,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0029992795040192127,
+      "loss": 1.3463,
+      "step": 1104
+    },
+    {
+      "epoch": 0.0969853891072462,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0029992655829758253,
+      "loss": 1.3948,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09707315869014868,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002999251528765264,
+      "loss": 1.3679,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09716092827305117,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0029992373413889146,
+      "loss": 1.3756,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09724869785595366,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0029992230208481787,
+      "loss": 1.4107,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09733646743885614,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00299920856714447,
+      "loss": 1.3858,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09742423702175863,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002999193980279213,
+      "loss": 1.3736,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09751200660466111,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0029991792602538495,
+      "loss": 1.4024,
+      "step": 1111
+    },
+    {
+      "epoch": 0.0975997761875636,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0029991644070698317,
+      "loss": 1.342,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09768754577046608,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0029991494207286247,
+      "loss": 1.3804,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09777531535336857,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002999134301231708,
+      "loss": 1.4002,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09786308493627105,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0029991190485805746,
+      "loss": 1.406,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09795085451917354,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0029991036627767293,
+      "loss": 1.4043,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09803862410207602,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0029990881438216902,
+      "loss": 1.3829,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09812639368497851,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00299907249171699,
+      "loss": 1.4053,
+      "step": 1118
+    },
+    {
+      "epoch": 0.098214163267881,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0029990567064641723,
+      "loss": 1.4058,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09830193285078348,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0029990407880647954,
+      "loss": 1.3976,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09838970243368597,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0029990247365204313,
+      "loss": 1.3871,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09847747201658845,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.002999008551832663,
+      "loss": 1.3627,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09856524159949094,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0029989922340030884,
+      "loss": 1.4073,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09865301118239342,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0029989757830333184,
+      "loss": 1.4222,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09874078076529591,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002998959198924976,
+      "loss": 1.416,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09882855034819839,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0029989424816796984,
+      "loss": 1.3457,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09891631993110088,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0029989256312991353,
+      "loss": 1.3658,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09900408951400336,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002998908647784949,
+      "loss": 1.4118,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09909185909690585,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002998891531138818,
+      "loss": 1.3455,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09917962867980833,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.002998874281362429,
+      "loss": 1.3987,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09926739826271082,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.002998856898457486,
+      "loss": 1.3671,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09935516784561331,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0029988393824257045,
+      "loss": 1.4391,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09944293742851579,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0029988217332688133,
+      "loss": 1.3633,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09953070701141828,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0029988039509885535,
+      "loss": 1.3812,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09961847659432076,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.002998786035586681,
+      "loss": 1.3976,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09970624617722325,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0029987679870649635,
+      "loss": 1.3556,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09979401576012573,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0029987498054251825,
+      "loss": 1.3307,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09988178534302822,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0029987314906691326,
+      "loss": 1.3842,
+      "step": 1138
+    },
+    {
+      "epoch": 0.0999695549259307,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0029987130427986206,
+      "loss": 1.3949,
+      "step": 1139
+    },
+    {
+      "epoch": 0.10005732450883319,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0029986944618154688,
+      "loss": 1.4001,
+      "step": 1140
+    },
+    {
+      "epoch": 0.10014509409173566,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0029986757477215094,
+      "loss": 1.3963,
+      "step": 1141
+    },
+    {
+      "epoch": 0.10023286367463816,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029986569005185904,
+      "loss": 1.3784,
+      "step": 1142
+    },
+    {
+      "epoch": 0.10032063325754065,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0029986379202085714,
+      "loss": 1.3445,
+      "step": 1143
+    },
+    {
+      "epoch": 0.10040840284044313,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0029986188067933264,
+      "loss": 1.3893,
+      "step": 1144
+    },
+    {
+      "epoch": 0.10049617242334562,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0029985995602747413,
+      "loss": 1.3833,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1005839420062481,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0029985801806547154,
+      "loss": 1.412,
+      "step": 1146
+    },
+    {
+      "epoch": 0.10067171158915059,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002998560667935162,
+      "loss": 1.3944,
+      "step": 1147
+    },
+    {
+      "epoch": 0.10075948117205306,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002998541022118007,
+      "loss": 1.3138,
+      "step": 1148
+    },
+    {
+      "epoch": 0.10084725075495556,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0029985212432051883,
+      "loss": 1.383,
+      "step": 1149
+    },
+    {
+      "epoch": 0.10093502033785803,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.00299850133119866,
+      "loss": 1.293,
+      "step": 1150
+    },
+    {
+      "epoch": 0.10102278992076053,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002998481286100385,
+      "loss": 1.464,
+      "step": 1151
+    },
+    {
+      "epoch": 0.101110559503663,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0029984611079123427,
+      "loss": 1.3577,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1011983290865655,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0029984407966365248,
+      "loss": 1.3531,
+      "step": 1153
+    },
+    {
+      "epoch": 0.10128609866946799,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.002998420352274936,
+      "loss": 1.3539,
+      "step": 1154
+    },
+    {
+      "epoch": 0.10137386825237046,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0029983997748295934,
+      "loss": 1.4027,
+      "step": 1155
+    },
+    {
+      "epoch": 0.10146163783527296,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.002998379064302529,
+      "loss": 1.3267,
+      "step": 1156
+    },
+    {
+      "epoch": 0.10154940741817543,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.002998358220695786,
+      "loss": 1.3353,
+      "step": 1157
+    },
+    {
+      "epoch": 0.10163717700107792,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0029983372440114215,
+      "loss": 1.3713,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1017249465839804,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002998316134251506,
+      "loss": 1.3866,
+      "step": 1159
+    },
+    {
+      "epoch": 0.1018127161668829,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0029982948914181237,
+      "loss": 1.418,
+      "step": 1160
+    },
+    {
+      "epoch": 0.10190048574978537,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00299827351551337,
+      "loss": 1.3636,
+      "step": 1161
+    },
+    {
+      "epoch": 0.10198825533268786,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0029982520065393557,
+      "loss": 1.3576,
+      "step": 1162
+    },
+    {
+      "epoch": 0.10207602491559034,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0029982303644982026,
+      "loss": 1.406,
+      "step": 1163
+    },
+    {
+      "epoch": 0.10216379449849283,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0029982085893920474,
+      "loss": 1.4094,
+      "step": 1164
+    },
+    {
+      "epoch": 0.10225156408139531,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0029981866812230393,
+      "loss": 1.3641,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1023393336642978,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0029981646399933402,
+      "loss": 1.3666,
+      "step": 1166
+    },
+    {
+      "epoch": 0.1024271032472003,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002998142465705125,
+      "loss": 1.3237,
+      "step": 1167
+    },
+    {
+      "epoch": 0.10251487283010277,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0029981201583605833,
+      "loss": 1.3855,
+      "step": 1168
+    },
+    {
+      "epoch": 0.10260264241300526,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.002998097717961916,
+      "loss": 1.3694,
+      "step": 1169
+    },
+    {
+      "epoch": 0.10269041199590774,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.002998075144511338,
+      "loss": 1.4139,
+      "step": 1170
+    },
+    {
+      "epoch": 0.10277818157881023,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0029980524380110777,
+      "loss": 1.3603,
+      "step": 1171
+    },
+    {
+      "epoch": 0.10286595116171271,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0029980295984633754,
+      "loss": 1.3754,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1029537207446152,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0029980066258704858,
+      "loss": 1.4013,
+      "step": 1173
+    },
+    {
+      "epoch": 0.10304149032751768,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0029979835202346757,
+      "loss": 1.3714,
+      "step": 1174
+    },
+    {
+      "epoch": 0.10312925991042017,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0029979602815582264,
+      "loss": 1.393,
+      "step": 1175
+    },
+    {
+      "epoch": 0.10321702949332265,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0029979369098434307,
+      "loss": 1.371,
+      "step": 1176
+    },
+    {
+      "epoch": 0.10330479907622514,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0029979134050925954,
+      "loss": 1.3506,
+      "step": 1177
+    },
+    {
+      "epoch": 0.10339256865912763,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00299788976730804,
+      "loss": 1.3718,
+      "step": 1178
+    },
+    {
+      "epoch": 0.10348033824203011,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0029978659964920986,
+      "loss": 1.3942,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1035681078249326,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0029978420926471167,
+      "loss": 1.3182,
+      "step": 1180
+    },
+    {
+      "epoch": 0.10365587740783508,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0029978180557754526,
+      "loss": 1.3925,
+      "step": 1181
+    },
+    {
+      "epoch": 0.10374364699073757,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00299779388587948,
+      "loss": 1.4286,
+      "step": 1182
+    },
+    {
+      "epoch": 0.10383141657364005,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0029977695829615835,
+      "loss": 1.3791,
+      "step": 1183
+    },
+    {
+      "epoch": 0.10391918615654254,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0029977451470241624,
+      "loss": 1.3654,
+      "step": 1184
+    },
+    {
+      "epoch": 0.10400695573944502,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0029977205780696275,
+      "loss": 1.3322,
+      "step": 1185
+    },
+    {
+      "epoch": 0.10409472532234751,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.002997695876100405,
+      "loss": 1.3478,
+      "step": 1186
+    },
+    {
+      "epoch": 0.10418249490524999,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0029976710411189315,
+      "loss": 1.4218,
+      "step": 1187
+    },
+    {
+      "epoch": 0.10427026448815248,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.002997646073127659,
+      "loss": 1.4149,
+      "step": 1188
+    },
+    {
+      "epoch": 0.10435803407105497,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002997620972129051,
+      "loss": 1.386,
+      "step": 1189
+    },
+    {
+      "epoch": 0.10444580365395745,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0029975957381255863,
+      "loss": 1.3308,
+      "step": 1190
+    },
+    {
+      "epoch": 0.10453357323685994,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0029975703711197536,
+      "loss": 1.3653,
+      "step": 1191
+    },
+    {
+      "epoch": 0.10462134281976242,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002997544871114057,
+      "loss": 1.3197,
+      "step": 1192
+    },
+    {
+      "epoch": 0.10470911240266491,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.002997519238111014,
+      "loss": 1.3795,
+      "step": 1193
+    },
+    {
+      "epoch": 0.10479688198556739,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0029974934721131545,
+      "loss": 1.3968,
+      "step": 1194
+    },
+    {
+      "epoch": 0.10488465156846988,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0029974675731230206,
+      "loss": 1.358,
+      "step": 1195
+    },
+    {
+      "epoch": 0.10497242115137236,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002997441541143169,
+      "loss": 1.3964,
+      "step": 1196
+    },
+    {
+      "epoch": 0.10506019073427485,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002997415376176169,
+      "loss": 1.3614,
+      "step": 1197
+    },
+    {
+      "epoch": 0.10514796031717732,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0029973890782246026,
+      "loss": 1.358,
+      "step": 1198
+    },
+    {
+      "epoch": 0.10523572990007982,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0029973626472910657,
+      "loss": 1.384,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1053234994829823,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.002997336083378167,
+      "loss": 1.401,
+      "step": 1200
+    },
+    {
+      "epoch": 0.10541126906588479,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002997309386488527,
+      "loss": 1.3944,
+      "step": 1201
+    },
+    {
+      "epoch": 0.10549903864878728,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0029972825566247823,
+      "loss": 1.4017,
+      "step": 1202
+    },
+    {
+      "epoch": 0.10558680823168975,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0029972555937895803,
+      "loss": 1.3158,
+      "step": 1203
+    },
+    {
+      "epoch": 0.10567457781459225,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002997228497985582,
+      "loss": 1.4139,
+      "step": 1204
+    },
+    {
+      "epoch": 0.10576234739749472,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0029972012692154612,
+      "loss": 1.3973,
+      "step": 1205
+    },
+    {
+      "epoch": 0.10585011698039722,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0029971739074819066,
+      "loss": 1.3658,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1059378865632997,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002997146412787617,
+      "loss": 1.3152,
+      "step": 1207
+    },
+    {
+      "epoch": 0.10602565614620219,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0029971187851353075,
+      "loss": 1.4202,
+      "step": 1208
+    },
+    {
+      "epoch": 0.10611342572910466,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0029970910245277036,
+      "loss": 1.3919,
+      "step": 1209
+    },
+    {
+      "epoch": 0.10620119531200715,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002997063130967546,
+      "loss": 1.3949,
+      "step": 1210
+    },
+    {
+      "epoch": 0.10628896489490963,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0029970351044575874,
+      "loss": 1.3882,
+      "step": 1211
+    },
+    {
+      "epoch": 0.10637673447781212,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.002997006945000594,
+      "loss": 1.3757,
+      "step": 1212
+    },
+    {
+      "epoch": 0.10646450406071462,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029969786525993448,
+      "loss": 1.4256,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1065522736436171,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0029969502272566327,
+      "loss": 1.4044,
+      "step": 1214
+    },
+    {
+      "epoch": 0.10664004322651958,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0029969216689752624,
+      "loss": 1.4365,
+      "step": 1215
+    },
+    {
+      "epoch": 0.10672781280942206,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0029968929777580526,
+      "loss": 1.3836,
+      "step": 1216
+    },
+    {
+      "epoch": 0.10681558239232455,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002996864153607836,
+      "loss": 1.3908,
+      "step": 1217
+    },
+    {
+      "epoch": 0.10690335197522703,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0029968351965274564,
+      "loss": 1.3522,
+      "step": 1218
+    },
+    {
+      "epoch": 0.10699112155812952,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0029968061065197725,
+      "loss": 1.391,
+      "step": 1219
+    },
+    {
+      "epoch": 0.107078891141032,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002996776883587654,
+      "loss": 1.389,
+      "step": 1220
+    },
+    {
+      "epoch": 0.10716666072393449,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0029967475277339864,
+      "loss": 1.376,
+      "step": 1221
+    },
+    {
+      "epoch": 0.10725443030683697,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.002996718038961667,
+      "loss": 1.3482,
+      "step": 1222
+    },
+    {
+      "epoch": 0.10734219988973946,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0029966884172736054,
+      "loss": 1.3258,
+      "step": 1223
+    },
+    {
+      "epoch": 0.10742996947264195,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002996658662672726,
+      "loss": 1.3769,
+      "step": 1224
+    },
+    {
+      "epoch": 0.10751773905554443,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0029966287751619644,
+      "loss": 1.3529,
+      "step": 1225
+    },
+    {
+      "epoch": 0.10760550863844692,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0029965987547442715,
+      "loss": 1.3632,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1076932782213494,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00299656860142261,
+      "loss": 1.3678,
+      "step": 1227
+    },
+    {
+      "epoch": 0.10778104780425189,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.002996538315199955,
+      "loss": 1.302,
+      "step": 1228
+    },
+    {
+      "epoch": 0.10786881738715437,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.002996507896079297,
+      "loss": 1.4266,
+      "step": 1229
+    },
+    {
+      "epoch": 0.10795658697005686,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002996477344063637,
+      "loss": 1.3641,
+      "step": 1230
+    },
+    {
+      "epoch": 0.10804435655295934,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.002996446659155991,
+      "loss": 1.3934,
+      "step": 1231
+    },
+    {
+      "epoch": 0.10813212613586183,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.002996415841359387,
+      "loss": 1.3668,
+      "step": 1232
+    },
+    {
+      "epoch": 0.10821989571876431,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0029963848906768677,
+      "loss": 1.445,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1083076653016668,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0029963538071114864,
+      "loss": 1.3484,
+      "step": 1234
+    },
+    {
+      "epoch": 0.10839543488456928,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.002996322590666312,
+      "loss": 1.3515,
+      "step": 1235
+    },
+    {
+      "epoch": 0.10848320446747177,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002996291241344425,
+      "loss": 1.3772,
+      "step": 1236
+    },
+    {
+      "epoch": 0.10857097405037426,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0029962597591489198,
+      "loss": 1.38,
+      "step": 1237
+    },
+    {
+      "epoch": 0.10865874363327674,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.002996228144082903,
+      "loss": 1.435,
+      "step": 1238
+    },
+    {
+      "epoch": 0.10874651321617923,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0029961963961494947,
+      "loss": 1.4508,
+      "step": 1239
+    },
+    {
+      "epoch": 0.10883428279908171,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.002996164515351829,
+      "loss": 1.3645,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1089220523819842,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002996132501693052,
+      "loss": 1.3712,
+      "step": 1241
+    },
+    {
+      "epoch": 0.10900982196488668,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0029961003551763235,
+      "loss": 1.3633,
+      "step": 1242
+    },
+    {
+      "epoch": 0.10909759154778917,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0029960680758048166,
+      "loss": 1.3638,
+      "step": 1243
+    },
+    {
+      "epoch": 0.10918536113069165,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0029960356635817162,
+      "loss": 1.3499,
+      "step": 1244
+    },
+    {
+      "epoch": 0.10927313071359414,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002996003118510222,
+      "loss": 1.3687,
+      "step": 1245
+    },
+    {
+      "epoch": 0.10936090029649662,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.002995970440593546,
+      "loss": 1.3434,
+      "step": 1246
+    },
+    {
+      "epoch": 0.10944866987939911,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0029959376298349136,
+      "loss": 1.3438,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1095364394623016,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.002995904686237562,
+      "loss": 1.3487,
+      "step": 1248
+    },
+    {
+      "epoch": 0.10962420904520408,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0029958716098047434,
+      "loss": 1.3443,
+      "step": 1249
+    },
+    {
+      "epoch": 0.10971197862810657,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0029958384005397227,
+      "loss": 1.4267,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10979974821100905,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.002995805058445777,
+      "loss": 1.4072,
+      "step": 1251
+    },
+    {
+      "epoch": 0.10988751779391154,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002995771583526197,
+      "loss": 1.4223,
+      "step": 1252
+    },
+    {
+      "epoch": 0.10997528737681402,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0029957379757842868,
+      "loss": 1.4047,
+      "step": 1253
+    },
+    {
+      "epoch": 0.11006305695971651,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0029957042352233634,
+      "loss": 1.3857,
+      "step": 1254
+    },
+    {
+      "epoch": 0.11015082654261898,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.002995670361846756,
+      "loss": 1.3768,
+      "step": 1255
+    },
+    {
+      "epoch": 0.11023859612552148,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0029956363556578086,
+      "loss": 1.3833,
+      "step": 1256
+    },
+    {
+      "epoch": 0.11032636570842395,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0029956022166598775,
+      "loss": 1.3569,
+      "step": 1257
+    },
+    {
+      "epoch": 0.11041413529132645,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0029955679448563323,
+      "loss": 1.3996,
+      "step": 1258
+    },
+    {
+      "epoch": 0.11050190487422894,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0029955335402505547,
+      "loss": 1.3399,
+      "step": 1259
+    },
+    {
+      "epoch": 0.11058967445713141,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00299549900284594,
+      "loss": 1.4107,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1106774440400339,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002995464332645899,
+      "loss": 1.3458,
+      "step": 1261
+    },
+    {
+      "epoch": 0.11076521362293638,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0029954295296538507,
+      "loss": 1.2937,
+      "step": 1262
+    },
+    {
+      "epoch": 0.11085298320583888,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0029953945938732325,
+      "loss": 1.3608,
+      "step": 1263
+    },
+    {
+      "epoch": 0.11094075278874135,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.002995359525307491,
+      "loss": 1.463,
+      "step": 1264
+    },
+    {
+      "epoch": 0.11102852237164385,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0029953243239600872,
+      "loss": 1.4157,
+      "step": 1265
+    },
+    {
+      "epoch": 0.11111629195454632,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0029952889898344965,
+      "loss": 1.3454,
+      "step": 1266
+    },
+    {
+      "epoch": 0.11120406153744881,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0029952535229342053,
+      "loss": 1.379,
+      "step": 1267
+    },
+    {
+      "epoch": 0.11129183112035129,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002995217923262714,
+      "loss": 1.3982,
+      "step": 1268
+    },
+    {
+      "epoch": 0.11137960070325378,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0029951821908235365,
+      "loss": 1.3906,
+      "step": 1269
+    },
+    {
+      "epoch": 0.11146737028615626,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0029951463256201994,
+      "loss": 1.3601,
+      "step": 1270
+    },
+    {
+      "epoch": 0.11155513986905875,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0029951103276562433,
+      "loss": 1.3265,
+      "step": 1271
+    },
+    {
+      "epoch": 0.11164290945196124,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.002995074196935219,
+      "loss": 1.3922,
+      "step": 1272
+    },
+    {
+      "epoch": 0.11173067903486372,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002995037933460694,
+      "loss": 1.3893,
+      "step": 1273
+    },
+    {
+      "epoch": 0.11181844861776621,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0029950015372362474,
+      "loss": 1.3614,
+      "step": 1274
+    },
+    {
+      "epoch": 0.11190621820066869,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002994965008265471,
+      "loss": 1.3425,
+      "step": 1275
+    },
+    {
+      "epoch": 0.11199398778357118,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00299492834655197,
+      "loss": 1.329,
+      "step": 1276
+    },
+    {
+      "epoch": 0.11208175736647366,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.002994891552099362,
+      "loss": 1.3674,
+      "step": 1277
+    },
+    {
+      "epoch": 0.11216952694937615,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0029948546249112798,
+      "loss": 1.3518,
+      "step": 1278
+    },
+    {
+      "epoch": 0.11225729653227863,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002994817564991368,
+      "loss": 1.3763,
+      "step": 1279
+    },
+    {
+      "epoch": 0.11234506611518112,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.002994780372343283,
+      "loss": 1.3855,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1124328356980836,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0029947430469706962,
+      "loss": 1.3847,
+      "step": 1281
+    },
+    {
+      "epoch": 0.11252060528098609,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.002994705588877292,
+      "loss": 1.3432,
+      "step": 1282
+    },
+    {
+      "epoch": 0.11260837486388858,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0029946679980667667,
+      "loss": 1.3996,
+      "step": 1283
+    },
+    {
+      "epoch": 0.11269614444679106,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0029946302745428306,
+      "loss": 1.3817,
+      "step": 1284
+    },
+    {
+      "epoch": 0.11278391402969355,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0029945924183092066,
+      "loss": 1.3386,
+      "step": 1285
+    },
+    {
+      "epoch": 0.11287168361259603,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.002994554429369632,
+      "loss": 1.4025,
+      "step": 1286
+    },
+    {
+      "epoch": 0.11295945319549852,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.002994516307727855,
+      "loss": 1.3984,
+      "step": 1287
+    },
+    {
+      "epoch": 0.113047222778401,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0029944780533876377,
+      "loss": 1.3728,
+      "step": 1288
+    },
+    {
+      "epoch": 0.11313499236130349,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0029944396663527566,
+      "loss": 1.3559,
+      "step": 1289
+    },
+    {
+      "epoch": 0.11322276194420597,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0029944011466270007,
+      "loss": 1.3149,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11331053152710846,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0029943624942141704,
+      "loss": 1.3477,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11339830111001094,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.002994323709118082,
+      "loss": 1.3523,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11348607069291343,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002994284791342563,
+      "loss": 1.3462,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11357384027581592,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0029942457408914536,
+      "loss": 1.3958,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1136616098587184,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0029942065577686083,
+      "loss": 1.3697,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11374937944162089,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0029941672419778946,
+      "loss": 1.3842,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11383714902452337,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0029941277935231925,
+      "loss": 1.3605,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11392491860742586,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002994088212408396,
+      "loss": 1.4183,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11401268819032834,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0029940484986374113,
+      "loss": 1.4284,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11410045777323083,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0029940086522141584,
+      "loss": 1.3648,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1141882273561333,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0029939686731425682,
+      "loss": 1.4495,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1142759969390358,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0029939285614265887,
+      "loss": 1.3307,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11436376652193828,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002993888317070178,
+      "loss": 1.3453,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11445153610484077,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029938479400773072,
+      "loss": 1.3246,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11453930568774325,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002993807430451963,
+      "loss": 1.3524,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11462707527064574,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0029937667881981414,
+      "loss": 1.3968,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11471484485354823,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0029937260133198556,
+      "loss": 1.3437,
+      "step": 1307
+    },
+    {
+      "epoch": 0.1148026144364507,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002993685105821129,
+      "loss": 1.4529,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1148903840193532,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002993644065705999,
+      "loss": 1.3231,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11497815360225568,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0029936028929785164,
+      "loss": 1.4442,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11506592318515817,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0029935615876427444,
+      "loss": 1.3429,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11515369276806064,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00299352014970276,
+      "loss": 1.3844,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11524146235096314,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0029934785791626527,
+      "loss": 1.3531,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11532923193386561,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002993436876026526,
+      "loss": 1.3314,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1154170015167681,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0029933950402984957,
+      "loss": 1.3913,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11550477109967058,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.002993353071982689,
+      "loss": 1.4322,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11559254068257308,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0029933109710832506,
+      "loss": 1.3768,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11568031026547557,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0029932687376043347,
+      "loss": 1.3952,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11576807984837804,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0029932263715501087,
+      "loss": 1.3741,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11585584943128054,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0029931838729247554,
+      "loss": 1.4054,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11594361901418301,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0029931412417324678,
+      "loss": 1.3497,
+      "step": 1321
+    },
+    {
+      "epoch": 0.1160313885970855,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.002993098477977455,
+      "loss": 1.404,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11611915817998798,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0029930555816639363,
+      "loss": 1.4049,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11620692776289047,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.002993012552796146,
+      "loss": 1.3405,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11629469734579295,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.002992969391378331,
+      "loss": 1.3917,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11638246692869544,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0029929260974147502,
+      "loss": 1.396,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11647023651159792,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.002992882670909678,
+      "loss": 1.3071,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11655800609450041,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.002992839111867399,
+      "loss": 1.3579,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1166457756774029,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.002992795420292214,
+      "loss": 1.3896,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11673354526030538,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0029927515961884338,
+      "loss": 1.3201,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11682131484320787,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.002992707639560384,
+      "loss": 1.3282,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11690908442611035,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0029926635504124027,
+      "loss": 1.3491,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11699685400901284,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002992619328748842,
+      "loss": 1.4268,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11708462359191532,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002992574974574066,
+      "loss": 1.3757,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11717239317481781,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002992530487892452,
+      "loss": 1.3592,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11726016275772029,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0029924858687083913,
+      "loss": 1.349,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11734793234062278,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002992441117026287,
+      "loss": 1.367,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11743570192352526,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0029923962328505566,
+      "loss": 1.3852,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11752347150642775,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0029923512161856296,
+      "loss": 1.3678,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11761124108933023,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0029923060670359485,
+      "loss": 1.3732,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11769901067223272,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00299226078540597,
+      "loss": 1.3959,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11778678025513521,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0029922153713001633,
+      "loss": 1.3595,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11787454983803769,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0029921698247230106,
+      "loss": 1.3821,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11796231942094018,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0029921241456790063,
+      "loss": 1.3625,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11805008900384266,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0029920783341726596,
+      "loss": 1.4052,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11813785858674515,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0029920323902084918,
+      "loss": 1.3745,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11822562816964763,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029919863137910364,
+      "loss": 1.3355,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11831339775255012,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002991940104924843,
+      "loss": 1.3472,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1184011673354526,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00299189376361447,
+      "loss": 1.3406,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11848893691835509,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002991847289864493,
+      "loss": 1.3594,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11857670650125757,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0029918006836794973,
+      "loss": 1.3208,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11866447608416006,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0029917539450640836,
+      "loss": 1.3718,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11875224566706255,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0029917070740228644,
+      "loss": 1.3373,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11884001524996503,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002991660070560466,
+      "loss": 1.3994,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11892778483286752,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0029916129346815273,
+      "loss": 1.3803,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11901555441577,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0029915656663907003,
+      "loss": 1.3329,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11910332399867249,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0029915182656926505,
+      "loss": 1.3175,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11919109358157497,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0029914707325920564,
+      "loss": 1.3346,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11927886316447746,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0029914230670936084,
+      "loss": 1.3422,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11936663274737994,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0029913752692020115,
+      "loss": 1.3332,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11945440233028243,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.002991327338921983,
+      "loss": 1.3248,
+      "step": 1361
+    },
+    {
+      "epoch": 0.1195421719131849,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0029912792762582536,
+      "loss": 1.3769,
+      "step": 1362
+    },
+    {
+      "epoch": 0.1196299414960874,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002991231081215567,
+      "loss": 1.3756,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11971771107898989,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0029911827537986795,
+      "loss": 1.362,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11980548066189237,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002991134294012361,
+      "loss": 1.3265,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11989325024479486,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.002991085701861394,
+      "loss": 1.3849,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11998101982769734,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0029910369773505753,
+      "loss": 1.3654,
+      "step": 1367
+    },
+    {
+      "epoch": 0.12006878941059983,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0029909881204847125,
+      "loss": 1.3789,
+      "step": 1368
+    },
+    {
+      "epoch": 0.1201565589935023,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002990939131268629,
+      "loss": 1.3777,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1202443285764048,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.002990890009707158,
+      "loss": 1.3323,
+      "step": 1370
+    },
+    {
+      "epoch": 0.12033209815930727,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0029908407558051496,
+      "loss": 1.3274,
+      "step": 1371
+    },
+    {
+      "epoch": 0.12041986774220977,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.002990791369567464,
+      "loss": 1.3721,
+      "step": 1372
+    },
+    {
+      "epoch": 0.12050763732511224,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0029907418509989747,
+      "loss": 1.3789,
+      "step": 1373
+    },
+    {
+      "epoch": 0.12059540690801474,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00299069220010457,
+      "loss": 1.3182,
+      "step": 1374
+    },
+    {
+      "epoch": 0.12068317649091721,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0029906424168891508,
+      "loss": 1.3797,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1207709460738197,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.002990592501357629,
+      "loss": 1.4241,
+      "step": 1376
+    },
+    {
+      "epoch": 0.1208587156567222,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002990542453514932,
+      "loss": 1.4091,
+      "step": 1377
+    },
+    {
+      "epoch": 0.12094648523962467,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0029904922733659987,
+      "loss": 1.3258,
+      "step": 1378
+    },
+    {
+      "epoch": 0.12103425482252717,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029904419609157826,
+      "loss": 1.3287,
+      "step": 1379
+    },
+    {
+      "epoch": 0.12112202440542964,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0029903915161692486,
+      "loss": 1.3019,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12120979398833213,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0029903409391313753,
+      "loss": 1.3934,
+      "step": 1381
+    },
+    {
+      "epoch": 0.12129756357123461,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002990290229807155,
+      "loss": 1.3374,
+      "step": 1382
+    },
+    {
+      "epoch": 0.1213853331541371,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0029902393882015924,
+      "loss": 1.3759,
+      "step": 1383
+    },
+    {
+      "epoch": 0.12147310273703958,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0029901884143197057,
+      "loss": 1.3208,
+      "step": 1384
+    },
+    {
+      "epoch": 0.12156087231994207,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0029901373081665248,
+      "loss": 1.3118,
+      "step": 1385
+    },
+    {
+      "epoch": 0.12164864190284455,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0029900860697470943,
+      "loss": 1.3979,
+      "step": 1386
+    },
+    {
+      "epoch": 0.12173641148574704,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002990034699066471,
+      "loss": 1.3388,
+      "step": 1387
+    },
+    {
+      "epoch": 0.12182418106864953,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002989983196129725,
+      "loss": 1.3613,
+      "step": 1388
+    },
+    {
+      "epoch": 0.12191195065155201,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0029899315609419404,
+      "loss": 1.4004,
+      "step": 1389
+    },
+    {
+      "epoch": 0.1219997202344545,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0029898797935082122,
+      "loss": 1.3582,
+      "step": 1390
+    },
+    {
+      "epoch": 0.12208748981735698,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00298982789383365,
+      "loss": 1.2863,
+      "step": 1391
+    },
+    {
+      "epoch": 0.12217525940025947,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.002989775861923376,
+      "loss": 1.352,
+      "step": 1392
+    },
+    {
+      "epoch": 0.12226302898316195,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0029897236977825255,
+      "loss": 1.3793,
+      "step": 1393
+    },
+    {
+      "epoch": 0.12235079856606444,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.002989671401416247,
+      "loss": 1.4031,
+      "step": 1394
+    },
+    {
+      "epoch": 0.12243856814896692,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029896189728297025,
+      "loss": 1.3277,
+      "step": 1395
+    },
+    {
+      "epoch": 0.12252633773186941,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0029895664120280657,
+      "loss": 1.3364,
+      "step": 1396
+    },
+    {
+      "epoch": 0.12261410731477189,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0029895137190165243,
+      "loss": 1.3499,
+      "step": 1397
+    },
+    {
+      "epoch": 0.12270187689767438,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0029894608938002792,
+      "loss": 1.2889,
+      "step": 1398
+    },
+    {
+      "epoch": 0.12278964648057687,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002989407936384544,
+      "loss": 1.374,
+      "step": 1399
+    },
+    {
+      "epoch": 0.12287741606347935,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.002989354846774545,
+      "loss": 1.3264,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12296518564638184,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0029893016249755225,
+      "loss": 1.3311,
+      "step": 1401
+    },
+    {
+      "epoch": 0.12305295522928432,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0029892482709927286,
+      "loss": 1.3802,
+      "step": 1402
+    },
+    {
+      "epoch": 0.12314072481218681,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0029891947848314296,
+      "loss": 1.3593,
+      "step": 1403
+    },
+    {
+      "epoch": 0.12322849439508929,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0029891411664969042,
+      "loss": 1.3706,
+      "step": 1404
+    },
+    {
+      "epoch": 0.12331626397799178,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0029890874159944447,
+      "loss": 1.3851,
+      "step": 1405
+    },
+    {
+      "epoch": 0.12340403356089426,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0029890335333293554,
+      "loss": 1.3679,
+      "step": 1406
+    },
+    {
+      "epoch": 0.12349180314379675,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0029889795185069545,
+      "loss": 1.4016,
+      "step": 1407
+    },
+    {
+      "epoch": 0.12357957272669923,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029889253715325737,
+      "loss": 1.359,
+      "step": 1408
+    },
+    {
+      "epoch": 0.12366734230960172,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0029888710924115564,
+      "loss": 1.3669,
+      "step": 1409
+    },
+    {
+      "epoch": 0.1237551118925042,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0029888166811492603,
+      "loss": 1.3607,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12384288147540669,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0029887621377510543,
+      "loss": 1.373,
+      "step": 1411
+    },
+    {
+      "epoch": 0.12393065105830918,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002988707462222323,
+      "loss": 1.3402,
+      "step": 1412
+    },
+    {
+      "epoch": 0.12401842064121166,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029886526545684624,
+      "loss": 1.3743,
+      "step": 1413
+    },
+    {
+      "epoch": 0.12410619022411415,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0029885977147948817,
+      "loss": 1.4382,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12419395980701663,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0029885426429070022,
+      "loss": 1.355,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12428172938991912,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029884874389102607,
+      "loss": 1.4003,
+      "step": 1416
+    },
+    {
+      "epoch": 0.1243694989728216,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0029884321028101055,
+      "loss": 1.3275,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12445726855572409,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.002988376634611997,
+      "loss": 1.3964,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12454503813862657,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0029883210343214107,
+      "loss": 1.3267,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12463280772152906,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002988265301943834,
+      "loss": 1.3603,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12472057730443153,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.002988209437484767,
+      "loss": 1.3409,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12480834688733403,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0029881534409497233,
+      "loss": 1.3781,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12489611647023652,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0029880973123442302,
+      "loss": 1.3592,
+      "step": 1423
+    },
+    {
+      "epoch": 0.124983886053139,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.002988041051673827,
+      "loss": 1.3418,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12507165563604147,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.002987984658944066,
+      "loss": 1.359,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12515942521894396,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002987928134160514,
+      "loss": 1.3482,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12524719480184646,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0029878714773287477,
+      "loss": 1.3765,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12533496438474895,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0029878146884543618,
+      "loss": 1.3548,
+      "step": 1428
+    },
+    {
+      "epoch": 0.1254227339676514,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002987757767542959,
+      "loss": 1.3192,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1255105035505539,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002987700714600157,
+      "loss": 1.3591,
+      "step": 1430
+    },
+    {
+      "epoch": 0.1255982731334564,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0029876435296315885,
+      "loss": 1.3376,
+      "step": 1431
+    },
+    {
+      "epoch": 0.1256860427163589,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002987586212642896,
+      "loss": 1.3401,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12577381229926135,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0029875287636397367,
+      "loss": 1.3824,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12586158188216384,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.002987471182627781,
+      "loss": 1.3772,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12594935146506633,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.002987413469612712,
+      "loss": 1.2982,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12603712104796883,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029873556246002254,
+      "loss": 1.3317,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12612489063087132,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00298729764759603,
+      "loss": 1.3751,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12621266021377378,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0029872395386058483,
+      "loss": 1.3396,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12630042979667627,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0029871812976354157,
+      "loss": 1.3549,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12638819937957876,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.002987122924690479,
+      "loss": 1.3742,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12647596896248126,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0029870644197768016,
+      "loss": 1.2948,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12656373854538372,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0029870057829001553,
+      "loss": 1.3143,
+      "step": 1442
+    },
+    {
+      "epoch": 0.1266515081282862,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.002986947014066329,
+      "loss": 1.3048,
+      "step": 1443
+    },
+    {
+      "epoch": 0.1267392777111887,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002986888113281123,
+      "loss": 1.3771,
+      "step": 1444
+    },
+    {
+      "epoch": 0.1268270472940912,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002986829080550349,
+      "loss": 1.3069,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12691481687699369,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.002986769915879835,
+      "loss": 1.3741,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12700258645989615,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.002986710619275419,
+      "loss": 1.3294,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12709035604279864,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0029866511907429547,
+      "loss": 1.3173,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12717812562570113,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0029865916302883066,
+      "loss": 1.328,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12726589520860362,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0029865319379173525,
+      "loss": 1.3647,
+      "step": 1450
+    },
+    {
+      "epoch": 0.1273536647915061,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002986472113635985,
+      "loss": 1.3442,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12744143437440858,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0029864121574501084,
+      "loss": 1.3617,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12752920395731107,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0029863520693656394,
+      "loss": 1.3258,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12761697354021356,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002986291849388509,
+      "loss": 1.326,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12770474312311603,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0029862314975246606,
+      "loss": 1.3416,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12779251270601852,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00298617101378005,
+      "loss": 1.3525,
+      "step": 1456
+    },
+    {
+      "epoch": 0.127880282288921,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002986110398160648,
+      "loss": 1.3637,
+      "step": 1457
+    },
+    {
+      "epoch": 0.1279680518718235,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0029860496506724362,
+      "loss": 1.3911,
+      "step": 1458
+    },
+    {
+      "epoch": 0.128055821454726,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00298598877132141,
+      "loss": 1.3446,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12814359103762846,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0029859277601135786,
+      "loss": 1.3192,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12823136062053095,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0029858666170549638,
+      "loss": 1.3302,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12831913020343344,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002985805342151599,
+      "loss": 1.3877,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12840689978633593,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0029857439354095323,
+      "loss": 1.3456,
+      "step": 1463
+    },
+    {
+      "epoch": 0.1284946693692384,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002985682396834825,
+      "loss": 1.3953,
+      "step": 1464
+    },
+    {
+      "epoch": 0.1285824389521409,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0029856207264335496,
+      "loss": 1.3483,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12867020853504338,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029855589242117936,
+      "loss": 1.3472,
+      "step": 1466
+    },
+    {
+      "epoch": 0.12875797811794587,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0029854969901756566,
+      "loss": 1.3891,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12884574770084833,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0029854349243312508,
+      "loss": 1.3622,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12893351728375083,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002985372726684702,
+      "loss": 1.2893,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12902128686665332,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.002985310397242148,
+      "loss": 1.3801,
+      "step": 1470
+    },
+    {
+      "epoch": 0.1291090564495558,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.002985247936009743,
+      "loss": 1.3415,
+      "step": 1471
+    },
+    {
+      "epoch": 0.1291968260324583,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0029851853429936487,
+      "loss": 1.3737,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12928459561536076,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0029851226182000448,
+      "loss": 1.371,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12937236519826326,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002985059761635121,
+      "loss": 1.3949,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12946013478116575,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0029849967733050816,
+      "loss": 1.411,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12954790436406824,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0029849336532161426,
+      "loss": 1.3172,
+      "step": 1476
+    },
+    {
+      "epoch": 0.1296356739469707,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002984870401374534,
+      "loss": 1.3485,
+      "step": 1477
+    },
+    {
+      "epoch": 0.1297234435298732,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0029848070177864987,
+      "loss": 1.3172,
+      "step": 1478
+    },
+    {
+      "epoch": 0.1298112131127757,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002984743502458292,
+      "loss": 1.3329,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12989898269567818,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002984679855396184,
+      "loss": 1.2902,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12998675227858067,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.002984616076606455,
+      "loss": 1.3072,
+      "step": 1481
+    },
+    {
+      "epoch": 0.13007452186148313,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0029845521660953993,
+      "loss": 1.4163,
+      "step": 1482
+    },
+    {
+      "epoch": 0.13016229144438562,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.002984488123869326,
+      "loss": 1.3862,
+      "step": 1483
+    },
+    {
+      "epoch": 0.13025006102728812,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002984423949934555,
+      "loss": 1.4167,
+      "step": 1484
+    },
+    {
+      "epoch": 0.1303378306101906,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0029843596442974203,
+      "loss": 1.33,
+      "step": 1485
+    },
+    {
+      "epoch": 0.13042560019309307,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0029842952069642683,
+      "loss": 1.3811,
+      "step": 1486
+    },
+    {
+      "epoch": 0.13051336977599556,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.002984230637941459,
+      "loss": 1.3511,
+      "step": 1487
+    },
+    {
+      "epoch": 0.13060113935889806,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.002984165937235365,
+      "loss": 1.3482,
+      "step": 1488
+    },
+    {
+      "epoch": 0.13068890894180055,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.002984101104852372,
+      "loss": 1.3123,
+      "step": 1489
+    },
+    {
+      "epoch": 0.130776678524703,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.002984036140798879,
+      "loss": 1.3456,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1308644481076055,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002983971045081297,
+      "loss": 1.2824,
+      "step": 1491
+    },
+    {
+      "epoch": 0.130952217690508,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.002983905817706051,
+      "loss": 1.3123,
+      "step": 1492
+    },
+    {
+      "epoch": 0.13103998727341049,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0029838404586795793,
+      "loss": 1.3536,
+      "step": 1493
+    },
+    {
+      "epoch": 0.13112775685631298,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0029837749680083317,
+      "loss": 1.3014,
+      "step": 1494
+    },
+    {
+      "epoch": 0.13121552643921544,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0029837093456987727,
+      "loss": 1.427,
+      "step": 1495
+    },
+    {
+      "epoch": 0.13130329602211793,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002983643591757378,
+      "loss": 1.3655,
+      "step": 1496
+    },
+    {
+      "epoch": 0.13139106560502042,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.002983577706190638,
+      "loss": 1.2918,
+      "step": 1497
+    },
+    {
+      "epoch": 0.13147883518792292,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.002983511689005055,
+      "loss": 1.3853,
+      "step": 1498
+    },
+    {
+      "epoch": 0.13156660477082538,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002983445540207145,
+      "loss": 1.3599,
+      "step": 1499
+    },
+    {
+      "epoch": 0.13165437435372787,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002983379259803436,
+      "loss": 1.3563,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13165437435372787,
+      "eval_loss": 1.3339033126831055,
+      "eval_runtime": 437.6205,
+      "eval_samples_per_second": 33.682,
+      "eval_steps_per_second": 8.421,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13174214393663036,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00298331284780047,
+      "loss": 1.3152,
+      "step": 1501
+    },
+    {
+      "epoch": 0.13182991351953285,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0029832463042048017,
+      "loss": 1.3866,
+      "step": 1502
+    },
+    {
+      "epoch": 0.13191768310243532,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.002983179629022999,
+      "loss": 1.3914,
+      "step": 1503
+    },
+    {
+      "epoch": 0.1320054526853378,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.002983112822261641,
+      "loss": 1.3485,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1320932222682403,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.002983045883927323,
+      "loss": 1.3199,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1321809918511428,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0029829788140266506,
+      "loss": 1.3547,
+      "step": 1506
+    },
+    {
+      "epoch": 0.13226876143404528,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002982911612566244,
+      "loss": 1.2855,
+      "step": 1507
+    },
+    {
+      "epoch": 0.13235653101694775,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0029828442795527352,
+      "loss": 1.3168,
+      "step": 1508
+    },
+    {
+      "epoch": 0.13244430059985024,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0029827768149927694,
+      "loss": 1.3254,
+      "step": 1509
+    },
+    {
+      "epoch": 0.13253207018275273,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029827092188930058,
+      "loss": 1.3223,
+      "step": 1510
+    },
+    {
+      "epoch": 0.13261983976565522,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029826414912601158,
+      "loss": 1.3559,
+      "step": 1511
+    },
+    {
+      "epoch": 0.1327076093485577,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0029825736321007833,
+      "loss": 1.3301,
+      "step": 1512
+    },
+    {
+      "epoch": 0.13279537893146018,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029825056414217056,
+      "loss": 1.3512,
+      "step": 1513
+    },
+    {
+      "epoch": 0.13288314851436267,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0029824375192295943,
+      "loss": 1.351,
+      "step": 1514
+    },
+    {
+      "epoch": 0.13297091809726516,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0029823692655311716,
+      "loss": 1.3494,
+      "step": 1515
+    },
+    {
+      "epoch": 0.13305868768016765,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0029823008803331746,
+      "loss": 1.3305,
+      "step": 1516
+    },
+    {
+      "epoch": 0.13314645726307012,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0029822323636423524,
+      "loss": 1.342,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1332342268459726,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002982163715465467,
+      "loss": 1.3594,
+      "step": 1518
+    },
+    {
+      "epoch": 0.1333219964288751,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0029820949358092944,
+      "loss": 1.3676,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1334097660117776,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029820260246806217,
+      "loss": 1.3476,
+      "step": 1520
+    },
+    {
+      "epoch": 0.13349753559468006,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0029819569820862517,
+      "loss": 1.3209,
+      "step": 1521
+    },
+    {
+      "epoch": 0.13358530517758255,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0029818878080329976,
+      "loss": 1.2706,
+      "step": 1522
+    },
+    {
+      "epoch": 0.13367307476048504,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0029818185025276872,
+      "loss": 1.3376,
+      "step": 1523
+    },
+    {
+      "epoch": 0.13376084434338753,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0029817490655771593,
+      "loss": 1.3505,
+      "step": 1524
+    },
+    {
+      "epoch": 0.13384861392629,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002981679497188269,
+      "loss": 1.3944,
+      "step": 1525
+    },
+    {
+      "epoch": 0.13393638350919249,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0029816097973678814,
+      "loss": 1.3454,
+      "step": 1526
+    },
+    {
+      "epoch": 0.13402415309209498,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0029815399661228753,
+      "loss": 1.3408,
+      "step": 1527
+    },
+    {
+      "epoch": 0.13411192267499747,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0029814700034601436,
+      "loss": 1.3401,
+      "step": 1528
+    },
+    {
+      "epoch": 0.13419969225789996,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002981399909386591,
+      "loss": 1.342,
+      "step": 1529
+    },
+    {
+      "epoch": 0.13428746184080242,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.002981329683909135,
+      "loss": 1.3279,
+      "step": 1530
+    },
+    {
+      "epoch": 0.13437523142370492,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.002981259327034707,
+      "loss": 1.3395,
+      "step": 1531
+    },
+    {
+      "epoch": 0.1344630010066074,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0029811888387702512,
+      "loss": 1.3474,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1345507705895099,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0029811182191227244,
+      "loss": 1.346,
+      "step": 1533
+    },
+    {
+      "epoch": 0.13463854017241236,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002981047468099096,
+      "loss": 1.3413,
+      "step": 1534
+    },
+    {
+      "epoch": 0.13472630975531485,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0029809765857063495,
+      "loss": 1.2592,
+      "step": 1535
+    },
+    {
+      "epoch": 0.13481407933821735,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0029809055719514804,
+      "loss": 1.4012,
+      "step": 1536
+    },
+    {
+      "epoch": 0.13490184892111984,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0029808344268414974,
+      "loss": 1.3536,
+      "step": 1537
+    },
+    {
+      "epoch": 0.1349896185040223,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.002980763150383422,
+      "loss": 1.3417,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1350773880869248,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00298069174258429,
+      "loss": 1.3449,
+      "step": 1539
+    },
+    {
+      "epoch": 0.13516515766982728,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0029806202034511472,
+      "loss": 1.3331,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13525292725272978,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.002980548532991056,
+      "loss": 1.3557,
+      "step": 1541
+    },
+    {
+      "epoch": 0.13534069683563227,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.002980476731211089,
+      "loss": 1.371,
+      "step": 1542
+    },
+    {
+      "epoch": 0.13542846641853473,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029804047981183334,
+      "loss": 1.358,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13551623600143722,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0029803327337198887,
+      "loss": 1.334,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13560400558433972,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.002980260538022866,
+      "loss": 1.3648,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1356917751672422,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0029801882110343926,
+      "loss": 1.3304,
+      "step": 1546
+    },
+    {
+      "epoch": 0.13577954475014467,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0029801157527616063,
+      "loss": 1.338,
+      "step": 1547
+    },
+    {
+      "epoch": 0.13586731433304716,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0029800431632116577,
+      "loss": 1.3448,
+      "step": 1548
+    },
+    {
+      "epoch": 0.13595508391594965,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.002979970442391711,
+      "loss": 1.3134,
+      "step": 1549
+    },
+    {
+      "epoch": 0.13604285349885215,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002979897590308945,
+      "loss": 1.3217,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13613062308175464,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0029798246069705483,
+      "loss": 1.32,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1362183926646571,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002979751492383725,
+      "loss": 1.3061,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1363061622475596,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0029796782465556916,
+      "loss": 1.3187,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13639393183046208,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.002979604869493676,
+      "loss": 1.3302,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13648170141336458,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029795313612049202,
+      "loss": 1.3247,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13656947099626704,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0029794577216966807,
+      "loss": 1.3718,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13665724057916953,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0029793839509762244,
+      "loss": 1.3535,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13674501016207202,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0029793100490508313,
+      "loss": 1.2986,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13683277974497451,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.002979236015927797,
+      "loss": 1.3521,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13692054932787698,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0029791618516144276,
+      "loss": 1.3595,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13700831891077947,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0029790875561180423,
+      "loss": 1.3786,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13709608849368196,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0029790131294459743,
+      "loss": 1.3616,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13718385807658445,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0029789385716055684,
+      "loss": 1.3719,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13727162765948694,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0029788638826041846,
+      "loss": 1.4084,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1373593972423894,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0029787890624491933,
+      "loss": 1.341,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1374471668252919,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0029787141111479796,
+      "loss": 1.3723,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1375349364081944,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00297863902870794,
+      "loss": 1.3145,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13762270599109688,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0029785638151364857,
+      "loss": 1.3448,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13771047557399935,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029784884704410403,
+      "loss": 1.3786,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13779824515690184,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0029784129946290386,
+      "loss": 1.3183,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13788601473980433,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002978337387707931,
+      "loss": 1.3635,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13797378432270682,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.002978261649685179,
+      "loss": 1.3174,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13806155390560929,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002978185780568258,
+      "loss": 1.35,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13814932348851178,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0029781097803646558,
+      "loss": 1.3759,
+      "step": 1574
+    },
+    {
+      "epoch": 0.13823709307141427,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0029780336490818735,
+      "loss": 1.394,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13832486265431676,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002977957386727425,
+      "loss": 1.2821,
+      "step": 1576
+    },
+    {
+      "epoch": 0.13841263223721925,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002977880993308836,
+      "loss": 1.3619,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13850040182012172,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.002977804468833648,
+      "loss": 1.3,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1385881714030242,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0029777278133094123,
+      "loss": 1.31,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1386759409859267,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0029776510267436954,
+      "loss": 1.3326,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1387637105688292,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.002977574109144075,
+      "loss": 1.4208,
+      "step": 1581
+    },
+    {
+      "epoch": 0.13885148015173165,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.002977497060518143,
+      "loss": 1.361,
+      "step": 1582
+    },
+    {
+      "epoch": 0.13893924973463415,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0029774198808735037,
+      "loss": 1.2836,
+      "step": 1583
+    },
+    {
+      "epoch": 0.13902701931753664,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002977342570217775,
+      "loss": 1.382,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13911478890043913,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029772651285585863,
+      "loss": 1.3055,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13920255848334162,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002977187555903581,
+      "loss": 1.4256,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13929032806624408,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0029771098522604154,
+      "loss": 1.3969,
+      "step": 1587
+    },
+    {
+      "epoch": 0.13937809764914658,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002977032017636759,
+      "loss": 1.3544,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13946586723204907,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0029769540520402926,
+      "loss": 1.3442,
+      "step": 1589
+    },
+    {
+      "epoch": 0.13955363681495156,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002976875955478712,
+      "loss": 1.398,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13964140639785402,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.002976797727959725,
+      "loss": 1.3587,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13972917598075651,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0029767193694910515,
+      "loss": 1.3348,
+      "step": 1592
+    },
+    {
+      "epoch": 0.139816945563659,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0029766408800804265,
+      "loss": 1.3375,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1399047151465615,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0029765622597355953,
+      "loss": 1.3374,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13999248472946396,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0029764835084643186,
+      "loss": 1.2843,
+      "step": 1595
+    },
+    {
+      "epoch": 0.14008025431236645,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002976404626274368,
+      "loss": 1.304,
+      "step": 1596
+    },
+    {
+      "epoch": 0.14016802389526895,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0029763256131735286,
+      "loss": 1.3601,
+      "step": 1597
+    },
+    {
+      "epoch": 0.14025579347817144,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0029762464691696,
+      "loss": 1.3425,
+      "step": 1598
+    },
+    {
+      "epoch": 0.14034356306107393,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0029761671942703924,
+      "loss": 1.3645,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1404313326439764,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00297608778848373,
+      "loss": 1.3378,
+      "step": 1600
+    },
+    {
+      "epoch": 0.14051910222687888,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0029760082518174493,
+      "loss": 1.2775,
+      "step": 1601
+    },
+    {
+      "epoch": 0.14060687180978138,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.002975928584279402,
+      "loss": 1.2902,
+      "step": 1602
+    },
+    {
+      "epoch": 0.14069464139268387,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002975848785877449,
+      "loss": 1.2893,
+      "step": 1603
+    },
+    {
+      "epoch": 0.14078241097558633,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0029757688566194674,
+      "loss": 1.3126,
+      "step": 1604
+    },
+    {
+      "epoch": 0.14087018055848882,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0029756887965133455,
+      "loss": 1.3202,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1409579501413913,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0029756086055669844,
+      "loss": 1.2834,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1410457197242938,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0029755282837882996,
+      "loss": 1.3656,
+      "step": 1607
+    },
+    {
+      "epoch": 0.14113348930719627,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002975447831185218,
+      "loss": 1.3275,
+      "step": 1608
+    },
+    {
+      "epoch": 0.14122125889009876,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00297536724776568,
+      "loss": 1.3322,
+      "step": 1609
+    },
+    {
+      "epoch": 0.14130902847300125,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0029752865335376387,
+      "loss": 1.3148,
+      "step": 1610
+    },
+    {
+      "epoch": 0.14139679805590374,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00297520568850906,
+      "loss": 1.3834,
+      "step": 1611
+    },
+    {
+      "epoch": 0.14148456763880624,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0029751247126879243,
+      "loss": 1.3611,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1415723372217087,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0029750436060822223,
+      "loss": 1.3554,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1416601068046112,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002974962368699959,
+      "loss": 1.3455,
+      "step": 1614
+    },
+    {
+      "epoch": 0.14174787638751368,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0029748810005491527,
+      "loss": 1.3609,
+      "step": 1615
+    },
+    {
+      "epoch": 0.14183564597041617,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0029747995016378348,
+      "loss": 1.2825,
+      "step": 1616
+    },
+    {
+      "epoch": 0.14192341555331864,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0029747178719740468,
+      "loss": 1.3947,
+      "step": 1617
+    },
+    {
+      "epoch": 0.14201118513622113,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002974636111565847,
+      "loss": 1.3454,
+      "step": 1618
+    },
+    {
+      "epoch": 0.14209895471912362,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0029745542204213044,
+      "loss": 1.3789,
+      "step": 1619
+    },
+    {
+      "epoch": 0.1421867243020261,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.002974472198548501,
+      "loss": 1.3436,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1422744938849286,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029743900459555327,
+      "loss": 1.3593,
+      "step": 1621
+    },
+    {
+      "epoch": 0.14236226346783107,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.002974307762650507,
+      "loss": 1.3492,
+      "step": 1622
+    },
+    {
+      "epoch": 0.14245003305073356,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029742253486415445,
+      "loss": 1.3053,
+      "step": 1623
+    },
+    {
+      "epoch": 0.14253780263363605,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.002974142803936781,
+      "loss": 1.3863,
+      "step": 1624
+    },
+    {
+      "epoch": 0.14262557221653854,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002974060128544361,
+      "loss": 1.2817,
+      "step": 1625
+    },
+    {
+      "epoch": 0.142713341799441,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002973977322472446,
+      "loss": 1.3204,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1428011113823435,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0029738943857292076,
+      "loss": 1.3363,
+      "step": 1627
+    },
+    {
+      "epoch": 0.142888880965246,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002973811318322832,
+      "loss": 1.3195,
+      "step": 1628
+    },
+    {
+      "epoch": 0.14297665054814848,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0029737281202615175,
+      "loss": 1.287,
+      "step": 1629
+    },
+    {
+      "epoch": 0.14306442013105095,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002973644791553475,
+      "loss": 1.3488,
+      "step": 1630
+    },
+    {
+      "epoch": 0.14315218971395344,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0029735613322069286,
+      "loss": 1.3186,
+      "step": 1631
+    },
+    {
+      "epoch": 0.14323995929685593,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0029734777422301168,
+      "loss": 1.3029,
+      "step": 1632
+    },
+    {
+      "epoch": 0.14332772887975842,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029733940216312878,
+      "loss": 1.3344,
+      "step": 1633
+    },
+    {
+      "epoch": 0.1434154984626609,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0029733101704187055,
+      "loss": 1.3609,
+      "step": 1634
+    },
+    {
+      "epoch": 0.14350326804556338,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0029732261886006453,
+      "loss": 1.3649,
+      "step": 1635
+    },
+    {
+      "epoch": 0.14359103762846587,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029731420761853963,
+      "loss": 1.3182,
+      "step": 1636
+    },
+    {
+      "epoch": 0.14367880721136836,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0029730578331812595,
+      "loss": 1.3874,
+      "step": 1637
+    },
+    {
+      "epoch": 0.14376657679427085,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00297297345959655,
+      "loss": 1.3422,
+      "step": 1638
+    },
+    {
+      "epoch": 0.14385434637717331,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002972888955439594,
+      "loss": 1.4337,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1439421159600758,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002972804320718733,
+      "loss": 1.3749,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1440298855429783,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0029727195554423197,
+      "loss": 1.3506,
+      "step": 1641
+    },
+    {
+      "epoch": 0.1441176551258808,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029726346596187204,
+      "loss": 1.3125,
+      "step": 1642
+    },
+    {
+      "epoch": 0.14420542470878325,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0029725496332563134,
+      "loss": 1.3447,
+      "step": 1643
+    },
+    {
+      "epoch": 0.14429319429168574,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00297246447636349,
+      "loss": 1.2968,
+      "step": 1644
+    },
+    {
+      "epoch": 0.14438096387458824,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002972379188948656,
+      "loss": 1.315,
+      "step": 1645
+    },
+    {
+      "epoch": 0.14446873345749073,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.002972293771020228,
+      "loss": 1.3592,
+      "step": 1646
+    },
+    {
+      "epoch": 0.14455650304039322,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029722082225866375,
+      "loss": 1.3075,
+      "step": 1647
+    },
+    {
+      "epoch": 0.14464427262329568,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.002972122543656327,
+      "loss": 1.3091,
+      "step": 1648
+    },
+    {
+      "epoch": 0.14473204220619817,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0029720367342377526,
+      "loss": 1.3056,
+      "step": 1649
+    },
+    {
+      "epoch": 0.14481981178910067,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002971950794339384,
+      "loss": 1.3202,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14490758137200316,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002971864723969702,
+      "loss": 1.3596,
+      "step": 1651
+    },
+    {
+      "epoch": 0.14499535095490562,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.002971778523137203,
+      "loss": 1.3292,
+      "step": 1652
+    },
+    {
+      "epoch": 0.1450831205378081,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002971692191850393,
+      "loss": 1.3106,
+      "step": 1653
+    },
+    {
+      "epoch": 0.1451708901207106,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0029716057301177935,
+      "loss": 1.3506,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1452586597036131,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.002971519137947938,
+      "loss": 1.3225,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1453464292865156,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029714324153493725,
+      "loss": 1.3785,
+      "step": 1656
+    },
+    {
+      "epoch": 0.14543419886941805,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0029713455623306562,
+      "loss": 1.328,
+      "step": 1657
+    },
+    {
+      "epoch": 0.14552196845232054,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.002971258578900361,
+      "loss": 1.3539,
+      "step": 1658
+    },
+    {
+      "epoch": 0.14560973803522304,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002971171465067072,
+      "loss": 1.3286,
+      "step": 1659
+    },
+    {
+      "epoch": 0.14569750761812553,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0029710842208393873,
+      "loss": 1.3292,
+      "step": 1660
+    },
+    {
+      "epoch": 0.145785277201028,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0029709968462259175,
+      "loss": 1.3318,
+      "step": 1661
+    },
+    {
+      "epoch": 0.14587304678393048,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0029709093412352856,
+      "loss": 1.304,
+      "step": 1662
+    },
+    {
+      "epoch": 0.14596081636683297,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.002970821705876128,
+      "loss": 1.3398,
+      "step": 1663
+    },
+    {
+      "epoch": 0.14604858594973547,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0029707339401570947,
+      "loss": 1.3355,
+      "step": 1664
+    },
+    {
+      "epoch": 0.14613635553263793,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002970646044086847,
+      "loss": 1.3437,
+      "step": 1665
+    },
+    {
+      "epoch": 0.14622412511554042,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0029705580176740604,
+      "loss": 1.3,
+      "step": 1666
+    },
+    {
+      "epoch": 0.1463118946984429,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0029704698609274226,
+      "loss": 1.3435,
+      "step": 1667
+    },
+    {
+      "epoch": 0.1463996642813454,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0029703815738556347,
+      "loss": 1.3546,
+      "step": 1668
+    },
+    {
+      "epoch": 0.1464874338642479,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00297029315646741,
+      "loss": 1.2785,
+      "step": 1669
+    },
+    {
+      "epoch": 0.14657520344715036,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029702046087714744,
+      "loss": 1.3239,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14666297303005285,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002970115930776568,
+      "loss": 1.3436,
+      "step": 1671
+    },
+    {
+      "epoch": 0.14675074261295534,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.002970027122491443,
+      "loss": 1.3654,
+      "step": 1672
+    },
+    {
+      "epoch": 0.14683851219585783,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029699381839248633,
+      "loss": 1.3356,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1469262817787603,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0029698491150856082,
+      "loss": 1.3234,
+      "step": 1674
+    },
+    {
+      "epoch": 0.1470140513616628,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0029697599159824676,
+      "loss": 1.3224,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14710182094456528,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0029696705866242455,
+      "loss": 1.3654,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14718959052746777,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0029695811270197578,
+      "loss": 1.2866,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14727736011037024,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002969491537177834,
+      "loss": 1.3362,
+      "step": 1678
+    },
+    {
+      "epoch": 0.14736512969327273,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.002969401817107317,
+      "loss": 1.337,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14745289927617522,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0029693119668170614,
+      "loss": 1.4001,
+      "step": 1680
+    },
+    {
+      "epoch": 0.1475406688590777,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029692219863159345,
+      "loss": 1.3351,
+      "step": 1681
+    },
+    {
+      "epoch": 0.1476284384419802,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029691318756128173,
+      "loss": 1.3084,
+      "step": 1682
+    },
+    {
+      "epoch": 0.14771620802488267,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.002969041634716604,
+      "loss": 1.3714,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14780397760778516,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0029689512636362006,
+      "loss": 1.2869,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14789174719068765,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.002968860762380526,
+      "loss": 1.3356,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14797951677359014,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0029687701309585126,
+      "loss": 1.3372,
+      "step": 1686
+    },
+    {
+      "epoch": 0.1480672863564926,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0029686793693791055,
+      "loss": 1.3333,
+      "step": 1687
+    },
+    {
+      "epoch": 0.1481550559393951,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002968588477651262,
+      "loss": 1.3643,
+      "step": 1688
+    },
+    {
+      "epoch": 0.1482428255222976,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0029684974557839536,
+      "loss": 1.3312,
+      "step": 1689
+    },
+    {
+      "epoch": 0.14833059510520008,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002968406303786163,
+      "loss": 1.3926,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14841836468810257,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0029683150216668875,
+      "loss": 1.3173,
+      "step": 1691
+    },
+    {
+      "epoch": 0.14850613427100504,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.002968223609435135,
+      "loss": 1.3306,
+      "step": 1692
+    },
+    {
+      "epoch": 0.14859390385390753,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0029681320670999285,
+      "loss": 1.3116,
+      "step": 1693
+    },
+    {
+      "epoch": 0.14868167343681002,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029680403946703024,
+      "loss": 1.3639,
+      "step": 1694
+    },
+    {
+      "epoch": 0.1487694430197125,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0029679485921553046,
+      "loss": 1.38,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14885721260261497,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0029678566595639964,
+      "loss": 1.293,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14894498218551747,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0029677645969054495,
+      "loss": 1.3172,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14903275176841996,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0029676724041887512,
+      "loss": 1.3452,
+      "step": 1698
+    },
+    {
+      "epoch": 0.14912052135132245,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0029675800814230003,
+      "loss": 1.3612,
+      "step": 1699
+    },
+    {
+      "epoch": 0.1492082909342249,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0029674876286173092,
+      "loss": 1.3398,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1492960605171274,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0029673950457808017,
+      "loss": 1.3592,
+      "step": 1701
+    },
+    {
+      "epoch": 0.1493838301000299,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.002967302332922616,
+      "loss": 1.375,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1494715996829324,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002967209490051903,
+      "loss": 1.3833,
+      "step": 1703
+    },
+    {
+      "epoch": 0.14955936926583488,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002967116517177824,
+      "loss": 1.3336,
+      "step": 1704
+    },
+    {
+      "epoch": 0.14964713884873734,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0029670234143095568,
+      "loss": 1.3037,
+      "step": 1705
+    },
+    {
+      "epoch": 0.14973490843163983,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00296693018145629,
+      "loss": 1.3435,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14982267801454233,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.002966836818627225,
+      "loss": 1.4144,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14991044759744482,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.002966743325831577,
+      "loss": 1.3661,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14999821718034728,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0029666497030785724,
+      "loss": 1.3676,
+      "step": 1709
+    },
+    {
+      "epoch": 0.15008598676324977,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029665559503774517,
+      "loss": 1.3522,
+      "step": 1710
+    },
+    {
+      "epoch": 0.15017375634615227,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.002966462067737468,
+      "loss": 1.3386,
+      "step": 1711
+    },
+    {
+      "epoch": 0.15026152592905476,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0029663680551678877,
+      "loss": 1.3055,
+      "step": 1712
+    },
+    {
+      "epoch": 0.15034929551195722,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002966273912677988,
+      "loss": 1.3126,
+      "step": 1713
+    },
+    {
+      "epoch": 0.1504370650948597,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002966179640277062,
+      "loss": 1.3661,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1505248346777622,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.002966085237974414,
+      "loss": 1.3166,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1506126042606647,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00296599070577936,
+      "loss": 1.2842,
+      "step": 1716
+    },
+    {
+      "epoch": 0.1507003738435672,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0029658960437012297,
+      "loss": 1.3629,
+      "step": 1717
+    },
+    {
+      "epoch": 0.15078814342646965,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002965801251749367,
+      "loss": 1.2343,
+      "step": 1718
+    },
+    {
+      "epoch": 0.15087591300937214,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0029657063299331275,
+      "loss": 1.3709,
+      "step": 1719
+    },
+    {
+      "epoch": 0.15096368259227463,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002965611278261879,
+      "loss": 1.3609,
+      "step": 1720
+    },
+    {
+      "epoch": 0.15105145217517713,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0029655160967450036,
+      "loss": 1.3305,
+      "step": 1721
+    },
+    {
+      "epoch": 0.1511392217580796,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0029654207853918937,
+      "loss": 1.3748,
+      "step": 1722
+    },
+    {
+      "epoch": 0.15122699134098208,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0029653253442119575,
+      "loss": 1.3154,
+      "step": 1723
+    },
+    {
+      "epoch": 0.15131476092388457,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.002965229773214614,
+      "loss": 1.3339,
+      "step": 1724
+    },
+    {
+      "epoch": 0.15140253050678706,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.002965134072409297,
+      "loss": 1.2991,
+      "step": 1725
+    },
+    {
+      "epoch": 0.15149030008968956,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00296503824180545,
+      "loss": 1.3629,
+      "step": 1726
+    },
+    {
+      "epoch": 0.15157806967259202,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029649422814125317,
+      "loss": 1.3489,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1516658392554945,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0029648461912400136,
+      "loss": 1.2968,
+      "step": 1728
+    },
+    {
+      "epoch": 0.151753608838397,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.002964749971297379,
+      "loss": 1.312,
+      "step": 1729
+    },
+    {
+      "epoch": 0.1518413784212995,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.002964653621594124,
+      "loss": 1.3319,
+      "step": 1730
+    },
+    {
+      "epoch": 0.15192914800420196,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0029645571421397588,
+      "loss": 1.3371,
+      "step": 1731
+    },
+    {
+      "epoch": 0.15201691758710445,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002964460532943805,
+      "loss": 1.3227,
+      "step": 1732
+    },
+    {
+      "epoch": 0.15210468717000694,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002964363794015798,
+      "loss": 1.3408,
+      "step": 1733
+    },
+    {
+      "epoch": 0.15219245675290943,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029642669253652845,
+      "loss": 1.285,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1522802263358119,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002964169927001826,
+      "loss": 1.3042,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1523679959187144,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0029640727989349957,
+      "loss": 1.3264,
+      "step": 1736
+    },
+    {
+      "epoch": 0.15245576550161688,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029639755411743794,
+      "loss": 1.3388,
+      "step": 1737
+    },
+    {
+      "epoch": 0.15254353508451937,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002963878153729577,
+      "loss": 1.2758,
+      "step": 1738
+    },
+    {
+      "epoch": 0.15263130466742186,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002963780636610199,
+      "loss": 1.3148,
+      "step": 1739
+    },
+    {
+      "epoch": 0.15271907425032433,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002963682989825871,
+      "loss": 1.3109,
+      "step": 1740
+    },
+    {
+      "epoch": 0.15280684383322682,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002963585213386229,
+      "loss": 1.3591,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1528946134161293,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0029634873073009255,
+      "loss": 1.3326,
+      "step": 1742
+    },
+    {
+      "epoch": 0.1529823829990318,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0029633892715796206,
+      "loss": 1.4084,
+      "step": 1743
+    },
+    {
+      "epoch": 0.15307015258193427,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0029632911062319922,
+      "loss": 1.3302,
+      "step": 1744
+    },
+    {
+      "epoch": 0.15315792216483676,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002963192811267728,
+      "loss": 1.323,
+      "step": 1745
+    },
+    {
+      "epoch": 0.15324569174773925,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002963094386696529,
+      "loss": 1.3406,
+      "step": 1746
+    },
+    {
+      "epoch": 0.15333346133064174,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0029629958325281098,
+      "loss": 1.3364,
+      "step": 1747
+    },
+    {
+      "epoch": 0.1534212309135442,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.002962897148772197,
+      "loss": 1.3756,
+      "step": 1748
+    },
+    {
+      "epoch": 0.1535090004964467,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002962798335438531,
+      "loss": 1.3815,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1535967700793492,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0029626993925368633,
+      "loss": 1.2959,
+      "step": 1750
+    },
+    {
+      "epoch": 0.15368453966225168,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0029626003200769603,
+      "loss": 1.3421,
+      "step": 1751
+    },
+    {
+      "epoch": 0.15377230924515417,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0029625011180685985,
+      "loss": 1.2874,
+      "step": 1752
+    },
+    {
+      "epoch": 0.15386007882805663,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00296240178652157,
+      "loss": 1.4442,
+      "step": 1753
+    },
+    {
+      "epoch": 0.15394784841095913,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0029623023254456786,
+      "loss": 1.3041,
+      "step": 1754
+    },
+    {
+      "epoch": 0.15403561799386162,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0029622027348507394,
+      "loss": 1.3858,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1541233875767641,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.002962103014746583,
+      "loss": 1.2826,
+      "step": 1756
+    },
+    {
+      "epoch": 0.15421115715966657,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.002962003165143051,
+      "loss": 1.3288,
+      "step": 1757
+    },
+    {
+      "epoch": 0.15429892674256906,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0029619031860499976,
+      "loss": 1.3348,
+      "step": 1758
+    },
+    {
+      "epoch": 0.15438669632547156,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00296180307747729,
+      "loss": 1.3248,
+      "step": 1759
+    },
+    {
+      "epoch": 0.15447446590837405,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00296170283943481,
+      "loss": 1.3364,
+      "step": 1760
+    },
+    {
+      "epoch": 0.15456223549127654,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0029616024719324495,
+      "loss": 1.2944,
+      "step": 1761
+    },
+    {
+      "epoch": 0.154650005074179,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.002961501974980115,
+      "loss": 1.3995,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1547377746570815,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002961401348587725,
+      "loss": 1.3033,
+      "step": 1763
+    },
+    {
+      "epoch": 0.154825544239984,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0029613005927652107,
+      "loss": 1.3228,
+      "step": 1764
+    },
+    {
+      "epoch": 0.15491331382288648,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0029611997075225166,
+      "loss": 1.3379,
+      "step": 1765
+    },
+    {
+      "epoch": 0.15500108340578894,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.002961098692869599,
+      "loss": 1.3414,
+      "step": 1766
+    },
+    {
+      "epoch": 0.15508885298869143,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0029609975488164284,
+      "loss": 1.2882,
+      "step": 1767
+    },
+    {
+      "epoch": 0.15517662257159393,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.002960896275372987,
+      "loss": 1.3297,
+      "step": 1768
+    },
+    {
+      "epoch": 0.15526439215449642,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002960794872549271,
+      "loss": 1.3558,
+      "step": 1769
+    },
+    {
+      "epoch": 0.15535216173739888,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002960693340355287,
+      "loss": 1.292,
+      "step": 1770
+    },
+    {
+      "epoch": 0.15543993132030137,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029605916788010564,
+      "loss": 1.2973,
+      "step": 1771
+    },
+    {
+      "epoch": 0.15552770090320386,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0029604898878966127,
+      "loss": 1.3015,
+      "step": 1772
+    },
+    {
+      "epoch": 0.15561547048610636,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002960387967652003,
+      "loss": 1.3052,
+      "step": 1773
+    },
+    {
+      "epoch": 0.15570324006900885,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.002960285918077285,
+      "loss": 1.3675,
+      "step": 1774
+    },
+    {
+      "epoch": 0.1557910096519113,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0029601837391825324,
+      "loss": 1.3843,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1558787792348138,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0029600814309778288,
+      "loss": 1.3903,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1559665488177163,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.002959978993473271,
+      "loss": 1.3131,
+      "step": 1777
+    },
+    {
+      "epoch": 0.15605431840061879,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00295987642667897,
+      "loss": 1.368,
+      "step": 1778
+    },
+    {
+      "epoch": 0.15614208798352125,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.002959773730605049,
+      "loss": 1.3273,
+      "step": 1779
+    },
+    {
+      "epoch": 0.15622985756642374,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0029596709052616427,
+      "loss": 1.2855,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15631762714932623,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0029595679506589007,
+      "loss": 1.3126,
+      "step": 1781
+    },
+    {
+      "epoch": 0.15640539673222872,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0029594648668069836,
+      "loss": 1.3176,
+      "step": 1782
+    },
+    {
+      "epoch": 0.1564931663151312,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0029593616537160654,
+      "loss": 1.2997,
+      "step": 1783
+    },
+    {
+      "epoch": 0.15658093589803368,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0029592583113963325,
+      "loss": 1.3273,
+      "step": 1784
+    },
+    {
+      "epoch": 0.15666870548093617,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.002959154839857985,
+      "loss": 1.2225,
+      "step": 1785
+    },
+    {
+      "epoch": 0.15675647506383866,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0029590512391112355,
+      "loss": 1.3383,
+      "step": 1786
+    },
+    {
+      "epoch": 0.15684424464674115,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.002958947509166307,
+      "loss": 1.2893,
+      "step": 1787
+    },
+    {
+      "epoch": 0.15693201422964362,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00295884365003344,
+      "loss": 1.3182,
+      "step": 1788
+    },
+    {
+      "epoch": 0.1570197838125461,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002958739661722883,
+      "loss": 1.3575,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1571075533954486,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0029586355442448994,
+      "loss": 1.3134,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1571953229783511,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.002958531297609766,
+      "loss": 1.2641,
+      "step": 1791
+    },
+    {
+      "epoch": 0.15728309256125356,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.002958426921827771,
+      "loss": 1.3329,
+      "step": 1792
+    },
+    {
+      "epoch": 0.15737086214415605,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002958322416909216,
+      "loss": 1.3176,
+      "step": 1793
+    },
+    {
+      "epoch": 0.15745863172705854,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.002958217782864415,
+      "loss": 1.3579,
+      "step": 1794
+    },
+    {
+      "epoch": 0.15754640130996103,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.002958113019703696,
+      "loss": 1.3458,
+      "step": 1795
+    },
+    {
+      "epoch": 0.15763417089286352,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029580081274373973,
+      "loss": 1.3286,
+      "step": 1796
+    },
+    {
+      "epoch": 0.157721940475766,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.002957903106075872,
+      "loss": 1.2778,
+      "step": 1797
+    },
+    {
+      "epoch": 0.15780971005866848,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0029577979556294856,
+      "loss": 1.4001,
+      "step": 1798
+    },
+    {
+      "epoch": 0.15789747964157097,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.002957692676108615,
+      "loss": 1.3583,
+      "step": 1799
+    },
+    {
+      "epoch": 0.15798524922447346,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0029575872675236523,
+      "loss": 1.3638,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15807301880737593,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0029574817298849993,
+      "loss": 1.3447,
+      "step": 1801
+    },
+    {
+      "epoch": 0.15816078839027842,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0029573760632030737,
+      "loss": 1.3347,
+      "step": 1802
+    },
+    {
+      "epoch": 0.1582485579731809,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0029572702674883033,
+      "loss": 1.338,
+      "step": 1803
+    },
+    {
+      "epoch": 0.1583363275560834,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0029571643427511304,
+      "loss": 1.3664,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15842409713898586,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0029570582890020088,
+      "loss": 1.3042,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15851186672188836,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002956952106251406,
+      "loss": 1.3055,
+      "step": 1806
+    },
+    {
+      "epoch": 0.15859963630479085,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0029568457945098017,
+      "loss": 1.3305,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15868740588769334,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0029567393537876883,
+      "loss": 1.2891,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15877517547059583,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029566327840955713,
+      "loss": 1.3102,
+      "step": 1809
+    },
+    {
+      "epoch": 0.1588629450534983,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0029565260854439684,
+      "loss": 1.3204,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1589507146364008,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0029564192578434103,
+      "loss": 1.3443,
+      "step": 1811
+    },
+    {
+      "epoch": 0.15903848421930328,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0029563123013044415,
+      "loss": 1.3179,
+      "step": 1812
+    },
+    {
+      "epoch": 0.15912625380220577,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002956205215837617,
+      "loss": 1.3436,
+      "step": 1813
+    },
+    {
+      "epoch": 0.15921402338510823,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029560980014535056,
+      "loss": 1.3426,
+      "step": 1814
+    },
+    {
+      "epoch": 0.15930179296801072,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00295599065816269,
+      "loss": 1.3132,
+      "step": 1815
+    },
+    {
+      "epoch": 0.15938956255091322,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002955883185975764,
+      "loss": 1.2819,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1594773321338157,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002955775584903334,
+      "loss": 1.276,
+      "step": 1817
+    },
+    {
+      "epoch": 0.15956510171671817,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0029556678549560213,
+      "loss": 1.2953,
+      "step": 1818
+    },
+    {
+      "epoch": 0.15965287129962066,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0029555599961444577,
+      "loss": 1.3201,
+      "step": 1819
+    },
+    {
+      "epoch": 0.15974064088252315,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0029554520084792875,
+      "loss": 1.3501,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15982841046542565,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00295534389197117,
+      "loss": 1.3125,
+      "step": 1821
+    },
+    {
+      "epoch": 0.15991618004832814,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002955235646630775,
+      "loss": 1.3271,
+      "step": 1822
+    },
+    {
+      "epoch": 0.1600039496312306,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0029551272724687863,
+      "loss": 1.3424,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1600917192141331,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029550187694959,
+      "loss": 1.3016,
+      "step": 1824
+    },
+    {
+      "epoch": 0.16017948879703559,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0029549101377228247,
+      "loss": 1.3332,
+      "step": 1825
+    },
+    {
+      "epoch": 0.16026725837993808,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0029548013771602826,
+      "loss": 1.2631,
+      "step": 1826
+    },
+    {
+      "epoch": 0.16035502796284054,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0029546924878190066,
+      "loss": 1.2696,
+      "step": 1827
+    },
+    {
+      "epoch": 0.16044279754574303,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029545834697097453,
+      "loss": 1.2918,
+      "step": 1828
+    },
+    {
+      "epoch": 0.16053056712864552,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0029544743228432572,
+      "loss": 1.2892,
+      "step": 1829
+    },
+    {
+      "epoch": 0.16061833671154802,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0029543650472303147,
+      "loss": 1.3321,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1607061062944505,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002954255642881704,
+      "loss": 1.3393,
+      "step": 1831
+    },
+    {
+      "epoch": 0.16079387587735297,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002954146109808221,
+      "loss": 1.3037,
+      "step": 1832
+    },
+    {
+      "epoch": 0.16088164546025546,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0029540364480206776,
+      "loss": 1.3358,
+      "step": 1833
+    },
+    {
+      "epoch": 0.16096941504315795,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002953926657529897,
+      "loss": 1.2961,
+      "step": 1834
+    },
+    {
+      "epoch": 0.16105718462606045,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0029538167383467142,
+      "loss": 1.3394,
+      "step": 1835
+    },
+    {
+      "epoch": 0.1611449542089629,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0029537066904819784,
+      "loss": 1.3531,
+      "step": 1836
+    },
+    {
+      "epoch": 0.1612327237918654,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0029535965139465504,
+      "loss": 1.2862,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1613204933747679,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002953486208751305,
+      "loss": 1.3085,
+      "step": 1838
+    },
+    {
+      "epoch": 0.16140826295767038,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0029533757749071283,
+      "loss": 1.3163,
+      "step": 1839
+    },
+    {
+      "epoch": 0.16149603254057285,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0029532652124249197,
+      "loss": 1.2983,
+      "step": 1840
+    },
+    {
+      "epoch": 0.16158380212347534,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0029531545213155918,
+      "loss": 1.3236,
+      "step": 1841
+    },
+    {
+      "epoch": 0.16167157170637783,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.002953043701590069,
+      "loss": 1.3114,
+      "step": 1842
+    },
+    {
+      "epoch": 0.16175934128928032,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002952932753259288,
+      "loss": 1.3014,
+      "step": 1843
+    },
+    {
+      "epoch": 0.16184711087218281,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0029528216763342,
+      "loss": 1.3481,
+      "step": 1844
+    },
+    {
+      "epoch": 0.16193488045508528,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0029527104708257676,
+      "loss": 1.2856,
+      "step": 1845
+    },
+    {
+      "epoch": 0.16202265003798777,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0029525991367449667,
+      "loss": 1.3673,
+      "step": 1846
+    },
+    {
+      "epoch": 0.16211041962089026,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0029524876741027846,
+      "loss": 1.3396,
+      "step": 1847
+    },
+    {
+      "epoch": 0.16219818920379275,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002952376082910223,
+      "loss": 1.2964,
+      "step": 1848
+    },
+    {
+      "epoch": 0.16228595878669522,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002952264363178295,
+      "loss": 1.2986,
+      "step": 1849
+    },
+    {
+      "epoch": 0.1623737283695977,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0029521525149180275,
+      "loss": 1.349,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1624614979525002,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.002952040538140459,
+      "loss": 1.3361,
+      "step": 1851
+    },
+    {
+      "epoch": 0.1625492675354027,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002951928432856641,
+      "loss": 1.2789,
+      "step": 1852
+    },
+    {
+      "epoch": 0.16263703711830516,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.002951816199077639,
+      "loss": 1.3311,
+      "step": 1853
+    },
+    {
+      "epoch": 0.16272480670120765,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029517038368145287,
+      "loss": 1.3419,
+      "step": 1854
+    },
+    {
+      "epoch": 0.16281257628411014,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0029515913460784,
+      "loss": 1.2982,
+      "step": 1855
+    },
+    {
+      "epoch": 0.16290034586701263,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0029514787268803554,
+      "loss": 1.3098,
+      "step": 1856
+    },
+    {
+      "epoch": 0.16298811544991512,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0029513659792315106,
+      "loss": 1.2828,
+      "step": 1857
+    },
+    {
+      "epoch": 0.16307588503281759,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0029512531031429933,
+      "loss": 1.3173,
+      "step": 1858
+    },
+    {
+      "epoch": 0.16316365461572008,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0029511400986259426,
+      "loss": 1.3546,
+      "step": 1859
+    },
+    {
+      "epoch": 0.16325142419862257,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002951026965691513,
+      "loss": 1.3284,
+      "step": 1860
+    },
+    {
+      "epoch": 0.16333919378152506,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00295091370435087,
+      "loss": 1.3001,
+      "step": 1861
+    },
+    {
+      "epoch": 0.16342696336442752,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0029508003146151915,
+      "loss": 1.3214,
+      "step": 1862
+    },
+    {
+      "epoch": 0.16351473294733002,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0029506867964956694,
+      "loss": 1.2816,
+      "step": 1863
+    },
+    {
+      "epoch": 0.1636025025302325,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0029505731500035063,
+      "loss": 1.3581,
+      "step": 1864
+    },
+    {
+      "epoch": 0.163690272113135,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0029504593751499197,
+      "loss": 1.33,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1637780416960375,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002950345471946139,
+      "loss": 1.318,
+      "step": 1866
+    },
+    {
+      "epoch": 0.16386581127893995,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0029502314404034045,
+      "loss": 1.2866,
+      "step": 1867
+    },
+    {
+      "epoch": 0.16395358086184245,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002950117280532972,
+      "loss": 1.3199,
+      "step": 1868
+    },
+    {
+      "epoch": 0.16404135044474494,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029500029923461087,
+      "loss": 1.307,
+      "step": 1869
+    },
+    {
+      "epoch": 0.16412912002764743,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0029498885758540935,
+      "loss": 1.2498,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1642168896105499,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0029497740310682192,
+      "loss": 1.3835,
+      "step": 1871
+    },
+    {
+      "epoch": 0.16430465919345238,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.002949659357999791,
+      "loss": 1.3665,
+      "step": 1872
+    },
+    {
+      "epoch": 0.16439242877635488,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0029495445566601265,
+      "loss": 1.3354,
+      "step": 1873
+    },
+    {
+      "epoch": 0.16448019835925737,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0029494296270605566,
+      "loss": 1.3732,
+      "step": 1874
+    },
+    {
+      "epoch": 0.16456796794215983,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002949314569212424,
+      "loss": 1.3693,
+      "step": 1875
+    },
+    {
+      "epoch": 0.16465573752506232,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0029491993831270847,
+      "loss": 1.3408,
+      "step": 1876
+    },
+    {
+      "epoch": 0.16474350710796482,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.002949084068815907,
+      "loss": 1.3244,
+      "step": 1877
+    },
+    {
+      "epoch": 0.1648312766908673,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.002948968626290271,
+      "loss": 1.32,
+      "step": 1878
+    },
+    {
+      "epoch": 0.1649190462737698,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0029488530555615722,
+      "loss": 1.2796,
+      "step": 1879
+    },
+    {
+      "epoch": 0.16500681585667226,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0029487373566412156,
+      "loss": 1.3104,
+      "step": 1880
+    },
+    {
+      "epoch": 0.16509458543957475,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0029486215295406216,
+      "loss": 1.3216,
+      "step": 1881
+    },
+    {
+      "epoch": 0.16518235502247725,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0029485055742712206,
+      "loss": 1.3015,
+      "step": 1882
+    },
+    {
+      "epoch": 0.16527012460537974,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002948389490844457,
+      "loss": 1.3027,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1653578941882822,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002948273279271789,
+      "loss": 1.2722,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1654456637711847,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0029481569395646843,
+      "loss": 1.2734,
+      "step": 1885
+    },
+    {
+      "epoch": 0.16553343335408718,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002948040471734627,
+      "loss": 1.3116,
+      "step": 1886
+    },
+    {
+      "epoch": 0.16562120293698968,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00294792387579311,
+      "loss": 1.2624,
+      "step": 1887
+    },
+    {
+      "epoch": 0.16570897251989214,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0029478071517516436,
+      "loss": 1.3011,
+      "step": 1888
+    },
+    {
+      "epoch": 0.16579674210279463,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.002947690299621746,
+      "loss": 1.3676,
+      "step": 1889
+    },
+    {
+      "epoch": 0.16588451168569712,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0029475733194149504,
+      "loss": 1.3082,
+      "step": 1890
+    },
+    {
+      "epoch": 0.16597228126859961,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0029474562111428022,
+      "loss": 1.3579,
+      "step": 1891
+    },
+    {
+      "epoch": 0.1660600508515021,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00294733897481686,
+      "loss": 1.3564,
+      "step": 1892
+    },
+    {
+      "epoch": 0.16614782043440457,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002947221610448694,
+      "loss": 1.3186,
+      "step": 1893
+    },
+    {
+      "epoch": 0.16623559001730706,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002947104118049888,
+      "loss": 1.3272,
+      "step": 1894
+    },
+    {
+      "epoch": 0.16632335960020955,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0029469864976320384,
+      "loss": 1.2855,
+      "step": 1895
+    },
+    {
+      "epoch": 0.16641112918311204,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.002946868749206753,
+      "loss": 1.3089,
+      "step": 1896
+    },
+    {
+      "epoch": 0.1664988987660145,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.002946750872785654,
+      "loss": 1.2869,
+      "step": 1897
+    },
+    {
+      "epoch": 0.166586668348917,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002946632868380374,
+      "loss": 1.38,
+      "step": 1898
+    },
+    {
+      "epoch": 0.1666744379318195,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002946514736002561,
+      "loss": 1.3421,
+      "step": 1899
+    },
+    {
+      "epoch": 0.16676220751472198,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.002946396475663873,
+      "loss": 1.2534,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16684997709762447,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0029462780873759833,
+      "loss": 1.3187,
+      "step": 1901
+    },
+    {
+      "epoch": 0.16693774668052694,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0029461595711505746,
+      "loss": 1.312,
+      "step": 1902
+    },
+    {
+      "epoch": 0.16702551626342943,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0029460409269993454,
+      "loss": 1.2773,
+      "step": 1903
+    },
+    {
+      "epoch": 0.16711328584633192,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.002945922154934005,
+      "loss": 1.3131,
+      "step": 1904
+    },
+    {
+      "epoch": 0.1672010554292344,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0029458032549662754,
+      "loss": 1.3198,
+      "step": 1905
+    },
+    {
+      "epoch": 0.16728882501213688,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029456842271078916,
+      "loss": 1.3323,
+      "step": 1906
+    },
+    {
+      "epoch": 0.16737659459503937,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0029455650713706015,
+      "loss": 1.3593,
+      "step": 1907
+    },
+    {
+      "epoch": 0.16746436417794186,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0029454457877661653,
+      "loss": 1.2929,
+      "step": 1908
+    },
+    {
+      "epoch": 0.16755213376084435,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0029453263763063557,
+      "loss": 1.3747,
+      "step": 1909
+    },
+    {
+      "epoch": 0.16763990334374682,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.002945206837002958,
+      "loss": 1.3061,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1677276729266493,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029450871698677703,
+      "loss": 1.2855,
+      "step": 1911
+    },
+    {
+      "epoch": 0.1678154425095518,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0029449673749126033,
+      "loss": 1.312,
+      "step": 1912
+    },
+    {
+      "epoch": 0.1679032120924543,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0029448474521492803,
+      "loss": 1.3613,
+      "step": 1913
+    },
+    {
+      "epoch": 0.16799098167535678,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029447274015896373,
+      "loss": 1.345,
+      "step": 1914
+    },
+    {
+      "epoch": 0.16807875125825925,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.002944607223245523,
+      "loss": 1.2973,
+      "step": 1915
+    },
+    {
+      "epoch": 0.16816652084116174,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0029444869171287983,
+      "loss": 1.3704,
+      "step": 1916
+    },
+    {
+      "epoch": 0.16825429042406423,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0029443664832513364,
+      "loss": 1.2957,
+      "step": 1917
+    },
+    {
+      "epoch": 0.16834206000696672,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002944245921625024,
+      "loss": 1.2621,
+      "step": 1918
+    },
+    {
+      "epoch": 0.16842982958986918,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0029441252322617607,
+      "loss": 1.382,
+      "step": 1919
+    },
+    {
+      "epoch": 0.16851759917277168,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0029440044151734573,
+      "loss": 1.2368,
+      "step": 1920
+    },
+    {
+      "epoch": 0.16860536875567417,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0029438834703720386,
+      "loss": 1.3038,
+      "step": 1921
+    },
+    {
+      "epoch": 0.16869313833857666,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.002943762397869441,
+      "loss": 1.2904,
+      "step": 1922
+    },
+    {
+      "epoch": 0.16878090792147912,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0029436411976776137,
+      "loss": 1.3509,
+      "step": 1923
+    },
+    {
+      "epoch": 0.16886867750438161,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0029435198698085185,
+      "loss": 1.3115,
+      "step": 1924
+    },
+    {
+      "epoch": 0.1689564470872841,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002943398414274131,
+      "loss": 1.3038,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1690442166701866,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002943276831086438,
+      "loss": 1.3011,
+      "step": 1926
+    },
+    {
+      "epoch": 0.1691319862530891,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.002943155120257438,
+      "loss": 1.2891,
+      "step": 1927
+    },
+    {
+      "epoch": 0.16921975583599155,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0029430332817991446,
+      "loss": 1.2734,
+      "step": 1928
+    },
+    {
+      "epoch": 0.16930752541889404,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002942911315723583,
+      "loss": 1.3928,
+      "step": 1929
+    },
+    {
+      "epoch": 0.16939529500179654,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00294278922204279,
+      "loss": 1.3,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16948306458469903,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0029426670007688164,
+      "loss": 1.2545,
+      "step": 1931
+    },
+    {
+      "epoch": 0.1695708341676015,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002942544651913724,
+      "loss": 1.3696,
+      "step": 1932
+    },
+    {
+      "epoch": 0.16965860375050398,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0029424221754895895,
+      "loss": 1.2651,
+      "step": 1933
+    },
+    {
+      "epoch": 0.16974637333340648,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0029422995715085,
+      "loss": 1.3701,
+      "step": 1934
+    },
+    {
+      "epoch": 0.16983414291630897,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0029421768399825565,
+      "loss": 1.2822,
+      "step": 1935
+    },
+    {
+      "epoch": 0.16992191249921146,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0029420539809238712,
+      "loss": 1.2544,
+      "step": 1936
+    },
+    {
+      "epoch": 0.17000968208211392,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002941930994344571,
+      "loss": 1.3308,
+      "step": 1937
+    },
+    {
+      "epoch": 0.1700974516650164,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002941807880256792,
+      "loss": 1.3318,
+      "step": 1938
+    },
+    {
+      "epoch": 0.1701852212479189,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0029416846386726885,
+      "loss": 1.3578,
+      "step": 1939
+    },
+    {
+      "epoch": 0.1702729908308214,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029415612696044213,
+      "loss": 1.3668,
+      "step": 1940
+    },
+    {
+      "epoch": 0.17036076041372386,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0029414377730641672,
+      "loss": 1.4032,
+      "step": 1941
+    },
+    {
+      "epoch": 0.17044852999662635,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0029413141490641148,
+      "loss": 1.3287,
+      "step": 1942
+    },
+    {
+      "epoch": 0.17053629957952884,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0029411903976164656,
+      "loss": 1.3229,
+      "step": 1943
+    },
+    {
+      "epoch": 0.17062406916243134,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0029410665187334334,
+      "loss": 1.2777,
+      "step": 1944
+    },
+    {
+      "epoch": 0.1707118387453338,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0029409425124272436,
+      "loss": 1.3409,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1707996083282363,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.002940818378710136,
+      "loss": 1.3148,
+      "step": 1946
+    },
+    {
+      "epoch": 0.17088737791113878,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0029406941175943624,
+      "loss": 1.2958,
+      "step": 1947
+    },
+    {
+      "epoch": 0.17097514749404127,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002940569729092186,
+      "loss": 1.37,
+      "step": 1948
+    },
+    {
+      "epoch": 0.17106291707694377,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.002940445213215884,
+      "loss": 1.2833,
+      "step": 1949
+    },
+    {
+      "epoch": 0.17115068665984623,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0029403205699777453,
+      "loss": 1.3383,
+      "step": 1950
+    },
+    {
+      "epoch": 0.17123845624274872,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0029401957993900724,
+      "loss": 1.3357,
+      "step": 1951
+    },
+    {
+      "epoch": 0.1713262258256512,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.002940070901465179,
+      "loss": 1.375,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1714139954085537,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0029399458762153917,
+      "loss": 1.28,
+      "step": 1953
+    },
+    {
+      "epoch": 0.17150176499145617,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0029398207236530512,
+      "loss": 1.3448,
+      "step": 1954
+    },
+    {
+      "epoch": 0.17158953457435866,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0029396954437905082,
+      "loss": 1.283,
+      "step": 1955
+    },
+    {
+      "epoch": 0.17167730415726115,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0029395700366401283,
+      "loss": 1.2932,
+      "step": 1956
+    },
+    {
+      "epoch": 0.17176507374016364,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002939444502214288,
+      "loss": 1.3213,
+      "step": 1957
+    },
+    {
+      "epoch": 0.17185284332306613,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0029393188405253774,
+      "loss": 1.2877,
+      "step": 1958
+    },
+    {
+      "epoch": 0.1719406129059686,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0029391930515857995,
+      "loss": 1.3419,
+      "step": 1959
+    },
+    {
+      "epoch": 0.1720283824888711,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0029390671354079683,
+      "loss": 1.4361,
+      "step": 1960
+    },
+    {
+      "epoch": 0.17211615207177358,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002938941092004311,
+      "loss": 1.335,
+      "step": 1961
+    },
+    {
+      "epoch": 0.17220392165467607,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.002938814921387268,
+      "loss": 1.3516,
+      "step": 1962
+    },
+    {
+      "epoch": 0.17229169123757854,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002938688623569292,
+      "loss": 1.3127,
+      "step": 1963
+    },
+    {
+      "epoch": 0.17237946082048103,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.002938562198562848,
+      "loss": 1.2745,
+      "step": 1964
+    },
+    {
+      "epoch": 0.17246723040338352,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0029384356463804133,
+      "loss": 1.3242,
+      "step": 1965
+    },
+    {
+      "epoch": 0.172554999986286,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0029383089670344787,
+      "loss": 1.3244,
+      "step": 1966
+    },
+    {
+      "epoch": 0.17264276956918848,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029381821605375465,
+      "loss": 1.3211,
+      "step": 1967
+    },
+    {
+      "epoch": 0.17273053915209097,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0029380552269021322,
+      "loss": 1.3583,
+      "step": 1968
+    },
+    {
+      "epoch": 0.17281830873499346,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002937928166140764,
+      "loss": 1.3176,
+      "step": 1969
+    },
+    {
+      "epoch": 0.17290607831789595,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0029378009782659816,
+      "loss": 1.4045,
+      "step": 1970
+    },
+    {
+      "epoch": 0.17299384790079844,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0029376736632903376,
+      "loss": 1.2959,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1730816174837009,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002937546221226399,
+      "loss": 1.3519,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1731693870666034,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0029374186520867425,
+      "loss": 1.3616,
+      "step": 1973
+    },
+    {
+      "epoch": 0.1732571566495059,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002937290955883959,
+      "loss": 1.2925,
+      "step": 1974
+    },
+    {
+      "epoch": 0.17334492623240838,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0029371631326306517,
+      "loss": 1.3424,
+      "step": 1975
+    },
+    {
+      "epoch": 0.17343269581531084,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0029370351823394363,
+      "loss": 1.2905,
+      "step": 1976
+    },
+    {
+      "epoch": 0.17352046539821334,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.002936907105022941,
+      "loss": 1.3576,
+      "step": 1977
+    },
+    {
+      "epoch": 0.17360823498111583,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002936778900693807,
+      "loss": 1.3014,
+      "step": 1978
+    },
+    {
+      "epoch": 0.17369600456401832,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0029366505693646863,
+      "loss": 1.3876,
+      "step": 1979
+    },
+    {
+      "epoch": 0.17378377414692078,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0029365221110482457,
+      "loss": 1.3465,
+      "step": 1980
+    },
+    {
+      "epoch": 0.17387154372982327,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002936393525757163,
+      "loss": 1.2636,
+      "step": 1981
+    },
+    {
+      "epoch": 0.17395931331272577,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0029362648135041296,
+      "loss": 1.3364,
+      "step": 1982
+    },
+    {
+      "epoch": 0.17404708289562826,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0029361359743018487,
+      "loss": 1.3762,
+      "step": 1983
+    },
+    {
+      "epoch": 0.17413485247853075,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002936007008163036,
+      "loss": 1.3732,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1742226220614332,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0029358779151004194,
+      "loss": 1.3531,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1743103916443357,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0029357486951267415,
+      "loss": 1.2781,
+      "step": 1986
+    },
+    {
+      "epoch": 0.1743981612272382,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002935619348254754,
+      "loss": 1.2857,
+      "step": 1987
+    },
+    {
+      "epoch": 0.1744859308101407,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0029354898744972245,
+      "loss": 1.2691,
+      "step": 1988
+    },
+    {
+      "epoch": 0.17457370039304315,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0029353602738669306,
+      "loss": 1.3292,
+      "step": 1989
+    },
+    {
+      "epoch": 0.17466146997594564,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0029352305463766636,
+      "loss": 1.2833,
+      "step": 1990
+    },
+    {
+      "epoch": 0.17474923955884814,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0029351006920392267,
+      "loss": 1.3149,
+      "step": 1991
+    },
+    {
+      "epoch": 0.17483700914175063,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002934970710867437,
+      "loss": 1.2937,
+      "step": 1992
+    },
+    {
+      "epoch": 0.17492477872465312,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0029348406028741223,
+      "loss": 1.2783,
+      "step": 1993
+    },
+    {
+      "epoch": 0.17501254830755558,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0029347103680721237,
+      "loss": 1.304,
+      "step": 1994
+    },
+    {
+      "epoch": 0.17510031789045807,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002934580006474295,
+      "loss": 1.3559,
+      "step": 1995
+    },
+    {
+      "epoch": 0.17518808747336057,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0029344495180935034,
+      "loss": 1.347,
+      "step": 1996
+    },
+    {
+      "epoch": 0.17527585705626306,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0029343189029426256,
+      "loss": 1.3305,
+      "step": 1997
+    },
+    {
+      "epoch": 0.17536362663916552,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0029341881610345545,
+      "loss": 1.2606,
+      "step": 1998
+    },
+    {
+      "epoch": 0.175451396222068,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.002934057292382193,
+      "loss": 1.2748,
+      "step": 1999
+    },
+    {
+      "epoch": 0.1755391658049705,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0029339262969984574,
+      "loss": 1.2638,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1755391658049705,
+      "eval_loss": 1.303299069404602,
+      "eval_runtime": 437.5784,
+      "eval_samples_per_second": 33.685,
+      "eval_steps_per_second": 8.421,
+      "step": 2000
+    },
+    {
+      "epoch": 0.175626935387873,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0029337951748962766,
+      "loss": 1.3049,
+      "step": 2001
+    },
+    {
+      "epoch": 0.17571470497077546,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0029336639260885916,
+      "loss": 1.3366,
+      "step": 2002
+    },
+    {
+      "epoch": 0.17580247455367795,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0029335325505883566,
+      "loss": 1.3756,
+      "step": 2003
+    },
+    {
+      "epoch": 0.17589024413658044,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0029334010484085376,
+      "loss": 1.3383,
+      "step": 2004
+    },
+    {
+      "epoch": 0.17597801371948293,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002933269419562113,
+      "loss": 1.3419,
+      "step": 2005
+    },
+    {
+      "epoch": 0.17606578330238543,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002933137664062074,
+      "loss": 1.3246,
+      "step": 2006
+    },
+    {
+      "epoch": 0.1761535528852879,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002933005781921426,
+      "loss": 1.2929,
+      "step": 2007
+    },
+    {
+      "epoch": 0.17624132246819038,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029328737731531824,
+      "loss": 1.3092,
+      "step": 2008
+    },
+    {
+      "epoch": 0.17632909205109287,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0029327416377703743,
+      "loss": 1.3364,
+      "step": 2009
+    },
+    {
+      "epoch": 0.17641686163399536,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0029326093757860417,
+      "loss": 1.3512,
+      "step": 2010
+    },
+    {
+      "epoch": 0.17650463121689783,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0029324769872132394,
+      "loss": 1.2838,
+      "step": 2011
+    },
+    {
+      "epoch": 0.17659240079980032,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0029323444720650324,
+      "loss": 1.272,
+      "step": 2012
+    },
+    {
+      "epoch": 0.1766801703827028,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0029322118303545,
+      "loss": 1.3825,
+      "step": 2013
+    },
+    {
+      "epoch": 0.1767679399656053,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0029320790620947336,
+      "loss": 1.2813,
+      "step": 2014
+    },
+    {
+      "epoch": 0.17685570954850777,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0029319461672988366,
+      "loss": 1.3582,
+      "step": 2015
+    },
+    {
+      "epoch": 0.17694347913141026,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0029318131459799253,
+      "loss": 1.2998,
+      "step": 2016
+    },
+    {
+      "epoch": 0.17703124871431275,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029316799981511286,
+      "loss": 1.3067,
+      "step": 2017
+    },
+    {
+      "epoch": 0.17711901829721524,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.002931546723825587,
+      "loss": 1.2548,
+      "step": 2018
+    },
+    {
+      "epoch": 0.17720678788011773,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0029314133230164546,
+      "loss": 1.3689,
+      "step": 2019
+    },
+    {
+      "epoch": 0.1772945574630202,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0029312797957368976,
+      "loss": 1.3229,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1773823270459227,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0029311461420000948,
+      "loss": 1.3033,
+      "step": 2021
+    },
+    {
+      "epoch": 0.17747009662882518,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.002931012361819237,
+      "loss": 1.2997,
+      "step": 2022
+    },
+    {
+      "epoch": 0.17755786621172767,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002930878455207527,
+      "loss": 1.3353,
+      "step": 2023
+    },
+    {
+      "epoch": 0.17764563579463014,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.002930744422178183,
+      "loss": 1.2805,
+      "step": 2024
+    },
+    {
+      "epoch": 0.17773340537753263,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002930610262744431,
+      "loss": 1.3085,
+      "step": 2025
+    },
+    {
+      "epoch": 0.17782117496043512,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0029304759769195137,
+      "loss": 1.304,
+      "step": 2026
+    },
+    {
+      "epoch": 0.1779089445433376,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0029303415647166844,
+      "loss": 1.2915,
+      "step": 2027
+    },
+    {
+      "epoch": 0.1779967141262401,
+      "grad_norm": 0.125,
+      "learning_rate": 0.002930207026149208,
+      "loss": 1.3411,
+      "step": 2028
+    },
+    {
+      "epoch": 0.17808448370914257,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0029300723612303643,
+      "loss": 1.3686,
+      "step": 2029
+    },
+    {
+      "epoch": 0.17817225329204506,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002929937569973443,
+      "loss": 1.283,
+      "step": 2030
+    },
+    {
+      "epoch": 0.17826002287494755,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0029298026523917483,
+      "loss": 1.3273,
+      "step": 2031
+    },
+    {
+      "epoch": 0.17834779245785004,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.002929667608498596,
+      "loss": 1.3028,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1784355620407525,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002929532438307314,
+      "loss": 1.3427,
+      "step": 2033
+    },
+    {
+      "epoch": 0.178523331623655,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0029293971418312427,
+      "loss": 1.3177,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1786111012065575,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.002929261719083736,
+      "loss": 1.3875,
+      "step": 2035
+    },
+    {
+      "epoch": 0.17869887078945998,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0029291261700781592,
+      "loss": 1.3412,
+      "step": 2036
+    },
+    {
+      "epoch": 0.17878664037236244,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0029289904948278916,
+      "loss": 1.3463,
+      "step": 2037
+    },
+    {
+      "epoch": 0.17887440995526493,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002928854693346322,
+      "loss": 1.3193,
+      "step": 2038
+    },
+    {
+      "epoch": 0.17896217953816743,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002928718765646855,
+      "loss": 1.3168,
+      "step": 2039
+    },
+    {
+      "epoch": 0.17904994912106992,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002928582711742905,
+      "loss": 1.3616,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1791377187039724,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029284465316479,
+      "loss": 1.301,
+      "step": 2041
+    },
+    {
+      "epoch": 0.17922548828687487,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0029283102253752817,
+      "loss": 1.299,
+      "step": 2042
+    },
+    {
+      "epoch": 0.17931325786977736,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002928173792938502,
+      "loss": 1.3139,
+      "step": 2043
+    },
+    {
+      "epoch": 0.17940102745267986,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0029280372343510265,
+      "loss": 1.3261,
+      "step": 2044
+    },
+    {
+      "epoch": 0.17948879703558235,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0029279005496263325,
+      "loss": 1.3821,
+      "step": 2045
+    },
+    {
+      "epoch": 0.1795765666184848,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0029277637387779107,
+      "loss": 1.3539,
+      "step": 2046
+    },
+    {
+      "epoch": 0.1796643362013873,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.002927626801819264,
+      "loss": 1.2577,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1797521057842898,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0029274897387639074,
+      "loss": 1.3158,
+      "step": 2048
+    },
+    {
+      "epoch": 0.1798398753671923,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002927352549625368,
+      "loss": 1.2814,
+      "step": 2049
+    },
+    {
+      "epoch": 0.17992764495009475,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.002927215234417186,
+      "loss": 1.3311,
+      "step": 2050
+    },
+    {
+      "epoch": 0.18001541453299724,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0029270777931529146,
+      "loss": 1.3424,
+      "step": 2051
+    },
+    {
+      "epoch": 0.18010318411589973,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002926940225846118,
+      "loss": 1.2519,
+      "step": 2052
+    },
+    {
+      "epoch": 0.18019095369880223,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002926802532510373,
+      "loss": 1.3654,
+      "step": 2053
+    },
+    {
+      "epoch": 0.18027872328170472,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002926664713159271,
+      "loss": 1.3042,
+      "step": 2054
+    },
+    {
+      "epoch": 0.18036649286460718,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.002926526767806413,
+      "loss": 1.2686,
+      "step": 2055
+    },
+    {
+      "epoch": 0.18045426244750967,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0029263886964654142,
+      "loss": 1.2894,
+      "step": 2056
+    },
+    {
+      "epoch": 0.18054203203041216,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002926250499149901,
+      "loss": 1.3748,
+      "step": 2057
+    },
+    {
+      "epoch": 0.18062980161331466,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002926112175873514,
+      "loss": 1.3336,
+      "step": 2058
+    },
+    {
+      "epoch": 0.18071757119621712,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0029259737266499044,
+      "loss": 1.3166,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1808053407791196,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0029258351514927367,
+      "loss": 1.3143,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1808931103620221,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0029256964504156876,
+      "loss": 1.2732,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1809808799449246,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002925557623432447,
+      "loss": 1.318,
+      "step": 2062
+    },
+    {
+      "epoch": 0.18106864952782709,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.002925418670556715,
+      "loss": 1.3617,
+      "step": 2063
+    },
+    {
+      "epoch": 0.18115641911072955,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0029252795918022077,
+      "loss": 1.301,
+      "step": 2064
+    },
+    {
+      "epoch": 0.18124418869363204,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0029251403871826506,
+      "loss": 1.3337,
+      "step": 2065
+    },
+    {
+      "epoch": 0.18133195827653453,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.002925001056711783,
+      "loss": 1.2517,
+      "step": 2066
+    },
+    {
+      "epoch": 0.18141972785943702,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0029248616004033555,
+      "loss": 1.3646,
+      "step": 2067
+    },
+    {
+      "epoch": 0.1815074974423395,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029247220182711323,
+      "loss": 1.2915,
+      "step": 2068
+    },
+    {
+      "epoch": 0.18159526702524198,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00292458231032889,
+      "loss": 1.2812,
+      "step": 2069
+    },
+    {
+      "epoch": 0.18168303660814447,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0029244424765904175,
+      "loss": 1.3036,
+      "step": 2070
+    },
+    {
+      "epoch": 0.18177080619104696,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0029243025170695145,
+      "loss": 1.2834,
+      "step": 2071
+    },
+    {
+      "epoch": 0.18185857577394943,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0029241624317799956,
+      "loss": 1.3161,
+      "step": 2072
+    },
+    {
+      "epoch": 0.18194634535685192,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002924022220735686,
+      "loss": 1.2891,
+      "step": 2073
+    },
+    {
+      "epoch": 0.1820341149397544,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0029238818839504247,
+      "loss": 1.3126,
+      "step": 2074
+    },
+    {
+      "epoch": 0.1821218845226569,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002923741421438061,
+      "loss": 1.3688,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1822096541055594,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0029236008332124597,
+      "loss": 1.3122,
+      "step": 2076
+    },
+    {
+      "epoch": 0.18229742368846186,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002923460119287496,
+      "loss": 1.3137,
+      "step": 2077
+    },
+    {
+      "epoch": 0.18238519327136435,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002923319279677057,
+      "loss": 1.2864,
+      "step": 2078
+    },
+    {
+      "epoch": 0.18247296285426684,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002923178314395043,
+      "loss": 1.2935,
+      "step": 2079
+    },
+    {
+      "epoch": 0.18256073243716933,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0029230372234553673,
+      "loss": 1.2647,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1826485020200718,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0029228960068719547,
+      "loss": 1.3071,
+      "step": 2081
+    },
+    {
+      "epoch": 0.1827362716029743,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0029227546646587426,
+      "loss": 1.3071,
+      "step": 2082
+    },
+    {
+      "epoch": 0.18282404118587678,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002922613196829681,
+      "loss": 1.3912,
+      "step": 2083
+    },
+    {
+      "epoch": 0.18291181076877927,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002922471603398733,
+      "loss": 1.3432,
+      "step": 2084
+    },
+    {
+      "epoch": 0.18299958035168173,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0029223298843798727,
+      "loss": 1.3597,
+      "step": 2085
+    },
+    {
+      "epoch": 0.18308734993458423,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0029221880397870873,
+      "loss": 1.2975,
+      "step": 2086
+    },
+    {
+      "epoch": 0.18317511951748672,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0029220460696343753,
+      "loss": 1.3389,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1832628891003892,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00292190397393575,
+      "loss": 1.32,
+      "step": 2088
+    },
+    {
+      "epoch": 0.1833506586832917,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.002921761752705235,
+      "loss": 1.336,
+      "step": 2089
+    },
+    {
+      "epoch": 0.18343842826619416,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002921619405956867,
+      "loss": 1.3098,
+      "step": 2090
+    },
+    {
+      "epoch": 0.18352619784909666,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.002921476933704696,
+      "loss": 1.2683,
+      "step": 2091
+    },
+    {
+      "epoch": 0.18361396743199915,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0029213343359627817,
+      "loss": 1.2809,
+      "step": 2092
+    },
+    {
+      "epoch": 0.18370173701490164,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0029211916127451993,
+      "loss": 1.2894,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1837895065978041,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0029210487640660343,
+      "loss": 1.2965,
+      "step": 2094
+    },
+    {
+      "epoch": 0.1838772761807066,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002920905789939386,
+      "loss": 1.3088,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1839650457636091,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0029207626903793643,
+      "loss": 1.2807,
+      "step": 2096
+    },
+    {
+      "epoch": 0.18405281534651158,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0029206194654000936,
+      "loss": 1.2842,
+      "step": 2097
+    },
+    {
+      "epoch": 0.18414058492941407,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002920476115015709,
+      "loss": 1.3278,
+      "step": 2098
+    },
+    {
+      "epoch": 0.18422835451231653,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0029203326392403597,
+      "loss": 1.3635,
+      "step": 2099
+    },
+    {
+      "epoch": 0.18431612409521902,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0029201890380882046,
+      "loss": 1.321,
+      "step": 2100
+    },
+    {
+      "epoch": 0.18440389367812152,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0029200453115734177,
+      "loss": 1.3196,
+      "step": 2101
+    },
+    {
+      "epoch": 0.184491663261024,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0029199014597101835,
+      "loss": 1.308,
+      "step": 2102
+    },
+    {
+      "epoch": 0.18457943284392647,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0029197574825126998,
+      "loss": 1.317,
+      "step": 2103
+    },
+    {
+      "epoch": 0.18466720242682896,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002919613379995177,
+      "loss": 1.2931,
+      "step": 2104
+    },
+    {
+      "epoch": 0.18475497200973146,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0029194691521718377,
+      "loss": 1.2302,
+      "step": 2105
+    },
+    {
+      "epoch": 0.18484274159263395,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029193247990569153,
+      "loss": 1.3406,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1849305111755364,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0029191803206646584,
+      "loss": 1.276,
+      "step": 2107
+    },
+    {
+      "epoch": 0.1850182807584389,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.002919035717009325,
+      "loss": 1.3149,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1851060503413414,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0029188909881051883,
+      "loss": 1.3338,
+      "step": 2109
+    },
+    {
+      "epoch": 0.18519381992424389,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0029187461339665317,
+      "loss": 1.3204,
+      "step": 2110
+    },
+    {
+      "epoch": 0.18528158950714638,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0029186011546076516,
+      "loss": 1.317,
+      "step": 2111
+    },
+    {
+      "epoch": 0.18536935909004884,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0029184560500428573,
+      "loss": 1.2885,
+      "step": 2112
+    },
+    {
+      "epoch": 0.18545712867295133,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0029183108202864697,
+      "loss": 1.3115,
+      "step": 2113
+    },
+    {
+      "epoch": 0.18554489825585382,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.002918165465352823,
+      "loss": 1.3109,
+      "step": 2114
+    },
+    {
+      "epoch": 0.18563266783875632,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002918019985256263,
+      "loss": 1.2992,
+      "step": 2115
+    },
+    {
+      "epoch": 0.18572043742165878,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002917874380011147,
+      "loss": 1.3017,
+      "step": 2116
+    },
+    {
+      "epoch": 0.18580820700456127,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029177286496318462,
+      "loss": 1.3196,
+      "step": 2117
+    },
+    {
+      "epoch": 0.18589597658746376,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0029175827941327445,
+      "loss": 1.2701,
+      "step": 2118
+    },
+    {
+      "epoch": 0.18598374617036625,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0029174368135282368,
+      "loss": 1.2962,
+      "step": 2119
+    },
+    {
+      "epoch": 0.18607151575326872,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0029172907078327303,
+      "loss": 1.2666,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1861592853361712,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002917144477060645,
+      "loss": 1.2587,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1862470549190737,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.002916998121226414,
+      "loss": 1.2867,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1863348245019762,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0029168516403444817,
+      "loss": 1.3342,
+      "step": 2123
+    },
+    {
+      "epoch": 0.18642259408487868,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0029167050344293052,
+      "loss": 1.3325,
+      "step": 2124
+    },
+    {
+      "epoch": 0.18651036366778115,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002916558303495354,
+      "loss": 1.3282,
+      "step": 2125
+    },
+    {
+      "epoch": 0.18659813325068364,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0029164114475571097,
+      "loss": 1.2693,
+      "step": 2126
+    },
+    {
+      "epoch": 0.18668590283358613,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.002916264466629066,
+      "loss": 1.2907,
+      "step": 2127
+    },
+    {
+      "epoch": 0.18677367241648862,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00291611736072573,
+      "loss": 1.3408,
+      "step": 2128
+    },
+    {
+      "epoch": 0.1868614419993911,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00291597012986162,
+      "loss": 1.345,
+      "step": 2129
+    },
+    {
+      "epoch": 0.18694921158229358,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002915822774051268,
+      "loss": 1.3048,
+      "step": 2130
+    },
+    {
+      "epoch": 0.18703698116519607,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0029156752933092167,
+      "loss": 1.3583,
+      "step": 2131
+    },
+    {
+      "epoch": 0.18712475074809856,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002915527687650022,
+      "loss": 1.3022,
+      "step": 2132
+    },
+    {
+      "epoch": 0.18721252033100105,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0029153799570882516,
+      "loss": 1.38,
+      "step": 2133
+    },
+    {
+      "epoch": 0.18730028991390352,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002915232101638487,
+      "loss": 1.31,
+      "step": 2134
+    },
+    {
+      "epoch": 0.187388059496806,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00291508412131532,
+      "loss": 1.2774,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1874758290797085,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002914936016133355,
+      "loss": 1.2977,
+      "step": 2136
+    },
+    {
+      "epoch": 0.187563598662611,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0029147877861072113,
+      "loss": 1.3102,
+      "step": 2137
+    },
+    {
+      "epoch": 0.18765136824551346,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029146394312515173,
+      "loss": 1.2452,
+      "step": 2138
+    },
+    {
+      "epoch": 0.18773913782841595,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0029144909515809152,
+      "loss": 1.3241,
+      "step": 2139
+    },
+    {
+      "epoch": 0.18782690741131844,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0029143423471100598,
+      "loss": 1.2632,
+      "step": 2140
+    },
+    {
+      "epoch": 0.18791467699422093,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0029141936178536175,
+      "loss": 1.3414,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1880024465771234,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0029140447638262667,
+      "loss": 1.3056,
+      "step": 2142
+    },
+    {
+      "epoch": 0.18809021616002589,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0029138957850427002,
+      "loss": 1.3506,
+      "step": 2143
+    },
+    {
+      "epoch": 0.18817798574292838,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00291374668151762,
+      "loss": 1.3226,
+      "step": 2144
+    },
+    {
+      "epoch": 0.18826575532583087,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.002913597453265743,
+      "loss": 1.3124,
+      "step": 2145
+    },
+    {
+      "epoch": 0.18835352490873336,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002913448100301797,
+      "loss": 1.3316,
+      "step": 2146
+    },
+    {
+      "epoch": 0.18844129449163582,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0029132986226405233,
+      "loss": 1.2576,
+      "step": 2147
+    },
+    {
+      "epoch": 0.18852906407453832,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0029131490202966735,
+      "loss": 1.347,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1886168336574408,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0029129992932850138,
+      "loss": 1.3147,
+      "step": 2149
+    },
+    {
+      "epoch": 0.1887046032403433,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0029128494416203214,
+      "loss": 1.3429,
+      "step": 2150
+    },
+    {
+      "epoch": 0.18879237282324576,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0029126994653173852,
+      "loss": 1.3109,
+      "step": 2151
+    },
+    {
+      "epoch": 0.18888014240614825,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0029125493643910086,
+      "loss": 1.2686,
+      "step": 2152
+    },
+    {
+      "epoch": 0.18896791198905075,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0029123991388560052,
+      "loss": 1.3486,
+      "step": 2153
+    },
+    {
+      "epoch": 0.18905568157195324,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002912248788727202,
+      "loss": 1.2511,
+      "step": 2154
+    },
+    {
+      "epoch": 0.1891434511548557,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0029120983140194377,
+      "loss": 1.3016,
+      "step": 2155
+    },
+    {
+      "epoch": 0.1892312207377582,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002911947714747564,
+      "loss": 1.259,
+      "step": 2156
+    },
+    {
+      "epoch": 0.18931899032066069,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0029117969909264437,
+      "loss": 1.2719,
+      "step": 2157
+    },
+    {
+      "epoch": 0.18940675990356318,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0029116461425709532,
+      "loss": 1.3267,
+      "step": 2158
+    },
+    {
+      "epoch": 0.18949452948646567,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002911495169695981,
+      "loss": 1.2936,
+      "step": 2159
+    },
+    {
+      "epoch": 0.18958229906936813,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0029113440723164265,
+      "loss": 1.3098,
+      "step": 2160
+    },
+    {
+      "epoch": 0.18967006865227062,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002911192850447203,
+      "loss": 1.2725,
+      "step": 2161
+    },
+    {
+      "epoch": 0.18975783823517312,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0029110415041032353,
+      "loss": 1.3344,
+      "step": 2162
+    },
+    {
+      "epoch": 0.1898456078180756,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002910890033299461,
+      "loss": 1.3558,
+      "step": 2163
+    },
+    {
+      "epoch": 0.18993337740097807,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00291073843805083,
+      "loss": 1.3024,
+      "step": 2164
+    },
+    {
+      "epoch": 0.19002114698388056,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0029105867183723028,
+      "loss": 1.2924,
+      "step": 2165
+    },
+    {
+      "epoch": 0.19010891656678305,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002910434874278855,
+      "loss": 1.344,
+      "step": 2166
+    },
+    {
+      "epoch": 0.19019668614968555,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0029102829057854724,
+      "loss": 1.3312,
+      "step": 2167
+    },
+    {
+      "epoch": 0.19028445573258804,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002910130812907153,
+      "loss": 1.2713,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1903722253154905,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002909978595658909,
+      "loss": 1.293,
+      "step": 2169
+    },
+    {
+      "epoch": 0.190459994898393,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002909826254055763,
+      "loss": 1.2687,
+      "step": 2170
+    },
+    {
+      "epoch": 0.19054776448129548,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029096737881127506,
+      "loss": 1.268,
+      "step": 2171
+    },
+    {
+      "epoch": 0.19063553406419798,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0029095211978449197,
+      "loss": 1.3009,
+      "step": 2172
+    },
+    {
+      "epoch": 0.19072330364710044,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0029093684832673305,
+      "loss": 1.2841,
+      "step": 2173
+    },
+    {
+      "epoch": 0.19081107323000293,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0029092156443950544,
+      "loss": 1.2893,
+      "step": 2174
+    },
+    {
+      "epoch": 0.19089884281290542,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002909062681243177,
+      "loss": 1.2805,
+      "step": 2175
+    },
+    {
+      "epoch": 0.19098661239580791,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002908909593826795,
+      "loss": 1.2638,
+      "step": 2176
+    },
+    {
+      "epoch": 0.19107438197871038,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002908756382161017,
+      "loss": 1.2976,
+      "step": 2177
+    },
+    {
+      "epoch": 0.19116215156161287,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0029086030462609647,
+      "loss": 1.3008,
+      "step": 2178
+    },
+    {
+      "epoch": 0.19124992114451536,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0029084495861417723,
+      "loss": 1.2945,
+      "step": 2179
+    },
+    {
+      "epoch": 0.19133769072741785,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0029082960018185847,
+      "loss": 1.3352,
+      "step": 2180
+    },
+    {
+      "epoch": 0.19142546031032034,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002908142293306561,
+      "loss": 1.3138,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1915132298932228,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0029079884606208707,
+      "loss": 1.3149,
+      "step": 2182
+    },
+    {
+      "epoch": 0.1916009994761253,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.002907834503776697,
+      "loss": 1.3493,
+      "step": 2183
+    },
+    {
+      "epoch": 0.1916887690590278,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002907680422789235,
+      "loss": 1.374,
+      "step": 2184
+    },
+    {
+      "epoch": 0.19177653864193028,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0029075262176736917,
+      "loss": 1.3089,
+      "step": 2185
+    },
+    {
+      "epoch": 0.19186430822483275,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0029073718884452863,
+      "loss": 1.3155,
+      "step": 2186
+    },
+    {
+      "epoch": 0.19195207780773524,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0029072174351192507,
+      "loss": 1.2978,
+      "step": 2187
+    },
+    {
+      "epoch": 0.19203984739063773,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002907062857710829,
+      "loss": 1.2919,
+      "step": 2188
+    },
+    {
+      "epoch": 0.19212761697354022,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002906908156235277,
+      "loss": 1.3122,
+      "step": 2189
+    },
+    {
+      "epoch": 0.19221538655644269,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.002906753330707863,
+      "loss": 1.3184,
+      "step": 2190
+    },
+    {
+      "epoch": 0.19230315613934518,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002906598381143869,
+      "loss": 1.3007,
+      "step": 2191
+    },
+    {
+      "epoch": 0.19239092572224767,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.002906443307558586,
+      "loss": 1.3458,
+      "step": 2192
+    },
+    {
+      "epoch": 0.19247869530515016,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0029062881099673203,
+      "loss": 1.314,
+      "step": 2193
+    },
+    {
+      "epoch": 0.19256646488805265,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0029061327883853887,
+      "loss": 1.3517,
+      "step": 2194
+    },
+    {
+      "epoch": 0.19265423447095512,
+      "grad_norm": 0.125,
+      "learning_rate": 0.002905977342828122,
+      "loss": 1.3017,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1927420040538576,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00290582177331086,
+      "loss": 1.2707,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1928297736367601,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002905666079848959,
+      "loss": 1.26,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1929175432196626,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.002905510262457784,
+      "loss": 1.2971,
+      "step": 2198
+    },
+    {
+      "epoch": 0.19300531280256505,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.002905354321152714,
+      "loss": 1.3218,
+      "step": 2199
+    },
+    {
+      "epoch": 0.19309308238546755,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0029051982559491394,
+      "loss": 1.2688,
+      "step": 2200
+    },
+    {
+      "epoch": 0.19318085196837004,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.002905042066862464,
+      "loss": 1.3785,
+      "step": 2201
+    },
+    {
+      "epoch": 0.19326862155127253,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002904885753908102,
+      "loss": 1.321,
+      "step": 2202
+    },
+    {
+      "epoch": 0.19335639113417502,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002904729317101482,
+      "loss": 1.323,
+      "step": 2203
+    },
+    {
+      "epoch": 0.19344416071707748,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0029045727564580427,
+      "loss": 1.2085,
+      "step": 2204
+    },
+    {
+      "epoch": 0.19353193029997998,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0029044160719932364,
+      "loss": 1.3641,
+      "step": 2205
+    },
+    {
+      "epoch": 0.19361969988288247,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002904259263722527,
+      "loss": 1.3244,
+      "step": 2206
+    },
+    {
+      "epoch": 0.19370746946578496,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002904102331661392,
+      "loss": 1.3476,
+      "step": 2207
+    },
+    {
+      "epoch": 0.19379523904868742,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0029039452758253185,
+      "loss": 1.343,
+      "step": 2208
+    },
+    {
+      "epoch": 0.19388300863158991,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002903788096229808,
+      "loss": 1.2731,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1939707782144924,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0029036307928903735,
+      "loss": 1.285,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1940585477973949,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0029034733658225404,
+      "loss": 1.2825,
+      "step": 2211
+    },
+    {
+      "epoch": 0.19414631738029736,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0029033158150418453,
+      "loss": 1.2815,
+      "step": 2212
+    },
+    {
+      "epoch": 0.19423408696319985,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002903158140563838,
+      "loss": 1.288,
+      "step": 2213
+    },
+    {
+      "epoch": 0.19432185654610235,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0029030003424040814,
+      "loss": 1.2565,
+      "step": 2214
+    },
+    {
+      "epoch": 0.19440962612900484,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002902842420578149,
+      "loss": 1.2907,
+      "step": 2215
+    },
+    {
+      "epoch": 0.19449739571190733,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0029026843751016264,
+      "loss": 1.3151,
+      "step": 2216
+    },
+    {
+      "epoch": 0.1945851652948098,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002902526205990113,
+      "loss": 1.2367,
+      "step": 2217
+    },
+    {
+      "epoch": 0.19467293487771228,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002902367913259219,
+      "loss": 1.4069,
+      "step": 2218
+    },
+    {
+      "epoch": 0.19476070446061478,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029022094969245672,
+      "loss": 1.2718,
+      "step": 2219
+    },
+    {
+      "epoch": 0.19484847404351727,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002902050957001793,
+      "loss": 1.3363,
+      "step": 2220
+    },
+    {
+      "epoch": 0.19493624362641973,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002901892293506543,
+      "loss": 1.3827,
+      "step": 2221
+    },
+    {
+      "epoch": 0.19502401320932222,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0029017335064544774,
+      "loss": 1.2886,
+      "step": 2222
+    },
+    {
+      "epoch": 0.19511178279222471,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002901574595861268,
+      "loss": 1.271,
+      "step": 2223
+    },
+    {
+      "epoch": 0.1951995523751272,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0029014155617425974,
+      "loss": 1.2646,
+      "step": 2224
+    },
+    {
+      "epoch": 0.19528732195802967,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0029012564041141628,
+      "loss": 1.34,
+      "step": 2225
+    },
+    {
+      "epoch": 0.19537509154093216,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0029010971229916725,
+      "loss": 1.2941,
+      "step": 2226
+    },
+    {
+      "epoch": 0.19546286112383465,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0029009377183908465,
+      "loss": 1.3545,
+      "step": 2227
+    },
+    {
+      "epoch": 0.19555063070673714,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0029007781903274167,
+      "loss": 1.2947,
+      "step": 2228
+    },
+    {
+      "epoch": 0.19563840028963964,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002900618538817129,
+      "loss": 1.3155,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1957261698725421,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00290045876387574,
+      "loss": 1.3905,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1958139394554446,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0029002988655190184,
+      "loss": 1.339,
+      "step": 2231
+    },
+    {
+      "epoch": 0.19590170903834708,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0029001388437627468,
+      "loss": 1.2732,
+      "step": 2232
+    },
+    {
+      "epoch": 0.19598947862124957,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0028999786986227175,
+      "loss": 1.3036,
+      "step": 2233
+    },
+    {
+      "epoch": 0.19607724820415204,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0028998184301147365,
+      "loss": 1.334,
+      "step": 2234
+    },
+    {
+      "epoch": 0.19616501778705453,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0028996580382546215,
+      "loss": 1.3219,
+      "step": 2235
+    },
+    {
+      "epoch": 0.19625278736995702,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0028994975230582027,
+      "loss": 1.2653,
+      "step": 2236
+    },
+    {
+      "epoch": 0.1963405569528595,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002899336884541323,
+      "loss": 1.3116,
+      "step": 2237
+    },
+    {
+      "epoch": 0.196428326535762,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002899176122719836,
+      "loss": 1.2569,
+      "step": 2238
+    },
+    {
+      "epoch": 0.19651609611866447,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002899015237609608,
+      "loss": 1.3395,
+      "step": 2239
+    },
+    {
+      "epoch": 0.19660386570156696,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002898854229226518,
+      "loss": 1.264,
+      "step": 2240
+    },
+    {
+      "epoch": 0.19669163528446945,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.002898693097586458,
+      "loss": 1.2882,
+      "step": 2241
+    },
+    {
+      "epoch": 0.19677940486737194,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0028985318427053293,
+      "loss": 1.3069,
+      "step": 2242
+    },
+    {
+      "epoch": 0.1968671744502744,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0028983704645990484,
+      "loss": 1.2961,
+      "step": 2243
+    },
+    {
+      "epoch": 0.1969549440331769,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0028982089632835417,
+      "loss": 1.3366,
+      "step": 2244
+    },
+    {
+      "epoch": 0.1970427136160794,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00289804733877475,
+      "loss": 1.328,
+      "step": 2245
+    },
+    {
+      "epoch": 0.19713048319898188,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002897885591088623,
+      "loss": 1.2685,
+      "step": 2246
+    },
+    {
+      "epoch": 0.19721825278188435,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0028977237202411268,
+      "loss": 1.3347,
+      "step": 2247
+    },
+    {
+      "epoch": 0.19730602236478684,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.002897561726248236,
+      "loss": 1.2983,
+      "step": 2248
+    },
+    {
+      "epoch": 0.19739379194768933,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0028973996091259395,
+      "loss": 1.2672,
+      "step": 2249
+    },
+    {
+      "epoch": 0.19748156153059182,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0028972373688902374,
+      "loss": 1.2969,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1975693311134943,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002897075005557142,
+      "loss": 1.2327,
+      "step": 2251
+    },
+    {
+      "epoch": 0.19765710069639678,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0028969125191426774,
+      "loss": 1.3086,
+      "step": 2252
+    },
+    {
+      "epoch": 0.19774487027929927,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.002896749909662881,
+      "loss": 1.295,
+      "step": 2253
+    },
+    {
+      "epoch": 0.19783263986220176,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0028965871771338024,
+      "loss": 1.3236,
+      "step": 2254
+    },
+    {
+      "epoch": 0.19792040944510425,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.002896424321571501,
+      "loss": 1.2835,
+      "step": 2255
+    },
+    {
+      "epoch": 0.19800817902800671,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0028962613429920517,
+      "loss": 1.3419,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1980959486109092,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0028960982414115385,
+      "loss": 1.329,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1981837181938117,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00289593501684606,
+      "loss": 1.3221,
+      "step": 2258
+    },
+    {
+      "epoch": 0.1982714877767142,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0028957716693117246,
+      "loss": 1.2858,
+      "step": 2259
+    },
+    {
+      "epoch": 0.19835925735961665,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002895608198824655,
+      "loss": 1.3049,
+      "step": 2260
+    },
+    {
+      "epoch": 0.19844702694251914,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002895444605400985,
+      "loss": 1.272,
+      "step": 2261
+    },
+    {
+      "epoch": 0.19853479652542164,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00289528088905686,
+      "loss": 1.3099,
+      "step": 2262
+    },
+    {
+      "epoch": 0.19862256610832413,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0028951170498084386,
+      "loss": 1.2792,
+      "step": 2263
+    },
+    {
+      "epoch": 0.19871033569122662,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002894953087671892,
+      "loss": 1.3152,
+      "step": 2264
+    },
+    {
+      "epoch": 0.19879810527412908,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002894789002663401,
+      "loss": 1.3203,
+      "step": 2265
+    },
+    {
+      "epoch": 0.19888587485703157,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0028946247947991607,
+      "loss": 1.301,
+      "step": 2266
+    },
+    {
+      "epoch": 0.19897364443993407,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0028944604640953778,
+      "loss": 1.3096,
+      "step": 2267
+    },
+    {
+      "epoch": 0.19906141402283656,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0028942960105682713,
+      "loss": 1.33,
+      "step": 2268
+    },
+    {
+      "epoch": 0.19914918360573902,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002894131434234072,
+      "loss": 1.3559,
+      "step": 2269
+    },
+    {
+      "epoch": 0.1992369531886415,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002893966735109023,
+      "loss": 1.276,
+      "step": 2270
+    },
+    {
+      "epoch": 0.199324722771544,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0028938019132093793,
+      "loss": 1.2809,
+      "step": 2271
+    },
+    {
+      "epoch": 0.1994124923544465,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.002893636968551408,
+      "loss": 1.2869,
+      "step": 2272
+    },
+    {
+      "epoch": 0.199500261937349,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.002893471901151389,
+      "loss": 1.217,
+      "step": 2273
+    },
+    {
+      "epoch": 0.19958803152025145,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.002893306711025614,
+      "loss": 1.3172,
+      "step": 2274
+    },
+    {
+      "epoch": 0.19967580110315394,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0028931413981903857,
+      "loss": 1.3734,
+      "step": 2275
+    },
+    {
+      "epoch": 0.19976357068605644,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0028929759626620203,
+      "loss": 1.2834,
+      "step": 2276
+    },
+    {
+      "epoch": 0.19985134026895893,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0028928104044568448,
+      "loss": 1.3359,
+      "step": 2277
+    },
+    {
+      "epoch": 0.1999391098518614,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0028926447235912015,
+      "loss": 1.2903,
+      "step": 2278
+    },
+    {
+      "epoch": 0.20002687943476388,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0028924789200814394,
+      "loss": 1.4082,
+      "step": 2279
+    },
+    {
+      "epoch": 0.20011464901766637,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.002892312993943925,
+      "loss": 1.2767,
+      "step": 2280
+    },
+    {
+      "epoch": 0.20020241860056887,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0028921469451950337,
+      "loss": 1.2858,
+      "step": 2281
+    },
+    {
+      "epoch": 0.20029018818347133,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0028919807738511535,
+      "loss": 1.3065,
+      "step": 2282
+    },
+    {
+      "epoch": 0.20037795776637382,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0028918144799286853,
+      "loss": 1.3137,
+      "step": 2283
+    },
+    {
+      "epoch": 0.2004657273492763,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0028916480634440414,
+      "loss": 1.2735,
+      "step": 2284
+    },
+    {
+      "epoch": 0.2005534969321788,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0028914815244136466,
+      "loss": 1.2725,
+      "step": 2285
+    },
+    {
+      "epoch": 0.2006412665150813,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002891314862853938,
+      "loss": 1.2455,
+      "step": 2286
+    },
+    {
+      "epoch": 0.20072903609798376,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0028911480787813634,
+      "loss": 1.2959,
+      "step": 2287
+    },
+    {
+      "epoch": 0.20081680568088625,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002890981172212385,
+      "loss": 1.2491,
+      "step": 2288
+    },
+    {
+      "epoch": 0.20090457526378874,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0028908141431634752,
+      "loss": 1.3692,
+      "step": 2289
+    },
+    {
+      "epoch": 0.20099234484669123,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0028906469916511186,
+      "loss": 1.3203,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2010801144295937,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.002890479717691814,
+      "loss": 1.2715,
+      "step": 2291
+    },
+    {
+      "epoch": 0.2011678840124962,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002890312321302068,
+      "loss": 1.2872,
+      "step": 2292
+    },
+    {
+      "epoch": 0.20125565359539868,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002890144802498405,
+      "loss": 1.3064,
+      "step": 2293
+    },
+    {
+      "epoch": 0.20134342317830117,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0028899771612973563,
+      "loss": 1.2813,
+      "step": 2294
+    },
+    {
+      "epoch": 0.20143119276120364,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002889809397715468,
+      "loss": 1.309,
+      "step": 2295
+    },
+    {
+      "epoch": 0.20151896234410613,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0028896415117692985,
+      "loss": 1.3348,
+      "step": 2296
+    },
+    {
+      "epoch": 0.20160673192700862,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0028894735034754163,
+      "loss": 1.2646,
+      "step": 2297
+    },
+    {
+      "epoch": 0.2016945015099111,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0028893053728504038,
+      "loss": 1.2591,
+      "step": 2298
+    },
+    {
+      "epoch": 0.2017822710928136,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002889137119910854,
+      "loss": 1.3292,
+      "step": 2299
+    },
+    {
+      "epoch": 0.20187004067571607,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0028889687446733746,
+      "loss": 1.2715,
+      "step": 2300
+    },
+    {
+      "epoch": 0.20195781025861856,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002888800247154582,
+      "loss": 1.3209,
+      "step": 2301
+    },
+    {
+      "epoch": 0.20204557984152105,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0028886316273711065,
+      "loss": 1.3558,
+      "step": 2302
+    },
+    {
+      "epoch": 0.20213334942442354,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0028884628853395908,
+      "loss": 1.2889,
+      "step": 2303
+    },
+    {
+      "epoch": 0.202221119007326,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0028882940210766886,
+      "loss": 1.319,
+      "step": 2304
+    },
+    {
+      "epoch": 0.2023088885902285,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0028881250345990654,
+      "loss": 1.3286,
+      "step": 2305
+    },
+    {
+      "epoch": 0.202396658173131,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0028879559259234015,
+      "loss": 1.2802,
+      "step": 2306
+    },
+    {
+      "epoch": 0.20248442775603348,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0028877866950663853,
+      "loss": 1.3011,
+      "step": 2307
+    },
+    {
+      "epoch": 0.20257219733893597,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00288761734204472,
+      "loss": 1.3207,
+      "step": 2308
+    },
+    {
+      "epoch": 0.20265996692183844,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0028874478668751203,
+      "loss": 1.2711,
+      "step": 2309
+    },
+    {
+      "epoch": 0.20274773650474093,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002887278269574312,
+      "loss": 1.2906,
+      "step": 2310
+    },
+    {
+      "epoch": 0.20283550608764342,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002887108550159035,
+      "loss": 1.2525,
+      "step": 2311
+    },
+    {
+      "epoch": 0.2029232756705459,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.002886938708646039,
+      "loss": 1.3109,
+      "step": 2312
+    },
+    {
+      "epoch": 0.20301104525344837,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002886768745052086,
+      "loss": 1.3077,
+      "step": 2313
+    },
+    {
+      "epoch": 0.20309881483635087,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002886598659393952,
+      "loss": 1.3287,
+      "step": 2314
+    },
+    {
+      "epoch": 0.20318658441925336,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0028864284516884227,
+      "loss": 1.3065,
+      "step": 2315
+    },
+    {
+      "epoch": 0.20327435400215585,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002886258121952298,
+      "loss": 1.2697,
+      "step": 2316
+    },
+    {
+      "epoch": 0.2033621235850583,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0028860876702023882,
+      "loss": 1.2881,
+      "step": 2317
+    },
+    {
+      "epoch": 0.2034498931679608,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0028859170964555167,
+      "loss": 1.3198,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2035376627508633,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002885746400728518,
+      "loss": 1.3082,
+      "step": 2319
+    },
+    {
+      "epoch": 0.2036254323337658,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002885575583038239,
+      "loss": 1.3452,
+      "step": 2320
+    },
+    {
+      "epoch": 0.20371320191666828,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0028854046434015385,
+      "loss": 1.3227,
+      "step": 2321
+    },
+    {
+      "epoch": 0.20380097149957074,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0028852335818352884,
+      "loss": 1.3368,
+      "step": 2322
+    },
+    {
+      "epoch": 0.20388874108247323,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002885062398356371,
+      "loss": 1.3182,
+      "step": 2323
+    },
+    {
+      "epoch": 0.20397651066537573,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0028848910929816824,
+      "loss": 1.2907,
+      "step": 2324
+    },
+    {
+      "epoch": 0.20406428024827822,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0028847196657281284,
+      "loss": 1.3292,
+      "step": 2325
+    },
+    {
+      "epoch": 0.20415204983118068,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0028845481166126295,
+      "loss": 1.3166,
+      "step": 2326
+    },
+    {
+      "epoch": 0.20423981941408317,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0028843764456521156,
+      "loss": 1.2836,
+      "step": 2327
+    },
+    {
+      "epoch": 0.20432758899698567,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0028842046528635315,
+      "loss": 1.3468,
+      "step": 2328
+    },
+    {
+      "epoch": 0.20441535857988816,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0028840327382638315,
+      "loss": 1.3004,
+      "step": 2329
+    },
+    {
+      "epoch": 0.20450312816279062,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0028838607018699825,
+      "loss": 1.2457,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2045908977456931,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002883688543698965,
+      "loss": 1.2878,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2046786673285956,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0028835162637677694,
+      "loss": 1.2904,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2047664369114981,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0028833438620933997,
+      "loss": 1.2577,
+      "step": 2333
+    },
+    {
+      "epoch": 0.2048542064944006,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00288317133869287,
+      "loss": 1.3141,
+      "step": 2334
+    },
+    {
+      "epoch": 0.20494197607730305,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0028829986935832094,
+      "loss": 1.2969,
+      "step": 2335
+    },
+    {
+      "epoch": 0.20502974566020554,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0028828259267814566,
+      "loss": 1.329,
+      "step": 2336
+    },
+    {
+      "epoch": 0.20511751524310803,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0028826530383046626,
+      "loss": 1.3423,
+      "step": 2337
+    },
+    {
+      "epoch": 0.20520528482601053,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0028824800281698907,
+      "loss": 1.2635,
+      "step": 2338
+    },
+    {
+      "epoch": 0.205293054408913,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0028823068963942174,
+      "loss": 1.2605,
+      "step": 2339
+    },
+    {
+      "epoch": 0.20538082399181548,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002882133642994729,
+      "loss": 1.2924,
+      "step": 2340
+    },
+    {
+      "epoch": 0.20546859357471797,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0028819602679885253,
+      "loss": 1.3222,
+      "step": 2341
+    },
+    {
+      "epoch": 0.20555636315762046,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002881786771392718,
+      "loss": 1.2976,
+      "step": 2342
+    },
+    {
+      "epoch": 0.20564413274052296,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00288161315322443,
+      "loss": 1.2954,
+      "step": 2343
+    },
+    {
+      "epoch": 0.20573190232342542,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0028814394135007974,
+      "loss": 1.3379,
+      "step": 2344
+    },
+    {
+      "epoch": 0.2058196719063279,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002881265552238967,
+      "loss": 1.2876,
+      "step": 2345
+    },
+    {
+      "epoch": 0.2059074414892304,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0028810915694560987,
+      "loss": 1.2907,
+      "step": 2346
+    },
+    {
+      "epoch": 0.2059952110721329,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0028809174651693633,
+      "loss": 1.3451,
+      "step": 2347
+    },
+    {
+      "epoch": 0.20608298065503536,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0028807432393959448,
+      "loss": 1.3329,
+      "step": 2348
+    },
+    {
+      "epoch": 0.20617075023793785,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.002880568892153038,
+      "loss": 1.3236,
+      "step": 2349
+    },
+    {
+      "epoch": 0.20625851982084034,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0028803944234578513,
+      "loss": 1.3613,
+      "step": 2350
+    },
+    {
+      "epoch": 0.20634628940374283,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.002880219833327603,
+      "loss": 1.2599,
+      "step": 2351
+    },
+    {
+      "epoch": 0.2064340589866453,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002880045121779525,
+      "loss": 1.3439,
+      "step": 2352
+    },
+    {
+      "epoch": 0.2065218285695478,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0028798702888308604,
+      "loss": 1.3908,
+      "step": 2353
+    },
+    {
+      "epoch": 0.20660959815245028,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002879695334498865,
+      "loss": 1.3702,
+      "step": 2354
+    },
+    {
+      "epoch": 0.20669736773535277,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0028795202588008054,
+      "loss": 1.2913,
+      "step": 2355
+    },
+    {
+      "epoch": 0.20678513731825526,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002879345061753961,
+      "loss": 1.2786,
+      "step": 2356
+    },
+    {
+      "epoch": 0.20687290690115773,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0028791697433756245,
+      "loss": 1.3243,
+      "step": 2357
+    },
+    {
+      "epoch": 0.20696067648406022,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0028789943036830966,
+      "loss": 1.3272,
+      "step": 2358
+    },
+    {
+      "epoch": 0.2070484460669627,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0028788187426936943,
+      "loss": 1.3094,
+      "step": 2359
+    },
+    {
+      "epoch": 0.2071362156498652,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0028786430604247444,
+      "loss": 1.2953,
+      "step": 2360
+    },
+    {
+      "epoch": 0.20722398523276767,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0028784672568935854,
+      "loss": 1.3535,
+      "step": 2361
+    },
+    {
+      "epoch": 0.20731175481567016,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00287829133211757,
+      "loss": 1.3137,
+      "step": 2362
+    },
+    {
+      "epoch": 0.20739952439857265,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002878115286114059,
+      "loss": 1.2906,
+      "step": 2363
+    },
+    {
+      "epoch": 0.20748729398147514,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0028779391189004294,
+      "loss": 1.3038,
+      "step": 2364
+    },
+    {
+      "epoch": 0.2075750635643776,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002877762830494067,
+      "loss": 1.296,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2076628331472801,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0028775864209123715,
+      "loss": 1.2616,
+      "step": 2366
+    },
+    {
+      "epoch": 0.2077506027301826,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0028774098901727537,
+      "loss": 1.2957,
+      "step": 2367
+    },
+    {
+      "epoch": 0.20783837231308508,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0028772332382926364,
+      "loss": 1.3068,
+      "step": 2368
+    },
+    {
+      "epoch": 0.20792614189598757,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0028770564652894543,
+      "loss": 1.2917,
+      "step": 2369
+    },
+    {
+      "epoch": 0.20801391147889003,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0028768795711806546,
+      "loss": 1.2785,
+      "step": 2370
+    },
+    {
+      "epoch": 0.20810168106179253,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.002876702555983695,
+      "loss": 1.2979,
+      "step": 2371
+    },
+    {
+      "epoch": 0.20818945064469502,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.002876525419716047,
+      "loss": 1.2484,
+      "step": 2372
+    },
+    {
+      "epoch": 0.2082772202275975,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0028763481623951936,
+      "loss": 1.3524,
+      "step": 2373
+    },
+    {
+      "epoch": 0.20836498981049997,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0028761707840386287,
+      "loss": 1.2403,
+      "step": 2374
+    },
+    {
+      "epoch": 0.20845275939340246,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00287599328466386,
+      "loss": 1.306,
+      "step": 2375
+    },
+    {
+      "epoch": 0.20854052897630496,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.002875815664288404,
+      "loss": 1.2855,
+      "step": 2376
+    },
+    {
+      "epoch": 0.20862829855920745,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002875637922929793,
+      "loss": 1.2874,
+      "step": 2377
+    },
+    {
+      "epoch": 0.20871606814210994,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002875460060605568,
+      "loss": 1.3003,
+      "step": 2378
+    },
+    {
+      "epoch": 0.2088038377250124,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0028752820773332846,
+      "loss": 1.2508,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2088916073079149,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0028751039731305075,
+      "loss": 1.2825,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2089793768908174,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0028749257480148163,
+      "loss": 1.3611,
+      "step": 2381
+    },
+    {
+      "epoch": 0.20906714647371988,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.002874747402003801,
+      "loss": 1.2617,
+      "step": 2382
+    },
+    {
+      "epoch": 0.20915491605662234,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0028745689351150627,
+      "loss": 1.3146,
+      "step": 2383
+    },
+    {
+      "epoch": 0.20924268563952483,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0028743903473662158,
+      "loss": 1.3093,
+      "step": 2384
+    },
+    {
+      "epoch": 0.20933045522242733,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0028742116387748865,
+      "loss": 1.348,
+      "step": 2385
+    },
+    {
+      "epoch": 0.20941822480532982,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0028740328093587123,
+      "loss": 1.2962,
+      "step": 2386
+    },
+    {
+      "epoch": 0.20950599438823228,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0028738538591353427,
+      "loss": 1.3539,
+      "step": 2387
+    },
+    {
+      "epoch": 0.20959376397113477,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0028736747881224402,
+      "loss": 1.3226,
+      "step": 2388
+    },
+    {
+      "epoch": 0.20968153355403726,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0028734955963376782,
+      "loss": 1.2698,
+      "step": 2389
+    },
+    {
+      "epoch": 0.20976930313693976,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.002873316283798741,
+      "loss": 1.2835,
+      "step": 2390
+    },
+    {
+      "epoch": 0.20985707271984225,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002873136850523328,
+      "loss": 1.2798,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2099448423027447,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0028729572965291474,
+      "loss": 1.3507,
+      "step": 2392
+    },
+    {
+      "epoch": 0.2100326118856472,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0028727776218339203,
+      "loss": 1.2885,
+      "step": 2393
+    },
+    {
+      "epoch": 0.2101203814685497,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.00287259782645538,
+      "loss": 1.314,
+      "step": 2394
+    },
+    {
+      "epoch": 0.21020815105145219,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0028724179104112724,
+      "loss": 1.3207,
+      "step": 2395
+    },
+    {
+      "epoch": 0.21029592063435465,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0028722378737193535,
+      "loss": 1.2802,
+      "step": 2396
+    },
+    {
+      "epoch": 0.21038369021725714,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0028720577163973927,
+      "loss": 1.2236,
+      "step": 2397
+    },
+    {
+      "epoch": 0.21047145980015963,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002871877438463171,
+      "loss": 1.2892,
+      "step": 2398
+    },
+    {
+      "epoch": 0.21055922938306212,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0028716970399344805,
+      "loss": 1.3224,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2106469989659646,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0028715165208291265,
+      "loss": 1.2556,
+      "step": 2400
+    },
+    {
+      "epoch": 0.21073476854886708,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.002871335881164925,
+      "loss": 1.3055,
+      "step": 2401
+    },
+    {
+      "epoch": 0.21082253813176957,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0028711551209597043,
+      "loss": 1.299,
+      "step": 2402
+    },
+    {
+      "epoch": 0.21091030771467206,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0028709742402313054,
+      "loss": 1.3579,
+      "step": 2403
+    },
+    {
+      "epoch": 0.21099807729757455,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0028707932389975804,
+      "loss": 1.3595,
+      "step": 2404
+    },
+    {
+      "epoch": 0.21108584688047702,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002870612117276393,
+      "loss": 1.2872,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2111736164633795,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0028704308750856188,
+      "loss": 1.2803,
+      "step": 2406
+    },
+    {
+      "epoch": 0.211261386046282,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002870249512443147,
+      "loss": 1.3525,
+      "step": 2407
+    },
+    {
+      "epoch": 0.2113491556291845,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002870068029366876,
+      "loss": 1.2667,
+      "step": 2408
+    },
+    {
+      "epoch": 0.21143692521208696,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002869886425874718,
+      "loss": 1.3334,
+      "step": 2409
+    },
+    {
+      "epoch": 0.21152469479498945,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.002869704701984597,
+      "loss": 1.2823,
+      "step": 2410
+    },
+    {
+      "epoch": 0.21161246437789194,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0028695228577144483,
+      "loss": 1.3074,
+      "step": 2411
+    },
+    {
+      "epoch": 0.21170023396079443,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0028693408930822186,
+      "loss": 1.2667,
+      "step": 2412
+    },
+    {
+      "epoch": 0.21178800354369692,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002869158808105867,
+      "loss": 1.3875,
+      "step": 2413
+    },
+    {
+      "epoch": 0.2118757731265994,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0028689766028033655,
+      "loss": 1.302,
+      "step": 2414
+    },
+    {
+      "epoch": 0.21196354270950188,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0028687942771926968,
+      "loss": 1.334,
+      "step": 2415
+    },
+    {
+      "epoch": 0.21205131229240437,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0028686118312918553,
+      "loss": 1.2637,
+      "step": 2416
+    },
+    {
+      "epoch": 0.21213908187530686,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028684292651188475,
+      "loss": 1.3055,
+      "step": 2417
+    },
+    {
+      "epoch": 0.21222685145820933,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0028682465786916927,
+      "loss": 1.2792,
+      "step": 2418
+    },
+    {
+      "epoch": 0.21231462104111182,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0028680637720284205,
+      "loss": 1.3173,
+      "step": 2419
+    },
+    {
+      "epoch": 0.2124023906240143,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002867880845147074,
+      "loss": 1.2788,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2124901602069168,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0028676977980657066,
+      "loss": 1.2482,
+      "step": 2421
+    },
+    {
+      "epoch": 0.21257792978981926,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0028675146308023854,
+      "loss": 1.2852,
+      "step": 2422
+    },
+    {
+      "epoch": 0.21266569937272176,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.002867331343375187,
+      "loss": 1.21,
+      "step": 2423
+    },
+    {
+      "epoch": 0.21275346895562425,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002867147935802202,
+      "loss": 1.2382,
+      "step": 2424
+    },
+    {
+      "epoch": 0.21284123853852674,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.002866964408101532,
+      "loss": 1.3486,
+      "step": 2425
+    },
+    {
+      "epoch": 0.21292900812142923,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0028667807602912898,
+      "loss": 1.272,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2130167777043317,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.002866596992389601,
+      "loss": 1.2788,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2131045472872342,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002866413104414603,
+      "loss": 1.2698,
+      "step": 2428
+    },
+    {
+      "epoch": 0.21319231687013668,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.002866229096384445,
+      "loss": 1.329,
+      "step": 2429
+    },
+    {
+      "epoch": 0.21328008645303917,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0028660449683172873,
+      "loss": 1.2794,
+      "step": 2430
+    },
+    {
+      "epoch": 0.21336785603594163,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002865860720231303,
+      "loss": 1.2735,
+      "step": 2431
+    },
+    {
+      "epoch": 0.21345562561884412,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002865676352144677,
+      "loss": 1.3595,
+      "step": 2432
+    },
+    {
+      "epoch": 0.21354339520174662,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0028654918640756044,
+      "loss": 1.2883,
+      "step": 2433
+    },
+    {
+      "epoch": 0.2136311647846491,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0028653072560422947,
+      "loss": 1.2778,
+      "step": 2434
+    },
+    {
+      "epoch": 0.21371893436755157,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0028651225280629674,
+      "loss": 1.35,
+      "step": 2435
+    },
+    {
+      "epoch": 0.21380670395045406,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0028649376801558545,
+      "loss": 1.2904,
+      "step": 2436
+    },
+    {
+      "epoch": 0.21389447353335655,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0028647527123391997,
+      "loss": 1.2926,
+      "step": 2437
+    },
+    {
+      "epoch": 0.21398224311625905,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002864567624631259,
+      "loss": 1.322,
+      "step": 2438
+    },
+    {
+      "epoch": 0.21407001269916154,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0028643824170502996,
+      "loss": 1.2998,
+      "step": 2439
+    },
+    {
+      "epoch": 0.214157782282064,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002864197089614601,
+      "loss": 1.2738,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2142455518649665,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0028640116423424536,
+      "loss": 1.3009,
+      "step": 2441
+    },
+    {
+      "epoch": 0.21433332144786899,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0028638260752521607,
+      "loss": 1.2854,
+      "step": 2442
+    },
+    {
+      "epoch": 0.21442109103077148,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0028636403883620367,
+      "loss": 1.2842,
+      "step": 2443
+    },
+    {
+      "epoch": 0.21450886061367394,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.002863454581690409,
+      "loss": 1.3257,
+      "step": 2444
+    },
+    {
+      "epoch": 0.21459663019657643,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0028632686552556162,
+      "loss": 1.2528,
+      "step": 2445
+    },
+    {
+      "epoch": 0.21468439977947892,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0028630826090760065,
+      "loss": 1.2651,
+      "step": 2446
+    },
+    {
+      "epoch": 0.21477216936238142,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0028628964431699436,
+      "loss": 1.3215,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2148599389452839,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002862710157555801,
+      "loss": 1.2939,
+      "step": 2448
+    },
+    {
+      "epoch": 0.21494770852818637,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028625237522519643,
+      "loss": 1.2623,
+      "step": 2449
+    },
+    {
+      "epoch": 0.21503547811108886,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002862337227276831,
+      "loss": 1.314,
+      "step": 2450
+    },
+    {
+      "epoch": 0.21512324769399135,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0028621505826488106,
+      "loss": 1.2892,
+      "step": 2451
+    },
+    {
+      "epoch": 0.21521101727689385,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028619638183863235,
+      "loss": 1.2987,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2152987868597963,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002861776934507803,
+      "loss": 1.2992,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2153865564426988,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0028615899310316943,
+      "loss": 1.31,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2154743260256013,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0028614028079764534,
+      "loss": 1.2462,
+      "step": 2455
+    },
+    {
+      "epoch": 0.21556209560850378,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002861215565360548,
+      "loss": 1.31,
+      "step": 2456
+    },
+    {
+      "epoch": 0.21564986519140625,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0028610282032024595,
+      "loss": 1.3042,
+      "step": 2457
+    },
+    {
+      "epoch": 0.21573763477430874,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0028608407215206786,
+      "loss": 1.2576,
+      "step": 2458
+    },
+    {
+      "epoch": 0.21582540435721123,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00286065312033371,
+      "loss": 1.3492,
+      "step": 2459
+    },
+    {
+      "epoch": 0.21591317394011372,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0028604653996600685,
+      "loss": 1.3167,
+      "step": 2460
+    },
+    {
+      "epoch": 0.21600094352301621,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0028602775595182816,
+      "loss": 1.2599,
+      "step": 2461
+    },
+    {
+      "epoch": 0.21608871310591868,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0028600895999268882,
+      "loss": 1.305,
+      "step": 2462
+    },
+    {
+      "epoch": 0.21617648268882117,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0028599015209044404,
+      "loss": 1.2478,
+      "step": 2463
+    },
+    {
+      "epoch": 0.21626425227172366,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0028597133224694984,
+      "loss": 1.3179,
+      "step": 2464
+    },
+    {
+      "epoch": 0.21635202185462615,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.002859525004640639,
+      "loss": 1.3279,
+      "step": 2465
+    },
+    {
+      "epoch": 0.21643979143752862,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0028593365674364473,
+      "loss": 1.291,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2165275610204311,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002859148010875521,
+      "loss": 1.2743,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2166153306033336,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0028589593349764714,
+      "loss": 1.2508,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2167031001862361,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002858770539757918,
+      "loss": 1.3308,
+      "step": 2469
+    },
+    {
+      "epoch": 0.21679086976913856,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0028585816252384958,
+      "loss": 1.2618,
+      "step": 2470
+    },
+    {
+      "epoch": 0.21687863935204105,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.002858392591436849,
+      "loss": 1.3408,
+      "step": 2471
+    },
+    {
+      "epoch": 0.21696640893494354,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002858203438371635,
+      "loss": 1.2779,
+      "step": 2472
+    },
+    {
+      "epoch": 0.21705417851784603,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0028580141660615225,
+      "loss": 1.2347,
+      "step": 2473
+    },
+    {
+      "epoch": 0.21714194810074852,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.002857824774525191,
+      "loss": 1.2368,
+      "step": 2474
+    },
+    {
+      "epoch": 0.21722971768365099,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002857635263781334,
+      "loss": 1.2678,
+      "step": 2475
+    },
+    {
+      "epoch": 0.21731748726655348,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0028574456338486547,
+      "loss": 1.2988,
+      "step": 2476
+    },
+    {
+      "epoch": 0.21740525684945597,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0028572558847458685,
+      "loss": 1.2508,
+      "step": 2477
+    },
+    {
+      "epoch": 0.21749302643235846,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002857066016491704,
+      "loss": 1.3486,
+      "step": 2478
+    },
+    {
+      "epoch": 0.21758079601526092,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0028568760291048994,
+      "loss": 1.252,
+      "step": 2479
+    },
+    {
+      "epoch": 0.21766856559816342,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002856685922604206,
+      "loss": 1.2619,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2177563351810659,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002856495697008387,
+      "loss": 1.3153,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2178441047639684,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0028563053523362166,
+      "loss": 1.3575,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2179318743468709,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002856114888606481,
+      "loss": 1.285,
+      "step": 2483
+    },
+    {
+      "epoch": 0.21801964392977335,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002855924305837978,
+      "loss": 1.269,
+      "step": 2484
+    },
+    {
+      "epoch": 0.21810741351267585,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0028557336040495175,
+      "loss": 1.3152,
+      "step": 2485
+    },
+    {
+      "epoch": 0.21819518309557834,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028555427832599213,
+      "loss": 1.2733,
+      "step": 2486
+    },
+    {
+      "epoch": 0.21828295267848083,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0028553518434880224,
+      "loss": 1.3076,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2183707222613833,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0028551607847526663,
+      "loss": 1.2777,
+      "step": 2488
+    },
+    {
+      "epoch": 0.21845849184428578,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0028549696070727087,
+      "loss": 1.2325,
+      "step": 2489
+    },
+    {
+      "epoch": 0.21854626142718828,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002854778310467019,
+      "loss": 1.2831,
+      "step": 2490
+    },
+    {
+      "epoch": 0.21863403101009077,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002854586894954478,
+      "loss": 1.2836,
+      "step": 2491
+    },
+    {
+      "epoch": 0.21872180059299323,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0028543953605539756,
+      "loss": 1.3051,
+      "step": 2492
+    },
+    {
+      "epoch": 0.21880957017589572,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002854203707284417,
+      "loss": 1.2443,
+      "step": 2493
+    },
+    {
+      "epoch": 0.21889733975879822,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002854011935164717,
+      "loss": 1.3067,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2189851093417007,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002853820044213803,
+      "loss": 1.3076,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2190728789246032,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0028536280344506147,
+      "loss": 1.3224,
+      "step": 2496
+    },
+    {
+      "epoch": 0.21916064850750566,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002853435905894101,
+      "loss": 1.278,
+      "step": 2497
+    },
+    {
+      "epoch": 0.21924841809040815,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.002853243658563226,
+      "loss": 1.3356,
+      "step": 2498
+    },
+    {
+      "epoch": 0.21933618767331065,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.002853051292476963,
+      "loss": 1.3006,
+      "step": 2499
+    },
+    {
+      "epoch": 0.21942395725621314,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0028528588076542965,
+      "loss": 1.2974,
+      "step": 2500
+    },
+    {
+      "epoch": 0.21942395725621314,
+      "eval_loss": 1.2801631689071655,
+      "eval_runtime": 437.6333,
+      "eval_samples_per_second": 33.681,
+      "eval_steps_per_second": 8.42,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2195117268391156,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002852666204114226,
+      "loss": 1.305,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2195994964220181,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0028524734818757597,
+      "loss": 1.3353,
+      "step": 2502
+    },
+    {
+      "epoch": 0.21968726600492058,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0028522806409579187,
+      "loss": 1.3063,
+      "step": 2503
+    },
+    {
+      "epoch": 0.21977503558782308,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0028520876813797353,
+      "loss": 1.3428,
+      "step": 2504
+    },
+    {
+      "epoch": 0.21986280517072554,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0028518946031602546,
+      "loss": 1.3381,
+      "step": 2505
+    },
+    {
+      "epoch": 0.21995057475362803,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.002851701406318532,
+      "loss": 1.2715,
+      "step": 2506
+    },
+    {
+      "epoch": 0.22003834433653052,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.002851508090873636,
+      "loss": 1.251,
+      "step": 2507
+    },
+    {
+      "epoch": 0.22012611391943301,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0028513146568446453,
+      "loss": 1.272,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2202138835023355,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002851121104250651,
+      "loss": 1.3294,
+      "step": 2509
+    },
+    {
+      "epoch": 0.22030165308523797,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0028509274331107565,
+      "loss": 1.2951,
+      "step": 2510
+    },
+    {
+      "epoch": 0.22038942266814046,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002850733643444076,
+      "loss": 1.2978,
+      "step": 2511
+    },
+    {
+      "epoch": 0.22047719225104295,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002850539735269736,
+      "loss": 1.2558,
+      "step": 2512
+    },
+    {
+      "epoch": 0.22056496183394544,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002850345708606875,
+      "loss": 1.3334,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2206527314168479,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002850151563474641,
+      "loss": 1.2744,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2207405009997504,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002849957299892197,
+      "loss": 1.3208,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2208282705826529,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.002849762917878716,
+      "loss": 1.2707,
+      "step": 2516
+    },
+    {
+      "epoch": 0.22091604016555538,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0028495684174533816,
+      "loss": 1.3309,
+      "step": 2517
+    },
+    {
+      "epoch": 0.22100380974845787,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.002849373798635391,
+      "loss": 1.2875,
+      "step": 2518
+    },
+    {
+      "epoch": 0.22109157933136034,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0028491790614439522,
+      "loss": 1.304,
+      "step": 2519
+    },
+    {
+      "epoch": 0.22117934891426283,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002848984205898285,
+      "loss": 1.2817,
+      "step": 2520
+    },
+    {
+      "epoch": 0.22126711849716532,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002848789232017621,
+      "loss": 1.3154,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2213548880800678,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0028485941398212027,
+      "loss": 1.2477,
+      "step": 2522
+    },
+    {
+      "epoch": 0.22144265766297028,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.002848398929328286,
+      "loss": 1.3148,
+      "step": 2523
+    },
+    {
+      "epoch": 0.22153042724587277,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0028482036005581363,
+      "loss": 1.3134,
+      "step": 2524
+    },
+    {
+      "epoch": 0.22161819682877526,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002848008153530033,
+      "loss": 1.2918,
+      "step": 2525
+    },
+    {
+      "epoch": 0.22170596641167775,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002847812588263264,
+      "loss": 1.2822,
+      "step": 2526
+    },
+    {
+      "epoch": 0.22179373599458022,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.002847616904777133,
+      "loss": 1.2444,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2218815055774827,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0028474211030909517,
+      "loss": 1.3081,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2219692751603852,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002847225183224046,
+      "loss": 1.2849,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2220570447432877,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0028470291451957513,
+      "loss": 1.2453,
+      "step": 2530
+    },
+    {
+      "epoch": 0.22214481432619018,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0028468329890254168,
+      "loss": 1.2833,
+      "step": 2531
+    },
+    {
+      "epoch": 0.22223258390909265,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0028466367147324018,
+      "loss": 1.2666,
+      "step": 2532
+    },
+    {
+      "epoch": 0.22232035349199514,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002846440322336078,
+      "loss": 1.3726,
+      "step": 2533
+    },
+    {
+      "epoch": 0.22240812307489763,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0028462438118558283,
+      "loss": 1.2715,
+      "step": 2534
+    },
+    {
+      "epoch": 0.22249589265780012,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0028460471833110473,
+      "loss": 1.2975,
+      "step": 2535
+    },
+    {
+      "epoch": 0.22258366224070258,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.002845850436721143,
+      "loss": 1.263,
+      "step": 2536
+    },
+    {
+      "epoch": 0.22267143182360508,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.002845653572105531,
+      "loss": 1.2012,
+      "step": 2537
+    },
+    {
+      "epoch": 0.22275920140650757,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002845456589483643,
+      "loss": 1.2756,
+      "step": 2538
+    },
+    {
+      "epoch": 0.22284697098941006,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00284525948887492,
+      "loss": 1.2775,
+      "step": 2539
+    },
+    {
+      "epoch": 0.22293474057231252,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0028450622702988143,
+      "loss": 1.2989,
+      "step": 2540
+    },
+    {
+      "epoch": 0.22302251015521501,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0028448649337747918,
+      "loss": 1.2722,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2231102797381175,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002844667479322328,
+      "loss": 1.2688,
+      "step": 2542
+    },
+    {
+      "epoch": 0.22319804932102,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002844469906960911,
+      "loss": 1.3229,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2232858189039225,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0028442722167100402,
+      "loss": 1.323,
+      "step": 2544
+    },
+    {
+      "epoch": 0.22337358848682495,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0028440744085892274,
+      "loss": 1.2572,
+      "step": 2545
+    },
+    {
+      "epoch": 0.22346135806972744,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.002843876482617995,
+      "loss": 1.262,
+      "step": 2546
+    },
+    {
+      "epoch": 0.22354912765262994,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0028436784388158774,
+      "loss": 1.2959,
+      "step": 2547
+    },
+    {
+      "epoch": 0.22363689723553243,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.002843480277202422,
+      "loss": 1.3413,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2237246668184349,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0028432819977971847,
+      "loss": 1.318,
+      "step": 2549
+    },
+    {
+      "epoch": 0.22381243640133738,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0028430836006197364,
+      "loss": 1.2518,
+      "step": 2550
+    },
+    {
+      "epoch": 0.22390020598423988,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002842885085689657,
+      "loss": 1.3169,
+      "step": 2551
+    },
+    {
+      "epoch": 0.22398797556714237,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00284268645302654,
+      "loss": 1.3088,
+      "step": 2552
+    },
+    {
+      "epoch": 0.22407574515004486,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0028424877026499893,
+      "loss": 1.2508,
+      "step": 2553
+    },
+    {
+      "epoch": 0.22416351473294732,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0028422888345796204,
+      "loss": 1.2534,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2242512843158498,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0028420898488350616,
+      "loss": 1.2965,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2243390538987523,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.002841890745435952,
+      "loss": 1.3284,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2244268234816548,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0028416915244019406,
+      "loss": 1.2437,
+      "step": 2557
+    },
+    {
+      "epoch": 0.22451459306455726,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.002841492185752692,
+      "loss": 1.3139,
+      "step": 2558
+    },
+    {
+      "epoch": 0.22460236264745975,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0028412927295078786,
+      "loss": 1.2503,
+      "step": 2559
+    },
+    {
+      "epoch": 0.22469013223036224,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.002841093155687187,
+      "loss": 1.2863,
+      "step": 2560
+    },
+    {
+      "epoch": 0.22477790181326474,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.002840893464310314,
+      "loss": 1.2805,
+      "step": 2561
+    },
+    {
+      "epoch": 0.2248656713961672,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002840693655396967,
+      "loss": 1.3579,
+      "step": 2562
+    },
+    {
+      "epoch": 0.2249534409790697,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002840493728966869,
+      "loss": 1.2618,
+      "step": 2563
+    },
+    {
+      "epoch": 0.22504121056197218,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0028402936850397496,
+      "loss": 1.3318,
+      "step": 2564
+    },
+    {
+      "epoch": 0.22512898014487467,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.002840093523635354,
+      "loss": 1.2483,
+      "step": 2565
+    },
+    {
+      "epoch": 0.22521674972777717,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002839893244773436,
+      "loss": 1.2382,
+      "step": 2566
+    },
+    {
+      "epoch": 0.22530451931067963,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.002839692848473763,
+      "loss": 1.2556,
+      "step": 2567
+    },
+    {
+      "epoch": 0.22539228889358212,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0028394923347561133,
+      "loss": 1.298,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2254800584764846,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002839291703640277,
+      "loss": 1.3292,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2255678280593871,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002839090955146055,
+      "loss": 1.3303,
+      "step": 2570
+    },
+    {
+      "epoch": 0.22565559764228957,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.002838890089293261,
+      "loss": 1.2944,
+      "step": 2571
+    },
+    {
+      "epoch": 0.22574336722519206,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0028386891061017187,
+      "loss": 1.3799,
+      "step": 2572
+    },
+    {
+      "epoch": 0.22583113680809455,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0028384880055912654,
+      "loss": 1.2495,
+      "step": 2573
+    },
+    {
+      "epoch": 0.22591890639099704,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002838286787781749,
+      "loss": 1.3019,
+      "step": 2574
+    },
+    {
+      "epoch": 0.2260066759738995,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002838085452693028,
+      "loss": 1.3039,
+      "step": 2575
+    },
+    {
+      "epoch": 0.226094445556802,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002837884000344974,
+      "loss": 1.2842,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2261822151397045,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0028376824307574694,
+      "loss": 1.312,
+      "step": 2577
+    },
+    {
+      "epoch": 0.22626998472260698,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0028374807439504085,
+      "loss": 1.3599,
+      "step": 2578
+    },
+    {
+      "epoch": 0.22635775430550947,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0028372789399436966,
+      "loss": 1.2641,
+      "step": 2579
+    },
+    {
+      "epoch": 0.22644552388841194,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002837077018757251,
+      "loss": 1.2717,
+      "step": 2580
+    },
+    {
+      "epoch": 0.22653329347131443,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0028368749804110007,
+      "loss": 1.2641,
+      "step": 2581
+    },
+    {
+      "epoch": 0.22662106305421692,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0028366728249248864,
+      "loss": 1.328,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2267088326371194,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0028364705523188593,
+      "loss": 1.3176,
+      "step": 2583
+    },
+    {
+      "epoch": 0.22679660222002188,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0028362681626128835,
+      "loss": 1.2509,
+      "step": 2584
+    },
+    {
+      "epoch": 0.22688437180292437,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002836065655826934,
+      "loss": 1.2126,
+      "step": 2585
+    },
+    {
+      "epoch": 0.22697214138582686,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002835863031980997,
+      "loss": 1.3478,
+      "step": 2586
+    },
+    {
+      "epoch": 0.22705991096872935,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002835660291095071,
+      "loss": 1.294,
+      "step": 2587
+    },
+    {
+      "epoch": 0.22714768055163184,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.002835457433189166,
+      "loss": 1.2759,
+      "step": 2588
+    },
+    {
+      "epoch": 0.2272354501345343,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.002835254458283303,
+      "loss": 1.2792,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2273232197174368,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.002835051366397514,
+      "loss": 1.317,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2274109893003393,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0028348481575518453,
+      "loss": 1.2758,
+      "step": 2591
+    },
+    {
+      "epoch": 0.22749875888324178,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0028346448317663506,
+      "loss": 1.3313,
+      "step": 2592
+    },
+    {
+      "epoch": 0.22758652846614424,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0028344413890610994,
+      "loss": 1.2772,
+      "step": 2593
+    },
+    {
+      "epoch": 0.22767429804904674,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002834237829456169,
+      "loss": 1.334,
+      "step": 2594
+    },
+    {
+      "epoch": 0.22776206763194923,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.002834034152971651,
+      "loss": 1.2577,
+      "step": 2595
+    },
+    {
+      "epoch": 0.22784983721485172,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0028338303596276465,
+      "loss": 1.2432,
+      "step": 2596
+    },
+    {
+      "epoch": 0.22793760679775418,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00283362644944427,
+      "loss": 1.2904,
+      "step": 2597
+    },
+    {
+      "epoch": 0.22802537638065667,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0028334224224416467,
+      "loss": 1.2992,
+      "step": 2598
+    },
+    {
+      "epoch": 0.22811314596355917,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0028332182786399126,
+      "loss": 1.2994,
+      "step": 2599
+    },
+    {
+      "epoch": 0.22820091554646166,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.002833014018059216,
+      "loss": 1.2872,
+      "step": 2600
+    },
+    {
+      "epoch": 0.22828868512936415,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002832809640719717,
+      "loss": 1.3066,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2283764547122666,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0028326051466415862,
+      "loss": 1.2929,
+      "step": 2602
+    },
+    {
+      "epoch": 0.2284642242951691,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002832400535845007,
+      "loss": 1.2974,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2285519938780716,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002832195808350174,
+      "loss": 1.281,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2286397634609741,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0028319909641772913,
+      "loss": 1.2723,
+      "step": 2605
+    },
+    {
+      "epoch": 0.22872753304387655,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0028317860033465784,
+      "loss": 1.2695,
+      "step": 2606
+    },
+    {
+      "epoch": 0.22881530262677904,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0028315809258782623,
+      "loss": 1.3227,
+      "step": 2607
+    },
+    {
+      "epoch": 0.22890307220968154,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0028313757317925846,
+      "loss": 1.2679,
+      "step": 2608
+    },
+    {
+      "epoch": 0.22899084179258403,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0028311704211097964,
+      "loss": 1.2505,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2290786113754865,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0028309649938501615,
+      "loss": 1.2932,
+      "step": 2610
+    },
+    {
+      "epoch": 0.22916638095838898,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0028307594500339545,
+      "loss": 1.3171,
+      "step": 2611
+    },
+    {
+      "epoch": 0.22925415054129147,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0028305537896814613,
+      "loss": 1.2893,
+      "step": 2612
+    },
+    {
+      "epoch": 0.22934192012419397,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002830348012812981,
+      "loss": 1.2851,
+      "step": 2613
+    },
+    {
+      "epoch": 0.22942968970709646,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0028301421194488214,
+      "loss": 1.2582,
+      "step": 2614
+    },
+    {
+      "epoch": 0.22951745928999892,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.002829936109609305,
+      "loss": 1.2551,
+      "step": 2615
+    },
+    {
+      "epoch": 0.2296052288729014,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002829729983314763,
+      "loss": 1.2668,
+      "step": 2616
+    },
+    {
+      "epoch": 0.2296929984558039,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00282952374058554,
+      "loss": 1.3086,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2297807680387064,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0028293173814419904,
+      "loss": 1.2954,
+      "step": 2618
+    },
+    {
+      "epoch": 0.22986853762160886,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0028291109059044818,
+      "loss": 1.3079,
+      "step": 2619
+    },
+    {
+      "epoch": 0.22995630720451135,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002828904313993392,
+      "loss": 1.3284,
+      "step": 2620
+    },
+    {
+      "epoch": 0.23004407678741384,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0028286976057291116,
+      "loss": 1.2443,
+      "step": 2621
+    },
+    {
+      "epoch": 0.23013184637031633,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0028284907811320413,
+      "loss": 1.2909,
+      "step": 2622
+    },
+    {
+      "epoch": 0.23021961595321883,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0028282838402225938,
+      "loss": 1.2969,
+      "step": 2623
+    },
+    {
+      "epoch": 0.2303073855361213,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002828076783021193,
+      "loss": 1.3221,
+      "step": 2624
+    },
+    {
+      "epoch": 0.23039515511902378,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0028278696095482762,
+      "loss": 1.2783,
+      "step": 2625
+    },
+    {
+      "epoch": 0.23048292470192627,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0028276623198242893,
+      "loss": 1.3408,
+      "step": 2626
+    },
+    {
+      "epoch": 0.23057069428482876,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0028274549138696905,
+      "loss": 1.2444,
+      "step": 2627
+    },
+    {
+      "epoch": 0.23065846386773123,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.002827247391704951,
+      "loss": 1.3747,
+      "step": 2628
+    },
+    {
+      "epoch": 0.23074623345063372,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002827039753350552,
+      "loss": 1.2593,
+      "step": 2629
+    },
+    {
+      "epoch": 0.2308340030335362,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0028268319988269874,
+      "loss": 1.219,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2309217726164387,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00282662412815476,
+      "loss": 1.2553,
+      "step": 2631
+    },
+    {
+      "epoch": 0.23100954219934117,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0028264161413543874,
+      "loss": 1.2944,
+      "step": 2632
+    },
+    {
+      "epoch": 0.23109731178224366,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.002826208038446396,
+      "loss": 1.2736,
+      "step": 2633
+    },
+    {
+      "epoch": 0.23118508136514615,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0028259998194513254,
+      "loss": 1.3153,
+      "step": 2634
+    },
+    {
+      "epoch": 0.23127285094804864,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0028257914843897265,
+      "loss": 1.2136,
+      "step": 2635
+    },
+    {
+      "epoch": 0.23136062053095113,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0028255830332821596,
+      "loss": 1.2581,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2314483901138536,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0028253744661491994,
+      "loss": 1.2819,
+      "step": 2637
+    },
+    {
+      "epoch": 0.2315361596967561,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0028251657830114297,
+      "loss": 1.2644,
+      "step": 2638
+    },
+    {
+      "epoch": 0.23162392927965858,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0028249569838894473,
+      "loss": 1.3334,
+      "step": 2639
+    },
+    {
+      "epoch": 0.23171169886256107,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0028247480688038595,
+      "loss": 1.252,
+      "step": 2640
+    },
+    {
+      "epoch": 0.23179946844546354,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002824539037775286,
+      "loss": 1.3316,
+      "step": 2641
+    },
+    {
+      "epoch": 0.23188723802836603,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0028243298908243565,
+      "loss": 1.2814,
+      "step": 2642
+    },
+    {
+      "epoch": 0.23197500761126852,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002824120627971713,
+      "loss": 1.2514,
+      "step": 2643
+    },
+    {
+      "epoch": 0.232062777194171,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00282391124923801,
+      "loss": 1.2649,
+      "step": 2644
+    },
+    {
+      "epoch": 0.23215054677707347,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0028237017546439117,
+      "loss": 1.2262,
+      "step": 2645
+    },
+    {
+      "epoch": 0.23223831635997597,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.002823492144210094,
+      "loss": 1.2959,
+      "step": 2646
+    },
+    {
+      "epoch": 0.23232608594287846,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002823282417957245,
+      "loss": 1.3092,
+      "step": 2647
+    },
+    {
+      "epoch": 0.23241385552578095,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.002823072575906064,
+      "loss": 1.3097,
+      "step": 2648
+    },
+    {
+      "epoch": 0.23250162510868344,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0028228626180772606,
+      "loss": 1.3089,
+      "step": 2649
+    },
+    {
+      "epoch": 0.2325893946915859,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028226525444915583,
+      "loss": 1.2939,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2326771642744884,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0028224423551696893,
+      "loss": 1.3127,
+      "step": 2651
+    },
+    {
+      "epoch": 0.2327649338573909,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0028222320501323996,
+      "loss": 1.2854,
+      "step": 2652
+    },
+    {
+      "epoch": 0.23285270344029338,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002822021629400444,
+      "loss": 1.3329,
+      "step": 2653
+    },
+    {
+      "epoch": 0.23294047302319584,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002821811092994592,
+      "loss": 1.3233,
+      "step": 2654
+    },
+    {
+      "epoch": 0.23302824260609833,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0028216004409356214,
+      "loss": 1.3286,
+      "step": 2655
+    },
+    {
+      "epoch": 0.23311601218900083,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002821389673244323,
+      "loss": 1.2255,
+      "step": 2656
+    },
+    {
+      "epoch": 0.23320378177190332,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0028211787899414986,
+      "loss": 1.2183,
+      "step": 2657
+    },
+    {
+      "epoch": 0.2332915513548058,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.002820967791047962,
+      "loss": 1.2769,
+      "step": 2658
+    },
+    {
+      "epoch": 0.23337932093770827,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0028207566765845377,
+      "loss": 1.206,
+      "step": 2659
+    },
+    {
+      "epoch": 0.23346709052061076,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0028205454465720613,
+      "loss": 1.2981,
+      "step": 2660
+    },
+    {
+      "epoch": 0.23355486010351326,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002820334101031382,
+      "loss": 1.2803,
+      "step": 2661
+    },
+    {
+      "epoch": 0.23364262968641575,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002820122639983357,
+      "loss": 1.2925,
+      "step": 2662
+    },
+    {
+      "epoch": 0.2337303992693182,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0028199110634488573,
+      "loss": 1.3133,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2338181688522207,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002819699371448765,
+      "loss": 1.2847,
+      "step": 2664
+    },
+    {
+      "epoch": 0.2339059384351232,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002819487564003973,
+      "loss": 1.2373,
+      "step": 2665
+    },
+    {
+      "epoch": 0.2339937080180257,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002819275641135386,
+      "loss": 1.2995,
+      "step": 2666
+    },
+    {
+      "epoch": 0.23408147760092815,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0028190636028639195,
+      "loss": 1.1893,
+      "step": 2667
+    },
+    {
+      "epoch": 0.23416924718383064,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0028188514492105015,
+      "loss": 1.2701,
+      "step": 2668
+    },
+    {
+      "epoch": 0.23425701676673313,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0028186391801960703,
+      "loss": 1.3307,
+      "step": 2669
+    },
+    {
+      "epoch": 0.23434478634963563,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0028184267958415755,
+      "loss": 1.3358,
+      "step": 2670
+    },
+    {
+      "epoch": 0.23443255593253812,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0028182142961679798,
+      "loss": 1.3149,
+      "step": 2671
+    },
+    {
+      "epoch": 0.23452032551544058,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0028180016811962558,
+      "loss": 1.3222,
+      "step": 2672
+    },
+    {
+      "epoch": 0.23460809509834307,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0028177889509473863,
+      "loss": 1.3285,
+      "step": 2673
+    },
+    {
+      "epoch": 0.23469586468124556,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.002817576105442369,
+      "loss": 1.2687,
+      "step": 2674
+    },
+    {
+      "epoch": 0.23478363426414806,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0028173631447022097,
+      "loss": 1.2757,
+      "step": 2675
+    },
+    {
+      "epoch": 0.23487140384705052,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0028171500687479267,
+      "loss": 1.2429,
+      "step": 2676
+    },
+    {
+      "epoch": 0.234959173429953,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0028169368776005503,
+      "loss": 1.2815,
+      "step": 2677
+    },
+    {
+      "epoch": 0.2350469430128555,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0028167235712811215,
+      "loss": 1.2862,
+      "step": 2678
+    },
+    {
+      "epoch": 0.235134712595758,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002816510149810693,
+      "loss": 1.2643,
+      "step": 2679
+    },
+    {
+      "epoch": 0.23522248217866046,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0028162966132103274,
+      "loss": 1.343,
+      "step": 2680
+    },
+    {
+      "epoch": 0.23531025176156295,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002816082961501102,
+      "loss": 1.342,
+      "step": 2681
+    },
+    {
+      "epoch": 0.23539802134446544,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0028158691947041017,
+      "loss": 1.1977,
+      "step": 2682
+    },
+    {
+      "epoch": 0.23548579092736793,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002815655312840425,
+      "loss": 1.3142,
+      "step": 2683
+    },
+    {
+      "epoch": 0.23557356051027042,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0028154413159311815,
+      "loss": 1.2546,
+      "step": 2684
+    },
+    {
+      "epoch": 0.2356613300931729,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.002815227203997491,
+      "loss": 1.2795,
+      "step": 2685
+    },
+    {
+      "epoch": 0.23574909967607538,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0028150129770604868,
+      "loss": 1.2788,
+      "step": 2686
+    },
+    {
+      "epoch": 0.23583686925897787,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0028147986351413105,
+      "loss": 1.2825,
+      "step": 2687
+    },
+    {
+      "epoch": 0.23592463884188036,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0028145841782611183,
+      "loss": 1.316,
+      "step": 2688
+    },
+    {
+      "epoch": 0.23601240842478283,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002814369606441076,
+      "loss": 1.2949,
+      "step": 2689
+    },
+    {
+      "epoch": 0.23610017800768532,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0028141549197023604,
+      "loss": 1.3241,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2361879475905878,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0028139401180661606,
+      "loss": 1.3208,
+      "step": 2691
+    },
+    {
+      "epoch": 0.2362757171734903,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0028137252015536765,
+      "loss": 1.2905,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2363634867563928,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00281351017018612,
+      "loss": 1.2847,
+      "step": 2693
+    },
+    {
+      "epoch": 0.23645125633929526,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002813295023984713,
+      "loss": 1.2918,
+      "step": 2694
+    },
+    {
+      "epoch": 0.23653902592219775,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0028130797629706905,
+      "loss": 1.2487,
+      "step": 2695
+    },
+    {
+      "epoch": 0.23662679550510024,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0028128643871652976,
+      "loss": 1.2858,
+      "step": 2696
+    },
+    {
+      "epoch": 0.23671456508800273,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0028126488965897903,
+      "loss": 1.2701,
+      "step": 2697
+    },
+    {
+      "epoch": 0.2368023346709052,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0028124332912654376,
+      "loss": 1.2544,
+      "step": 2698
+    },
+    {
+      "epoch": 0.2368901042538077,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0028122175712135184,
+      "loss": 1.2927,
+      "step": 2699
+    },
+    {
+      "epoch": 0.23697787383671018,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002812001736455324,
+      "loss": 1.2842,
+      "step": 2700
+    },
+    {
+      "epoch": 0.23706564341961267,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0028117857870121558,
+      "loss": 1.3013,
+      "step": 2701
+    },
+    {
+      "epoch": 0.23715341300251513,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002811569722905327,
+      "loss": 1.2799,
+      "step": 2702
+    },
+    {
+      "epoch": 0.23724118258541763,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002811353544156163,
+      "loss": 1.3579,
+      "step": 2703
+    },
+    {
+      "epoch": 0.23732895216832012,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0028111372507859996,
+      "loss": 1.2623,
+      "step": 2704
+    },
+    {
+      "epoch": 0.2374167217512226,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0028109208428161838,
+      "loss": 1.3012,
+      "step": 2705
+    },
+    {
+      "epoch": 0.2375044913341251,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002810704320268074,
+      "loss": 1.2589,
+      "step": 2706
+    },
+    {
+      "epoch": 0.23759226091702756,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002810487683163041,
+      "loss": 1.2362,
+      "step": 2707
+    },
+    {
+      "epoch": 0.23768003049993006,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0028102709315224656,
+      "loss": 1.2701,
+      "step": 2708
+    },
+    {
+      "epoch": 0.23776780008283255,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0028100540653677396,
+      "loss": 1.316,
+      "step": 2709
+    },
+    {
+      "epoch": 0.23785556966573504,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002809837084720268,
+      "loss": 1.2557,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2379433392486375,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0028096199896014648,
+      "loss": 1.2801,
+      "step": 2711
+    },
+    {
+      "epoch": 0.23803110883154,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0028094027800327575,
+      "loss": 1.3095,
+      "step": 2712
+    },
+    {
+      "epoch": 0.2381188784144425,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0028091854560355835,
+      "loss": 1.2497,
+      "step": 2713
+    },
+    {
+      "epoch": 0.23820664799734498,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002808968017631391,
+      "loss": 1.3031,
+      "step": 2714
+    },
+    {
+      "epoch": 0.23829441758024744,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002808750464841641,
+      "loss": 1.279,
+      "step": 2715
+    },
+    {
+      "epoch": 0.23838218716314993,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002808532797687806,
+      "loss": 1.3033,
+      "step": 2716
+    },
+    {
+      "epoch": 0.23846995674605242,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002808315016191367,
+      "loss": 1.2536,
+      "step": 2717
+    },
+    {
+      "epoch": 0.23855772632895492,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0028080971203738193,
+      "loss": 1.238,
+      "step": 2718
+    },
+    {
+      "epoch": 0.2386454959118574,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0028078791102566685,
+      "loss": 1.2741,
+      "step": 2719
+    },
+    {
+      "epoch": 0.23873326549475987,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0028076609858614307,
+      "loss": 1.3111,
+      "step": 2720
+    },
+    {
+      "epoch": 0.23882103507766236,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0028074427472096343,
+      "loss": 1.2899,
+      "step": 2721
+    },
+    {
+      "epoch": 0.23890880466056486,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002807224394322819,
+      "loss": 1.2425,
+      "step": 2722
+    },
+    {
+      "epoch": 0.23899657424346735,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002807005927222534,
+      "loss": 1.3177,
+      "step": 2723
+    },
+    {
+      "epoch": 0.2390843438263698,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002806787345930343,
+      "loss": 1.2674,
+      "step": 2724
+    },
+    {
+      "epoch": 0.2391721134092723,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0028065686504678174,
+      "loss": 1.2979,
+      "step": 2725
+    },
+    {
+      "epoch": 0.2392598829921748,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002806349840856542,
+      "loss": 1.3585,
+      "step": 2726
+    },
+    {
+      "epoch": 0.23934765257507729,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0028061309171181132,
+      "loss": 1.297,
+      "step": 2727
+    },
+    {
+      "epoch": 0.23943542215797978,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0028059118792741373,
+      "loss": 1.2898,
+      "step": 2728
+    },
+    {
+      "epoch": 0.23952319174088224,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0028056927273462334,
+      "loss": 1.2496,
+      "step": 2729
+    },
+    {
+      "epoch": 0.23961096132378473,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00280547346135603,
+      "loss": 1.2823,
+      "step": 2730
+    },
+    {
+      "epoch": 0.23969873090668722,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002805254081325167,
+      "loss": 1.2302,
+      "step": 2731
+    },
+    {
+      "epoch": 0.23978650048958972,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002805034587275298,
+      "loss": 1.2376,
+      "step": 2732
+    },
+    {
+      "epoch": 0.23987427007249218,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0028048149792280854,
+      "loss": 1.3058,
+      "step": 2733
+    },
+    {
+      "epoch": 0.23996203965539467,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0028045952572052033,
+      "loss": 1.3163,
+      "step": 2734
+    },
+    {
+      "epoch": 0.24004980923829716,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0028043754212283382,
+      "loss": 1.2277,
+      "step": 2735
+    },
+    {
+      "epoch": 0.24013757882119965,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002804155471319187,
+      "loss": 1.2385,
+      "step": 2736
+    },
+    {
+      "epoch": 0.24022534840410212,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002803935407499457,
+      "loss": 1.2289,
+      "step": 2737
+    },
+    {
+      "epoch": 0.2403131179870046,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0028037152297908687,
+      "loss": 1.2196,
+      "step": 2738
+    },
+    {
+      "epoch": 0.2404008875699071,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0028034949382151515,
+      "loss": 1.2046,
+      "step": 2739
+    },
+    {
+      "epoch": 0.2404886571528096,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0028032745327940486,
+      "loss": 1.2787,
+      "step": 2740
+    },
+    {
+      "epoch": 0.24057642673571208,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0028030540135493124,
+      "loss": 1.2919,
+      "step": 2741
+    },
+    {
+      "epoch": 0.24066419631861455,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0028028333805027073,
+      "loss": 1.2882,
+      "step": 2742
+    },
+    {
+      "epoch": 0.24075196590151704,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0028026126336760096,
+      "loss": 1.3052,
+      "step": 2743
+    },
+    {
+      "epoch": 0.24083973548441953,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0028023917730910053,
+      "loss": 1.3639,
+      "step": 2744
+    },
+    {
+      "epoch": 0.24092750506732202,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0028021707987694925,
+      "loss": 1.2574,
+      "step": 2745
+    },
+    {
+      "epoch": 0.2410152746502245,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0028019497107332814,
+      "loss": 1.2762,
+      "step": 2746
+    },
+    {
+      "epoch": 0.24110304423312698,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002801728509004191,
+      "loss": 1.3217,
+      "step": 2747
+    },
+    {
+      "epoch": 0.24119081381602947,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002801507193604054,
+      "loss": 1.337,
+      "step": 2748
+    },
+    {
+      "epoch": 0.24127858339893196,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0028012857645547144,
+      "loss": 1.249,
+      "step": 2749
+    },
+    {
+      "epoch": 0.24136635298183443,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0028010642218780244,
+      "loss": 1.289,
+      "step": 2750
+    },
+    {
+      "epoch": 0.24145412256473692,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0028008425655958505,
+      "loss": 1.3001,
+      "step": 2751
+    },
+    {
+      "epoch": 0.2415418921476394,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.002800620795730069,
+      "loss": 1.2688,
+      "step": 2752
+    },
+    {
+      "epoch": 0.2416296617305419,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0028003989123025676,
+      "loss": 1.3052,
+      "step": 2753
+    },
+    {
+      "epoch": 0.2417174313134444,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0028001769153352457,
+      "loss": 1.345,
+      "step": 2754
+    },
+    {
+      "epoch": 0.24180520089634686,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002799954804850013,
+      "loss": 1.2633,
+      "step": 2755
+    },
+    {
+      "epoch": 0.24189297047924935,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0027997325808687915,
+      "loss": 1.2989,
+      "step": 2756
+    },
+    {
+      "epoch": 0.24198074006215184,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002799510243413513,
+      "loss": 1.3136,
+      "step": 2757
+    },
+    {
+      "epoch": 0.24206850964505433,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0027992877925061225,
+      "loss": 1.2612,
+      "step": 2758
+    },
+    {
+      "epoch": 0.2421562792279568,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002799065228168574,
+      "loss": 1.3486,
+      "step": 2759
+    },
+    {
+      "epoch": 0.24224404881085929,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0027988425504228342,
+      "loss": 1.2895,
+      "step": 2760
+    },
+    {
+      "epoch": 0.24233181839376178,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0027986197592908806,
+      "loss": 1.3472,
+      "step": 2761
+    },
+    {
+      "epoch": 0.24241958797666427,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002798396854794701,
+      "loss": 1.2866,
+      "step": 2762
+    },
+    {
+      "epoch": 0.24250735755956676,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002798173836956296,
+      "loss": 1.291,
+      "step": 2763
+    },
+    {
+      "epoch": 0.24259512714246922,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002797950705797677,
+      "loss": 1.2936,
+      "step": 2764
+    },
+    {
+      "epoch": 0.24268289672537172,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.002797727461340865,
+      "loss": 1.2728,
+      "step": 2765
+    },
+    {
+      "epoch": 0.2427706663082742,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002797504103607894,
+      "loss": 1.2896,
+      "step": 2766
+    },
+    {
+      "epoch": 0.2428584358911767,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0027972806326208087,
+      "loss": 1.2608,
+      "step": 2767
+    },
+    {
+      "epoch": 0.24294620547407916,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0027970570484016645,
+      "loss": 1.2772,
+      "step": 2768
+    },
+    {
+      "epoch": 0.24303397505698165,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002796833350972528,
+      "loss": 1.2883,
+      "step": 2769
+    },
+    {
+      "epoch": 0.24312174463988415,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0027966095403554776,
+      "loss": 1.3528,
+      "step": 2770
+    },
+    {
+      "epoch": 0.24320951422278664,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002796385616572602,
+      "loss": 1.228,
+      "step": 2771
+    },
+    {
+      "epoch": 0.2432972838056891,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0027961615796460028,
+      "loss": 1.3179,
+      "step": 2772
+    },
+    {
+      "epoch": 0.2433850533885916,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0027959374295977906,
+      "loss": 1.2832,
+      "step": 2773
+    },
+    {
+      "epoch": 0.24347282297149409,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0027957131664500885,
+      "loss": 1.2588,
+      "step": 2774
+    },
+    {
+      "epoch": 0.24356059255439658,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0027954887902250304,
+      "loss": 1.2755,
+      "step": 2775
+    },
+    {
+      "epoch": 0.24364836213729907,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0027952643009447604,
+      "loss": 1.2313,
+      "step": 2776
+    },
+    {
+      "epoch": 0.24373613172020153,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.002795039698631436,
+      "loss": 1.331,
+      "step": 2777
+    },
+    {
+      "epoch": 0.24382390130310402,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.002794814983307224,
+      "loss": 1.2873,
+      "step": 2778
+    },
+    {
+      "epoch": 0.24391167088600652,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0027945901549943026,
+      "loss": 1.2511,
+      "step": 2779
+    },
+    {
+      "epoch": 0.243999440468909,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0027943652137148623,
+      "loss": 1.2692,
+      "step": 2780
+    },
+    {
+      "epoch": 0.24408721005181147,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002794140159491103,
+      "loss": 1.2691,
+      "step": 2781
+    },
+    {
+      "epoch": 0.24417497963471396,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002793914992345238,
+      "loss": 1.2996,
+      "step": 2782
+    },
+    {
+      "epoch": 0.24426274921761645,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0027936897122994887,
+      "loss": 1.2517,
+      "step": 2783
+    },
+    {
+      "epoch": 0.24435051880051895,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00279346431937609,
+      "loss": 1.2767,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2444382883834214,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002793238813597288,
+      "loss": 1.2524,
+      "step": 2785
+    },
+    {
+      "epoch": 0.2445260579663239,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002793013194985338,
+      "loss": 1.294,
+      "step": 2786
+    },
+    {
+      "epoch": 0.2446138275492264,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0027927874635625093,
+      "loss": 1.2917,
+      "step": 2787
+    },
+    {
+      "epoch": 0.24470159713212888,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002792561619351079,
+      "loss": 1.2642,
+      "step": 2788
+    },
+    {
+      "epoch": 0.24478936671503138,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002792335662373338,
+      "loss": 1.3205,
+      "step": 2789
+    },
+    {
+      "epoch": 0.24487713629793384,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.002792109592651587,
+      "loss": 1.2848,
+      "step": 2790
+    },
+    {
+      "epoch": 0.24496490588083633,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002791883410208138,
+      "loss": 1.3269,
+      "step": 2791
+    },
+    {
+      "epoch": 0.24505267546373882,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0027916571150653157,
+      "loss": 1.2572,
+      "step": 2792
+    },
+    {
+      "epoch": 0.24514044504664131,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0027914307072454524,
+      "loss": 1.2437,
+      "step": 2793
+    },
+    {
+      "epoch": 0.24522821462954378,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002791204186770895,
+      "loss": 1.2579,
+      "step": 2794
+    },
+    {
+      "epoch": 0.24531598421244627,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0027909775536639997,
+      "loss": 1.2755,
+      "step": 2795
+    },
+    {
+      "epoch": 0.24540375379534876,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.002790750807947135,
+      "loss": 1.3189,
+      "step": 2796
+    },
+    {
+      "epoch": 0.24549152337825125,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0027905239496426788,
+      "loss": 1.3053,
+      "step": 2797
+    },
+    {
+      "epoch": 0.24557929296115374,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002790296978773022,
+      "loss": 1.3057,
+      "step": 2798
+    },
+    {
+      "epoch": 0.2456670625440562,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0027900698953605653,
+      "loss": 1.3154,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2457548321269587,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0027898426994277204,
+      "loss": 1.3037,
+      "step": 2800
+    },
+    {
+      "epoch": 0.2458426017098612,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0027896153909969116,
+      "loss": 1.2648,
+      "step": 2801
+    },
+    {
+      "epoch": 0.24593037129276368,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.002789387970090573,
+      "loss": 1.2515,
+      "step": 2802
+    },
+    {
+      "epoch": 0.24601814087566615,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00278916043673115,
+      "loss": 1.2333,
+      "step": 2803
+    },
+    {
+      "epoch": 0.24610591045856864,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002788932790941099,
+      "loss": 1.2861,
+      "step": 2804
+    },
+    {
+      "epoch": 0.24619368004147113,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0027887050327428885,
+      "loss": 1.2821,
+      "step": 2805
+    },
+    {
+      "epoch": 0.24628144962437362,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.002788477162158997,
+      "loss": 1.2608,
+      "step": 2806
+    },
+    {
+      "epoch": 0.24636921920727609,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002788249179211914,
+      "loss": 1.2737,
+      "step": 2807
+    },
+    {
+      "epoch": 0.24645698879017858,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0027880210839241405,
+      "loss": 1.2698,
+      "step": 2808
+    },
+    {
+      "epoch": 0.24654475837308107,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0027877928763181898,
+      "loss": 1.2469,
+      "step": 2809
+    },
+    {
+      "epoch": 0.24663252795598356,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0027875645564165836,
+      "loss": 1.2851,
+      "step": 2810
+    },
+    {
+      "epoch": 0.24672029753888605,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002787336124241857,
+      "loss": 1.3066,
+      "step": 2811
+    },
+    {
+      "epoch": 0.24680806712178852,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002787107579816555,
+      "loss": 1.2742,
+      "step": 2812
+    },
+    {
+      "epoch": 0.246895836704691,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0027868789231632342,
+      "loss": 1.2275,
+      "step": 2813
+    },
+    {
+      "epoch": 0.2469836062875935,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0027866501543044623,
+      "loss": 1.2392,
+      "step": 2814
+    },
+    {
+      "epoch": 0.247071375870496,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0027864212732628176,
+      "loss": 1.2716,
+      "step": 2815
+    },
+    {
+      "epoch": 0.24715914545339845,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00278619228006089,
+      "loss": 1.3828,
+      "step": 2816
+    },
+    {
+      "epoch": 0.24724691503630095,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00278596317472128,
+      "loss": 1.288,
+      "step": 2817
+    },
+    {
+      "epoch": 0.24733468461920344,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0027857339572665988,
+      "loss": 1.3191,
+      "step": 2818
+    },
+    {
+      "epoch": 0.24742245420210593,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.002785504627719471,
+      "loss": 1.2204,
+      "step": 2819
+    },
+    {
+      "epoch": 0.2475102237850084,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002785275186102528,
+      "loss": 1.2903,
+      "step": 2820
+    },
+    {
+      "epoch": 0.24759799336791088,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0027850456324384175,
+      "loss": 1.2806,
+      "step": 2821
+    },
+    {
+      "epoch": 0.24768576295081338,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0027848159667497934,
+      "loss": 1.2507,
+      "step": 2822
+    },
+    {
+      "epoch": 0.24777353253371587,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002784586189059324,
+      "loss": 1.2892,
+      "step": 2823
+    },
+    {
+      "epoch": 0.24786130211661836,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002784356299389687,
+      "loss": 1.3613,
+      "step": 2824
+    },
+    {
+      "epoch": 0.24794907169952082,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0027841262977635717,
+      "loss": 1.295,
+      "step": 2825
+    },
+    {
+      "epoch": 0.24803684128242331,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002783896184203678,
+      "loss": 1.2566,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2481246108653258,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002783665958732718,
+      "loss": 1.2489,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2482123804482283,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002783435621373413,
+      "loss": 1.3062,
+      "step": 2828
+    },
+    {
+      "epoch": 0.24830015003113076,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0027832051721484972,
+      "loss": 1.2788,
+      "step": 2829
+    },
+    {
+      "epoch": 0.24838791961403325,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0027829746110807152,
+      "loss": 1.2482,
+      "step": 2830
+    },
+    {
+      "epoch": 0.24847568919693575,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002782743938192822,
+      "loss": 1.3907,
+      "step": 2831
+    },
+    {
+      "epoch": 0.24856345877983824,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0027825131535075837,
+      "loss": 1.3046,
+      "step": 2832
+    },
+    {
+      "epoch": 0.24865122836274073,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0027822822570477785,
+      "loss": 1.3073,
+      "step": 2833
+    },
+    {
+      "epoch": 0.2487389979456432,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002782051248836195,
+      "loss": 1.2559,
+      "step": 2834
+    },
+    {
+      "epoch": 0.24882676752854568,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0027818201288956315,
+      "loss": 1.3289,
+      "step": 2835
+    },
+    {
+      "epoch": 0.24891453711144818,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002781588897248901,
+      "loss": 1.2602,
+      "step": 2836
+    },
+    {
+      "epoch": 0.24900230669435067,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002781357553918823,
+      "loss": 1.2663,
+      "step": 2837
+    },
+    {
+      "epoch": 0.24909007627725313,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0027811260989282316,
+      "loss": 1.3229,
+      "step": 2838
+    },
+    {
+      "epoch": 0.24917784586015562,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.002780894532299969,
+      "loss": 1.2647,
+      "step": 2839
+    },
+    {
+      "epoch": 0.24926561544305811,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0027806628540568913,
+      "loss": 1.3106,
+      "step": 2840
+    },
+    {
+      "epoch": 0.2493533850259606,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002780431064221864,
+      "loss": 1.3421,
+      "step": 2841
+    },
+    {
+      "epoch": 0.24944115460886307,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002780199162817763,
+      "loss": 1.3106,
+      "step": 2842
+    },
+    {
+      "epoch": 0.24952892419176556,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.002779967149867477,
+      "loss": 1.2301,
+      "step": 2843
+    },
+    {
+      "epoch": 0.24961669377466805,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0027797350253939038,
+      "loss": 1.307,
+      "step": 2844
+    },
+    {
+      "epoch": 0.24970446335757054,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.002779502789419954,
+      "loss": 1.3101,
+      "step": 2845
+    },
+    {
+      "epoch": 0.24979223294047304,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0027792704419685485,
+      "loss": 1.2597,
+      "step": 2846
+    },
+    {
+      "epoch": 0.2498800025233755,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002779037983062618,
+      "loss": 1.2363,
+      "step": 2847
+    },
+    {
+      "epoch": 0.249967772106278,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0027788054127251068,
+      "loss": 1.2775,
+      "step": 2848
+    },
+    {
+      "epoch": 0.2500555416891805,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0027785727309789674,
+      "loss": 1.2657,
+      "step": 2849
+    },
+    {
+      "epoch": 0.25014331127208295,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0027783399378471653,
+      "loss": 1.3093,
+      "step": 2850
+    },
+    {
+      "epoch": 0.25023108085498547,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002778107033352675,
+      "loss": 1.2647,
+      "step": 2851
+    },
+    {
+      "epoch": 0.25031885043788793,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0027778740175184854,
+      "loss": 1.2843,
+      "step": 2852
+    },
+    {
+      "epoch": 0.2504066200207904,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002777640890367592,
+      "loss": 1.3074,
+      "step": 2853
+    },
+    {
+      "epoch": 0.2504943896036929,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002777407651923005,
+      "loss": 1.3022,
+      "step": 2854
+    },
+    {
+      "epoch": 0.2505821591865954,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002777174302207744,
+      "loss": 1.3172,
+      "step": 2855
+    },
+    {
+      "epoch": 0.2506699287694979,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002776940841244839,
+      "loss": 1.3158,
+      "step": 2856
+    },
+    {
+      "epoch": 0.25075769835240036,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002776707269057332,
+      "loss": 1.257,
+      "step": 2857
+    },
+    {
+      "epoch": 0.2508454679353028,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0027764735856682763,
+      "loss": 1.3238,
+      "step": 2858
+    },
+    {
+      "epoch": 0.25093323751820534,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0027762397911007338,
+      "loss": 1.2507,
+      "step": 2859
+    },
+    {
+      "epoch": 0.2510210071011078,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0027760058853777813,
+      "loss": 1.3555,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2511087766840103,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002775771868522503,
+      "loss": 1.3293,
+      "step": 2861
+    },
+    {
+      "epoch": 0.2511965462669128,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0027755377405579954,
+      "loss": 1.2397,
+      "step": 2862
+    },
+    {
+      "epoch": 0.25128431584981525,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0027753035015073667,
+      "loss": 1.3477,
+      "step": 2863
+    },
+    {
+      "epoch": 0.2513720854327178,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0027750691513937346,
+      "loss": 1.2326,
+      "step": 2864
+    },
+    {
+      "epoch": 0.25145985501562024,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0027748346902402295,
+      "loss": 1.2899,
+      "step": 2865
+    },
+    {
+      "epoch": 0.2515476245985227,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0027746001180699907,
+      "loss": 1.3252,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2516353941814252,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00277436543490617,
+      "loss": 1.2278,
+      "step": 2867
+    },
+    {
+      "epoch": 0.2517231637643277,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0027741306407719304,
+      "loss": 1.2682,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2518109333472302,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.002773895735690444,
+      "loss": 1.2905,
+      "step": 2869
+    },
+    {
+      "epoch": 0.25189870293013267,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0027736607196848957,
+      "loss": 1.3347,
+      "step": 2870
+    },
+    {
+      "epoch": 0.25198647251303513,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0027734255927784803,
+      "loss": 1.2919,
+      "step": 2871
+    },
+    {
+      "epoch": 0.25207424209593765,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0027731903549944046,
+      "loss": 1.2876,
+      "step": 2872
+    },
+    {
+      "epoch": 0.2521620116788401,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0027729550063558843,
+      "loss": 1.2204,
+      "step": 2873
+    },
+    {
+      "epoch": 0.25224978126174263,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0027727195468861487,
+      "loss": 1.3083,
+      "step": 2874
+    },
+    {
+      "epoch": 0.2523375508446451,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002772483976608436,
+      "loss": 1.3388,
+      "step": 2875
+    },
+    {
+      "epoch": 0.25242532042754756,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0027722482955459967,
+      "loss": 1.3027,
+      "step": 2876
+    },
+    {
+      "epoch": 0.2525130900104501,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0027720125037220904,
+      "loss": 1.3228,
+      "step": 2877
+    },
+    {
+      "epoch": 0.25260085959335254,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0027717766011599904,
+      "loss": 1.3671,
+      "step": 2878
+    },
+    {
+      "epoch": 0.25268862917625506,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.002771540587882978,
+      "loss": 1.3272,
+      "step": 2879
+    },
+    {
+      "epoch": 0.25277639875915753,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002771304463914347,
+      "loss": 1.2853,
+      "step": 2880
+    },
+    {
+      "epoch": 0.25286416834206,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0027710682292774024,
+      "loss": 1.3132,
+      "step": 2881
+    },
+    {
+      "epoch": 0.2529519379249625,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00277083188399546,
+      "loss": 1.3118,
+      "step": 2882
+    },
+    {
+      "epoch": 0.253039707507865,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002770595428091845,
+      "loss": 1.2235,
+      "step": 2883
+    },
+    {
+      "epoch": 0.25312747709076744,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0027703588615898955,
+      "loss": 1.2472,
+      "step": 2884
+    },
+    {
+      "epoch": 0.25321524667366996,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002770122184512959,
+      "loss": 1.2688,
+      "step": 2885
+    },
+    {
+      "epoch": 0.2533030162565724,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0027698853968843953,
+      "loss": 1.3174,
+      "step": 2886
+    },
+    {
+      "epoch": 0.25339078583947494,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0027696484987275736,
+      "loss": 1.321,
+      "step": 2887
+    },
+    {
+      "epoch": 0.2534785554223774,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0027694114900658757,
+      "loss": 1.2987,
+      "step": 2888
+    },
+    {
+      "epoch": 0.25356632500527987,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0027691743709226933,
+      "loss": 1.2478,
+      "step": 2889
+    },
+    {
+      "epoch": 0.2536540945881824,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0027689371413214285,
+      "loss": 1.2137,
+      "step": 2890
+    },
+    {
+      "epoch": 0.25374186417108485,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0027686998012854953,
+      "loss": 1.3501,
+      "step": 2891
+    },
+    {
+      "epoch": 0.25382963375398737,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0027684623508383186,
+      "loss": 1.2621,
+      "step": 2892
+    },
+    {
+      "epoch": 0.25391740333688984,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002768224790003333,
+      "loss": 1.3532,
+      "step": 2893
+    },
+    {
+      "epoch": 0.2540051729197923,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0027679871188039853,
+      "loss": 1.2514,
+      "step": 2894
+    },
+    {
+      "epoch": 0.2540929425026948,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0027677493372637325,
+      "loss": 1.2835,
+      "step": 2895
+    },
+    {
+      "epoch": 0.2541807120855973,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0027675114454060426,
+      "loss": 1.2727,
+      "step": 2896
+    },
+    {
+      "epoch": 0.25426848166849975,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002767273443254395,
+      "loss": 1.2444,
+      "step": 2897
+    },
+    {
+      "epoch": 0.25435625125140227,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0027670353308322797,
+      "loss": 1.3285,
+      "step": 2898
+    },
+    {
+      "epoch": 0.25444402083430473,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0027667971081631972,
+      "loss": 1.21,
+      "step": 2899
+    },
+    {
+      "epoch": 0.25453179041720725,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002766558775270658,
+      "loss": 1.2441,
+      "step": 2900
+    },
+    {
+      "epoch": 0.2546195600001097,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002766320332178187,
+      "loss": 1.2806,
+      "step": 2901
+    },
+    {
+      "epoch": 0.2547073295830122,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0027660817789093156,
+      "loss": 1.276,
+      "step": 2902
+    },
+    {
+      "epoch": 0.2547950991659147,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002765843115487589,
+      "loss": 1.2851,
+      "step": 2903
+    },
+    {
+      "epoch": 0.25488286874881716,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002765604341936562,
+      "loss": 1.3001,
+      "step": 2904
+    },
+    {
+      "epoch": 0.2549706383317197,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0027653654582798005,
+      "loss": 1.2329,
+      "step": 2905
+    },
+    {
+      "epoch": 0.25505840791462214,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0027651264645408816,
+      "loss": 1.2928,
+      "step": 2906
+    },
+    {
+      "epoch": 0.2551461774975246,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002764887360743393,
+      "loss": 1.3153,
+      "step": 2907
+    },
+    {
+      "epoch": 0.2552339470804271,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0027646481469109333,
+      "loss": 1.2155,
+      "step": 2908
+    },
+    {
+      "epoch": 0.2553217166633296,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002764408823067112,
+      "loss": 1.3383,
+      "step": 2909
+    },
+    {
+      "epoch": 0.25540948624623205,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0027641693892355495,
+      "loss": 1.3011,
+      "step": 2910
+    },
+    {
+      "epoch": 0.2554972558291346,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0027639298454398766,
+      "loss": 1.2554,
+      "step": 2911
+    },
+    {
+      "epoch": 0.25558502541203704,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.002763690191703736,
+      "loss": 1.2817,
+      "step": 2912
+    },
+    {
+      "epoch": 0.25567279499493956,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0027634504280507795,
+      "loss": 1.2574,
+      "step": 2913
+    },
+    {
+      "epoch": 0.255760564577842,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0027632105545046712,
+      "loss": 1.3318,
+      "step": 2914
+    },
+    {
+      "epoch": 0.2558483341607445,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0027629705710890866,
+      "loss": 1.3059,
+      "step": 2915
+    },
+    {
+      "epoch": 0.255936103743647,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0027627304778277107,
+      "loss": 1.3332,
+      "step": 2916
+    },
+    {
+      "epoch": 0.25602387332654947,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002762490274744239,
+      "loss": 1.2467,
+      "step": 2917
+    },
+    {
+      "epoch": 0.256111642909452,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002762249961862379,
+      "loss": 1.2977,
+      "step": 2918
+    },
+    {
+      "epoch": 0.25619941249235445,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.002762009539205849,
+      "loss": 1.2804,
+      "step": 2919
+    },
+    {
+      "epoch": 0.2562871820752569,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.002761769006798377,
+      "loss": 1.2428,
+      "step": 2920
+    },
+    {
+      "epoch": 0.25637495165815943,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0027615283646637034,
+      "loss": 1.2527,
+      "step": 2921
+    },
+    {
+      "epoch": 0.2564627212410619,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0027612876128255783,
+      "loss": 1.2966,
+      "step": 2922
+    },
+    {
+      "epoch": 0.25655049082396436,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002761046751307763,
+      "loss": 1.2712,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2566382604068669,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0027608057801340297,
+      "loss": 1.2714,
+      "step": 2924
+    },
+    {
+      "epoch": 0.25672602998976934,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002760564699328161,
+      "loss": 1.3367,
+      "step": 2925
+    },
+    {
+      "epoch": 0.25681379957267186,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0027603235089139503,
+      "loss": 1.2798,
+      "step": 2926
+    },
+    {
+      "epoch": 0.2569015691555743,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0027600822089152034,
+      "loss": 1.3074,
+      "step": 2927
+    },
+    {
+      "epoch": 0.2569893387384768,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0027598407993557343,
+      "loss": 1.267,
+      "step": 2928
+    },
+    {
+      "epoch": 0.2570771083213793,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0027595992802593697,
+      "loss": 1.2576,
+      "step": 2929
+    },
+    {
+      "epoch": 0.2571648779042818,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0027593576516499467,
+      "loss": 1.2105,
+      "step": 2930
+    },
+    {
+      "epoch": 0.2572526474871843,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0027591159135513135,
+      "loss": 1.245,
+      "step": 2931
+    },
+    {
+      "epoch": 0.25734041707008676,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0027588740659873277,
+      "loss": 1.2673,
+      "step": 2932
+    },
+    {
+      "epoch": 0.2574281866529892,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002758632108981859,
+      "loss": 1.2934,
+      "step": 2933
+    },
+    {
+      "epoch": 0.25751595623589174,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002758390042558788,
+      "loss": 1.2124,
+      "step": 2934
+    },
+    {
+      "epoch": 0.2576037258187942,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0027581478667420053,
+      "loss": 1.2355,
+      "step": 2935
+    },
+    {
+      "epoch": 0.25769149540169667,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0027579055815554134,
+      "loss": 1.2205,
+      "step": 2936
+    },
+    {
+      "epoch": 0.2577792649845992,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0027576631870229237,
+      "loss": 1.2628,
+      "step": 2937
+    },
+    {
+      "epoch": 0.25786703456750165,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0027574206831684606,
+      "loss": 1.2896,
+      "step": 2938
+    },
+    {
+      "epoch": 0.25795480415040417,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002757178070015958,
+      "loss": 1.2801,
+      "step": 2939
+    },
+    {
+      "epoch": 0.25804257373330663,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.002756935347589361,
+      "loss": 1.2961,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2581303433162091,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0027566925159126245,
+      "loss": 1.2878,
+      "step": 2941
+    },
+    {
+      "epoch": 0.2582181128991116,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.002756449575009716,
+      "loss": 1.2064,
+      "step": 2942
+    },
+    {
+      "epoch": 0.2583058824820141,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0027562065249046126,
+      "loss": 1.2946,
+      "step": 2943
+    },
+    {
+      "epoch": 0.2583936520649166,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.002755963365621301,
+      "loss": 1.3275,
+      "step": 2944
+    },
+    {
+      "epoch": 0.25848142164781907,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0027557200971837833,
+      "loss": 1.3336,
+      "step": 2945
+    },
+    {
+      "epoch": 0.25856919123072153,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0027554767196160657,
+      "loss": 1.332,
+      "step": 2946
+    },
+    {
+      "epoch": 0.25865696081362405,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0027552332329421705,
+      "loss": 1.2901,
+      "step": 2947
+    },
+    {
+      "epoch": 0.2587447303965265,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.002754989637186128,
+      "loss": 1.3124,
+      "step": 2948
+    },
+    {
+      "epoch": 0.25883249997942903,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0027547459323719816,
+      "loss": 1.3086,
+      "step": 2949
+    },
+    {
+      "epoch": 0.2589202695623315,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0027545021185237813,
+      "loss": 1.3149,
+      "step": 2950
+    },
+    {
+      "epoch": 0.25900803914523396,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0027542581956655934,
+      "loss": 1.2337,
+      "step": 2951
+    },
+    {
+      "epoch": 0.2590958087281365,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002754014163821491,
+      "loss": 1.2668,
+      "step": 2952
+    },
+    {
+      "epoch": 0.25918357831103894,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0027537700230155586,
+      "loss": 1.3238,
+      "step": 2953
+    },
+    {
+      "epoch": 0.2592713478939414,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002753525773271892,
+      "loss": 1.299,
+      "step": 2954
+    },
+    {
+      "epoch": 0.2593591174768439,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0027532814146145985,
+      "loss": 1.2783,
+      "step": 2955
+    },
+    {
+      "epoch": 0.2594468870597464,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0027530369470677944,
+      "loss": 1.299,
+      "step": 2956
+    },
+    {
+      "epoch": 0.2595346566426489,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.002752792370655608,
+      "loss": 1.306,
+      "step": 2957
+    },
+    {
+      "epoch": 0.2596224262255514,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0027525476854021787,
+      "loss": 1.2516,
+      "step": 2958
+    },
+    {
+      "epoch": 0.25971019580845384,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0027523028913316555,
+      "loss": 1.2497,
+      "step": 2959
+    },
+    {
+      "epoch": 0.25979796539135636,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002752057988468199,
+      "loss": 1.2862,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2598857349742588,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002751812976835978,
+      "loss": 1.2235,
+      "step": 2961
+    },
+    {
+      "epoch": 0.25997350455716134,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002751567856459177,
+      "loss": 1.2284,
+      "step": 2962
+    },
+    {
+      "epoch": 0.2600612741400638,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002751322627361987,
+      "loss": 1.2561,
+      "step": 2963
+    },
+    {
+      "epoch": 0.26014904372296627,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0027510772895686115,
+      "loss": 1.2054,
+      "step": 2964
+    },
+    {
+      "epoch": 0.2602368133058688,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002750831843103264,
+      "loss": 1.2635,
+      "step": 2965
+    },
+    {
+      "epoch": 0.26032458288877125,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00275058628799017,
+      "loss": 1.2572,
+      "step": 2966
+    },
+    {
+      "epoch": 0.2604123524716737,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0027503406242535638,
+      "loss": 1.2678,
+      "step": 2967
+    },
+    {
+      "epoch": 0.26050012205457623,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0027500948519176916,
+      "loss": 1.3122,
+      "step": 2968
+    },
+    {
+      "epoch": 0.2605878916374787,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00274984897100681,
+      "loss": 1.3071,
+      "step": 2969
+    },
+    {
+      "epoch": 0.2606756612203812,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002749602981545188,
+      "loss": 1.3317,
+      "step": 2970
+    },
+    {
+      "epoch": 0.2607634308032837,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0027493568835571017,
+      "loss": 1.2548,
+      "step": 2971
+    },
+    {
+      "epoch": 0.26085120038618614,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002749110677066841,
+      "loss": 1.2289,
+      "step": 2972
+    },
+    {
+      "epoch": 0.26093896996908866,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002748864362098706,
+      "loss": 1.2861,
+      "step": 2973
+    },
+    {
+      "epoch": 0.2610267395519911,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002748617938677006,
+      "loss": 1.3302,
+      "step": 2974
+    },
+    {
+      "epoch": 0.26111450913489365,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0027483714068260627,
+      "loss": 1.3258,
+      "step": 2975
+    },
+    {
+      "epoch": 0.2612022787177961,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0027481247665702073,
+      "loss": 1.2715,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2612900483006986,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0027478780179337825,
+      "loss": 1.2598,
+      "step": 2977
+    },
+    {
+      "epoch": 0.2613778178836011,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002747631160941142,
+      "loss": 1.3123,
+      "step": 2978
+    },
+    {
+      "epoch": 0.26146558746650356,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0027473841956166486,
+      "loss": 1.2426,
+      "step": 2979
+    },
+    {
+      "epoch": 0.261553357049406,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002747137121984677,
+      "loss": 1.2666,
+      "step": 2980
+    },
+    {
+      "epoch": 0.26164112663230854,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0027468899400696126,
+      "loss": 1.2684,
+      "step": 2981
+    },
+    {
+      "epoch": 0.261728896215211,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002746642649895851,
+      "loss": 1.2418,
+      "step": 2982
+    },
+    {
+      "epoch": 0.2618166657981135,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0027463952514878,
+      "loss": 1.229,
+      "step": 2983
+    },
+    {
+      "epoch": 0.261904435381016,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002746147744869876,
+      "loss": 1.3024,
+      "step": 2984
+    },
+    {
+      "epoch": 0.26199220496391845,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002745900130066506,
+      "loss": 1.2839,
+      "step": 2985
+    },
+    {
+      "epoch": 0.26207997454682097,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0027456524071021293,
+      "loss": 1.2519,
+      "step": 2986
+    },
+    {
+      "epoch": 0.26216774412972343,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0027454045760011963,
+      "loss": 1.2891,
+      "step": 2987
+    },
+    {
+      "epoch": 0.26225551371262595,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002745156636788165,
+      "loss": 1.3284,
+      "step": 2988
+    },
+    {
+      "epoch": 0.2623432832955284,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0027449085894875073,
+      "loss": 1.1907,
+      "step": 2989
+    },
+    {
+      "epoch": 0.2624310528784309,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0027446604341237043,
+      "loss": 1.2706,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2625188224613334,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0027444121707212474,
+      "loss": 1.2473,
+      "step": 2991
+    },
+    {
+      "epoch": 0.26260659204423586,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0027441637993046403,
+      "loss": 1.2635,
+      "step": 2992
+    },
+    {
+      "epoch": 0.26269436162713833,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002743915319898395,
+      "loss": 1.33,
+      "step": 2993
+    },
+    {
+      "epoch": 0.26278213121004085,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002743666732527036,
+      "loss": 1.237,
+      "step": 2994
+    },
+    {
+      "epoch": 0.2628699007929433,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0027434180372150983,
+      "loss": 1.261,
+      "step": 2995
+    },
+    {
+      "epoch": 0.26295767037584583,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0027431692339871267,
+      "loss": 1.2457,
+      "step": 2996
+    },
+    {
+      "epoch": 0.2630454399587483,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0027429203228676773,
+      "loss": 1.2615,
+      "step": 2997
+    },
+    {
+      "epoch": 0.26313320954165076,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0027426713038813163,
+      "loss": 1.2716,
+      "step": 2998
+    },
+    {
+      "epoch": 0.2632209791245533,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.002742422177052621,
+      "loss": 1.3092,
+      "step": 2999
+    },
+    {
+      "epoch": 0.26330874870745574,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002742172942406179,
+      "loss": 1.3333,
+      "step": 3000
+    },
+    {
+      "epoch": 0.26330874870745574,
+      "eval_loss": 1.2623319625854492,
+      "eval_runtime": 437.7402,
+      "eval_samples_per_second": 33.673,
+      "eval_steps_per_second": 8.418,
+      "step": 3000
+    },
+    {
+      "epoch": 0.26339651829035826,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0027419235999665897,
+      "loss": 1.3216,
+      "step": 3001
+    },
+    {
+      "epoch": 0.2634842878732607,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002741674149758461,
+      "loss": 1.2587,
+      "step": 3002
+    },
+    {
+      "epoch": 0.2635720574561632,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002741424591806414,
+      "loss": 1.2656,
+      "step": 3003
+    },
+    {
+      "epoch": 0.2636598270390657,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.002741174926135078,
+      "loss": 1.2652,
+      "step": 3004
+    },
+    {
+      "epoch": 0.26374759662196817,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.002740925152769094,
+      "loss": 1.2813,
+      "step": 3005
+    },
+    {
+      "epoch": 0.26383536620487064,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0027406752717331145,
+      "loss": 1.1798,
+      "step": 3006
+    },
+    {
+      "epoch": 0.26392313578777316,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002740425283051801,
+      "loss": 1.1798,
+      "step": 3007
+    },
+    {
+      "epoch": 0.2640109053706756,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0027401751867498266,
+      "loss": 1.3219,
+      "step": 3008
+    },
+    {
+      "epoch": 0.26409867495357814,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0027399249828518753,
+      "loss": 1.2289,
+      "step": 3009
+    },
+    {
+      "epoch": 0.2641864445364806,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00273967467138264,
+      "loss": 1.2492,
+      "step": 3010
+    },
+    {
+      "epoch": 0.26427421411938307,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.002739424252366827,
+      "loss": 1.2659,
+      "step": 3011
+    },
+    {
+      "epoch": 0.2643619837022856,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0027391737258291513,
+      "loss": 1.2757,
+      "step": 3012
+    },
+    {
+      "epoch": 0.26444975328518805,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0027389230917943376,
+      "loss": 1.2943,
+      "step": 3013
+    },
+    {
+      "epoch": 0.26453752286809057,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0027386723502871237,
+      "loss": 1.2918,
+      "step": 3014
+    },
+    {
+      "epoch": 0.26462529245099303,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0027384215013322565,
+      "loss": 1.3467,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2647130620338955,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0027381705449544937,
+      "loss": 1.1855,
+      "step": 3016
+    },
+    {
+      "epoch": 0.264800831616798,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002737919481178604,
+      "loss": 1.2759,
+      "step": 3017
+    },
+    {
+      "epoch": 0.2648886011997005,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002737668310029366,
+      "loss": 1.3151,
+      "step": 3018
+    },
+    {
+      "epoch": 0.264976370782603,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0027374170315315697,
+      "loss": 1.2375,
+      "step": 3019
+    },
+    {
+      "epoch": 0.26506414036550546,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0027371656457100147,
+      "loss": 1.3008,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2651519099484079,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0027369141525895122,
+      "loss": 1.2934,
+      "step": 3021
+    },
+    {
+      "epoch": 0.26523967953131045,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0027366625521948844,
+      "loss": 1.2363,
+      "step": 3022
+    },
+    {
+      "epoch": 0.2653274491142129,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002736410844550961,
+      "loss": 1.2215,
+      "step": 3023
+    },
+    {
+      "epoch": 0.2654152186971154,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0027361590296825873,
+      "loss": 1.2465,
+      "step": 3024
+    },
+    {
+      "epoch": 0.2655029882800179,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002735907107614614,
+      "loss": 1.2531,
+      "step": 3025
+    },
+    {
+      "epoch": 0.26559075786292036,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0027356550783719067,
+      "loss": 1.2992,
+      "step": 3026
+    },
+    {
+      "epoch": 0.2656785274458229,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0027354029419793388,
+      "loss": 1.2623,
+      "step": 3027
+    },
+    {
+      "epoch": 0.26576629702872534,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0027351506984617945,
+      "loss": 1.2807,
+      "step": 3028
+    },
+    {
+      "epoch": 0.2658540666116278,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0027348983478441705,
+      "loss": 1.2909,
+      "step": 3029
+    },
+    {
+      "epoch": 0.2659418361945303,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0027346458901513727,
+      "loss": 1.2481,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2660296057774328,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0027343933254083165,
+      "loss": 1.3352,
+      "step": 3031
+    },
+    {
+      "epoch": 0.2661173753603353,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0027341406536399303,
+      "loss": 1.3,
+      "step": 3032
+    },
+    {
+      "epoch": 0.26620514494323777,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002733887874871151,
+      "loss": 1.2858,
+      "step": 3033
+    },
+    {
+      "epoch": 0.26629291452614023,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0027336349891269275,
+      "loss": 1.2806,
+      "step": 3034
+    },
+    {
+      "epoch": 0.26638068410904275,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002733381996432218,
+      "loss": 1.3209,
+      "step": 3035
+    },
+    {
+      "epoch": 0.2664684536919452,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0027331288968119928,
+      "loss": 1.2558,
+      "step": 3036
+    },
+    {
+      "epoch": 0.2665562232748477,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0027328756902912305,
+      "loss": 1.2399,
+      "step": 3037
+    },
+    {
+      "epoch": 0.2666439928577502,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002732622376894922,
+      "loss": 1.2427,
+      "step": 3038
+    },
+    {
+      "epoch": 0.26673176244065266,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0027323689566480696,
+      "loss": 1.2434,
+      "step": 3039
+    },
+    {
+      "epoch": 0.2668195320235552,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0027321154295756834,
+      "loss": 1.2715,
+      "step": 3040
+    },
+    {
+      "epoch": 0.26690730160645765,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0027318617957027857,
+      "loss": 1.2983,
+      "step": 3041
+    },
+    {
+      "epoch": 0.2669950711893601,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00273160805505441,
+      "loss": 1.306,
+      "step": 3042
+    },
+    {
+      "epoch": 0.26708284077226263,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002731354207655599,
+      "loss": 1.2572,
+      "step": 3043
+    },
+    {
+      "epoch": 0.2671706103551651,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002731100253531406,
+      "loss": 1.3614,
+      "step": 3044
+    },
+    {
+      "epoch": 0.2672583799380676,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0027308461927068963,
+      "loss": 1.2965,
+      "step": 3045
+    },
+    {
+      "epoch": 0.2673461495209701,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0027305920252071437,
+      "loss": 1.2385,
+      "step": 3046
+    },
+    {
+      "epoch": 0.26743391910387254,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0027303377510572335,
+      "loss": 1.3075,
+      "step": 3047
+    },
+    {
+      "epoch": 0.26752168868677506,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0027300833702822627,
+      "loss": 1.299,
+      "step": 3048
+    },
+    {
+      "epoch": 0.2676094582696775,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0027298288829073368,
+      "loss": 1.2657,
+      "step": 3049
+    },
+    {
+      "epoch": 0.26769722785258,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0027295742889575725,
+      "loss": 1.2794,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2677849974354825,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0027293195884580974,
+      "loss": 1.2349,
+      "step": 3051
+    },
+    {
+      "epoch": 0.26787276701838497,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00272906478143405,
+      "loss": 1.288,
+      "step": 3052
+    },
+    {
+      "epoch": 0.2679605366012875,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002728809867910578,
+      "loss": 1.2928,
+      "step": 3053
+    },
+    {
+      "epoch": 0.26804830618418996,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002728554847912841,
+      "loss": 1.2692,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2681360757670924,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0027282997214660076,
+      "loss": 1.2758,
+      "step": 3055
+    },
+    {
+      "epoch": 0.26822384534999494,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0027280444885952587,
+      "loss": 1.303,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2683116149328974,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0027277891493257844,
+      "loss": 1.2385,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2683993845157999,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0027275337036827855,
+      "loss": 1.2298,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2684871540987024,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0027272781516914733,
+      "loss": 1.3027,
+      "step": 3059
+    },
+    {
+      "epoch": 0.26857492368160485,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002727022493377071,
+      "loss": 1.3254,
+      "step": 3060
+    },
+    {
+      "epoch": 0.26866269326450737,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0027267667287648096,
+      "loss": 1.2489,
+      "step": 3061
+    },
+    {
+      "epoch": 0.26875046284740983,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0027265108578799326,
+      "loss": 1.2781,
+      "step": 3062
+    },
+    {
+      "epoch": 0.2688382324303123,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.002726254880747694,
+      "loss": 1.2501,
+      "step": 3063
+    },
+    {
+      "epoch": 0.2689260020132148,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0027259987973933565,
+      "loss": 1.2445,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2690137715961173,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0027257426078421956,
+      "loss": 1.3197,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2691015411790198,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0027254863121194957,
+      "loss": 1.2904,
+      "step": 3066
+    },
+    {
+      "epoch": 0.26918931076192226,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0027252299102505525,
+      "loss": 1.2786,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2692770803448247,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002724973402260671,
+      "loss": 1.2729,
+      "step": 3068
+    },
+    {
+      "epoch": 0.26936484992772725,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0027247167881751687,
+      "loss": 1.2518,
+      "step": 3069
+    },
+    {
+      "epoch": 0.2694526195106297,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.002724460068019371,
+      "loss": 1.3055,
+      "step": 3070
+    },
+    {
+      "epoch": 0.26954038909353223,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0027242032418186173,
+      "loss": 1.2591,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2696281586764347,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0027239463095982534,
+      "loss": 1.2607,
+      "step": 3072
+    },
+    {
+      "epoch": 0.26971592825933716,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002723689271383638,
+      "loss": 1.3159,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2698036978422397,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00272343212720014,
+      "loss": 1.2002,
+      "step": 3074
+    },
+    {
+      "epoch": 0.26989146742514214,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0027231748770731386,
+      "loss": 1.316,
+      "step": 3075
+    },
+    {
+      "epoch": 0.2699792370080446,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002722917521028023,
+      "loss": 1.2967,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2700670065909471,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0027226600590901933,
+      "loss": 1.3347,
+      "step": 3077
+    },
+    {
+      "epoch": 0.2701547761738496,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0027224024912850604,
+      "loss": 1.3171,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2702425457567521,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0027221448176380446,
+      "loss": 1.299,
+      "step": 3079
+    },
+    {
+      "epoch": 0.27033031533965457,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0027218870381745777,
+      "loss": 1.2303,
+      "step": 3080
+    },
+    {
+      "epoch": 0.27041808492255703,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002721629152920101,
+      "loss": 1.332,
+      "step": 3081
+    },
+    {
+      "epoch": 0.27050585450545955,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0027213711619000678,
+      "loss": 1.221,
+      "step": 3082
+    },
+    {
+      "epoch": 0.270593624088362,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0027211130651399397,
+      "loss": 1.2766,
+      "step": 3083
+    },
+    {
+      "epoch": 0.27068139367126454,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0027208548626651906,
+      "loss": 1.3172,
+      "step": 3084
+    },
+    {
+      "epoch": 0.270769163254167,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.002720596554501303,
+      "loss": 1.2946,
+      "step": 3085
+    },
+    {
+      "epoch": 0.27085693283706946,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0027203381406737726,
+      "loss": 1.2846,
+      "step": 3086
+    },
+    {
+      "epoch": 0.270944702419972,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002720079621208102,
+      "loss": 1.3227,
+      "step": 3087
+    },
+    {
+      "epoch": 0.27103247200287445,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0027198209961298075,
+      "loss": 1.2751,
+      "step": 3088
+    },
+    {
+      "epoch": 0.27112024158577697,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002719562265464413,
+      "loss": 1.2337,
+      "step": 3089
+    },
+    {
+      "epoch": 0.27120801116867943,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0027193034292374556,
+      "loss": 1.3,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2712957807515819,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0027190444874744804,
+      "loss": 1.2456,
+      "step": 3091
+    },
+    {
+      "epoch": 0.2713835503344844,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002718785440201044,
+      "loss": 1.2532,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2714713199173869,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002718526287442715,
+      "loss": 1.249,
+      "step": 3093
+    },
+    {
+      "epoch": 0.27155908950028934,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002718267029225068,
+      "loss": 1.2369,
+      "step": 3094
+    },
+    {
+      "epoch": 0.27164685908319186,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0027180076655736927,
+      "loss": 1.3229,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2717346286660943,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0027177481965141863,
+      "loss": 1.2477,
+      "step": 3096
+    },
+    {
+      "epoch": 0.27182239824899684,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002717488622072158,
+      "loss": 1.2562,
+      "step": 3097
+    },
+    {
+      "epoch": 0.2719101678318993,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0027172289422732264,
+      "loss": 1.2525,
+      "step": 3098
+    },
+    {
+      "epoch": 0.27199793741480177,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0027169691571430213,
+      "loss": 1.3141,
+      "step": 3099
+    },
+    {
+      "epoch": 0.2720857069977043,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0027167092667071823,
+      "loss": 1.2962,
+      "step": 3100
+    },
+    {
+      "epoch": 0.27217347658060675,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0027164492709913592,
+      "loss": 1.2762,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2722612461635093,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0027161891700212123,
+      "loss": 1.243,
+      "step": 3102
+    },
+    {
+      "epoch": 0.27234901574641174,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0027159289638224136,
+      "loss": 1.2123,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2724367853293142,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002715668652420644,
+      "loss": 1.3245,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2725245549122167,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0027154082358415944,
+      "loss": 1.2756,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2726123244951192,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002715147714110968,
+      "loss": 1.3757,
+      "step": 3106
+    },
+    {
+      "epoch": 0.27270009407802165,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0027148870872544767,
+      "loss": 1.2723,
+      "step": 3107
+    },
+    {
+      "epoch": 0.27278786366092417,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002714626355297843,
+      "loss": 1.2623,
+      "step": 3108
+    },
+    {
+      "epoch": 0.27287563324382663,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0027143655182668013,
+      "loss": 1.3115,
+      "step": 3109
+    },
+    {
+      "epoch": 0.27296340282672915,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0027141045761870943,
+      "loss": 1.2493,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2730511724096316,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0027138435290844755,
+      "loss": 1.2701,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2731389419925341,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0027135823769847103,
+      "loss": 1.274,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2732267115754366,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002713321119913573,
+      "loss": 1.2382,
+      "step": 3113
+    },
+    {
+      "epoch": 0.27331448115833906,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002713059757896848,
+      "loss": 1.3094,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2734022507412416,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0027127982909603316,
+      "loss": 1.2698,
+      "step": 3115
+    },
+    {
+      "epoch": 0.27349002032414405,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.002712536719129829,
+      "loss": 1.2649,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2735777899070465,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0027122750424311563,
+      "loss": 1.2478,
+      "step": 3117
+    },
+    {
+      "epoch": 0.27366555948994903,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0027120132608901403,
+      "loss": 1.2676,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2737533290728515,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0027117513745326177,
+      "loss": 1.2763,
+      "step": 3119
+    },
+    {
+      "epoch": 0.27384109865575396,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0027114893833844354,
+      "loss": 1.2725,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2739288682386565,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0027112272874714512,
+      "loss": 1.2625,
+      "step": 3121
+    },
+    {
+      "epoch": 0.27401663782155894,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002710965086819533,
+      "loss": 1.2014,
+      "step": 3122
+    },
+    {
+      "epoch": 0.27410440740446146,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0027107027814545588,
+      "loss": 1.2619,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2741921769873639,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002710440371402416,
+      "loss": 1.2834,
+      "step": 3124
+    },
+    {
+      "epoch": 0.2742799465702664,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0027101778566890056,
+      "loss": 1.2103,
+      "step": 3125
+    },
+    {
+      "epoch": 0.2743677161531689,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0027099152373402354,
+      "loss": 1.2863,
+      "step": 3126
+    },
+    {
+      "epoch": 0.27445548573607137,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.002709652513382025,
+      "loss": 1.2799,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2745432553189739,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0027093896848403047,
+      "loss": 1.3309,
+      "step": 3128
+    },
+    {
+      "epoch": 0.27463102490187635,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0027091267517410137,
+      "loss": 1.2526,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2747187944847788,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0027088637141101037,
+      "loss": 1.2656,
+      "step": 3130
+    },
+    {
+      "epoch": 0.27480656406768134,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0027086005719735348,
+      "loss": 1.2612,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2748943336505838,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0027083373253572775,
+      "loss": 1.318,
+      "step": 3132
+    },
+    {
+      "epoch": 0.27498210323348626,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002708073974287314,
+      "loss": 1.2395,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2750698728163888,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002707810518789636,
+      "loss": 1.2935,
+      "step": 3134
+    },
+    {
+      "epoch": 0.27515764239929125,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0027075469588902458,
+      "loss": 1.2433,
+      "step": 3135
+    },
+    {
+      "epoch": 0.27524541198219377,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0027072832946151548,
+      "loss": 1.3055,
+      "step": 3136
+    },
+    {
+      "epoch": 0.27533318156509623,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0027070195259903855,
+      "loss": 1.2839,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2754209511479987,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0027067556530419724,
+      "loss": 1.2903,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2755087207309012,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002706491675795957,
+      "loss": 1.2727,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2755964903138037,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0027062275942783937,
+      "loss": 1.3098,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2756842598967062,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002705963408515347,
+      "loss": 1.2358,
+      "step": 3141
+    },
+    {
+      "epoch": 0.27577202947960866,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002705699118532889,
+      "loss": 1.2854,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2758597990625111,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0027054347243571056,
+      "loss": 1.3004,
+      "step": 3143
+    },
+    {
+      "epoch": 0.27594756864541364,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002705170226014091,
+      "loss": 1.2557,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2760353382283161,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00270490562352995,
+      "loss": 1.343,
+      "step": 3145
+    },
+    {
+      "epoch": 0.27612310781121857,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002704640916930798,
+      "loss": 1.332,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2762108773941211,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0027043761062427608,
+      "loss": 1.2694,
+      "step": 3147
+    },
+    {
+      "epoch": 0.27629864697702355,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002704111191491974,
+      "loss": 1.225,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2763864165599261,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002703846172704583,
+      "loss": 1.2331,
+      "step": 3149
+    },
+    {
+      "epoch": 0.27647418614282854,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002703581049906745,
+      "loss": 1.2785,
+      "step": 3150
+    },
+    {
+      "epoch": 0.276561955725731,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002703315823124626,
+      "loss": 1.3095,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2766497253086335,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002703050492384404,
+      "loss": 1.2248,
+      "step": 3152
+    },
+    {
+      "epoch": 0.276737494891536,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002702785057712264,
+      "loss": 1.3222,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2768252644744385,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002702519519134404,
+      "loss": 1.3349,
+      "step": 3154
+    },
+    {
+      "epoch": 0.27691303405734097,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0027022538766770334,
+      "loss": 1.2534,
+      "step": 3155
+    },
+    {
+      "epoch": 0.27700080364024343,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002701988130366368,
+      "loss": 1.2315,
+      "step": 3156
+    },
+    {
+      "epoch": 0.27708857322314595,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0027017222802286365,
+      "loss": 1.2524,
+      "step": 3157
+    },
+    {
+      "epoch": 0.2771763428060484,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002701456326290078,
+      "loss": 1.2928,
+      "step": 3158
+    },
+    {
+      "epoch": 0.27726411238895093,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0027011902685769398,
+      "loss": 1.2641,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2773518819718534,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002700924107115482,
+      "loss": 1.2903,
+      "step": 3160
+    },
+    {
+      "epoch": 0.27743965155475586,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0027006578419319725,
+      "loss": 1.3015,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2775274211376584,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0027003914730526915,
+      "loss": 1.272,
+      "step": 3162
+    },
+    {
+      "epoch": 0.27761519072056084,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0027001250005039288,
+      "loss": 1.2366,
+      "step": 3163
+    },
+    {
+      "epoch": 0.2777029603034633,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002699858424311983,
+      "loss": 1.2801,
+      "step": 3164
+    },
+    {
+      "epoch": 0.27779072988636583,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.002699591744503165,
+      "loss": 1.2869,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2778784994692683,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0026993249611037954,
+      "loss": 1.2579,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2779662690521708,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002699058074140204,
+      "loss": 1.2847,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2780540386350733,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002698791083638731,
+      "loss": 1.2227,
+      "step": 3168
+    },
+    {
+      "epoch": 0.27814180821797574,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0026985239896257277,
+      "loss": 1.3268,
+      "step": 3169
+    },
+    {
+      "epoch": 0.27822957780087826,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002698256792127556,
+      "loss": 1.2644,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2783173473837807,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.002697989491170587,
+      "loss": 1.2894,
+      "step": 3171
+    },
+    {
+      "epoch": 0.27840511696668324,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.002697722086781201,
+      "loss": 1.2656,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2784928865495857,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0026974545789857913,
+      "loss": 1.2645,
+      "step": 3173
+    },
+    {
+      "epoch": 0.27858065613248817,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0026971869678107585,
+      "loss": 1.2479,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2786684257153907,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0026969192532825164,
+      "loss": 1.2105,
+      "step": 3175
+    },
+    {
+      "epoch": 0.27875619529829315,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002696651435427486,
+      "loss": 1.2568,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2788439648811956,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0026963835142721007,
+      "loss": 1.2287,
+      "step": 3177
+    },
+    {
+      "epoch": 0.27893173446409814,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002696115489842803,
+      "loss": 1.2107,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2790195040470006,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0026958473621660445,
+      "loss": 1.2313,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2791072736299031,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0026955791312682905,
+      "loss": 1.228,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2791950432128056,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0026953107971760134,
+      "loss": 1.2692,
+      "step": 3181
+    },
+    {
+      "epoch": 0.27928281279570805,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0026950423599156966,
+      "loss": 1.3474,
+      "step": 3182
+    },
+    {
+      "epoch": 0.27937058237861057,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0026947738195138337,
+      "loss": 1.2893,
+      "step": 3183
+    },
+    {
+      "epoch": 0.27945835196151303,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0026945051759969295,
+      "loss": 1.299,
+      "step": 3184
+    },
+    {
+      "epoch": 0.27954612154441555,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0026942364293914963,
+      "loss": 1.2903,
+      "step": 3185
+    },
+    {
+      "epoch": 0.279633891127318,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00269396757972406,
+      "loss": 1.2591,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2797216607102205,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002693698627021154,
+      "loss": 1.2394,
+      "step": 3187
+    },
+    {
+      "epoch": 0.279809430293123,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.002693429571309324,
+      "loss": 1.2616,
+      "step": 3188
+    },
+    {
+      "epoch": 0.27989719987602546,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002693160412615123,
+      "loss": 1.2913,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2799849694589279,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002692891150965117,
+      "loss": 1.2628,
+      "step": 3190
+    },
+    {
+      "epoch": 0.28007273904183044,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0026926217863858814,
+      "loss": 1.2134,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2801605086247329,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0026923523189040003,
+      "loss": 1.2175,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2802482782076354,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0026920827485460703,
+      "loss": 1.2723,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2803360477905379,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0026918130753386957,
+      "loss": 1.3215,
+      "step": 3194
+    },
+    {
+      "epoch": 0.28042381737344035,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0026915432993084936,
+      "loss": 1.1965,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2805115869563429,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0026912734204820885,
+      "loss": 1.2367,
+      "step": 3196
+    },
+    {
+      "epoch": 0.28059935653924534,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.002691003438886117,
+      "loss": 1.2584,
+      "step": 3197
+    },
+    {
+      "epoch": 0.28068712612214786,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002690733354547225,
+      "loss": 1.2969,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2807748957050503,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0026904631674920694,
+      "loss": 1.2951,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2808626652879528,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0026901928777473155,
+      "loss": 1.2894,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2809504348708553,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002689922485339641,
+      "loss": 1.2851,
+      "step": 3201
+    },
+    {
+      "epoch": 0.28103820445375777,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002689651990295731,
+      "loss": 1.3346,
+      "step": 3202
+    },
+    {
+      "epoch": 0.28112597403666023,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0026893813926422844,
+      "loss": 1.3104,
+      "step": 3203
+    },
+    {
+      "epoch": 0.28121374361956275,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0026891106924060064,
+      "loss": 1.2237,
+      "step": 3204
+    },
+    {
+      "epoch": 0.2813015132024652,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0026888398896136143,
+      "loss": 1.2676,
+      "step": 3205
+    },
+    {
+      "epoch": 0.28138928278536773,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002688568984291836,
+      "loss": 1.2077,
+      "step": 3206
+    },
+    {
+      "epoch": 0.2814770523682702,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0026882979764674083,
+      "loss": 1.3397,
+      "step": 3207
+    },
+    {
+      "epoch": 0.28156482195117266,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0026880268661670786,
+      "loss": 1.2755,
+      "step": 3208
+    },
+    {
+      "epoch": 0.2816525915340752,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0026877556534176043,
+      "loss": 1.2845,
+      "step": 3209
+    },
+    {
+      "epoch": 0.28174036111697764,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0026874843382457534,
+      "loss": 1.3066,
+      "step": 3210
+    },
+    {
+      "epoch": 0.28182813069988016,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0026872129206783036,
+      "loss": 1.2582,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2819159002827826,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0026869414007420424,
+      "loss": 1.2557,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2820036698656851,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.002686669778463768,
+      "loss": 1.3134,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2820914394485876,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0026863980538702884,
+      "loss": 1.2484,
+      "step": 3214
+    },
+    {
+      "epoch": 0.2821792090314901,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002686126226988421,
+      "loss": 1.2878,
+      "step": 3215
+    },
+    {
+      "epoch": 0.28226697861439254,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0026858542978449957,
+      "loss": 1.3048,
+      "step": 3216
+    },
+    {
+      "epoch": 0.28235474819729506,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0026855822664668496,
+      "loss": 1.2652,
+      "step": 3217
+    },
+    {
+      "epoch": 0.2824425177801975,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0026853101328808316,
+      "loss": 1.2367,
+      "step": 3218
+    },
+    {
+      "epoch": 0.28253028736310004,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0026850378971137994,
+      "loss": 1.2976,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2826180569460025,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0026847655591926223,
+      "loss": 1.2665,
+      "step": 3220
+    },
+    {
+      "epoch": 0.28270582652890497,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002684493119144179,
+      "loss": 1.2501,
+      "step": 3221
+    },
+    {
+      "epoch": 0.2827935961118075,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0026842205769953584,
+      "loss": 1.3064,
+      "step": 3222
+    },
+    {
+      "epoch": 0.28288136569470995,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0026839479327730588,
+      "loss": 1.2508,
+      "step": 3223
+    },
+    {
+      "epoch": 0.28296913527761247,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0026836751865041896,
+      "loss": 1.3441,
+      "step": 3224
+    },
+    {
+      "epoch": 0.28305690486051494,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002683402338215669,
+      "loss": 1.2478,
+      "step": 3225
+    },
+    {
+      "epoch": 0.2831446744434174,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002683129387934427,
+      "loss": 1.2065,
+      "step": 3226
+    },
+    {
+      "epoch": 0.2832324440263199,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002682856335687402,
+      "loss": 1.2661,
+      "step": 3227
+    },
+    {
+      "epoch": 0.2833202136092224,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002682583181501544,
+      "loss": 1.2668,
+      "step": 3228
+    },
+    {
+      "epoch": 0.2834079831921249,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.002682309925403811,
+      "loss": 1.2687,
+      "step": 3229
+    },
+    {
+      "epoch": 0.28349575277502737,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002682036567421174,
+      "loss": 1.2489,
+      "step": 3230
+    },
+    {
+      "epoch": 0.28358352235792983,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.002681763107580611,
+      "loss": 1.2499,
+      "step": 3231
+    },
+    {
+      "epoch": 0.28367129194083235,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002681489545909111,
+      "loss": 1.2641,
+      "step": 3232
+    },
+    {
+      "epoch": 0.2837590615237348,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0026812158824336746,
+      "loss": 1.3063,
+      "step": 3233
+    },
+    {
+      "epoch": 0.2838468311066373,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0026809421171813115,
+      "loss": 1.318,
+      "step": 3234
+    },
+    {
+      "epoch": 0.2839346006895398,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0026806682501790404,
+      "loss": 1.3138,
+      "step": 3235
+    },
+    {
+      "epoch": 0.28402237027244226,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0026803942814538906,
+      "loss": 1.3088,
+      "step": 3236
+    },
+    {
+      "epoch": 0.2841101398553448,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0026801202110329028,
+      "loss": 1.2563,
+      "step": 3237
+    },
+    {
+      "epoch": 0.28419790943824724,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0026798460389431257,
+      "loss": 1.2921,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2842856790211497,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00267957176521162,
+      "loss": 1.2426,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2843734486040522,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0026792973898654548,
+      "loss": 1.306,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2844612181869547,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00267902291293171,
+      "loss": 1.2449,
+      "step": 3241
+    },
+    {
+      "epoch": 0.2845489877698572,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0026787483344374746,
+      "loss": 1.2781,
+      "step": 3242
+    },
+    {
+      "epoch": 0.2846367573527597,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00267847365440985,
+      "loss": 1.2482,
+      "step": 3243
+    },
+    {
+      "epoch": 0.28472452693566214,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002678198872875945,
+      "loss": 1.3168,
+      "step": 3244
+    },
+    {
+      "epoch": 0.28481229651856466,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0026779239898628796,
+      "loss": 1.2142,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2849000661014671,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.002677649005397783,
+      "loss": 1.2748,
+      "step": 3246
+    },
+    {
+      "epoch": 0.2849878356843696,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0026773739195077965,
+      "loss": 1.2776,
+      "step": 3247
+    },
+    {
+      "epoch": 0.2850756052672721,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0026770987322200695,
+      "loss": 1.2221,
+      "step": 3248
+    },
+    {
+      "epoch": 0.28516337485017457,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002676823443561761,
+      "loss": 1.2471,
+      "step": 3249
+    },
+    {
+      "epoch": 0.2852511444330771,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0026765480535600418,
+      "loss": 1.2988,
+      "step": 3250
+    },
+    {
+      "epoch": 0.28533891401597955,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002676272562242092,
+      "loss": 1.2773,
+      "step": 3251
+    },
+    {
+      "epoch": 0.285426683598882,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0026759969696351004,
+      "loss": 1.2824,
+      "step": 3252
+    },
+    {
+      "epoch": 0.28551445318178453,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0026757212757662677,
+      "loss": 1.2609,
+      "step": 3253
+    },
+    {
+      "epoch": 0.285602222764687,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002675445480662804,
+      "loss": 1.2466,
+      "step": 3254
+    },
+    {
+      "epoch": 0.2856899923475895,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002675169584351928,
+      "loss": 1.2627,
+      "step": 3255
+    },
+    {
+      "epoch": 0.285777761930492,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026748935868608713,
+      "loss": 1.2983,
+      "step": 3256
+    },
+    {
+      "epoch": 0.28586553151339444,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002674617488216873,
+      "loss": 1.2704,
+      "step": 3257
+    },
+    {
+      "epoch": 0.28595330109629696,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.002674341288447182,
+      "loss": 1.2435,
+      "step": 3258
+    },
+    {
+      "epoch": 0.2860410706791994,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0026740649875790596,
+      "loss": 1.2771,
+      "step": 3259
+    },
+    {
+      "epoch": 0.2861288402621019,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002673788585639775,
+      "loss": 1.2363,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2862166098450044,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002673512082656607,
+      "loss": 1.2083,
+      "step": 3261
+    },
+    {
+      "epoch": 0.2863043794279069,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002673235478656847,
+      "loss": 1.3354,
+      "step": 3262
+    },
+    {
+      "epoch": 0.2863921490108094,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002672958773667794,
+      "loss": 1.278,
+      "step": 3263
+    },
+    {
+      "epoch": 0.28647991859371186,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002672681967716758,
+      "loss": 1.2476,
+      "step": 3264
+    },
+    {
+      "epoch": 0.2865676881766143,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0026724050608310573,
+      "loss": 1.3382,
+      "step": 3265
+    },
+    {
+      "epoch": 0.28665545775951684,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0026721280530380234,
+      "loss": 1.308,
+      "step": 3266
+    },
+    {
+      "epoch": 0.2867432273424193,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002671850944364995,
+      "loss": 1.3281,
+      "step": 3267
+    },
+    {
+      "epoch": 0.2868309969253218,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0026715737348393213,
+      "loss": 1.2219,
+      "step": 3268
+    },
+    {
+      "epoch": 0.2869187665082243,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002671296424488362,
+      "loss": 1.2588,
+      "step": 3269
+    },
+    {
+      "epoch": 0.28700653609112675,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0026710190133394872,
+      "loss": 1.1849,
+      "step": 3270
+    },
+    {
+      "epoch": 0.28709430567402927,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0026707415014200755,
+      "loss": 1.292,
+      "step": 3271
+    },
+    {
+      "epoch": 0.28718207525693173,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.002670463888757517,
+      "loss": 1.2453,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2872698448398342,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.002670186175379209,
+      "loss": 1.2285,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2873576144227367,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0026699083613125636,
+      "loss": 1.2668,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2874453840056392,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002669630446584998,
+      "loss": 1.2596,
+      "step": 3275
+    },
+    {
+      "epoch": 0.2875331535885417,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002669352431223942,
+      "loss": 1.272,
+      "step": 3276
+    },
+    {
+      "epoch": 0.28762092317144416,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0026690743152568337,
+      "loss": 1.1951,
+      "step": 3277
+    },
+    {
+      "epoch": 0.28770869275434663,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0026687960987111237,
+      "loss": 1.2812,
+      "step": 3278
+    },
+    {
+      "epoch": 0.28779646233724915,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0026685177816142694,
+      "loss": 1.2518,
+      "step": 3279
+    },
+    {
+      "epoch": 0.2878842319201516,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0026682393639937403,
+      "loss": 1.2975,
+      "step": 3280
+    },
+    {
+      "epoch": 0.28797200150305413,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0026679608458770147,
+      "loss": 1.1963,
+      "step": 3281
+    },
+    {
+      "epoch": 0.2880597710859566,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0026676822272915816,
+      "loss": 1.2776,
+      "step": 3282
+    },
+    {
+      "epoch": 0.28814754066885906,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0026674035082649397,
+      "loss": 1.206,
+      "step": 3283
+    },
+    {
+      "epoch": 0.2882353102517616,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0026671246888245966,
+      "loss": 1.2309,
+      "step": 3284
+    },
+    {
+      "epoch": 0.28832307983466404,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.002666845768998072,
+      "loss": 1.2657,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2884108494175665,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002666566748812893,
+      "loss": 1.2762,
+      "step": 3286
+    },
+    {
+      "epoch": 0.288498619000469,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0026662876282965986,
+      "loss": 1.1997,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2885863885833715,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0026660084074767365,
+      "loss": 1.2974,
+      "step": 3288
+    },
+    {
+      "epoch": 0.288674158166274,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0026657290863808646,
+      "loss": 1.2317,
+      "step": 3289
+    },
+    {
+      "epoch": 0.2887619277491765,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0026654496650365504,
+      "loss": 1.2428,
+      "step": 3290
+    },
+    {
+      "epoch": 0.28884969733207894,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002665170143471373,
+      "loss": 1.2552,
+      "step": 3291
+    },
+    {
+      "epoch": 0.28893746691498146,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0026648905217129186,
+      "loss": 1.2861,
+      "step": 3292
+    },
+    {
+      "epoch": 0.2890252364978839,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0026646107997887855,
+      "loss": 1.2968,
+      "step": 3293
+    },
+    {
+      "epoch": 0.28911300608078644,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0026643309777265815,
+      "loss": 1.3059,
+      "step": 3294
+    },
+    {
+      "epoch": 0.2892007756636889,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002664051055553923,
+      "loss": 1.3011,
+      "step": 3295
+    },
+    {
+      "epoch": 0.28928854524659137,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0026637710332984388,
+      "loss": 1.2771,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2893763148294939,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0026634909109877637,
+      "loss": 1.3054,
+      "step": 3297
+    },
+    {
+      "epoch": 0.28946408441239635,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002663210688649546,
+      "loss": 1.2648,
+      "step": 3298
+    },
+    {
+      "epoch": 0.28955185399529887,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002662930366311443,
+      "loss": 1.2607,
+      "step": 3299
+    },
+    {
+      "epoch": 0.28963962357820133,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0026626499440011203,
+      "loss": 1.2232,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2897273931611038,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002662369421746255,
+      "loss": 1.2023,
+      "step": 3301
+    },
+    {
+      "epoch": 0.2898151627440063,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.002662088799574534,
+      "loss": 1.2609,
+      "step": 3302
+    },
+    {
+      "epoch": 0.2899029323269088,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002661808077513653,
+      "loss": 1.2576,
+      "step": 3303
+    },
+    {
+      "epoch": 0.28999070190981124,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0026615272555913184,
+      "loss": 1.2719,
+      "step": 3304
+    },
+    {
+      "epoch": 0.29007847149271376,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.002661246333835246,
+      "loss": 1.263,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2901662410756162,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002660965312273161,
+      "loss": 1.2305,
+      "step": 3306
+    },
+    {
+      "epoch": 0.29025401065851875,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0026606841909328,
+      "loss": 1.2543,
+      "step": 3307
+    },
+    {
+      "epoch": 0.2903417802414212,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0026604029698419087,
+      "loss": 1.2504,
+      "step": 3308
+    },
+    {
+      "epoch": 0.2904295498243237,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0026601216490282427,
+      "loss": 1.2578,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2905173194072262,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0026598402285195662,
+      "loss": 1.2319,
+      "step": 3310
+    },
+    {
+      "epoch": 0.29060508899012866,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0026595587083436552,
+      "loss": 1.2379,
+      "step": 3311
+    },
+    {
+      "epoch": 0.2906928585730312,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0026592770885282943,
+      "loss": 1.229,
+      "step": 3312
+    },
+    {
+      "epoch": 0.29078062815593364,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002658995369101278,
+      "loss": 1.2887,
+      "step": 3313
+    },
+    {
+      "epoch": 0.2908683977388361,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0026587135500904116,
+      "loss": 1.2987,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2909561673217386,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002658431631523509,
+      "loss": 1.2273,
+      "step": 3315
+    },
+    {
+      "epoch": 0.2910439369046411,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0026581496134283943,
+      "loss": 1.2565,
+      "step": 3316
+    },
+    {
+      "epoch": 0.29113170648754355,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002657867495832902,
+      "loss": 1.2901,
+      "step": 3317
+    },
+    {
+      "epoch": 0.29121947607044607,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002657585278764876,
+      "loss": 1.2591,
+      "step": 3318
+    },
+    {
+      "epoch": 0.29130724565334853,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002657302962252169,
+      "loss": 1.2941,
+      "step": 3319
+    },
+    {
+      "epoch": 0.29139501523625105,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.002657020546322646,
+      "loss": 1.3409,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2914827848191535,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0026567380310041793,
+      "loss": 1.2749,
+      "step": 3321
+    },
+    {
+      "epoch": 0.291570554402056,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0026564554163246525,
+      "loss": 1.2671,
+      "step": 3322
+    },
+    {
+      "epoch": 0.2916583239849585,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0026561727023119586,
+      "loss": 1.2847,
+      "step": 3323
+    },
+    {
+      "epoch": 0.29174609356786096,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002655889888994,
+      "loss": 1.2533,
+      "step": 3324
+    },
+    {
+      "epoch": 0.2918338631507635,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00265560697639869,
+      "loss": 1.3036,
+      "step": 3325
+    },
+    {
+      "epoch": 0.29192163273366595,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026553239645539495,
+      "loss": 1.2192,
+      "step": 3326
+    },
+    {
+      "epoch": 0.2920094023165684,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0026550408534877124,
+      "loss": 1.2501,
+      "step": 3327
+    },
+    {
+      "epoch": 0.29209717189947093,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0026547576432279195,
+      "loss": 1.2671,
+      "step": 3328
+    },
+    {
+      "epoch": 0.2921849414823734,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002654474333802523,
+      "loss": 1.2793,
+      "step": 3329
+    },
+    {
+      "epoch": 0.29227271106527586,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002654190925239484,
+      "loss": 1.2288,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2923604806481784,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002653907417566774,
+      "loss": 1.2461,
+      "step": 3331
+    },
+    {
+      "epoch": 0.29244825023108084,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002653623810812374,
+      "loss": 1.2731,
+      "step": 3332
+    },
+    {
+      "epoch": 0.29253601981398336,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0026533401050042754,
+      "loss": 1.2592,
+      "step": 3333
+    },
+    {
+      "epoch": 0.2926237893968858,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0026530563001704787,
+      "loss": 1.2732,
+      "step": 3334
+    },
+    {
+      "epoch": 0.2927115589797883,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0026527723963389936,
+      "loss": 1.2814,
+      "step": 3335
+    },
+    {
+      "epoch": 0.2927993285626908,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002652488393537841,
+      "loss": 1.2276,
+      "step": 3336
+    },
+    {
+      "epoch": 0.29288709814559327,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0026522042917950504,
+      "loss": 1.2368,
+      "step": 3337
+    },
+    {
+      "epoch": 0.2929748677284958,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002651920091138662,
+      "loss": 1.2755,
+      "step": 3338
+    },
+    {
+      "epoch": 0.29306263731139826,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0026516357915967246,
+      "loss": 1.2489,
+      "step": 3339
+    },
+    {
+      "epoch": 0.2931504068943007,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002651351393197298,
+      "loss": 1.2415,
+      "step": 3340
+    },
+    {
+      "epoch": 0.29323817647720324,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002651066895968451,
+      "loss": 1.2498,
+      "step": 3341
+    },
+    {
+      "epoch": 0.2933259460601057,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002650782299938263,
+      "loss": 1.2397,
+      "step": 3342
+    },
+    {
+      "epoch": 0.29341371564300817,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002650497605134821,
+      "loss": 1.2603,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2935014852259107,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0026502128115862243,
+      "loss": 1.3261,
+      "step": 3344
+    },
+    {
+      "epoch": 0.29358925480881315,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0026499279193205813,
+      "loss": 1.3283,
+      "step": 3345
+    },
+    {
+      "epoch": 0.29367702439171567,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002649642928366009,
+      "loss": 1.2498,
+      "step": 3346
+    },
+    {
+      "epoch": 0.29376479397461813,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0026493578387506342,
+      "loss": 1.2134,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2938525635575206,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002649072650502596,
+      "loss": 1.2576,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2939403331404231,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00264878736365004,
+      "loss": 1.2196,
+      "step": 3349
+    },
+    {
+      "epoch": 0.2940281027233256,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002648501978221123,
+      "loss": 1.2563,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2941158723062281,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002648216494244012,
+      "loss": 1.3172,
+      "step": 3351
+    },
+    {
+      "epoch": 0.29420364188913056,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0026479309117468826,
+      "loss": 1.2335,
+      "step": 3352
+    },
+    {
+      "epoch": 0.294291411472033,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0026476452307579207,
+      "loss": 1.3113,
+      "step": 3353
+    },
+    {
+      "epoch": 0.29437918105493555,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0026473594513053223,
+      "loss": 1.2783,
+      "step": 3354
+    },
+    {
+      "epoch": 0.294466950637838,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0026470735734172926,
+      "loss": 1.2683,
+      "step": 3355
+    },
+    {
+      "epoch": 0.2945547202207405,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0026467875971220452,
+      "loss": 1.2716,
+      "step": 3356
+    },
+    {
+      "epoch": 0.294642489803643,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.002646501522447807,
+      "loss": 1.3174,
+      "step": 3357
+    },
+    {
+      "epoch": 0.29473025938654546,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0026462153494228114,
+      "loss": 1.21,
+      "step": 3358
+    },
+    {
+      "epoch": 0.294818028969448,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.002645929078075303,
+      "loss": 1.2611,
+      "step": 3359
+    },
+    {
+      "epoch": 0.29490579855235044,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0026456427084335346,
+      "loss": 1.2069,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2949935681352529,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002645356240525771,
+      "loss": 1.3464,
+      "step": 3361
+    },
+    {
+      "epoch": 0.2950813377181554,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0026450696743802846,
+      "loss": 1.2969,
+      "step": 3362
+    },
+    {
+      "epoch": 0.2951691073010579,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0026447830100253585,
+      "loss": 1.2954,
+      "step": 3363
+    },
+    {
+      "epoch": 0.2952568768839604,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0026444962474892857,
+      "loss": 1.301,
+      "step": 3364
+    },
+    {
+      "epoch": 0.29534464646686287,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0026442093868003684,
+      "loss": 1.2894,
+      "step": 3365
+    },
+    {
+      "epoch": 0.29543241604976533,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0026439224279869184,
+      "loss": 1.313,
+      "step": 3366
+    },
+    {
+      "epoch": 0.29552018563266785,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002643635371077257,
+      "loss": 1.2582,
+      "step": 3367
+    },
+    {
+      "epoch": 0.2956079552155703,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.002643348216099717,
+      "loss": 1.2957,
+      "step": 3368
+    },
+    {
+      "epoch": 0.29569572479847284,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002643060963082638,
+      "loss": 1.2871,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2957834943813753,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0026427736120543718,
+      "loss": 1.307,
+      "step": 3370
+    },
+    {
+      "epoch": 0.29587126396427776,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0026424861630432775,
+      "loss": 1.2383,
+      "step": 3371
+    },
+    {
+      "epoch": 0.2959590335471803,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0026421986160777263,
+      "loss": 1.244,
+      "step": 3372
+    },
+    {
+      "epoch": 0.29604680313008275,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0026419109711860976,
+      "loss": 1.2831,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2961345727129852,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002641623228396781,
+      "loss": 1.2764,
+      "step": 3374
+    },
+    {
+      "epoch": 0.29622234229588773,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002641335387738175,
+      "loss": 1.2788,
+      "step": 3375
+    },
+    {
+      "epoch": 0.2963101118787902,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0026410474492386893,
+      "loss": 1.2458,
+      "step": 3376
+    },
+    {
+      "epoch": 0.2963978814616927,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002640759412926742,
+      "loss": 1.2588,
+      "step": 3377
+    },
+    {
+      "epoch": 0.2964856510445952,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0026404712788307603,
+      "loss": 1.2749,
+      "step": 3378
+    },
+    {
+      "epoch": 0.29657342062749764,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0026401830469791826,
+      "loss": 1.3258,
+      "step": 3379
+    },
+    {
+      "epoch": 0.29666119021040016,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002639894717400456,
+      "loss": 1.2941,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2967489597933026,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.002639606290123038,
+      "loss": 1.2627,
+      "step": 3381
+    },
+    {
+      "epoch": 0.29683672937620514,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0026393177651753944,
+      "loss": 1.2329,
+      "step": 3382
+    },
+    {
+      "epoch": 0.2969244989591076,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002639029142586002,
+      "loss": 1.2544,
+      "step": 3383
+    },
+    {
+      "epoch": 0.29701226854201007,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002638740422383347,
+      "loss": 1.2586,
+      "step": 3384
+    },
+    {
+      "epoch": 0.2971000381249126,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002638451604595924,
+      "loss": 1.2558,
+      "step": 3385
+    },
+    {
+      "epoch": 0.29718780770781505,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002638162689252239,
+      "loss": 1.2762,
+      "step": 3386
+    },
+    {
+      "epoch": 0.2972755772907175,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026378736763808066,
+      "loss": 1.2893,
+      "step": 3387
+    },
+    {
+      "epoch": 0.29736334687362004,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.002637584566010151,
+      "loss": 1.2352,
+      "step": 3388
+    },
+    {
+      "epoch": 0.2974511164565225,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0026372953581688062,
+      "loss": 1.25,
+      "step": 3389
+    },
+    {
+      "epoch": 0.297538886039425,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0026370060528853167,
+      "loss": 1.2761,
+      "step": 3390
+    },
+    {
+      "epoch": 0.2976266556223275,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0026367166501882347,
+      "loss": 1.3038,
+      "step": 3391
+    },
+    {
+      "epoch": 0.29771442520522995,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.002636427150106124,
+      "loss": 1.2012,
+      "step": 3392
+    },
+    {
+      "epoch": 0.29780219478813247,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002636137552667556,
+      "loss": 1.2477,
+      "step": 3393
+    },
+    {
+      "epoch": 0.29788996437103493,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.002635847857901114,
+      "loss": 1.2499,
+      "step": 3394
+    },
+    {
+      "epoch": 0.29797773395393745,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0026355580658353893,
+      "loss": 1.2579,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2980655035368399,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002635268176498983,
+      "loss": 1.2835,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2981532731197424,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.002634978189920506,
+      "loss": 1.3178,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2982410427026449,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002634688106128579,
+      "loss": 1.2855,
+      "step": 3398
+    },
+    {
+      "epoch": 0.29832881228554736,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0026343979251518324,
+      "loss": 1.2734,
+      "step": 3399
+    },
+    {
+      "epoch": 0.2984165818684498,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0026341076470189056,
+      "loss": 1.2831,
+      "step": 3400
+    },
+    {
+      "epoch": 0.29850435145135235,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.002633817271758448,
+      "loss": 1.271,
+      "step": 3401
+    },
+    {
+      "epoch": 0.2985921210342548,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0026335267993991183,
+      "loss": 1.209,
+      "step": 3402
+    },
+    {
+      "epoch": 0.29867989061715733,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.002633236229969585,
+      "loss": 1.2606,
+      "step": 3403
+    },
+    {
+      "epoch": 0.2987676602000598,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0026329455634985267,
+      "loss": 1.3079,
+      "step": 3404
+    },
+    {
+      "epoch": 0.29885542978296226,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0026326548000146304,
+      "loss": 1.2775,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2989431993658648,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0026323639395465933,
+      "loss": 1.2821,
+      "step": 3406
+    },
+    {
+      "epoch": 0.29903096894876724,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0026320729821231225,
+      "loss": 1.2218,
+      "step": 3407
+    },
+    {
+      "epoch": 0.29911873853166976,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0026317819277729346,
+      "loss": 1.2085,
+      "step": 3408
+    },
+    {
+      "epoch": 0.2992065081145722,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002631490776524755,
+      "loss": 1.1999,
+      "step": 3409
+    },
+    {
+      "epoch": 0.2992942776974747,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0026311995284073186,
+      "loss": 1.2382,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2993820472803772,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002630908183449372,
+      "loss": 1.2776,
+      "step": 3411
+    },
+    {
+      "epoch": 0.29946981686327967,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0026306167416796683,
+      "loss": 1.2725,
+      "step": 3412
+    },
+    {
+      "epoch": 0.29955758644618213,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002630325203126973,
+      "loss": 1.2999,
+      "step": 3413
+    },
+    {
+      "epoch": 0.29964535602908465,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0026300335678200587,
+      "loss": 1.2479,
+      "step": 3414
+    },
+    {
+      "epoch": 0.2997331256119871,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002629741835787709,
+      "loss": 1.2683,
+      "step": 3415
+    },
+    {
+      "epoch": 0.29982089519488964,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0026294500070587166,
+      "loss": 1.2631,
+      "step": 3416
+    },
+    {
+      "epoch": 0.2999086647777921,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002629158081661885,
+      "loss": 1.2528,
+      "step": 3417
+    },
+    {
+      "epoch": 0.29999643436069456,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002628866059626024,
+      "loss": 1.2879,
+      "step": 3418
+    },
+    {
+      "epoch": 0.3000842039435971,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0026285739409799566,
+      "loss": 1.2246,
+      "step": 3419
+    },
+    {
+      "epoch": 0.30017197352649955,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0026282817257525133,
+      "loss": 1.3056,
+      "step": 3420
+    },
+    {
+      "epoch": 0.30025974310940207,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0026279894139725336,
+      "loss": 1.2958,
+      "step": 3421
+    },
+    {
+      "epoch": 0.30034751269230453,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00262769700566887,
+      "loss": 1.2394,
+      "step": 3422
+    },
+    {
+      "epoch": 0.300435282275207,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0026274045008703793,
+      "loss": 1.2265,
+      "step": 3423
+    },
+    {
+      "epoch": 0.3005230518581095,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002627111899605932,
+      "loss": 1.2423,
+      "step": 3424
+    },
+    {
+      "epoch": 0.300610821441012,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0026268192019044062,
+      "loss": 1.2272,
+      "step": 3425
+    },
+    {
+      "epoch": 0.30069859102391444,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0026265264077946914,
+      "loss": 1.2883,
+      "step": 3426
+    },
+    {
+      "epoch": 0.30078636060681696,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0026262335173056826,
+      "loss": 1.2927,
+      "step": 3427
+    },
+    {
+      "epoch": 0.3008741301897194,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0026259405304662896,
+      "loss": 1.2348,
+      "step": 3428
+    },
+    {
+      "epoch": 0.30096189977262194,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002625647447305427,
+      "loss": 1.3007,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3010496693555244,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002625354267852022,
+      "loss": 1.2587,
+      "step": 3430
+    },
+    {
+      "epoch": 0.30113743893842687,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0026250609921350107,
+      "loss": 1.2667,
+      "step": 3431
+    },
+    {
+      "epoch": 0.3012252085213294,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002624767620183337,
+      "loss": 1.2568,
+      "step": 3432
+    },
+    {
+      "epoch": 0.30131297810423185,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026244741520259564,
+      "loss": 1.2506,
+      "step": 3433
+    },
+    {
+      "epoch": 0.3014007476871344,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0026241805876918325,
+      "loss": 1.2402,
+      "step": 3434
+    },
+    {
+      "epoch": 0.30148851727003684,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0026238869272099392,
+      "loss": 1.2475,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3015762868529393,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00262359317060926,
+      "loss": 1.2976,
+      "step": 3436
+    },
+    {
+      "epoch": 0.3016640564358418,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002623299317918787,
+      "loss": 1.2157,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3017518260187443,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0026230053691675224,
+      "loss": 1.2716,
+      "step": 3438
+    },
+    {
+      "epoch": 0.3018395956016468,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002622711324384478,
+      "loss": 1.2368,
+      "step": 3439
+    },
+    {
+      "epoch": 0.30192736518454927,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002622417183598674,
+      "loss": 1.2473,
+      "step": 3440
+    },
+    {
+      "epoch": 0.30201513476745173,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0026221229468391424,
+      "loss": 1.2557,
+      "step": 3441
+    },
+    {
+      "epoch": 0.30210290435035425,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0026218286141349227,
+      "loss": 1.2355,
+      "step": 3442
+    },
+    {
+      "epoch": 0.3021906739332567,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026215341855150634,
+      "loss": 1.308,
+      "step": 3443
+    },
+    {
+      "epoch": 0.3022784435161592,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0026212396610086246,
+      "loss": 1.2519,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3023662130990617,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0026209450406446742,
+      "loss": 1.223,
+      "step": 3445
+    },
+    {
+      "epoch": 0.30245398268196416,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00262065032445229,
+      "loss": 1.2368,
+      "step": 3446
+    },
+    {
+      "epoch": 0.3025417522648667,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0026203555124605597,
+      "loss": 1.2802,
+      "step": 3447
+    },
+    {
+      "epoch": 0.30262952184776915,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0026200606046985794,
+      "loss": 1.2761,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3027172914306716,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0026197656011954563,
+      "loss": 1.2377,
+      "step": 3449
+    },
+    {
+      "epoch": 0.30280506101357413,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0026194705019803046,
+      "loss": 1.2664,
+      "step": 3450
+    },
+    {
+      "epoch": 0.3028928305964766,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002619175307082251,
+      "loss": 1.28,
+      "step": 3451
+    },
+    {
+      "epoch": 0.3029806001793791,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0026188800165304294,
+      "loss": 1.2825,
+      "step": 3452
+    },
+    {
+      "epoch": 0.3030683697622816,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002618584630353984,
+      "loss": 1.2874,
+      "step": 3453
+    },
+    {
+      "epoch": 0.30315613934518404,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.002618289148582068,
+      "loss": 1.2591,
+      "step": 3454
+    },
+    {
+      "epoch": 0.30324390892808656,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002617993571243844,
+      "loss": 1.2472,
+      "step": 3455
+    },
+    {
+      "epoch": 0.303331678510989,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0026176978983684855,
+      "loss": 1.2675,
+      "step": 3456
+    },
+    {
+      "epoch": 0.3034194480938915,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002617402129985173,
+      "loss": 1.2374,
+      "step": 3457
+    },
+    {
+      "epoch": 0.303507217676794,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0026171062661230976,
+      "loss": 1.2856,
+      "step": 3458
+    },
+    {
+      "epoch": 0.30359498725969647,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002616810306811461,
+      "loss": 1.3043,
+      "step": 3459
+    },
+    {
+      "epoch": 0.303682756842599,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002616514252079472,
+      "loss": 1.2512,
+      "step": 3460
+    },
+    {
+      "epoch": 0.30377052642550145,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.002616218101956351,
+      "loss": 1.3342,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3038582960084039,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0026159218564713264,
+      "loss": 1.239,
+      "step": 3462
+    },
+    {
+      "epoch": 0.30394606559130644,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.002615625515653637,
+      "loss": 1.2244,
+      "step": 3463
+    },
+    {
+      "epoch": 0.3040338351742089,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002615329079532529,
+      "loss": 1.2765,
+      "step": 3464
+    },
+    {
+      "epoch": 0.3041216047571114,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0026150325481372615,
+      "loss": 1.2157,
+      "step": 3465
+    },
+    {
+      "epoch": 0.3042093743400139,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002614735921497099,
+      "loss": 1.3023,
+      "step": 3466
+    },
+    {
+      "epoch": 0.30429714392291635,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0026144391996413188,
+      "loss": 1.2464,
+      "step": 3467
+    },
+    {
+      "epoch": 0.30438491350581887,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0026141423825992056,
+      "loss": 1.2427,
+      "step": 3468
+    },
+    {
+      "epoch": 0.30447268308872133,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002613845470400054,
+      "loss": 1.2725,
+      "step": 3469
+    },
+    {
+      "epoch": 0.3045604526716238,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002613548463073169,
+      "loss": 1.2515,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3046482222545263,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0026132513606478623,
+      "loss": 1.2545,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3047359918374288,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002612954163153458,
+      "loss": 1.2926,
+      "step": 3472
+    },
+    {
+      "epoch": 0.3048237614203313,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002612656870619288,
+      "loss": 1.1954,
+      "step": 3473
+    },
+    {
+      "epoch": 0.30491153100323376,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0026123594830746943,
+      "loss": 1.2229,
+      "step": 3474
+    },
+    {
+      "epoch": 0.3049993005861362,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0026120620005490275,
+      "loss": 1.2445,
+      "step": 3475
+    },
+    {
+      "epoch": 0.30508707016903874,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0026117644230716475,
+      "loss": 1.2419,
+      "step": 3476
+    },
+    {
+      "epoch": 0.3051748397519412,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0026114667506719246,
+      "loss": 1.3121,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3052626093348437,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0026111689833792375,
+      "loss": 1.2266,
+      "step": 3478
+    },
+    {
+      "epoch": 0.3053503789177462,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002610871121222975,
+      "loss": 1.2657,
+      "step": 3479
+    },
+    {
+      "epoch": 0.30543814850064865,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0026105731642325353,
+      "loss": 1.2014,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3055259180835512,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.002610275112437325,
+      "loss": 1.2845,
+      "step": 3481
+    },
+    {
+      "epoch": 0.30561368766645364,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0026099769658667602,
+      "loss": 1.2589,
+      "step": 3482
+    },
+    {
+      "epoch": 0.3057014572493561,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0026096787245502673,
+      "loss": 1.3053,
+      "step": 3483
+    },
+    {
+      "epoch": 0.3057892268322586,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002609380388517282,
+      "loss": 1.2686,
+      "step": 3484
+    },
+    {
+      "epoch": 0.3058769964151611,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0026090819577972483,
+      "loss": 1.297,
+      "step": 3485
+    },
+    {
+      "epoch": 0.3059647659980636,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0026087834324196204,
+      "loss": 1.2643,
+      "step": 3486
+    },
+    {
+      "epoch": 0.30605253558096607,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.002608484812413861,
+      "loss": 1.2466,
+      "step": 3487
+    },
+    {
+      "epoch": 0.30614030516386853,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0026081860978094435,
+      "loss": 1.2822,
+      "step": 3488
+    },
+    {
+      "epoch": 0.30622807474677105,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.002607887288635849,
+      "loss": 1.2487,
+      "step": 3489
+    },
+    {
+      "epoch": 0.3063158443296735,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0026075883849225696,
+      "loss": 1.3226,
+      "step": 3490
+    },
+    {
+      "epoch": 0.30640361391257603,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002607289386699106,
+      "loss": 1.2883,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3064913834954785,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002606990293994967,
+      "loss": 1.307,
+      "step": 3492
+    },
+    {
+      "epoch": 0.30657915307838096,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0026066911068396733,
+      "loss": 1.2099,
+      "step": 3493
+    },
+    {
+      "epoch": 0.3066669226612835,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0026063918252627522,
+      "loss": 1.2483,
+      "step": 3494
+    },
+    {
+      "epoch": 0.30675469224418594,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0026060924492937424,
+      "loss": 1.2298,
+      "step": 3495
+    },
+    {
+      "epoch": 0.3068424618270884,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0026057929789621913,
+      "loss": 1.3231,
+      "step": 3496
+    },
+    {
+      "epoch": 0.30693023140999093,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0026054934142976547,
+      "loss": 1.2646,
+      "step": 3497
+    },
+    {
+      "epoch": 0.3070180009928934,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002605193755329699,
+      "loss": 1.2266,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3071057705757959,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002604894002087899,
+      "loss": 1.2036,
+      "step": 3499
+    },
+    {
+      "epoch": 0.3071935401586984,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002604594154601839,
+      "loss": 1.254,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3071935401586984,
+      "eval_loss": 1.246621012687683,
+      "eval_runtime": 437.6804,
+      "eval_samples_per_second": 33.678,
+      "eval_steps_per_second": 8.419,
+      "step": 3500
+    },
+    {
+      "epoch": 0.30728130974160084,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0026042942129011135,
+      "loss": 1.1991,
+      "step": 3501
+    },
+    {
+      "epoch": 0.30736907932450336,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0026039941770153252,
+      "loss": 1.2096,
+      "step": 3502
+    },
+    {
+      "epoch": 0.3074568489074058,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0026036940469740862,
+      "loss": 1.2317,
+      "step": 3503
+    },
+    {
+      "epoch": 0.30754461849030834,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002603393822807018,
+      "loss": 1.2436,
+      "step": 3504
+    },
+    {
+      "epoch": 0.3076323880732108,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0026030935045437526,
+      "loss": 1.2749,
+      "step": 3505
+    },
+    {
+      "epoch": 0.30772015765611327,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.002602793092213929,
+      "loss": 1.2819,
+      "step": 3506
+    },
+    {
+      "epoch": 0.3078079272390158,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0026024925858471977,
+      "loss": 1.2134,
+      "step": 3507
+    },
+    {
+      "epoch": 0.30789569682191825,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0026021919854732163,
+      "loss": 1.2651,
+      "step": 3508
+    },
+    {
+      "epoch": 0.30798346640482077,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002601891291121654,
+      "loss": 1.2901,
+      "step": 3509
+    },
+    {
+      "epoch": 0.30807123598772324,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.002601590502822188,
+      "loss": 1.2187,
+      "step": 3510
+    },
+    {
+      "epoch": 0.3081590055706257,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002601289620604504,
+      "loss": 1.2439,
+      "step": 3511
+    },
+    {
+      "epoch": 0.3082467751535282,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002600988644498299,
+      "loss": 1.2828,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3083345447364307,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0026006875745332777,
+      "loss": 1.2665,
+      "step": 3513
+    },
+    {
+      "epoch": 0.30842231431933315,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002600386410739154,
+      "loss": 1.2554,
+      "step": 3514
+    },
+    {
+      "epoch": 0.30851008390223567,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0026000851531456527,
+      "loss": 1.2307,
+      "step": 3515
+    },
+    {
+      "epoch": 0.30859785348513813,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.002599783801782505,
+      "loss": 1.2609,
+      "step": 3516
+    },
+    {
+      "epoch": 0.30868562306804065,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.002599482356679455,
+      "loss": 1.2158,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3087733926509431,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.002599180817866253,
+      "loss": 1.2437,
+      "step": 3518
+    },
+    {
+      "epoch": 0.3088611622338456,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00259887918537266,
+      "loss": 1.2729,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3089489318167481,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.002598577459228447,
+      "loss": 1.2513,
+      "step": 3520
+    },
+    {
+      "epoch": 0.30903670139965056,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002598275639463391,
+      "loss": 1.2668,
+      "step": 3521
+    },
+    {
+      "epoch": 0.3091244709825531,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002597973726107281,
+      "loss": 1.2166,
+      "step": 3522
+    },
+    {
+      "epoch": 0.30921224056545554,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.002597671719189916,
+      "loss": 1.3082,
+      "step": 3523
+    },
+    {
+      "epoch": 0.309300010148358,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0025973696187411026,
+      "loss": 1.2779,
+      "step": 3524
+    },
+    {
+      "epoch": 0.3093877797312605,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0025970674247906556,
+      "loss": 1.312,
+      "step": 3525
+    },
+    {
+      "epoch": 0.309475549314163,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002596765137368401,
+      "loss": 1.2172,
+      "step": 3526
+    },
+    {
+      "epoch": 0.30956331889706545,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002596462756504174,
+      "loss": 1.2589,
+      "step": 3527
+    },
+    {
+      "epoch": 0.309651088479968,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002596160282227817,
+      "loss": 1.3159,
+      "step": 3528
+    },
+    {
+      "epoch": 0.30973885806287044,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002595857714569185,
+      "loss": 1.27,
+      "step": 3529
+    },
+    {
+      "epoch": 0.30982662764577296,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0025955550535581387,
+      "loss": 1.3142,
+      "step": 3530
+    },
+    {
+      "epoch": 0.3099143972286754,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0025952522992245502,
+      "loss": 1.2416,
+      "step": 3531
+    },
+    {
+      "epoch": 0.3100021668115779,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0025949494515982996,
+      "loss": 1.2216,
+      "step": 3532
+    },
+    {
+      "epoch": 0.3100899363944804,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002594646510709277,
+      "loss": 1.3537,
+      "step": 3533
+    },
+    {
+      "epoch": 0.31017770597738287,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0025943434765873822,
+      "loss": 1.2775,
+      "step": 3534
+    },
+    {
+      "epoch": 0.3102654755602854,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0025940403492625223,
+      "loss": 1.2768,
+      "step": 3535
+    },
+    {
+      "epoch": 0.31035324514318785,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0025937371287646153,
+      "loss": 1.2531,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3104410147260903,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002593433815123588,
+      "loss": 1.258,
+      "step": 3537
+    },
+    {
+      "epoch": 0.31052878430899283,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002593130408369376,
+      "loss": 1.2718,
+      "step": 3538
+    },
+    {
+      "epoch": 0.3106165538918953,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002592826908531925,
+      "loss": 1.243,
+      "step": 3539
+    },
+    {
+      "epoch": 0.31070432347479776,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0025925233156411883,
+      "loss": 1.2287,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3107920930577003,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0025922196297271296,
+      "loss": 1.2506,
+      "step": 3541
+    },
+    {
+      "epoch": 0.31087986264060274,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0025919158508197217,
+      "loss": 1.3043,
+      "step": 3542
+    },
+    {
+      "epoch": 0.31096763222350526,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0025916119789489464,
+      "loss": 1.2521,
+      "step": 3543
+    },
+    {
+      "epoch": 0.3110554018064077,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0025913080141447943,
+      "loss": 1.275,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3111431713893102,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0025910039564372663,
+      "loss": 1.2509,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3112309409722127,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002590699805856371,
+      "loss": 1.2654,
+      "step": 3546
+    },
+    {
+      "epoch": 0.3113187105551152,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.002590395562432127,
+      "loss": 1.2131,
+      "step": 3547
+    },
+    {
+      "epoch": 0.3114064801380177,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0025900912261945614,
+      "loss": 1.2613,
+      "step": 3548
+    },
+    {
+      "epoch": 0.31149424972092016,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.002589786797173712,
+      "loss": 1.3005,
+      "step": 3549
+    },
+    {
+      "epoch": 0.3115820193038226,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0025894822753996243,
+      "loss": 1.244,
+      "step": 3550
+    },
+    {
+      "epoch": 0.31166978888672514,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0025891776609023535,
+      "loss": 1.2597,
+      "step": 3551
+    },
+    {
+      "epoch": 0.3117575584696276,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0025888729537119634,
+      "loss": 1.2941,
+      "step": 3552
+    },
+    {
+      "epoch": 0.31184532805253007,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0025885681538585278,
+      "loss": 1.2666,
+      "step": 3553
+    },
+    {
+      "epoch": 0.3119330976354326,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002588263261372129,
+      "loss": 1.222,
+      "step": 3554
+    },
+    {
+      "epoch": 0.31202086721833505,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.002587958276282859,
+      "loss": 1.282,
+      "step": 3555
+    },
+    {
+      "epoch": 0.31210863680123757,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0025876531986208187,
+      "loss": 1.2138,
+      "step": 3556
+    },
+    {
+      "epoch": 0.31219640638414003,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0025873480284161173,
+      "loss": 1.2079,
+      "step": 3557
+    },
+    {
+      "epoch": 0.3122841759670425,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0025870427656988744,
+      "loss": 1.2885,
+      "step": 3558
+    },
+    {
+      "epoch": 0.312371945549945,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002586737410499219,
+      "loss": 1.2305,
+      "step": 3559
+    },
+    {
+      "epoch": 0.3124597151328475,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0025864319628472864,
+      "loss": 1.2768,
+      "step": 3560
+    },
+    {
+      "epoch": 0.31254748471575,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0025861264227732255,
+      "loss": 1.1599,
+      "step": 3561
+    },
+    {
+      "epoch": 0.31263525429865247,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0025858207903071903,
+      "loss": 1.3309,
+      "step": 3562
+    },
+    {
+      "epoch": 0.31272302388155493,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002585515065479346,
+      "loss": 1.227,
+      "step": 3563
+    },
+    {
+      "epoch": 0.31281079346445745,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002585209248319866,
+      "loss": 1.3205,
+      "step": 3564
+    },
+    {
+      "epoch": 0.3128985630473599,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0025849033388589343,
+      "loss": 1.2751,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3129863326302624,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002584597337126742,
+      "loss": 1.2963,
+      "step": 3566
+    },
+    {
+      "epoch": 0.3130741022131649,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00258429124315349,
+      "loss": 1.3122,
+      "step": 3567
+    },
+    {
+      "epoch": 0.31316187179606736,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0025839850569693897,
+      "loss": 1.2886,
+      "step": 3568
+    },
+    {
+      "epoch": 0.3132496413789699,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0025836787786046598,
+      "loss": 1.2859,
+      "step": 3569
+    },
+    {
+      "epoch": 0.31333741096187234,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0025833724080895288,
+      "loss": 1.2981,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3134251805447748,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002583065945454234,
+      "loss": 1.224,
+      "step": 3571
+    },
+    {
+      "epoch": 0.3135129501276773,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002582759390729023,
+      "loss": 1.2456,
+      "step": 3572
+    },
+    {
+      "epoch": 0.3136007197105798,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0025824527439441505,
+      "loss": 1.2585,
+      "step": 3573
+    },
+    {
+      "epoch": 0.3136884892934823,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0025821460051298813,
+      "loss": 1.2229,
+      "step": 3574
+    },
+    {
+      "epoch": 0.3137762588763848,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0025818391743164903,
+      "loss": 1.2405,
+      "step": 3575
+    },
+    {
+      "epoch": 0.31386402845928724,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0025815322515342597,
+      "loss": 1.1832,
+      "step": 3576
+    },
+    {
+      "epoch": 0.31395179804218976,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002581225236813482,
+      "loss": 1.2483,
+      "step": 3577
+    },
+    {
+      "epoch": 0.3140395676250922,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0025809181301844576,
+      "loss": 1.2632,
+      "step": 3578
+    },
+    {
+      "epoch": 0.31412733720799474,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0025806109316774974,
+      "loss": 1.2861,
+      "step": 3579
+    },
+    {
+      "epoch": 0.3142151067908972,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0025803036413229205,
+      "loss": 1.2116,
+      "step": 3580
+    },
+    {
+      "epoch": 0.31430287637379967,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0025799962591510554,
+      "loss": 1.2423,
+      "step": 3581
+    },
+    {
+      "epoch": 0.3143906459567022,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0025796887851922395,
+      "loss": 1.2125,
+      "step": 3582
+    },
+    {
+      "epoch": 0.31447841553960465,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0025793812194768186,
+      "loss": 1.2657,
+      "step": 3583
+    },
+    {
+      "epoch": 0.3145661851225071,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0025790735620351488,
+      "loss": 1.2553,
+      "step": 3584
+    },
+    {
+      "epoch": 0.31465395470540963,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0025787658128975945,
+      "loss": 1.2947,
+      "step": 3585
+    },
+    {
+      "epoch": 0.3147417242883121,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00257845797209453,
+      "loss": 1.1999,
+      "step": 3586
+    },
+    {
+      "epoch": 0.3148294938712146,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0025781500396563368,
+      "loss": 1.2308,
+      "step": 3587
+    },
+    {
+      "epoch": 0.3149172634541171,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002577842015613407,
+      "loss": 1.2637,
+      "step": 3588
+    },
+    {
+      "epoch": 0.31500503303701954,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0025775338999961418,
+      "loss": 1.2466,
+      "step": 3589
+    },
+    {
+      "epoch": 0.31509280261992206,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002577225692834951,
+      "loss": 1.2522,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3151805722028245,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0025769173941602532,
+      "loss": 1.3115,
+      "step": 3591
+    },
+    {
+      "epoch": 0.31526834178572705,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002576609004002476,
+      "loss": 1.2314,
+      "step": 3592
+    },
+    {
+      "epoch": 0.3153561113686295,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002576300522392057,
+      "loss": 1.2635,
+      "step": 3593
+    },
+    {
+      "epoch": 0.315443880951532,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0025759919493594406,
+      "loss": 1.2546,
+      "step": 3594
+    },
+    {
+      "epoch": 0.3155316505344345,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0025756832849350834,
+      "loss": 1.2556,
+      "step": 3595
+    },
+    {
+      "epoch": 0.31561942011733696,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0025753745291494487,
+      "loss": 1.2822,
+      "step": 3596
+    },
+    {
+      "epoch": 0.3157071897002394,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0025750656820330096,
+      "loss": 1.243,
+      "step": 3597
+    },
+    {
+      "epoch": 0.31579495928314194,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002574756743616248,
+      "loss": 1.2971,
+      "step": 3598
+    },
+    {
+      "epoch": 0.3158827288660444,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0025744477139296542,
+      "loss": 1.2585,
+      "step": 3599
+    },
+    {
+      "epoch": 0.3159704984489469,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0025741385930037296,
+      "loss": 1.2598,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3160582680318494,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0025738293808689823,
+      "loss": 1.2685,
+      "step": 3601
+    },
+    {
+      "epoch": 0.31614603761475185,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002573520077555931,
+      "loss": 1.2809,
+      "step": 3602
+    },
+    {
+      "epoch": 0.31623380719765437,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0025732106830951016,
+      "loss": 1.2518,
+      "step": 3603
+    },
+    {
+      "epoch": 0.31632157678055683,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0025729011975170308,
+      "loss": 1.2493,
+      "step": 3604
+    },
+    {
+      "epoch": 0.31640934636345935,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0025725916208522636,
+      "loss": 1.2884,
+      "step": 3605
+    },
+    {
+      "epoch": 0.3164971159463618,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002572281953131354,
+      "loss": 1.21,
+      "step": 3606
+    },
+    {
+      "epoch": 0.3165848855292643,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002571972194384865,
+      "loss": 1.3251,
+      "step": 3607
+    },
+    {
+      "epoch": 0.3166726551121668,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0025716623446433683,
+      "loss": 1.288,
+      "step": 3608
+    },
+    {
+      "epoch": 0.31676042469506926,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002571352403937445,
+      "loss": 1.2037,
+      "step": 3609
+    },
+    {
+      "epoch": 0.31684819427797173,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0025710423722976852,
+      "loss": 1.2591,
+      "step": 3610
+    },
+    {
+      "epoch": 0.31693596386087425,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0025707322497546875,
+      "loss": 1.2781,
+      "step": 3611
+    },
+    {
+      "epoch": 0.3170237334437767,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00257042203633906,
+      "loss": 1.353,
+      "step": 3612
+    },
+    {
+      "epoch": 0.31711150302667923,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0025701117320814193,
+      "loss": 1.2505,
+      "step": 3613
+    },
+    {
+      "epoch": 0.3171992726095817,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0025698013370123915,
+      "loss": 1.2587,
+      "step": 3614
+    },
+    {
+      "epoch": 0.31728704219248416,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002569490851162611,
+      "loss": 1.2979,
+      "step": 3615
+    },
+    {
+      "epoch": 0.3173748117753867,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002569180274562722,
+      "loss": 1.3164,
+      "step": 3616
+    },
+    {
+      "epoch": 0.31746258135828914,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0025688696072433768,
+      "loss": 1.1939,
+      "step": 3617
+    },
+    {
+      "epoch": 0.31755035094119166,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002568558849235237,
+      "loss": 1.1871,
+      "step": 3618
+    },
+    {
+      "epoch": 0.3176381205240941,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002568248000568974,
+      "loss": 1.2734,
+      "step": 3619
+    },
+    {
+      "epoch": 0.3177258901069966,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002567937061275266,
+      "loss": 1.2382,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3178136596898991,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0025676260313848022,
+      "loss": 1.2273,
+      "step": 3621
+    },
+    {
+      "epoch": 0.3179014292728016,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0025673149109282805,
+      "loss": 1.2013,
+      "step": 3622
+    },
+    {
+      "epoch": 0.31798919885570404,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002567003699936406,
+      "loss": 1.2256,
+      "step": 3623
+    },
+    {
+      "epoch": 0.31807696843860656,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002566692398439895,
+      "loss": 1.2752,
+      "step": 3624
+    },
+    {
+      "epoch": 0.318164738021509,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0025663810064694716,
+      "loss": 1.2639,
+      "step": 3625
+    },
+    {
+      "epoch": 0.31825250760441154,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0025660695240558684,
+      "loss": 1.192,
+      "step": 3626
+    },
+    {
+      "epoch": 0.318340277187314,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0025657579512298285,
+      "loss": 1.2568,
+      "step": 3627
+    },
+    {
+      "epoch": 0.31842804677021647,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0025654462880221017,
+      "loss": 1.2127,
+      "step": 3628
+    },
+    {
+      "epoch": 0.318515816353119,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0025651345344634485,
+      "loss": 1.2482,
+      "step": 3629
+    },
+    {
+      "epoch": 0.31860358593602145,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002564822690584638,
+      "loss": 1.2515,
+      "step": 3630
+    },
+    {
+      "epoch": 0.31869135551892397,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0025645107564164476,
+      "loss": 1.2398,
+      "step": 3631
+    },
+    {
+      "epoch": 0.31877912510182643,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0025641987319896644,
+      "loss": 1.2943,
+      "step": 3632
+    },
+    {
+      "epoch": 0.3188668946847289,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0025638866173350837,
+      "loss": 1.2504,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3189546642676314,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002563574412483509,
+      "loss": 1.2551,
+      "step": 3634
+    },
+    {
+      "epoch": 0.3190424338505339,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0025632621174657556,
+      "loss": 1.21,
+      "step": 3635
+    },
+    {
+      "epoch": 0.31913020343343634,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0025629497323126447,
+      "loss": 1.2433,
+      "step": 3636
+    },
+    {
+      "epoch": 0.31921797301633886,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0025626372570550075,
+      "loss": 1.2733,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3193057425992413,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.002562324691723684,
+      "loss": 1.2792,
+      "step": 3638
+    },
+    {
+      "epoch": 0.31939351218214385,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0025620120363495226,
+      "loss": 1.2842,
+      "step": 3639
+    },
+    {
+      "epoch": 0.3194812817650463,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002561699290963383,
+      "loss": 1.23,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3195690513479488,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0025613864555961304,
+      "loss": 1.221,
+      "step": 3641
+    },
+    {
+      "epoch": 0.3196568209308513,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002561073530278641,
+      "loss": 1.292,
+      "step": 3642
+    },
+    {
+      "epoch": 0.31974459051375376,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0025607605150417987,
+      "loss": 1.2336,
+      "step": 3643
+    },
+    {
+      "epoch": 0.3198323600966563,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002560447409916497,
+      "loss": 1.2731,
+      "step": 3644
+    },
+    {
+      "epoch": 0.31992012967955874,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0025601342149336394,
+      "loss": 1.2563,
+      "step": 3645
+    },
+    {
+      "epoch": 0.3200078992624612,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0025598209301241352,
+      "loss": 1.2311,
+      "step": 3646
+    },
+    {
+      "epoch": 0.3200956688453637,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002559507555518906,
+      "loss": 1.2805,
+      "step": 3647
+    },
+    {
+      "epoch": 0.3201834384282662,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002559194091148879,
+      "loss": 1.2659,
+      "step": 3648
+    },
+    {
+      "epoch": 0.3202712080111687,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0025588805370449926,
+      "loss": 1.3416,
+      "step": 3649
+    },
+    {
+      "epoch": 0.32035897759407117,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002558566893238194,
+      "loss": 1.2211,
+      "step": 3650
+    },
+    {
+      "epoch": 0.32044674717697363,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002558253159759438,
+      "loss": 1.2193,
+      "step": 3651
+    },
+    {
+      "epoch": 0.32053451675987615,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0025579393366396887,
+      "loss": 1.264,
+      "step": 3652
+    },
+    {
+      "epoch": 0.3206222863427786,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0025576254239099193,
+      "loss": 1.3171,
+      "step": 3653
+    },
+    {
+      "epoch": 0.3207100559256811,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002557311421601112,
+      "loss": 1.3002,
+      "step": 3654
+    },
+    {
+      "epoch": 0.3207978255085836,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0025569973297442576,
+      "loss": 1.2464,
+      "step": 3655
+    },
+    {
+      "epoch": 0.32088559509148606,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0025566831483703556,
+      "loss": 1.1842,
+      "step": 3656
+    },
+    {
+      "epoch": 0.3209733646743886,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002556368877510414,
+      "loss": 1.277,
+      "step": 3657
+    },
+    {
+      "epoch": 0.32106113425729105,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002556054517195451,
+      "loss": 1.3148,
+      "step": 3658
+    },
+    {
+      "epoch": 0.3211489038401935,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002555740067456492,
+      "loss": 1.2964,
+      "step": 3659
+    },
+    {
+      "epoch": 0.32123667342309603,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002555425528324572,
+      "loss": 1.269,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3213244430059985,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0025551108998307353,
+      "loss": 1.2718,
+      "step": 3661
+    },
+    {
+      "epoch": 0.321412212588901,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0025547961820060337,
+      "loss": 1.2604,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3214999821718035,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00255448137488153,
+      "loss": 1.2569,
+      "step": 3663
+    },
+    {
+      "epoch": 0.32158775175470594,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002554166478488292,
+      "loss": 1.2961,
+      "step": 3664
+    },
+    {
+      "epoch": 0.32167552133760846,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002553851492857401,
+      "loss": 1.2614,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3217632909205109,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002553536418019944,
+      "loss": 1.219,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3218510605034134,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0025532212540070173,
+      "loss": 1.2551,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3219388300863159,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002552906000849727,
+      "loss": 1.2483,
+      "step": 3668
+    },
+    {
+      "epoch": 0.32202659966921837,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0025525906585791873,
+      "loss": 1.2635,
+      "step": 3669
+    },
+    {
+      "epoch": 0.3221143692521209,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0025522752272265207,
+      "loss": 1.2612,
+      "step": 3670
+    },
+    {
+      "epoch": 0.32220213883502336,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0025519597068228593,
+      "loss": 1.2294,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3222899084179258,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0025516440973993434,
+      "loss": 1.2335,
+      "step": 3672
+    },
+    {
+      "epoch": 0.32237767800082834,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002551328398987123,
+      "loss": 1.2709,
+      "step": 3673
+    },
+    {
+      "epoch": 0.3224654475837308,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0025510126116173562,
+      "loss": 1.1987,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3225532171666333,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0025506967353212096,
+      "loss": 1.2795,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3226409867495358,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0025503807701298593,
+      "loss": 1.2423,
+      "step": 3676
+    },
+    {
+      "epoch": 0.32272875633243825,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0025500647160744895,
+      "loss": 1.316,
+      "step": 3677
+    },
+    {
+      "epoch": 0.32281652591534077,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0025497485731862933,
+      "loss": 1.3196,
+      "step": 3678
+    },
+    {
+      "epoch": 0.32290429549824323,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002549432341496473,
+      "loss": 1.2743,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3229920650811457,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0025491160210362397,
+      "loss": 1.2443,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3230798346640482,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002548799611836813,
+      "loss": 1.2591,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3231676042469507,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002548483113929421,
+      "loss": 1.2808,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3232553738298532,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0025481665273453004,
+      "loss": 1.2841,
+      "step": 3683
+    },
+    {
+      "epoch": 0.32334314341275566,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002547849852115697,
+      "loss": 1.2251,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3234309129956581,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002547533088271866,
+      "loss": 1.2241,
+      "step": 3685
+    },
+    {
+      "epoch": 0.32351868257856065,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0025472162358450715,
+      "loss": 1.2915,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3236064521614631,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.002546899294866584,
+      "loss": 1.2427,
+      "step": 3687
+    },
+    {
+      "epoch": 0.32369422174436563,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002546582265367685,
+      "loss": 1.2149,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3237819913272681,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002546265147379664,
+      "loss": 1.2542,
+      "step": 3689
+    },
+    {
+      "epoch": 0.32386976091017056,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0025459479409338194,
+      "loss": 1.2619,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3239575304930731,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0025456306460614575,
+      "loss": 1.3024,
+      "step": 3691
+    },
+    {
+      "epoch": 0.32404530007597554,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002545313262793895,
+      "loss": 1.2757,
+      "step": 3692
+    },
+    {
+      "epoch": 0.324133069658878,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0025449957911624563,
+      "loss": 1.2264,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3242208392417805,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0025446782311984747,
+      "loss": 1.3051,
+      "step": 3694
+    },
+    {
+      "epoch": 0.324308608824683,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0025443605829332905,
+      "loss": 1.2478,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3243963784075855,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002544042846398257,
+      "loss": 1.2214,
+      "step": 3696
+    },
+    {
+      "epoch": 0.32448414799048797,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0025437250216247312,
+      "loss": 1.2114,
+      "step": 3697
+    },
+    {
+      "epoch": 0.32457191757339043,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002543407108644082,
+      "loss": 1.2714,
+      "step": 3698
+    },
+    {
+      "epoch": 0.32465968715629295,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.002543089107487687,
+      "loss": 1.2723,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3247474567391954,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00254277101818693,
+      "loss": 1.2854,
+      "step": 3700
+    },
+    {
+      "epoch": 0.32483522632209794,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0025424528407732064,
+      "loss": 1.2899,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3249229959050004,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0025421345752779188,
+      "loss": 1.2391,
+      "step": 3702
+    },
+    {
+      "epoch": 0.32501076548790286,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0025418162217324784,
+      "loss": 1.251,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3250985350708054,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.002541497780168306,
+      "loss": 1.2149,
+      "step": 3704
+    },
+    {
+      "epoch": 0.32518630465370785,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0025411792506168296,
+      "loss": 1.2652,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3252740742366103,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0025408606331094874,
+      "loss": 1.258,
+      "step": 3706
+    },
+    {
+      "epoch": 0.32536184381951283,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002540541927677726,
+      "loss": 1.2684,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3254496134024153,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0025402231343530004,
+      "loss": 1.2957,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3255373829853178,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0025399042531667735,
+      "loss": 1.293,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3256251525682203,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0025395852841505182,
+      "loss": 1.2639,
+      "step": 3710
+    },
+    {
+      "epoch": 0.32571292215112274,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0025392662273357146,
+      "loss": 1.2412,
+      "step": 3711
+    },
+    {
+      "epoch": 0.32580069173402526,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0025389470827538534,
+      "loss": 1.2482,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3258884613169277,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0025386278504364325,
+      "loss": 1.225,
+      "step": 3713
+    },
+    {
+      "epoch": 0.32597623089983024,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0025383085304149593,
+      "loss": 1.2702,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3260640004827327,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002537989122720949,
+      "loss": 1.2097,
+      "step": 3715
+    },
+    {
+      "epoch": 0.32615177006563517,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002537669627385925,
+      "loss": 1.2717,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3262395396485377,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0025373500444414217,
+      "loss": 1.2309,
+      "step": 3717
+    },
+    {
+      "epoch": 0.32632730923144015,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00253703037391898,
+      "loss": 1.2578,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3264150788143427,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00253671061585015,
+      "loss": 1.3245,
+      "step": 3719
+    },
+    {
+      "epoch": 0.32650284839724514,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002536390770266491,
+      "loss": 1.2825,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3265906179801476,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00253607083719957,
+      "loss": 1.3416,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3266783875630501,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002535750816680964,
+      "loss": 1.1654,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3267661571459526,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0025354307087422567,
+      "loss": 1.2169,
+      "step": 3723
+    },
+    {
+      "epoch": 0.32685392672885505,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002535110513415042,
+      "loss": 1.2615,
+      "step": 3724
+    },
+    {
+      "epoch": 0.32694169631175757,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0025347902307309217,
+      "loss": 1.2649,
+      "step": 3725
+    },
+    {
+      "epoch": 0.32702946589466003,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002534469860721507,
+      "loss": 1.2465,
+      "step": 3726
+    },
+    {
+      "epoch": 0.32711723547756255,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002534149403418416,
+      "loss": 1.214,
+      "step": 3727
+    },
+    {
+      "epoch": 0.327205005060465,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0025338288588532777,
+      "loss": 1.2594,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3272927746433675,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0025335082270577284,
+      "loss": 1.2803,
+      "step": 3729
+    },
+    {
+      "epoch": 0.32738054422627,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0025331875080634126,
+      "loss": 1.2449,
+      "step": 3730
+    },
+    {
+      "epoch": 0.32746831380917246,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0025328667019019844,
+      "loss": 1.2621,
+      "step": 3731
+    },
+    {
+      "epoch": 0.327556083392075,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.002532545808605106,
+      "loss": 1.203,
+      "step": 3732
+    },
+    {
+      "epoch": 0.32764385297497745,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.002532224828204448,
+      "loss": 1.2359,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3277316225578799,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0025319037607316915,
+      "loss": 1.1992,
+      "step": 3734
+    },
+    {
+      "epoch": 0.32781939214078243,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0025315826062185217,
+      "loss": 1.3304,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3279071617236849,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0025312613646966383,
+      "loss": 1.2537,
+      "step": 3736
+    },
+    {
+      "epoch": 0.32799493130658736,
+      "grad_norm": 0.125,
+      "learning_rate": 0.002530940036197744,
+      "loss": 1.2027,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3280827008894899,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002530618620753555,
+      "loss": 1.2073,
+      "step": 3738
+    },
+    {
+      "epoch": 0.32817047047239234,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.002530297118395792,
+      "loss": 1.2433,
+      "step": 3739
+    },
+    {
+      "epoch": 0.32825824005529486,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0025299755291561866,
+      "loss": 1.2212,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3283460096381973,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.002529653853066478,
+      "loss": 1.222,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3284337792210998,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002529332090158415,
+      "loss": 1.2265,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3285215488040023,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0025290102404637543,
+      "loss": 1.1789,
+      "step": 3743
+    },
+    {
+      "epoch": 0.32860931838690477,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0025286883040142607,
+      "loss": 1.2608,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3286970879698073,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002528366280841708,
+      "loss": 1.287,
+      "step": 3745
+    },
+    {
+      "epoch": 0.32878485755270975,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0025280441709778795,
+      "loss": 1.2626,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3288726271356122,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.002527721974454565,
+      "loss": 1.2261,
+      "step": 3747
+    },
+    {
+      "epoch": 0.32896039671851474,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0025273996913035652,
+      "loss": 1.2195,
+      "step": 3748
+    },
+    {
+      "epoch": 0.3290481663014172,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0025270773215566875,
+      "loss": 1.2956,
+      "step": 3749
+    },
+    {
+      "epoch": 0.32913593588431966,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002526754865245748,
+      "loss": 1.2899,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3292237054672222,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002526432322402573,
+      "loss": 1.2792,
+      "step": 3751
+    },
+    {
+      "epoch": 0.32931147505012465,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002526109693058995,
+      "loss": 1.2438,
+      "step": 3752
+    },
+    {
+      "epoch": 0.32939924463302717,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0025257869772468576,
+      "loss": 1.3401,
+      "step": 3753
+    },
+    {
+      "epoch": 0.32948701421592963,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0025254641749980107,
+      "loss": 1.3386,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3295747837988321,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0025251412863443136,
+      "loss": 1.2555,
+      "step": 3755
+    },
+    {
+      "epoch": 0.3296625533817346,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002524818311317634,
+      "loss": 1.2637,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3297503229646371,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002524495249949849,
+      "loss": 1.2575,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3298380925475396,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002524172102272843,
+      "loss": 1.2357,
+      "step": 3758
+    },
+    {
+      "epoch": 0.32992586213044206,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0025238488683185095,
+      "loss": 1.238,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3300136317133445,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00252352554811875,
+      "loss": 1.2975,
+      "step": 3760
+    },
+    {
+      "epoch": 0.33010140129624704,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002523202141705475,
+      "loss": 1.2368,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3301891708791495,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0025228786491106043,
+      "loss": 1.2287,
+      "step": 3762
+    },
+    {
+      "epoch": 0.33027694046205197,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0025225550703660646,
+      "loss": 1.2193,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3303647100449545,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002522231405503792,
+      "loss": 1.2226,
+      "step": 3764
+    },
+    {
+      "epoch": 0.33045247962785695,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002521907654555731,
+      "loss": 1.3005,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3305402492107595,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0025215838175538345,
+      "loss": 1.2229,
+      "step": 3766
+    },
+    {
+      "epoch": 0.33062801879366194,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002521259894530064,
+      "loss": 1.262,
+      "step": 3767
+    },
+    {
+      "epoch": 0.3307157883765644,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0025209358855163887,
+      "loss": 1.2642,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3308035579594669,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.002520611790544788,
+      "loss": 1.2599,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3308913275423694,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0025202876096472484,
+      "loss": 1.2076,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3309790971252719,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002519963342855765,
+      "loss": 1.2606,
+      "step": 3771
+    },
+    {
+      "epoch": 0.33106686670817437,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0025196389902023427,
+      "loss": 1.2664,
+      "step": 3772
+    },
+    {
+      "epoch": 0.33115463629107683,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0025193145517189925,
+      "loss": 1.2137,
+      "step": 3773
+    },
+    {
+      "epoch": 0.33124240587397935,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0025189900274377357,
+      "loss": 1.2974,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3313301754568818,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0025186654173906015,
+      "loss": 1.3374,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3314179450397843,
+      "grad_norm": 0.125,
+      "learning_rate": 0.002518340721609628,
+      "loss": 1.2484,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3315057146226868,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002518015940126861,
+      "loss": 1.3006,
+      "step": 3777
+    },
+    {
+      "epoch": 0.33159348420558926,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0025176910729743556,
+      "loss": 1.2739,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3316812537884918,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0025173661201841743,
+      "loss": 1.2801,
+      "step": 3779
+    },
+    {
+      "epoch": 0.33176902337139424,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0025170410817883892,
+      "loss": 1.2251,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3318567929542967,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0025167159578190794,
+      "loss": 1.2788,
+      "step": 3781
+    },
+    {
+      "epoch": 0.33194456253719923,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0025163907483083344,
+      "loss": 1.2707,
+      "step": 3782
+    },
+    {
+      "epoch": 0.3320323321201017,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0025160654532882507,
+      "loss": 1.2502,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3321201017030042,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0025157400727909337,
+      "loss": 1.2705,
+      "step": 3784
+    },
+    {
+      "epoch": 0.3322078712859067,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002515414606848497,
+      "loss": 1.2243,
+      "step": 3785
+    },
+    {
+      "epoch": 0.33229564086880914,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0025150890554930627,
+      "loss": 1.2388,
+      "step": 3786
+    },
+    {
+      "epoch": 0.33238341045171166,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0025147634187567616,
+      "loss": 1.1998,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3324711800346141,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0025144376966717333,
+      "loss": 1.2723,
+      "step": 3788
+    },
+    {
+      "epoch": 0.33255894961751664,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0025141118892701244,
+      "loss": 1.2278,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3326467192004191,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.002513785996584091,
+      "loss": 1.2397,
+      "step": 3790
+    },
+    {
+      "epoch": 0.33273448878332157,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0025134600186457975,
+      "loss": 1.199,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3328222583662241,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0025131339554874173,
+      "loss": 1.2355,
+      "step": 3792
+    },
+    {
+      "epoch": 0.33291002794912655,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00251280780714113,
+      "loss": 1.2621,
+      "step": 3793
+    },
+    {
+      "epoch": 0.332997797532029,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002512481573639127,
+      "loss": 1.3712,
+      "step": 3794
+    },
+    {
+      "epoch": 0.33308556711493154,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0025121552550136043,
+      "loss": 1.2313,
+      "step": 3795
+    },
+    {
+      "epoch": 0.333173336697834,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0025118288512967707,
+      "loss": 1.2302,
+      "step": 3796
+    },
+    {
+      "epoch": 0.3332611062807365,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0025115023625208385,
+      "loss": 1.2087,
+      "step": 3797
+    },
+    {
+      "epoch": 0.333348875863639,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0025111757887180323,
+      "loss": 1.3146,
+      "step": 3798
+    },
+    {
+      "epoch": 0.33343664544654145,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0025108491299205833,
+      "loss": 1.2671,
+      "step": 3799
+    },
+    {
+      "epoch": 0.33352441502944397,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0025105223861607313,
+      "loss": 1.2379,
+      "step": 3800
+    },
+    {
+      "epoch": 0.33361218461234643,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0025101955574707248,
+      "loss": 1.2303,
+      "step": 3801
+    },
+    {
+      "epoch": 0.33369995419524895,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0025098686438828205,
+      "loss": 1.3177,
+      "step": 3802
+    },
+    {
+      "epoch": 0.3337877237781514,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.002509541645429283,
+      "loss": 1.3069,
+      "step": 3803
+    },
+    {
+      "epoch": 0.3338754933610539,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0025092145621423857,
+      "loss": 1.2034,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3339632629439564,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002508887394054412,
+      "loss": 1.2386,
+      "step": 3805
+    },
+    {
+      "epoch": 0.33405103252685886,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0025085601411976503,
+      "loss": 1.2443,
+      "step": 3806
+    },
+    {
+      "epoch": 0.3341388021097613,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0025082328036043996,
+      "loss": 1.2725,
+      "step": 3807
+    },
+    {
+      "epoch": 0.33422657169266384,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0025079053813069677,
+      "loss": 1.2294,
+      "step": 3808
+    },
+    {
+      "epoch": 0.3343143412755663,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0025075778743376687,
+      "loss": 1.1634,
+      "step": 3809
+    },
+    {
+      "epoch": 0.3344021108584688,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0025072502827288273,
+      "loss": 1.2462,
+      "step": 3810
+    },
+    {
+      "epoch": 0.3344898804413713,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0025069226065127744,
+      "loss": 1.2371,
+      "step": 3811
+    },
+    {
+      "epoch": 0.33457765002427375,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0025065948457218506,
+      "loss": 1.3102,
+      "step": 3812
+    },
+    {
+      "epoch": 0.3346654196071763,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002506267000388406,
+      "loss": 1.216,
+      "step": 3813
+    },
+    {
+      "epoch": 0.33475318919007874,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0025059390705447957,
+      "loss": 1.2178,
+      "step": 3814
+    },
+    {
+      "epoch": 0.33484095877298126,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0025056110562233862,
+      "loss": 1.261,
+      "step": 3815
+    },
+    {
+      "epoch": 0.3349287283558837,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0025052829574565513,
+      "loss": 1.2298,
+      "step": 3816
+    },
+    {
+      "epoch": 0.3350164979387862,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.002504954774276672,
+      "loss": 1.262,
+      "step": 3817
+    },
+    {
+      "epoch": 0.3351042675216887,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0025046265067161398,
+      "loss": 1.232,
+      "step": 3818
+    },
+    {
+      "epoch": 0.33519203710459117,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002504298154807353,
+      "loss": 1.2552,
+      "step": 3819
+    },
+    {
+      "epoch": 0.33527980668749363,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002503969718582718,
+      "loss": 1.1982,
+      "step": 3820
+    },
+    {
+      "epoch": 0.33536757627039615,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0025036411980746513,
+      "loss": 1.2486,
+      "step": 3821
+    },
+    {
+      "epoch": 0.3354553458532986,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0025033125933155756,
+      "loss": 1.2,
+      "step": 3822
+    },
+    {
+      "epoch": 0.33554311543620113,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002502983904337924,
+      "loss": 1.2906,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3356308850191036,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0025026551311741357,
+      "loss": 1.2372,
+      "step": 3824
+    },
+    {
+      "epoch": 0.33571865460200606,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.00250232627385666,
+      "loss": 1.2027,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3358064241849086,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0025019973324179536,
+      "loss": 1.2718,
+      "step": 3826
+    },
+    {
+      "epoch": 0.33589419376781104,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0025016683068904816,
+      "loss": 1.1979,
+      "step": 3827
+    },
+    {
+      "epoch": 0.33598196335071356,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002501339197306717,
+      "loss": 1.2185,
+      "step": 3828
+    },
+    {
+      "epoch": 0.336069732933616,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0025010100036991427,
+      "loss": 1.2562,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3361575025165185,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0025006807261002484,
+      "loss": 1.1998,
+      "step": 3830
+    },
+    {
+      "epoch": 0.336245272099421,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002500351364542532,
+      "loss": 1.2845,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3363330416823235,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002500021919058501,
+      "loss": 1.3241,
+      "step": 3832
+    },
+    {
+      "epoch": 0.33642081126522594,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00249969238968067,
+      "loss": 1.2449,
+      "step": 3833
+    },
+    {
+      "epoch": 0.33650858084812846,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002499362776441562,
+      "loss": 1.3042,
+      "step": 3834
+    },
+    {
+      "epoch": 0.3365963504310309,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0024990330793737093,
+      "loss": 1.2679,
+      "step": 3835
+    },
+    {
+      "epoch": 0.33668412001393344,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00249870329850965,
+      "loss": 1.246,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3367718895968359,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0024983734338819338,
+      "loss": 1.2822,
+      "step": 3837
+    },
+    {
+      "epoch": 0.33685965917973837,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0024980434855231166,
+      "loss": 1.178,
+      "step": 3838
+    },
+    {
+      "epoch": 0.3369474287626409,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002497713453465763,
+      "loss": 1.3513,
+      "step": 3839
+    },
+    {
+      "epoch": 0.33703519834554335,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0024973833377424457,
+      "loss": 1.198,
+      "step": 3840
+    },
+    {
+      "epoch": 0.33712296792844587,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002497053138385746,
+      "loss": 1.2593,
+      "step": 3841
+    },
+    {
+      "epoch": 0.33721073751134834,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002496722855428253,
+      "loss": 1.2104,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3372985070942508,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0024963924889025648,
+      "loss": 1.2327,
+      "step": 3843
+    },
+    {
+      "epoch": 0.3373862766771533,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0024960620388412864,
+      "loss": 1.2432,
+      "step": 3844
+    },
+    {
+      "epoch": 0.3374740462600558,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0024957315052770326,
+      "loss": 1.2453,
+      "step": 3845
+    },
+    {
+      "epoch": 0.33756181584295825,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0024954008882424254,
+      "loss": 1.2313,
+      "step": 3846
+    },
+    {
+      "epoch": 0.33764958542586077,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002495070187770096,
+      "loss": 1.2881,
+      "step": 3847
+    },
+    {
+      "epoch": 0.33773735500876323,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0024947394038926833,
+      "loss": 1.2224,
+      "step": 3848
+    },
+    {
+      "epoch": 0.33782512459166575,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0024944085366428335,
+      "loss": 1.2479,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3379128941745682,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0024940775860532024,
+      "loss": 1.2176,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3380006637574707,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0024937465521564527,
+      "loss": 1.2383,
+      "step": 3851
+    },
+    {
+      "epoch": 0.3380884333403732,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0024934154349852576,
+      "loss": 1.2914,
+      "step": 3852
+    },
+    {
+      "epoch": 0.33817620292327566,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002493084234572296,
+      "loss": 1.2776,
+      "step": 3853
+    },
+    {
+      "epoch": 0.3382639725061782,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002492752950950256,
+      "loss": 1.2459,
+      "step": 3854
+    },
+    {
+      "epoch": 0.33835174208908064,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0024924215841518353,
+      "loss": 1.2197,
+      "step": 3855
+    },
+    {
+      "epoch": 0.3384395116719831,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002492090134209737,
+      "loss": 1.276,
+      "step": 3856
+    },
+    {
+      "epoch": 0.3385272812548856,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002491758601156674,
+      "loss": 1.2357,
+      "step": 3857
+    },
+    {
+      "epoch": 0.3386150508377881,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0024914269850253683,
+      "loss": 1.21,
+      "step": 3858
+    },
+    {
+      "epoch": 0.3387028204206906,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0024910952858485484,
+      "loss": 1.2655,
+      "step": 3859
+    },
+    {
+      "epoch": 0.3387905900035931,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.002490763503658952,
+      "loss": 1.2379,
+      "step": 3860
+    },
+    {
+      "epoch": 0.33887835958649554,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0024904316384893243,
+      "loss": 1.2728,
+      "step": 3861
+    },
+    {
+      "epoch": 0.33896612916939806,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0024900996903724193,
+      "loss": 1.2496,
+      "step": 3862
+    },
+    {
+      "epoch": 0.3390538987523005,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0024897676593409993,
+      "loss": 1.2273,
+      "step": 3863
+    },
+    {
+      "epoch": 0.339141668335203,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0024894355454278335,
+      "loss": 1.2697,
+      "step": 3864
+    },
+    {
+      "epoch": 0.3392294379181055,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0024891033486657012,
+      "loss": 1.282,
+      "step": 3865
+    },
+    {
+      "epoch": 0.33931720750100797,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0024887710690873885,
+      "loss": 1.2563,
+      "step": 3866
+    },
+    {
+      "epoch": 0.3394049770839105,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0024884387067256905,
+      "loss": 1.1868,
+      "step": 3867
+    },
+    {
+      "epoch": 0.33949274666681295,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0024881062616134092,
+      "loss": 1.1902,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3395805162497154,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0024877737337833565,
+      "loss": 1.2901,
+      "step": 3869
+    },
+    {
+      "epoch": 0.33966828583261793,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0024874411232683508,
+      "loss": 1.2237,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3397560554155204,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00248710843010122,
+      "loss": 1.2257,
+      "step": 3871
+    },
+    {
+      "epoch": 0.3398438249984229,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002486775654314799,
+      "loss": 1.2259,
+      "step": 3872
+    },
+    {
+      "epoch": 0.3399315945813254,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0024864427959419323,
+      "loss": 1.2533,
+      "step": 3873
+    },
+    {
+      "epoch": 0.34001936416422784,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002486109855015472,
+      "loss": 1.2246,
+      "step": 3874
+    },
+    {
+      "epoch": 0.34010713374713036,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0024857768315682763,
+      "loss": 1.265,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3401949033300328,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0024854437256332145,
+      "loss": 1.2477,
+      "step": 3876
+    },
+    {
+      "epoch": 0.3402826729129353,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0024851105372431623,
+      "loss": 1.2539,
+      "step": 3877
+    },
+    {
+      "epoch": 0.3403704424958378,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0024847772664310054,
+      "loss": 1.2641,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3404582120787403,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0024844439132296345,
+      "loss": 1.2266,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3405459816616428,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002484110477671952,
+      "loss": 1.2298,
+      "step": 3880
+    },
+    {
+      "epoch": 0.34063375124454526,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0024837769597908644,
+      "loss": 1.223,
+      "step": 3881
+    },
+    {
+      "epoch": 0.3407215208274477,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0024834433596192906,
+      "loss": 1.2716,
+      "step": 3882
+    },
+    {
+      "epoch": 0.34080929041035024,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002483109677190156,
+      "loss": 1.2197,
+      "step": 3883
+    },
+    {
+      "epoch": 0.3408970599932527,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0024827759125363915,
+      "loss": 1.2683,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3409848295761552,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0024824420656909397,
+      "loss": 1.2819,
+      "step": 3885
+    },
+    {
+      "epoch": 0.3410725991590577,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00248210813668675,
+      "loss": 1.2713,
+      "step": 3886
+    },
+    {
+      "epoch": 0.34116036874196015,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0024817741255567798,
+      "loss": 1.2244,
+      "step": 3887
+    },
+    {
+      "epoch": 0.34124813832486267,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0024814400323339945,
+      "loss": 1.2013,
+      "step": 3888
+    },
+    {
+      "epoch": 0.34133590790776513,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0024811058570513678,
+      "loss": 1.2884,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3414236774906676,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0024807715997418817,
+      "loss": 1.2264,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3415114470735701,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0024804372604385256,
+      "loss": 1.1952,
+      "step": 3891
+    },
+    {
+      "epoch": 0.3415992166564726,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002480102839174298,
+      "loss": 1.2594,
+      "step": 3892
+    },
+    {
+      "epoch": 0.3416869862393751,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002479768335982205,
+      "loss": 1.2732,
+      "step": 3893
+    },
+    {
+      "epoch": 0.34177475582227757,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0024794337508952597,
+      "loss": 1.1971,
+      "step": 3894
+    },
+    {
+      "epoch": 0.34186252540518003,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.002479099083946486,
+      "loss": 1.3004,
+      "step": 3895
+    },
+    {
+      "epoch": 0.34195029498808255,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.002478764335168912,
+      "loss": 1.2729,
+      "step": 3896
+    },
+    {
+      "epoch": 0.342038064570985,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0024784295045955787,
+      "loss": 1.2337,
+      "step": 3897
+    },
+    {
+      "epoch": 0.34212583415388753,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0024780945922595306,
+      "loss": 1.2409,
+      "step": 3898
+    },
+    {
+      "epoch": 0.34221360373679,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002477759598193823,
+      "loss": 1.2426,
+      "step": 3899
+    },
+    {
+      "epoch": 0.34230137331969246,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0024774245224315184,
+      "loss": 1.2438,
+      "step": 3900
+    },
+    {
+      "epoch": 0.342389142902595,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002477089365005687,
+      "loss": 1.2612,
+      "step": 3901
+    },
+    {
+      "epoch": 0.34247691248549744,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0024767541259494087,
+      "loss": 1.2024,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3425646820683999,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002476418805295768,
+      "loss": 1.201,
+      "step": 3903
+    },
+    {
+      "epoch": 0.3426524516513024,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0024760834030778628,
+      "loss": 1.2228,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3427402212342049,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002475747919328794,
+      "loss": 1.2355,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3428279908171074,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0024754123540816726,
+      "loss": 1.2358,
+      "step": 3906
+    },
+    {
+      "epoch": 0.3429157604000099,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002475076707369618,
+      "loss": 1.2443,
+      "step": 3907
+    },
+    {
+      "epoch": 0.34300352998291234,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0024747409792257565,
+      "loss": 1.2954,
+      "step": 3908
+    },
+    {
+      "epoch": 0.34309129956581486,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0024744051696832243,
+      "loss": 1.2233,
+      "step": 3909
+    },
+    {
+      "epoch": 0.3431790691487173,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0024740692787751633,
+      "loss": 1.2086,
+      "step": 3910
+    },
+    {
+      "epoch": 0.34326683873161984,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002473733306534726,
+      "loss": 1.2927,
+      "step": 3911
+    },
+    {
+      "epoch": 0.3433546083145223,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0024733972529950698,
+      "loss": 1.2207,
+      "step": 3912
+    },
+    {
+      "epoch": 0.34344237789742477,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0024730611181893636,
+      "loss": 1.2145,
+      "step": 3913
+    },
+    {
+      "epoch": 0.3435301474803273,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002472724902150781,
+      "loss": 1.2483,
+      "step": 3914
+    },
+    {
+      "epoch": 0.34361791706322975,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0024723886049125058,
+      "loss": 1.2293,
+      "step": 3915
+    },
+    {
+      "epoch": 0.34370568664613227,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00247205222650773,
+      "loss": 1.2995,
+      "step": 3916
+    },
+    {
+      "epoch": 0.34379345622903473,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002471715766969652,
+      "loss": 1.2245,
+      "step": 3917
+    },
+    {
+      "epoch": 0.3438812258119372,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002471379226331479,
+      "loss": 1.2291,
+      "step": 3918
+    },
+    {
+      "epoch": 0.3439689953948397,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0024710426046264263,
+      "loss": 1.2931,
+      "step": 3919
+    },
+    {
+      "epoch": 0.3440567649777422,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0024707059018877174,
+      "loss": 1.2876,
+      "step": 3920
+    },
+    {
+      "epoch": 0.34414453456064464,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0024703691181485834,
+      "loss": 1.1831,
+      "step": 3921
+    },
+    {
+      "epoch": 0.34423230414354716,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0024700322534422633,
+      "loss": 1.2332,
+      "step": 3922
+    },
+    {
+      "epoch": 0.3443200737264496,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002469695307802005,
+      "loss": 1.2615,
+      "step": 3923
+    },
+    {
+      "epoch": 0.34440784330935215,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0024693582812610622,
+      "loss": 1.2372,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3444956128922546,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0024690211738526992,
+      "loss": 1.2727,
+      "step": 3925
+    },
+    {
+      "epoch": 0.3445833824751571,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002468683985610188,
+      "loss": 1.2459,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3446711520580596,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.002468346716566806,
+      "loss": 1.2815,
+      "step": 3927
+    },
+    {
+      "epoch": 0.34475892164096206,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002468009366755841,
+      "loss": 1.199,
+      "step": 3928
+    },
+    {
+      "epoch": 0.3448466912238646,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002467671936210589,
+      "loss": 1.1999,
+      "step": 3929
+    },
+    {
+      "epoch": 0.34493446080676704,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0024673344249643517,
+      "loss": 1.2399,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3450222303896695,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002466996833050441,
+      "loss": 1.2403,
+      "step": 3931
+    },
+    {
+      "epoch": 0.345109999972572,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0024666591605021753,
+      "loss": 1.2722,
+      "step": 3932
+    },
+    {
+      "epoch": 0.3451977695554745,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0024663214073528824,
+      "loss": 1.2229,
+      "step": 3933
+    },
+    {
+      "epoch": 0.34528553913837695,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002465983573635896,
+      "loss": 1.2623,
+      "step": 3934
+    },
+    {
+      "epoch": 0.34537330872127947,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0024656456593845594,
+      "loss": 1.2396,
+      "step": 3935
+    },
+    {
+      "epoch": 0.34546107830418193,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0024653076646322237,
+      "loss": 1.2818,
+      "step": 3936
+    },
+    {
+      "epoch": 0.34554884788708445,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0024649695894122478,
+      "loss": 1.1482,
+      "step": 3937
+    },
+    {
+      "epoch": 0.3456366174699869,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002464631433757998,
+      "loss": 1.2432,
+      "step": 3938
+    },
+    {
+      "epoch": 0.3457243870528894,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.002464293197702849,
+      "loss": 1.2769,
+      "step": 3939
+    },
+    {
+      "epoch": 0.3458121566357919,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0024639548812801833,
+      "loss": 1.2549,
+      "step": 3940
+    },
+    {
+      "epoch": 0.34589992621869436,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.002463616484523392,
+      "loss": 1.2159,
+      "step": 3941
+    },
+    {
+      "epoch": 0.3459876958015969,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002463278007465872,
+      "loss": 1.2639,
+      "step": 3942
+    },
+    {
+      "epoch": 0.34607546538449935,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0024629394501410316,
+      "loss": 1.2568,
+      "step": 3943
+    },
+    {
+      "epoch": 0.3461632349674018,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0024626008125822833,
+      "loss": 1.2463,
+      "step": 3944
+    },
+    {
+      "epoch": 0.34625100455030433,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0024622620948230507,
+      "loss": 1.2087,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3463387741332068,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0024619232968967632,
+      "loss": 1.268,
+      "step": 3946
+    },
+    {
+      "epoch": 0.34642654371610926,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0024615844188368586,
+      "loss": 1.2481,
+      "step": 3947
+    },
+    {
+      "epoch": 0.3465143132990118,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.002461245460676783,
+      "loss": 1.3106,
+      "step": 3948
+    },
+    {
+      "epoch": 0.34660208288191424,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0024609064224499913,
+      "loss": 1.3123,
+      "step": 3949
+    },
+    {
+      "epoch": 0.34668985246481676,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0024605673041899435,
+      "loss": 1.1776,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3467776220477192,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0024602281059301107,
+      "loss": 1.2278,
+      "step": 3951
+    },
+    {
+      "epoch": 0.3468653916306217,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002459888827703969,
+      "loss": 1.236,
+      "step": 3952
+    },
+    {
+      "epoch": 0.3469531612135242,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002459549469545005,
+      "loss": 1.2221,
+      "step": 3953
+    },
+    {
+      "epoch": 0.34704093079642667,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002459210031486711,
+      "loss": 1.2717,
+      "step": 3954
+    },
+    {
+      "epoch": 0.3471287003793292,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0024588705135625897,
+      "loss": 1.258,
+      "step": 3955
+    },
+    {
+      "epoch": 0.34721646996223166,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0024585309158061484,
+      "loss": 1.31,
+      "step": 3956
+    },
+    {
+      "epoch": 0.3473042395451341,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0024581912382509053,
+      "loss": 1.2284,
+      "step": 3957
+    },
+    {
+      "epoch": 0.34739200912803664,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002457851480930385,
+      "loss": 1.234,
+      "step": 3958
+    },
+    {
+      "epoch": 0.3474797787109391,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00245751164387812,
+      "loss": 1.2052,
+      "step": 3959
+    },
+    {
+      "epoch": 0.34756754829384157,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0024571717271276504,
+      "loss": 1.3198,
+      "step": 3960
+    },
+    {
+      "epoch": 0.3476553178767441,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0024568317307125257,
+      "loss": 1.2388,
+      "step": 3961
+    },
+    {
+      "epoch": 0.34774308745964655,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0024564916546663015,
+      "loss": 1.2584,
+      "step": 3962
+    },
+    {
+      "epoch": 0.34783085704254907,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.002456151499022542,
+      "loss": 1.2585,
+      "step": 3963
+    },
+    {
+      "epoch": 0.34791862662545153,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0024558112638148194,
+      "loss": 1.2513,
+      "step": 3964
+    },
+    {
+      "epoch": 0.348006396208354,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.002455470949076714,
+      "loss": 1.2528,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3480941657912565,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002455130554841812,
+      "loss": 1.2199,
+      "step": 3966
+    },
+    {
+      "epoch": 0.348181935374159,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0024547900811437108,
+      "loss": 1.1906,
+      "step": 3967
+    },
+    {
+      "epoch": 0.3482697049570615,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0024544495280160125,
+      "loss": 1.1924,
+      "step": 3968
+    },
+    {
+      "epoch": 0.34835747453996396,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002454108895492329,
+      "loss": 1.2129,
+      "step": 3969
+    },
+    {
+      "epoch": 0.3484452441228664,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0024537681836062793,
+      "loss": 1.2634,
+      "step": 3970
+    },
+    {
+      "epoch": 0.34853301370576895,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00245342739239149,
+      "loss": 1.2644,
+      "step": 3971
+    },
+    {
+      "epoch": 0.3486207832886714,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0024530865218815963,
+      "loss": 1.2404,
+      "step": 3972
+    },
+    {
+      "epoch": 0.3487085528715739,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0024527455721102404,
+      "loss": 1.2037,
+      "step": 3973
+    },
+    {
+      "epoch": 0.3487963224544764,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0024524045431110735,
+      "loss": 1.2171,
+      "step": 3974
+    },
+    {
+      "epoch": 0.34888409203737886,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002452063434917752,
+      "loss": 1.2733,
+      "step": 3975
+    },
+    {
+      "epoch": 0.3489718616202814,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0024517222475639437,
+      "loss": 1.2401,
+      "step": 3976
+    },
+    {
+      "epoch": 0.34905963120318384,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0024513809810833218,
+      "loss": 1.3198,
+      "step": 3977
+    },
+    {
+      "epoch": 0.3491474007860863,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0024510396355095676,
+      "loss": 1.2371,
+      "step": 3978
+    },
+    {
+      "epoch": 0.3492351703689888,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0024506982108763718,
+      "loss": 1.2926,
+      "step": 3979
+    },
+    {
+      "epoch": 0.3493229399518913,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.00245035670721743,
+      "loss": 1.2289,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3494107095347938,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0024500151245664486,
+      "loss": 1.2357,
+      "step": 3981
+    },
+    {
+      "epoch": 0.34949847911769627,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002449673462957139,
+      "loss": 1.1955,
+      "step": 3982
+    },
+    {
+      "epoch": 0.34958624870059873,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002449331722423224,
+      "loss": 1.3234,
+      "step": 3983
+    },
+    {
+      "epoch": 0.34967401828350125,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00244898990299843,
+      "loss": 1.264,
+      "step": 3984
+    },
+    {
+      "epoch": 0.3497617878664037,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002448648004716494,
+      "loss": 1.2428,
+      "step": 3985
+    },
+    {
+      "epoch": 0.34984955744930624,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0024483060276111602,
+      "loss": 1.2085,
+      "step": 3986
+    },
+    {
+      "epoch": 0.3499373270322087,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0024479639717161802,
+      "loss": 1.2003,
+      "step": 3987
+    },
+    {
+      "epoch": 0.35002509661511116,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0024476218370653134,
+      "loss": 1.2714,
+      "step": 3988
+    },
+    {
+      "epoch": 0.3501128661980137,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002447279623692328,
+      "loss": 1.2583,
+      "step": 3989
+    },
+    {
+      "epoch": 0.35020063578091615,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002446937331630998,
+      "loss": 1.2962,
+      "step": 3990
+    },
+    {
+      "epoch": 0.3502884053638186,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0024465949609151065,
+      "loss": 1.2294,
+      "step": 3991
+    },
+    {
+      "epoch": 0.35037617494672113,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002446252511578444,
+      "loss": 1.2503,
+      "step": 3992
+    },
+    {
+      "epoch": 0.3504639445296236,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00244590998365481,
+      "loss": 1.222,
+      "step": 3993
+    },
+    {
+      "epoch": 0.3505517141125261,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0024455673771780094,
+      "loss": 1.2856,
+      "step": 3994
+    },
+    {
+      "epoch": 0.3506394836954286,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002445224692181857,
+      "loss": 1.2407,
+      "step": 3995
+    },
+    {
+      "epoch": 0.35072725327833104,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0024448819287001735,
+      "loss": 1.2514,
+      "step": 3996
+    },
+    {
+      "epoch": 0.35081502286123356,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002444539086766789,
+      "loss": 1.2559,
+      "step": 3997
+    },
+    {
+      "epoch": 0.350902792444136,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0024441961664155402,
+      "loss": 1.235,
+      "step": 3998
+    },
+    {
+      "epoch": 0.35099056202703854,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0024438531676802726,
+      "loss": 1.2386,
+      "step": 3999
+    },
+    {
+      "epoch": 0.351078331609941,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002443510090594839,
+      "loss": 1.2591,
+      "step": 4000
+    },
+    {
+      "epoch": 0.351078331609941,
+      "eval_loss": 1.231809139251709,
+      "eval_runtime": 437.7571,
+      "eval_samples_per_second": 33.672,
+      "eval_steps_per_second": 8.418,
+      "step": 4000
+    },
+    {
+      "epoch": 0.35116610119284347,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002443166935193099,
+      "loss": 1.245,
+      "step": 4001
+    },
+    {
+      "epoch": 0.351253870775746,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00244282370150892,
+      "loss": 1.224,
+      "step": 4002
+    },
+    {
+      "epoch": 0.35134164035864845,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0024424803895761794,
+      "loss": 1.2363,
+      "step": 4003
+    },
+    {
+      "epoch": 0.3514294099415509,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0024421369994287595,
+      "loss": 1.2285,
+      "step": 4004
+    },
+    {
+      "epoch": 0.35151717952445344,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0024417935311005527,
+      "loss": 1.2317,
+      "step": 4005
+    },
+    {
+      "epoch": 0.3516049491073559,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002441449984625457,
+      "loss": 1.2532,
+      "step": 4006
+    },
+    {
+      "epoch": 0.3516927186902584,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00244110636003738,
+      "loss": 1.2155,
+      "step": 4007
+    },
+    {
+      "epoch": 0.3517804882731609,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.002440762657370235,
+      "loss": 1.2521,
+      "step": 4008
+    },
+    {
+      "epoch": 0.35186825785606335,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0024404188766579448,
+      "loss": 1.26,
+      "step": 4009
+    },
+    {
+      "epoch": 0.35195602743896587,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002440075017934439,
+      "loss": 1.2432,
+      "step": 4010
+    },
+    {
+      "epoch": 0.35204379702186833,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0024397310812336545,
+      "loss": 1.2226,
+      "step": 4011
+    },
+    {
+      "epoch": 0.35213156660477085,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002439387066589538,
+      "loss": 1.2395,
+      "step": 4012
+    },
+    {
+      "epoch": 0.3522193361876733,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002439042974036041,
+      "loss": 1.2156,
+      "step": 4013
+    },
+    {
+      "epoch": 0.3523071057705758,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0024386988036071244,
+      "loss": 1.2281,
+      "step": 4014
+    },
+    {
+      "epoch": 0.3523948753534783,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002438354555336757,
+      "loss": 1.2281,
+      "step": 4015
+    },
+    {
+      "epoch": 0.35248264493638076,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002438010229258914,
+      "loss": 1.2613,
+      "step": 4016
+    },
+    {
+      "epoch": 0.3525704145192832,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0024376658254075794,
+      "loss": 1.2074,
+      "step": 4017
+    },
+    {
+      "epoch": 0.35265818410218575,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0024373213438167443,
+      "loss": 1.224,
+      "step": 4018
+    },
+    {
+      "epoch": 0.3527459536850882,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0024369767845204087,
+      "loss": 1.2414,
+      "step": 4019
+    },
+    {
+      "epoch": 0.35283372326799073,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0024366321475525776,
+      "loss": 1.2477,
+      "step": 4020
+    },
+    {
+      "epoch": 0.3529214928508932,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0024362874329472667,
+      "loss": 1.1795,
+      "step": 4021
+    },
+    {
+      "epoch": 0.35300926243379566,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0024359426407384965,
+      "loss": 1.2714,
+      "step": 4022
+    },
+    {
+      "epoch": 0.3530970320166982,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002435597770960298,
+      "loss": 1.3021,
+      "step": 4023
+    },
+    {
+      "epoch": 0.35318480159960064,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0024352528236467083,
+      "loss": 1.2611,
+      "step": 4024
+    },
+    {
+      "epoch": 0.35327257118250316,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002434907798831771,
+      "loss": 1.2246,
+      "step": 4025
+    },
+    {
+      "epoch": 0.3533603407654056,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0024345626965495406,
+      "loss": 1.2471,
+      "step": 4026
+    },
+    {
+      "epoch": 0.3534481103483081,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.002434217516834076,
+      "loss": 1.2298,
+      "step": 4027
+    },
+    {
+      "epoch": 0.3535358799312106,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0024338722597194457,
+      "loss": 1.2836,
+      "step": 4028
+    },
+    {
+      "epoch": 0.35362364951411307,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0024335269252397248,
+      "loss": 1.2222,
+      "step": 4029
+    },
+    {
+      "epoch": 0.35371141909701553,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0024331815134289968,
+      "loss": 1.2627,
+      "step": 4030
+    },
+    {
+      "epoch": 0.35379918867991805,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0024328360243213525,
+      "loss": 1.2837,
+      "step": 4031
+    },
+    {
+      "epoch": 0.3538869582628205,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0024324904579508893,
+      "loss": 1.2379,
+      "step": 4032
+    },
+    {
+      "epoch": 0.35397472784572304,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0024321448143517148,
+      "loss": 1.2551,
+      "step": 4033
+    },
+    {
+      "epoch": 0.3540624974286255,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.002431799093557942,
+      "loss": 1.2164,
+      "step": 4034
+    },
+    {
+      "epoch": 0.35415026701152796,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0024314532956036914,
+      "loss": 1.2226,
+      "step": 4035
+    },
+    {
+      "epoch": 0.3542380365944305,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0024311074205230934,
+      "loss": 1.2372,
+      "step": 4036
+    },
+    {
+      "epoch": 0.35432580617733295,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0024307614683502826,
+      "loss": 1.3279,
+      "step": 4037
+    },
+    {
+      "epoch": 0.35441357576023547,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002430415439119405,
+      "loss": 1.2822,
+      "step": 4038
+    },
+    {
+      "epoch": 0.35450134534313793,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0024300693328646117,
+      "loss": 1.2036,
+      "step": 4039
+    },
+    {
+      "epoch": 0.3545891149260404,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002429723149620061,
+      "loss": 1.2444,
+      "step": 4040
+    },
+    {
+      "epoch": 0.3546768845089429,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002429376889419922,
+      "loss": 1.2753,
+      "step": 4041
+    },
+    {
+      "epoch": 0.3547646540918454,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0024290305522983663,
+      "loss": 1.2824,
+      "step": 4042
+    },
+    {
+      "epoch": 0.35485242367474784,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002428684138289578,
+      "loss": 1.2159,
+      "step": 4043
+    },
+    {
+      "epoch": 0.35494019325765036,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002428337647427747,
+      "loss": 1.2922,
+      "step": 4044
+    },
+    {
+      "epoch": 0.3550279628405528,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.002427991079747069,
+      "loss": 1.2521,
+      "step": 4045
+    },
+    {
+      "epoch": 0.35511573242345534,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002427644435281751,
+      "loss": 1.2519,
+      "step": 4046
+    },
+    {
+      "epoch": 0.3552035020063578,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0024272977140660035,
+      "loss": 1.3237,
+      "step": 4047
+    },
+    {
+      "epoch": 0.35529127158926027,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0024269509161340473,
+      "loss": 1.215,
+      "step": 4048
+    },
+    {
+      "epoch": 0.3553790411721628,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00242660404152011,
+      "loss": 1.2304,
+      "step": 4049
+    },
+    {
+      "epoch": 0.35546681075506525,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0024262570902584264,
+      "loss": 1.3023,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3555545803379678,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0024259100623832395,
+      "loss": 1.2725,
+      "step": 4051
+    },
+    {
+      "epoch": 0.35564234992087024,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0024255629579287997,
+      "loss": 1.1885,
+      "step": 4052
+    },
+    {
+      "epoch": 0.3557301195037727,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0024252157769293647,
+      "loss": 1.2408,
+      "step": 4053
+    },
+    {
+      "epoch": 0.3558178890866752,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0024248685194191997,
+      "loss": 1.2666,
+      "step": 4054
+    },
+    {
+      "epoch": 0.3559056586695777,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002424521185432578,
+      "loss": 1.2805,
+      "step": 4055
+    },
+    {
+      "epoch": 0.3559934282524802,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00242417377500378,
+      "loss": 1.2469,
+      "step": 4056
+    },
+    {
+      "epoch": 0.35608119783538267,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0024238262881670935,
+      "loss": 1.2282,
+      "step": 4057
+    },
+    {
+      "epoch": 0.35616896741828513,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0024234787249568144,
+      "loss": 1.2184,
+      "step": 4058
+    },
+    {
+      "epoch": 0.35625673700118765,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0024231310854072458,
+      "loss": 1.3039,
+      "step": 4059
+    },
+    {
+      "epoch": 0.3563445065840901,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002422783369552698,
+      "loss": 1.2545,
+      "step": 4060
+    },
+    {
+      "epoch": 0.3564322761669926,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002422435577427489,
+      "loss": 1.2207,
+      "step": 4061
+    },
+    {
+      "epoch": 0.3565200457498951,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002422087709065945,
+      "loss": 1.2464,
+      "step": 4062
+    },
+    {
+      "epoch": 0.35660781533279756,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0024217397645023998,
+      "loss": 1.2484,
+      "step": 4063
+    },
+    {
+      "epoch": 0.3566955849157001,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0024213917437711932,
+      "loss": 1.2868,
+      "step": 4064
+    },
+    {
+      "epoch": 0.35678335449860255,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002421043646906673,
+      "loss": 1.2181,
+      "step": 4065
+    },
+    {
+      "epoch": 0.356871124081505,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002420695473943197,
+      "loss": 1.28,
+      "step": 4066
+    },
+    {
+      "epoch": 0.35695889366440753,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0024203472249151262,
+      "loss": 1.2527,
+      "step": 4067
+    },
+    {
+      "epoch": 0.35704666324731,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0024199988998568323,
+      "loss": 1.1978,
+      "step": 4068
+    },
+    {
+      "epoch": 0.3571344328302125,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0024196504988026946,
+      "loss": 1.2356,
+      "step": 4069
+    },
+    {
+      "epoch": 0.357222202413115,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002419302021787097,
+      "loss": 1.2529,
+      "step": 4070
+    },
+    {
+      "epoch": 0.35730997199601744,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0024189534688444344,
+      "loss": 1.2146,
+      "step": 4071
+    },
+    {
+      "epoch": 0.35739774157891996,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0024186048400091074,
+      "loss": 1.2665,
+      "step": 4072
+    },
+    {
+      "epoch": 0.3574855111618224,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0024182561353155232,
+      "loss": 1.2542,
+      "step": 4073
+    },
+    {
+      "epoch": 0.3575732807447249,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0024179073547980984,
+      "loss": 1.1837,
+      "step": 4074
+    },
+    {
+      "epoch": 0.3576610503276274,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0024175584984912563,
+      "loss": 1.255,
+      "step": 4075
+    },
+    {
+      "epoch": 0.35774881991052987,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0024172095664294277,
+      "loss": 1.2374,
+      "step": 4076
+    },
+    {
+      "epoch": 0.3578365894934324,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0024168605586470503,
+      "loss": 1.2498,
+      "step": 4077
+    },
+    {
+      "epoch": 0.35792435907633485,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00241651147517857,
+      "loss": 1.2039,
+      "step": 4078
+    },
+    {
+      "epoch": 0.3580121286592373,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00241616231605844,
+      "loss": 1.199,
+      "step": 4079
+    },
+    {
+      "epoch": 0.35809989824213984,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0024158130813211214,
+      "loss": 1.2975,
+      "step": 4080
+    },
+    {
+      "epoch": 0.3581876678250423,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0024154637710010816,
+      "loss": 1.2954,
+      "step": 4081
+    },
+    {
+      "epoch": 0.3582754374079448,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0024151143851327966,
+      "loss": 1.2048,
+      "step": 4082
+    },
+    {
+      "epoch": 0.3583632069908473,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0024147649237507484,
+      "loss": 1.2391,
+      "step": 4083
+    },
+    {
+      "epoch": 0.35845097657374975,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002414415386889429,
+      "loss": 1.2503,
+      "step": 4084
+    },
+    {
+      "epoch": 0.35853874615665227,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0024140657745833354,
+      "loss": 1.2714,
+      "step": 4085
+    },
+    {
+      "epoch": 0.35862651573955473,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002413716086866973,
+      "loss": 1.198,
+      "step": 4086
+    },
+    {
+      "epoch": 0.3587142853224572,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0024133663237748547,
+      "loss": 1.1928,
+      "step": 4087
+    },
+    {
+      "epoch": 0.3588020549053597,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0024130164853415005,
+      "loss": 1.2913,
+      "step": 4088
+    },
+    {
+      "epoch": 0.3588898244882622,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002412666571601439,
+      "loss": 1.2147,
+      "step": 4089
+    },
+    {
+      "epoch": 0.3589775940711647,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002412316582589204,
+      "loss": 1.263,
+      "step": 4090
+    },
+    {
+      "epoch": 0.35906536365406716,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002411966518339339,
+      "loss": 1.2843,
+      "step": 4091
+    },
+    {
+      "epoch": 0.3591531332369696,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0024116163788863933,
+      "loss": 1.2045,
+      "step": 4092
+    },
+    {
+      "epoch": 0.35924090281987214,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002411266164264925,
+      "loss": 1.2543,
+      "step": 4093
+    },
+    {
+      "epoch": 0.3593286724027746,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002410915874509498,
+      "loss": 1.161,
+      "step": 4094
+    },
+    {
+      "epoch": 0.3594164419856771,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0024105655096546847,
+      "loss": 1.2318,
+      "step": 4095
+    },
+    {
+      "epoch": 0.3595042115685796,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002410215069735066,
+      "loss": 1.217,
+      "step": 4096
+    },
+    {
+      "epoch": 0.35959198115148205,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002409864554785227,
+      "loss": 1.2116,
+      "step": 4097
+    },
+    {
+      "epoch": 0.3596797507343846,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002409513964839764,
+      "loss": 1.2582,
+      "step": 4098
+    },
+    {
+      "epoch": 0.35976752031728704,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002409163299933277,
+      "loss": 1.221,
+      "step": 4099
+    },
+    {
+      "epoch": 0.3598552899001895,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0024088125601003764,
+      "loss": 1.1779,
+      "step": 4100
+    },
+    {
+      "epoch": 0.359943059483092,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0024084617453756785,
+      "loss": 1.2669,
+      "step": 4101
+    },
+    {
+      "epoch": 0.3600308290659945,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002408110855793808,
+      "loss": 1.2777,
+      "step": 4102
+    },
+    {
+      "epoch": 0.360118598648897,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002407759891389395,
+      "loss": 1.2349,
+      "step": 4103
+    },
+    {
+      "epoch": 0.36020636823179947,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0024074088521970796,
+      "loss": 1.2044,
+      "step": 4104
+    },
+    {
+      "epoch": 0.36029413781470193,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002407057738251507,
+      "loss": 1.1874,
+      "step": 4105
+    },
+    {
+      "epoch": 0.36038190739760445,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002406706549587331,
+      "loss": 1.1814,
+      "step": 4106
+    },
+    {
+      "epoch": 0.3604696769805069,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0024063552862392135,
+      "loss": 1.2386,
+      "step": 4107
+    },
+    {
+      "epoch": 0.36055744656340943,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0024060039482418213,
+      "loss": 1.2671,
+      "step": 4108
+    },
+    {
+      "epoch": 0.3606452161463119,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002405652535629831,
+      "loss": 1.1842,
+      "step": 4109
+    },
+    {
+      "epoch": 0.36073298572921436,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0024053010484379254,
+      "loss": 1.2425,
+      "step": 4110
+    },
+    {
+      "epoch": 0.3608207553121169,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0024049494867007947,
+      "loss": 1.2358,
+      "step": 4111
+    },
+    {
+      "epoch": 0.36090852489501934,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0024045978504531377,
+      "loss": 1.3355,
+      "step": 4112
+    },
+    {
+      "epoch": 0.3609962944779218,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0024042461397296576,
+      "loss": 1.2573,
+      "step": 4113
+    },
+    {
+      "epoch": 0.36108406406082433,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0024038943545650684,
+      "loss": 1.2416,
+      "step": 4114
+    },
+    {
+      "epoch": 0.3611718336437268,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002403542494994089,
+      "loss": 1.28,
+      "step": 4115
+    },
+    {
+      "epoch": 0.3612596032266293,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0024031905610514474,
+      "loss": 1.26,
+      "step": 4116
+    },
+    {
+      "epoch": 0.3613473728095318,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0024028385527718778,
+      "loss": 1.2549,
+      "step": 4117
+    },
+    {
+      "epoch": 0.36143514239243424,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0024024864701901213,
+      "loss": 1.2275,
+      "step": 4118
+    },
+    {
+      "epoch": 0.36152291197533676,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002402134313340928,
+      "loss": 1.2519,
+      "step": 4119
+    },
+    {
+      "epoch": 0.3616106815582392,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002401782082259054,
+      "loss": 1.3033,
+      "step": 4120
+    },
+    {
+      "epoch": 0.36169845114114174,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002401429776979263,
+      "loss": 1.2646,
+      "step": 4121
+    },
+    {
+      "epoch": 0.3617862207240442,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0024010773975363265,
+      "loss": 1.2315,
+      "step": 4122
+    },
+    {
+      "epoch": 0.36187399030694667,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002400724943965022,
+      "loss": 1.2356,
+      "step": 4123
+    },
+    {
+      "epoch": 0.3619617598898492,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0024003724163001367,
+      "loss": 1.2315,
+      "step": 4124
+    },
+    {
+      "epoch": 0.36204952947275165,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0024000198145764633,
+      "loss": 1.2166,
+      "step": 4125
+    },
+    {
+      "epoch": 0.36213729905565417,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002399667138828801,
+      "loss": 1.2421,
+      "step": 4126
+    },
+    {
+      "epoch": 0.36222506863855664,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0023993143890919592,
+      "loss": 1.3185,
+      "step": 4127
+    },
+    {
+      "epoch": 0.3623128382214591,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002398961565400751,
+      "loss": 1.2335,
+      "step": 4128
+    },
+    {
+      "epoch": 0.3624006078043616,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0023986086677900012,
+      "loss": 1.2527,
+      "step": 4129
+    },
+    {
+      "epoch": 0.3624883773872641,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002398255696294537,
+      "loss": 1.211,
+      "step": 4130
+    },
+    {
+      "epoch": 0.36257614697016655,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0023979026509491964,
+      "loss": 1.2183,
+      "step": 4131
+    },
+    {
+      "epoch": 0.36266391655306907,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002397549531788824,
+      "loss": 1.295,
+      "step": 4132
+    },
+    {
+      "epoch": 0.36275168613597153,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0023971963388482705,
+      "loss": 1.2466,
+      "step": 4133
+    },
+    {
+      "epoch": 0.36283945571887405,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002396843072162394,
+      "loss": 1.1935,
+      "step": 4134
+    },
+    {
+      "epoch": 0.3629272253017765,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0023964897317660626,
+      "loss": 1.2423,
+      "step": 4135
+    },
+    {
+      "epoch": 0.363014994884679,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0023961363176941476,
+      "loss": 1.2789,
+      "step": 4136
+    },
+    {
+      "epoch": 0.3631027644675815,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0023957828299815305,
+      "loss": 1.3001,
+      "step": 4137
+    },
+    {
+      "epoch": 0.36319053405048396,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0023954292686630993,
+      "loss": 1.2591,
+      "step": 4138
+    },
+    {
+      "epoch": 0.3632783036333865,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0023950756337737483,
+      "loss": 1.2397,
+      "step": 4139
+    },
+    {
+      "epoch": 0.36336607321628894,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002394721925348381,
+      "loss": 1.2071,
+      "step": 4140
+    },
+    {
+      "epoch": 0.3634538427991914,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002394368143421906,
+      "loss": 1.2525,
+      "step": 4141
+    },
+    {
+      "epoch": 0.3635416123820939,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00239401428802924,
+      "loss": 1.2249,
+      "step": 4142
+    },
+    {
+      "epoch": 0.3636293819649964,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002393660359205309,
+      "loss": 1.2805,
+      "step": 4143
+    },
+    {
+      "epoch": 0.36371715154789885,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0023933063569850418,
+      "loss": 1.2329,
+      "step": 4144
+    },
+    {
+      "epoch": 0.3638049211308014,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.002392952281403379,
+      "loss": 1.2505,
+      "step": 4145
+    },
+    {
+      "epoch": 0.36389269071370384,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0023925981324952653,
+      "loss": 1.2752,
+      "step": 4146
+    },
+    {
+      "epoch": 0.36398046029660636,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002392243910295655,
+      "loss": 1.2574,
+      "step": 4147
+    },
+    {
+      "epoch": 0.3640682298795088,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0023918896148395073,
+      "loss": 1.2714,
+      "step": 4148
+    },
+    {
+      "epoch": 0.3641559994624113,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00239153524616179,
+      "loss": 1.2251,
+      "step": 4149
+    },
+    {
+      "epoch": 0.3642437690453138,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.002391180804297478,
+      "loss": 1.2744,
+      "step": 4150
+    },
+    {
+      "epoch": 0.36433153862821627,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002390826289281553,
+      "loss": 1.278,
+      "step": 4151
+    },
+    {
+      "epoch": 0.3644193082111188,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002390471701149005,
+      "loss": 1.1991,
+      "step": 4152
+    },
+    {
+      "epoch": 0.36450707779402125,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00239011703993483,
+      "loss": 1.2535,
+      "step": 4153
+    },
+    {
+      "epoch": 0.3645948473769237,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002389762305674032,
+      "loss": 1.1837,
+      "step": 4154
+    },
+    {
+      "epoch": 0.36468261695982623,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002389407498401621,
+      "loss": 1.1859,
+      "step": 4155
+    },
+    {
+      "epoch": 0.3647703865427287,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0023890526181526155,
+      "loss": 1.2375,
+      "step": 4156
+    },
+    {
+      "epoch": 0.36485815612563116,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002388697664962041,
+      "loss": 1.2295,
+      "step": 4157
+    },
+    {
+      "epoch": 0.3649459257085337,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0023883426388649296,
+      "loss": 1.2605,
+      "step": 4158
+    },
+    {
+      "epoch": 0.36503369529143614,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0023879875398963215,
+      "loss": 1.2524,
+      "step": 4159
+    },
+    {
+      "epoch": 0.36512146487433866,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0023876323680912634,
+      "loss": 1.2105,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3652092344572411,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002387277123484809,
+      "loss": 1.2107,
+      "step": 4161
+    },
+    {
+      "epoch": 0.3652970040401436,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00238692180611202,
+      "loss": 1.2112,
+      "step": 4162
+    },
+    {
+      "epoch": 0.3653847736230461,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0023865664160079643,
+      "loss": 1.2372,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3654725432059486,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0023862109532077178,
+      "loss": 1.225,
+      "step": 4164
+    },
+    {
+      "epoch": 0.3655603127888511,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0023858554177463636,
+      "loss": 1.2303,
+      "step": 4165
+    },
+    {
+      "epoch": 0.36564808237175356,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002385499809658991,
+      "loss": 1.2478,
+      "step": 4166
+    },
+    {
+      "epoch": 0.365735851954656,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0023851441289806977,
+      "loss": 1.2062,
+      "step": 4167
+    },
+    {
+      "epoch": 0.36582362153755854,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002384788375746587,
+      "loss": 1.288,
+      "step": 4168
+    },
+    {
+      "epoch": 0.365911391120461,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.002384432549991772,
+      "loss": 1.1998,
+      "step": 4169
+    },
+    {
+      "epoch": 0.36599916070336347,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0023840766517513696,
+      "loss": 1.1492,
+      "step": 4170
+    },
+    {
+      "epoch": 0.366086930286266,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0023837206810605072,
+      "loss": 1.2296,
+      "step": 4171
+    },
+    {
+      "epoch": 0.36617469986916845,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.002383364637954316,
+      "loss": 1.255,
+      "step": 4172
+    },
+    {
+      "epoch": 0.36626246945207097,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0023830085224679364,
+      "loss": 1.2206,
+      "step": 4173
+    },
+    {
+      "epoch": 0.36635023903497344,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0023826523346365167,
+      "loss": 1.2389,
+      "step": 4174
+    },
+    {
+      "epoch": 0.3664380086178759,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0023822960744952102,
+      "loss": 1.2005,
+      "step": 4175
+    },
+    {
+      "epoch": 0.3665257782007784,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002381939742079179,
+      "loss": 1.2614,
+      "step": 4176
+    },
+    {
+      "epoch": 0.3666135477836809,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0023815833374235913,
+      "loss": 1.2504,
+      "step": 4177
+    },
+    {
+      "epoch": 0.3667013173665834,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0023812268605636227,
+      "loss": 1.2672,
+      "step": 4178
+    },
+    {
+      "epoch": 0.36678908694948587,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002380870311534457,
+      "loss": 1.218,
+      "step": 4179
+    },
+    {
+      "epoch": 0.36687685653238833,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002380513690371282,
+      "loss": 1.2576,
+      "step": 4180
+    },
+    {
+      "epoch": 0.36696462611529085,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0023801569971092975,
+      "loss": 1.2075,
+      "step": 4181
+    },
+    {
+      "epoch": 0.3670523956981933,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0023798002317837064,
+      "loss": 1.1908,
+      "step": 4182
+    },
+    {
+      "epoch": 0.3671401652810958,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0023794433944297195,
+      "loss": 1.2342,
+      "step": 4183
+    },
+    {
+      "epoch": 0.3672279348639983,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0023790864850825565,
+      "loss": 1.2217,
+      "step": 4184
+    },
+    {
+      "epoch": 0.36731570444690076,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0023787295037774416,
+      "loss": 1.2061,
+      "step": 4185
+    },
+    {
+      "epoch": 0.3674034740298033,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002378372450549609,
+      "loss": 1.2169,
+      "step": 4186
+    },
+    {
+      "epoch": 0.36749124361270574,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002378015325434297,
+      "loss": 1.2281,
+      "step": 4187
+    },
+    {
+      "epoch": 0.3675790131956082,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002377658128466753,
+      "loss": 1.254,
+      "step": 4188
+    },
+    {
+      "epoch": 0.3676667827785107,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0023773008596822305,
+      "loss": 1.248,
+      "step": 4189
+    },
+    {
+      "epoch": 0.3677545523614132,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0023769435191159914,
+      "loss": 1.2221,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3678423219443157,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002376586106803303,
+      "loss": 1.235,
+      "step": 4191
+    },
+    {
+      "epoch": 0.3679300915272182,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.002376228622779441,
+      "loss": 1.277,
+      "step": 4192
+    },
+    {
+      "epoch": 0.36801786111012064,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002375871067079688,
+      "loss": 1.2448,
+      "step": 4193
+    },
+    {
+      "epoch": 0.36810563069302316,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.002375513439739332,
+      "loss": 1.2459,
+      "step": 4194
+    },
+    {
+      "epoch": 0.3681934002759256,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002375155740793671,
+      "loss": 1.2039,
+      "step": 4195
+    },
+    {
+      "epoch": 0.36828116985882814,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0023747979702780074,
+      "loss": 1.2531,
+      "step": 4196
+    },
+    {
+      "epoch": 0.3683689394417306,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002374440128227652,
+      "loss": 1.1952,
+      "step": 4197
+    },
+    {
+      "epoch": 0.36845670902463307,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.002374082214677922,
+      "loss": 1.2274,
+      "step": 4198
+    },
+    {
+      "epoch": 0.3685444786075356,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002373724229664143,
+      "loss": 1.2187,
+      "step": 4199
+    },
+    {
+      "epoch": 0.36863224819043805,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0023733661732216454,
+      "loss": 1.2288,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3687200177733405,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0023730080453857696,
+      "loss": 1.2258,
+      "step": 4201
+    },
+    {
+      "epoch": 0.36880778735624303,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00237264984619186,
+      "loss": 1.2233,
+      "step": 4202
+    },
+    {
+      "epoch": 0.3688955569391455,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00237229157567527,
+      "loss": 1.2184,
+      "step": 4203
+    },
+    {
+      "epoch": 0.368983326522048,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002371933233871359,
+      "loss": 1.236,
+      "step": 4204
+    },
+    {
+      "epoch": 0.3690710961049505,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0023715748208154954,
+      "loss": 1.267,
+      "step": 4205
+    },
+    {
+      "epoch": 0.36915886568785294,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002371216336543052,
+      "loss": 1.229,
+      "step": 4206
+    },
+    {
+      "epoch": 0.36924663527075546,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0023708577810894094,
+      "loss": 1.2855,
+      "step": 4207
+    },
+    {
+      "epoch": 0.3693344048536579,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0023704991544899563,
+      "loss": 1.1887,
+      "step": 4208
+    },
+    {
+      "epoch": 0.36942217443656045,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.002370140456780087,
+      "loss": 1.3129,
+      "step": 4209
+    },
+    {
+      "epoch": 0.3695099440194629,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0023697816879952047,
+      "loss": 1.2225,
+      "step": 4210
+    },
+    {
+      "epoch": 0.3695977136023654,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0023694228481707176,
+      "loss": 1.2051,
+      "step": 4211
+    },
+    {
+      "epoch": 0.3696854831852679,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002369063937342042,
+      "loss": 1.233,
+      "step": 4212
+    },
+    {
+      "epoch": 0.36977325276817036,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0023687049555446015,
+      "loss": 1.2264,
+      "step": 4213
+    },
+    {
+      "epoch": 0.3698610223510728,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0023683459028138246,
+      "loss": 1.2295,
+      "step": 4214
+    },
+    {
+      "epoch": 0.36994879193397534,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0023679867791851503,
+      "loss": 1.2069,
+      "step": 4215
+    },
+    {
+      "epoch": 0.3700365615168778,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0023676275846940213,
+      "loss": 1.1983,
+      "step": 4216
+    },
+    {
+      "epoch": 0.3701243310997803,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0023672683193758895,
+      "loss": 1.209,
+      "step": 4217
+    },
+    {
+      "epoch": 0.3702121006826828,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002366908983266213,
+      "loss": 1.249,
+      "step": 4218
+    },
+    {
+      "epoch": 0.37029987026558525,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.002366549576400456,
+      "loss": 1.3026,
+      "step": 4219
+    },
+    {
+      "epoch": 0.37038763984848777,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0023661900988140912,
+      "loss": 1.2213,
+      "step": 4220
+    },
+    {
+      "epoch": 0.37047540943139023,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0023658305505425974,
+      "loss": 1.2299,
+      "step": 4221
+    },
+    {
+      "epoch": 0.37056317901429275,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0023654709316214603,
+      "loss": 1.2675,
+      "step": 4222
+    },
+    {
+      "epoch": 0.3706509485971952,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002365111242086174,
+      "loss": 1.2052,
+      "step": 4223
+    },
+    {
+      "epoch": 0.3707387181800977,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0023647514819722373,
+      "loss": 1.2471,
+      "step": 4224
+    },
+    {
+      "epoch": 0.3708264877630002,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002364391651315157,
+      "loss": 1.2203,
+      "step": 4225
+    },
+    {
+      "epoch": 0.37091425734590266,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0023640317501504486,
+      "loss": 1.2612,
+      "step": 4226
+    },
+    {
+      "epoch": 0.37100202692880513,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0023636717785136306,
+      "loss": 1.2579,
+      "step": 4227
+    },
+    {
+      "epoch": 0.37108979651170765,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002363311736440232,
+      "loss": 1.2167,
+      "step": 4228
+    },
+    {
+      "epoch": 0.3711775660946101,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0023629516239657886,
+      "loss": 1.25,
+      "step": 4229
+    },
+    {
+      "epoch": 0.37126533567751263,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.00236259144112584,
+      "loss": 1.3032,
+      "step": 4230
+    },
+    {
+      "epoch": 0.3713531052604151,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0023622311879559358,
+      "loss": 1.2754,
+      "step": 4231
+    },
+    {
+      "epoch": 0.37144087484331756,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0023618708644916313,
+      "loss": 1.2418,
+      "step": 4232
+    },
+    {
+      "epoch": 0.3715286444262201,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0023615104707684896,
+      "loss": 1.2513,
+      "step": 4233
+    },
+    {
+      "epoch": 0.37161641400912254,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.002361150006822079,
+      "loss": 1.2833,
+      "step": 4234
+    },
+    {
+      "epoch": 0.37170418359202506,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0023607894726879776,
+      "loss": 1.3101,
+      "step": 4235
+    },
+    {
+      "epoch": 0.3717919531749275,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002360428868401767,
+      "loss": 1.2007,
+      "step": 4236
+    },
+    {
+      "epoch": 0.37187972275783,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002360068193999039,
+      "loss": 1.2395,
+      "step": 4237
+    },
+    {
+      "epoch": 0.3719674923407325,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0023597074495153893,
+      "loss": 1.1942,
+      "step": 4238
+    },
+    {
+      "epoch": 0.372055261923635,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0023593466349864224,
+      "loss": 1.2727,
+      "step": 4239
+    },
+    {
+      "epoch": 0.37214303150653744,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0023589857504477496,
+      "loss": 1.2548,
+      "step": 4240
+    },
+    {
+      "epoch": 0.37223080108943996,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002358624795934988,
+      "loss": 1.2749,
+      "step": 4241
+    },
+    {
+      "epoch": 0.3723185706723424,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0023582637714837634,
+      "loss": 1.2743,
+      "step": 4242
+    },
+    {
+      "epoch": 0.37240634025524494,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0023579026771297067,
+      "loss": 1.2603,
+      "step": 4243
+    },
+    {
+      "epoch": 0.3724941098381474,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002357541512908457,
+      "loss": 1.2175,
+      "step": 4244
+    },
+    {
+      "epoch": 0.37258187942104987,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00235718027885566,
+      "loss": 1.2388,
+      "step": 4245
+    },
+    {
+      "epoch": 0.3726696490039524,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.002356818975006967,
+      "loss": 1.2107,
+      "step": 4246
+    },
+    {
+      "epoch": 0.37275741858685485,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0023564576013980383,
+      "loss": 1.2806,
+      "step": 4247
+    },
+    {
+      "epoch": 0.37284518816975737,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0023560961580645396,
+      "loss": 1.2628,
+      "step": 4248
+    },
+    {
+      "epoch": 0.37293295775265983,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0023557346450421433,
+      "loss": 1.2424,
+      "step": 4249
+    },
+    {
+      "epoch": 0.3730207273355623,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.002355373062366531,
+      "loss": 1.2533,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3731084969184648,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0023550114100733877,
+      "loss": 1.1894,
+      "step": 4251
+    },
+    {
+      "epoch": 0.3731962665013673,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0023546496881984087,
+      "loss": 1.2222,
+      "step": 4252
+    },
+    {
+      "epoch": 0.37328403608426974,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.002354287896777293,
+      "loss": 1.1487,
+      "step": 4253
+    },
+    {
+      "epoch": 0.37337180566717226,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002353926035845749,
+      "loss": 1.2176,
+      "step": 4254
+    },
+    {
+      "epoch": 0.3734595752500747,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00235356410543949,
+      "loss": 1.2821,
+      "step": 4255
+    },
+    {
+      "epoch": 0.37354734483297725,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0023532021055942378,
+      "loss": 1.209,
+      "step": 4256
+    },
+    {
+      "epoch": 0.3736351144158797,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00235284003634572,
+      "loss": 1.2476,
+      "step": 4257
+    },
+    {
+      "epoch": 0.3737228839987822,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0023524778977296727,
+      "loss": 1.2887,
+      "step": 4258
+    },
+    {
+      "epoch": 0.3738106535816847,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002352115689781835,
+      "loss": 1.2239,
+      "step": 4259
+    },
+    {
+      "epoch": 0.37389842316458716,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0023517534125379576,
+      "loss": 1.2232,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3739861927474897,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002351391066033795,
+      "loss": 1.2658,
+      "step": 4261
+    },
+    {
+      "epoch": 0.37407396233039214,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00235102865030511,
+      "loss": 1.2163,
+      "step": 4262
+    },
+    {
+      "epoch": 0.3741617319132946,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0023506661653876705,
+      "loss": 1.2392,
+      "step": 4263
+    },
+    {
+      "epoch": 0.3742495014961971,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0023503036113172527,
+      "loss": 1.238,
+      "step": 4264
+    },
+    {
+      "epoch": 0.3743372710790996,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0023499409881296398,
+      "loss": 1.1915,
+      "step": 4265
+    },
+    {
+      "epoch": 0.3744250406620021,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0023495782958606205,
+      "loss": 1.1968,
+      "step": 4266
+    },
+    {
+      "epoch": 0.37451281024490457,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002349215534545992,
+      "loss": 1.2149,
+      "step": 4267
+    },
+    {
+      "epoch": 0.37460057982780703,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0023488527042215563,
+      "loss": 1.2106,
+      "step": 4268
+    },
+    {
+      "epoch": 0.37468834941070955,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0023484898049231243,
+      "loss": 1.2169,
+      "step": 4269
+    },
+    {
+      "epoch": 0.374776118993612,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0023481268366865125,
+      "loss": 1.1899,
+      "step": 4270
+    },
+    {
+      "epoch": 0.3748638885765145,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0023477637995475437,
+      "loss": 1.315,
+      "step": 4271
+    },
+    {
+      "epoch": 0.374951658159417,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0023474006935420492,
+      "loss": 1.1852,
+      "step": 4272
+    },
+    {
+      "epoch": 0.37503942774231946,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0023470375187058658,
+      "loss": 1.2195,
+      "step": 4273
+    },
+    {
+      "epoch": 0.375127197325222,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002346674275074837,
+      "loss": 1.2612,
+      "step": 4274
+    },
+    {
+      "epoch": 0.37521496690812445,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.002346310962684814,
+      "loss": 1.2131,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3753027364910269,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0023459475815716543,
+      "loss": 1.2199,
+      "step": 4276
+    },
+    {
+      "epoch": 0.37539050607392943,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002345584131771222,
+      "loss": 1.2523,
+      "step": 4277
+    },
+    {
+      "epoch": 0.3754782756568319,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002345220613319388,
+      "loss": 1.234,
+      "step": 4278
+    },
+    {
+      "epoch": 0.3755660452397344,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0023448570262520306,
+      "loss": 1.2495,
+      "step": 4279
+    },
+    {
+      "epoch": 0.3756538148226369,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0023444933706050337,
+      "loss": 1.2104,
+      "step": 4280
+    },
+    {
+      "epoch": 0.37574158440553934,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0023441296464142896,
+      "loss": 1.2604,
+      "step": 4281
+    },
+    {
+      "epoch": 0.37582935398844186,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002343765853715695,
+      "loss": 1.2364,
+      "step": 4282
+    },
+    {
+      "epoch": 0.3759171235713443,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0023434019925451573,
+      "loss": 1.2391,
+      "step": 4283
+    },
+    {
+      "epoch": 0.3760048931542468,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0023430380629385853,
+      "loss": 1.2663,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3760926627371493,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.002342674064931899,
+      "loss": 1.2488,
+      "step": 4285
+    },
+    {
+      "epoch": 0.37618043232005177,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0023423099985610232,
+      "loss": 1.2324,
+      "step": 4286
+    },
+    {
+      "epoch": 0.3762682019029543,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0023419458638618902,
+      "loss": 1.1891,
+      "step": 4287
+    },
+    {
+      "epoch": 0.37635597148585676,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0023415816608704384,
+      "loss": 1.1995,
+      "step": 4288
+    },
+    {
+      "epoch": 0.3764437410687592,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.002341217389622613,
+      "loss": 1.2687,
+      "step": 4289
+    },
+    {
+      "epoch": 0.37653151065166174,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002340853050154366,
+      "loss": 1.2461,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3766192802345642,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0023404886425016564,
+      "loss": 1.2639,
+      "step": 4291
+    },
+    {
+      "epoch": 0.3767070498174667,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00234012416670045,
+      "loss": 1.2472,
+      "step": 4292
+    },
+    {
+      "epoch": 0.3767948194003692,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0023397596227867196,
+      "loss": 1.246,
+      "step": 4293
+    },
+    {
+      "epoch": 0.37688258898327165,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.002339395010796443,
+      "loss": 1.2845,
+      "step": 4294
+    },
+    {
+      "epoch": 0.37697035856617417,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0023390303307656065,
+      "loss": 1.2574,
+      "step": 4295
+    },
+    {
+      "epoch": 0.37705812814907663,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0023386655827302033,
+      "loss": 1.2736,
+      "step": 4296
+    },
+    {
+      "epoch": 0.3771458977319791,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002338300766726232,
+      "loss": 1.1969,
+      "step": 4297
+    },
+    {
+      "epoch": 0.3772336673148816,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0023379358827896983,
+      "loss": 1.2467,
+      "step": 4298
+    },
+    {
+      "epoch": 0.3773214368977841,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.002337570930956615,
+      "loss": 1.2285,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3774092064806866,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002337205911263001,
+      "loss": 1.2546,
+      "step": 4300
+    },
+    {
+      "epoch": 0.37749697606358906,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0023368408237448833,
+      "loss": 1.2204,
+      "step": 4301
+    },
+    {
+      "epoch": 0.3775847456464915,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0023364756684382936,
+      "loss": 1.3143,
+      "step": 4302
+    },
+    {
+      "epoch": 0.37767251522939405,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0023361104453792723,
+      "loss": 1.2787,
+      "step": 4303
+    },
+    {
+      "epoch": 0.3777602848122965,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0023357451546038637,
+      "loss": 1.2391,
+      "step": 4304
+    },
+    {
+      "epoch": 0.37784805439519903,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002335379796148123,
+      "loss": 1.273,
+      "step": 4305
+    },
+    {
+      "epoch": 0.3779358239781015,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002335014370048108,
+      "loss": 1.2159,
+      "step": 4306
+    },
+    {
+      "epoch": 0.37802359356100396,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0023346488763398843,
+      "loss": 1.2172,
+      "step": 4307
+    },
+    {
+      "epoch": 0.3781113631439065,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0023342833150595266,
+      "loss": 1.2985,
+      "step": 4308
+    },
+    {
+      "epoch": 0.37819913272680894,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0023339176862431123,
+      "loss": 1.2719,
+      "step": 4309
+    },
+    {
+      "epoch": 0.3782869023097114,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.002333551989926729,
+      "loss": 1.2209,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3783746718926139,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002333186226146469,
+      "loss": 1.2568,
+      "step": 4311
+    },
+    {
+      "epoch": 0.3784624414755164,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0023328203949384316,
+      "loss": 1.3053,
+      "step": 4312
+    },
+    {
+      "epoch": 0.3785502110584189,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002332454496338723,
+      "loss": 1.1982,
+      "step": 4313
+    },
+    {
+      "epoch": 0.37863798064132137,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0023320885303834563,
+      "loss": 1.2471,
+      "step": 4314
+    },
+    {
+      "epoch": 0.37872575022422383,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0023317224971087496,
+      "loss": 1.1985,
+      "step": 4315
+    },
+    {
+      "epoch": 0.37881351980712635,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0023313563965507308,
+      "loss": 1.2197,
+      "step": 4316
+    },
+    {
+      "epoch": 0.3789012893900288,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0023309902287455314,
+      "loss": 1.239,
+      "step": 4317
+    },
+    {
+      "epoch": 0.37898905897293134,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002330623993729291,
+      "loss": 1.2255,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3790768285558338,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0023302576915381554,
+      "loss": 1.2284,
+      "step": 4319
+    },
+    {
+      "epoch": 0.37916459813873626,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0023298913222082776,
+      "loss": 1.2321,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3792523677216388,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0023295248857758155,
+      "loss": 1.2198,
+      "step": 4321
+    },
+    {
+      "epoch": 0.37934013730454125,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002329158382276937,
+      "loss": 1.2978,
+      "step": 4322
+    },
+    {
+      "epoch": 0.3794279068874437,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0023287918117478133,
+      "loss": 1.1858,
+      "step": 4323
+    },
+    {
+      "epoch": 0.37951567647034623,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0023284251742246233,
+      "loss": 1.2114,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3796034460532487,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002328058469743553,
+      "loss": 1.2275,
+      "step": 4325
+    },
+    {
+      "epoch": 0.3796912156361512,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0023276916983407947,
+      "loss": 1.2222,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3797789852190537,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0023273248600525474,
+      "loss": 1.205,
+      "step": 4327
+    },
+    {
+      "epoch": 0.37986675480195614,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0023269579549150166,
+      "loss": 1.22,
+      "step": 4328
+    },
+    {
+      "epoch": 0.37995452438485866,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.002326590982964414,
+      "loss": 1.2259,
+      "step": 4329
+    },
+    {
+      "epoch": 0.3800422939677611,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0023262239442369587,
+      "loss": 1.2888,
+      "step": 4330
+    },
+    {
+      "epoch": 0.38013006355066364,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0023258568387688754,
+      "loss": 1.1986,
+      "step": 4331
+    },
+    {
+      "epoch": 0.3802178331335661,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0023254896665963966,
+      "loss": 1.2498,
+      "step": 4332
+    },
+    {
+      "epoch": 0.38030560271646857,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0023251224277557603,
+      "loss": 1.1899,
+      "step": 4333
+    },
+    {
+      "epoch": 0.3803933722993711,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0023247551222832126,
+      "loss": 1.2421,
+      "step": 4334
+    },
+    {
+      "epoch": 0.38048114188227355,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0023243877502150028,
+      "loss": 1.2499,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3805689114651761,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0023240203115873913,
+      "loss": 1.2779,
+      "step": 4336
+    },
+    {
+      "epoch": 0.38065668104807854,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002323652806436642,
+      "loss": 1.1817,
+      "step": 4337
+    },
+    {
+      "epoch": 0.380744450630981,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002323285234799026,
+      "loss": 1.226,
+      "step": 4338
+    },
+    {
+      "epoch": 0.3808322202138835,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0023229175967108217,
+      "loss": 1.1708,
+      "step": 4339
+    },
+    {
+      "epoch": 0.380919989796786,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0023225498922083126,
+      "loss": 1.2724,
+      "step": 4340
+    },
+    {
+      "epoch": 0.38100775937968845,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0023221821213277906,
+      "loss": 1.2119,
+      "step": 4341
+    },
+    {
+      "epoch": 0.38109552896259097,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002321814284105553,
+      "loss": 1.2239,
+      "step": 4342
+    },
+    {
+      "epoch": 0.38118329854549343,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002321446380577904,
+      "loss": 1.2583,
+      "step": 4343
+    },
+    {
+      "epoch": 0.38127106812839595,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0023210784107811537,
+      "loss": 1.2536,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3813588377112984,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002320710374751619,
+      "loss": 1.3168,
+      "step": 4345
+    },
+    {
+      "epoch": 0.3814466072942009,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0023203422725256257,
+      "loss": 1.1542,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3815343768771034,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0023199741041395015,
+      "loss": 1.2565,
+      "step": 4347
+    },
+    {
+      "epoch": 0.38162214646000586,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0023196058696295842,
+      "loss": 1.2397,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3817099160429084,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002319237569032217,
+      "loss": 1.258,
+      "step": 4349
+    },
+    {
+      "epoch": 0.38179768562581085,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0023188692023837503,
+      "loss": 1.2145,
+      "step": 4350
+    },
+    {
+      "epoch": 0.3818854552087133,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0023185007697205393,
+      "loss": 1.2465,
+      "step": 4351
+    },
+    {
+      "epoch": 0.38197322479161583,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.002318132271078948,
+      "loss": 1.2854,
+      "step": 4352
+    },
+    {
+      "epoch": 0.3820609943745183,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002317763706495345,
+      "loss": 1.229,
+      "step": 4353
+    },
+    {
+      "epoch": 0.38214876395742076,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0023173950760061062,
+      "loss": 1.2323,
+      "step": 4354
+    },
+    {
+      "epoch": 0.3822365335403233,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002317026379647614,
+      "loss": 1.196,
+      "step": 4355
+    },
+    {
+      "epoch": 0.38232430312322574,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0023166576174562577,
+      "loss": 1.224,
+      "step": 4356
+    },
+    {
+      "epoch": 0.38241207270612826,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0023162887894684326,
+      "loss": 1.2266,
+      "step": 4357
+    },
+    {
+      "epoch": 0.3824998422890307,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0023159198957205403,
+      "loss": 1.3065,
+      "step": 4358
+    },
+    {
+      "epoch": 0.3825876118719332,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002315550936248989,
+      "loss": 1.1871,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3826753814548357,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0023151819110901937,
+      "loss": 1.2751,
+      "step": 4360
+    },
+    {
+      "epoch": 0.38276315103773817,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0023148128202805755,
+      "loss": 1.2969,
+      "step": 4361
+    },
+    {
+      "epoch": 0.3828509206206407,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.002314443663856563,
+      "loss": 1.2331,
+      "step": 4362
+    },
+    {
+      "epoch": 0.38293869020354315,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0023140744418545895,
+      "loss": 1.257,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3830264597864456,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002313705154311097,
+      "loss": 1.2506,
+      "step": 4364
+    },
+    {
+      "epoch": 0.38311422936934814,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0023133358012625313,
+      "loss": 1.2421,
+      "step": 4365
+    },
+    {
+      "epoch": 0.3832019989522506,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0023129663827453468,
+      "loss": 1.3356,
+      "step": 4366
+    },
+    {
+      "epoch": 0.38328976853515306,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0023125968987960034,
+      "loss": 1.307,
+      "step": 4367
+    },
+    {
+      "epoch": 0.3833775381180556,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002312227349450968,
+      "loss": 1.2527,
+      "step": 4368
+    },
+    {
+      "epoch": 0.38346530770095805,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0023118577347467136,
+      "loss": 1.2796,
+      "step": 4369
+    },
+    {
+      "epoch": 0.38355307728386057,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002311488054719719,
+      "loss": 1.2212,
+      "step": 4370
+    },
+    {
+      "epoch": 0.38364084686676303,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.002311118309406472,
+      "loss": 1.2321,
+      "step": 4371
+    },
+    {
+      "epoch": 0.3837286164496655,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002310748498843463,
+      "loss": 1.2463,
+      "step": 4372
+    },
+    {
+      "epoch": 0.383816386032568,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0023103786230671917,
+      "loss": 1.2729,
+      "step": 4373
+    },
+    {
+      "epoch": 0.3839041556154705,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0023100086821141634,
+      "loss": 1.2719,
+      "step": 4374
+    },
+    {
+      "epoch": 0.383991925198373,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0023096386760208896,
+      "loss": 1.2278,
+      "step": 4375
+    },
+    {
+      "epoch": 0.38407969478127546,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.002309268604823888,
+      "loss": 1.2568,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3841674643641779,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0023088984685596847,
+      "loss": 1.2376,
+      "step": 4377
+    },
+    {
+      "epoch": 0.38425523394708044,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0023085282672648086,
+      "loss": 1.2862,
+      "step": 4378
+    },
+    {
+      "epoch": 0.3843430035299829,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002308158000975799,
+      "loss": 1.2044,
+      "step": 4379
+    },
+    {
+      "epoch": 0.38443077311288537,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0023077876697291987,
+      "loss": 1.2221,
+      "step": 4380
+    },
+    {
+      "epoch": 0.3845185426957879,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0023074172735615584,
+      "loss": 1.1765,
+      "step": 4381
+    },
+    {
+      "epoch": 0.38460631227869035,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0023070468125094335,
+      "loss": 1.1834,
+      "step": 4382
+    },
+    {
+      "epoch": 0.3846940818615929,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002306676286609389,
+      "loss": 1.2694,
+      "step": 4383
+    },
+    {
+      "epoch": 0.38478185144449534,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0023063056958979936,
+      "loss": 1.1859,
+      "step": 4384
+    },
+    {
+      "epoch": 0.3848696210273978,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0023059350404118217,
+      "loss": 1.2325,
+      "step": 4385
+    },
+    {
+      "epoch": 0.3849573906103003,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0023055643201874572,
+      "loss": 1.2329,
+      "step": 4386
+    },
+    {
+      "epoch": 0.3850451601932028,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002305193535261489,
+      "loss": 1.3328,
+      "step": 4387
+    },
+    {
+      "epoch": 0.3851329297761053,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0023048226856705106,
+      "loss": 1.2262,
+      "step": 4388
+    },
+    {
+      "epoch": 0.38522069935900777,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0023044517714511242,
+      "loss": 1.2544,
+      "step": 4389
+    },
+    {
+      "epoch": 0.38530846894191023,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002304080792639937,
+      "loss": 1.2504,
+      "step": 4390
+    },
+    {
+      "epoch": 0.38539623852481275,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0023037097492735648,
+      "loss": 1.2199,
+      "step": 4391
+    },
+    {
+      "epoch": 0.3854840081077152,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0023033386413886256,
+      "loss": 1.263,
+      "step": 4392
+    },
+    {
+      "epoch": 0.3855717776906177,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002302967469021749,
+      "loss": 1.2093,
+      "step": 4393
+    },
+    {
+      "epoch": 0.3856595472735202,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0023025962322095663,
+      "loss": 1.1909,
+      "step": 4394
+    },
+    {
+      "epoch": 0.38574731685642266,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002302224930988717,
+      "loss": 1.2085,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3858350864393252,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0023018535653958486,
+      "loss": 1.2716,
+      "step": 4396
+    },
+    {
+      "epoch": 0.38592285602222764,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0023014821354676125,
+      "loss": 1.2248,
+      "step": 4397
+    },
+    {
+      "epoch": 0.3860106256051301,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002301110641240667,
+      "loss": 1.2211,
+      "step": 4398
+    },
+    {
+      "epoch": 0.38609839518803263,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002300739082751678,
+      "loss": 1.2716,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3861861647709351,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0023003674600373153,
+      "loss": 1.1989,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3862739343538376,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002299995773134259,
+      "loss": 1.1621,
+      "step": 4401
+    },
+    {
+      "epoch": 0.3863617039367401,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.002299624022079191,
+      "loss": 1.2333,
+      "step": 4402
+    },
+    {
+      "epoch": 0.38644947351964254,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.002299252206908803,
+      "loss": 1.2239,
+      "step": 4403
+    },
+    {
+      "epoch": 0.38653724310254506,
+      "grad_norm": 0.125,
+      "learning_rate": 0.002298880327659791,
+      "loss": 1.1831,
+      "step": 4404
+    },
+    {
+      "epoch": 0.3866250126854475,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0022985083843688574,
+      "loss": 1.197,
+      "step": 4405
+    },
+    {
+      "epoch": 0.38671278226835004,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002298136377072713,
+      "loss": 1.2152,
+      "step": 4406
+    },
+    {
+      "epoch": 0.3868005518512525,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0022977643058080728,
+      "loss": 1.2183,
+      "step": 4407
+    },
+    {
+      "epoch": 0.38688832143415497,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002297392170611658,
+      "loss": 1.2229,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3869760910170575,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002297019971520199,
+      "loss": 1.2607,
+      "step": 4409
+    },
+    {
+      "epoch": 0.38706386059995995,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0022966477085704273,
+      "loss": 1.2776,
+      "step": 4410
+    },
+    {
+      "epoch": 0.3871516301828624,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0022962753817990862,
+      "loss": 1.1781,
+      "step": 4411
+    },
+    {
+      "epoch": 0.38723939976576494,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0022959029912429222,
+      "loss": 1.2315,
+      "step": 4412
+    },
+    {
+      "epoch": 0.3873271693486674,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0022955305369386884,
+      "loss": 1.2476,
+      "step": 4413
+    },
+    {
+      "epoch": 0.3874149389315699,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002295158018923145,
+      "loss": 1.1712,
+      "step": 4414
+    },
+    {
+      "epoch": 0.3875027085144724,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0022947854372330576,
+      "loss": 1.1969,
+      "step": 4415
+    },
+    {
+      "epoch": 0.38759047809737485,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0022944127919051994,
+      "loss": 1.27,
+      "step": 4416
+    },
+    {
+      "epoch": 0.38767824768027737,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002294040082976348,
+      "loss": 1.2638,
+      "step": 4417
+    },
+    {
+      "epoch": 0.38776601726317983,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.002293667310483289,
+      "loss": 1.2421,
+      "step": 4418
+    },
+    {
+      "epoch": 0.38785378684608235,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.002293294474462813,
+      "loss": 1.2792,
+      "step": 4419
+    },
+    {
+      "epoch": 0.3879415564289848,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002292921574951718,
+      "loss": 1.2174,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3880293260118873,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0022925486119868073,
+      "loss": 1.2498,
+      "step": 4421
+    },
+    {
+      "epoch": 0.3881170955947898,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002292175585604891,
+      "loss": 1.3241,
+      "step": 4422
+    },
+    {
+      "epoch": 0.38820486517769226,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002291802495842786,
+      "loss": 1.256,
+      "step": 4423
+    },
+    {
+      "epoch": 0.3882926347605947,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0022914293427373136,
+      "loss": 1.2652,
+      "step": 4424
+    },
+    {
+      "epoch": 0.38838040434349724,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0022910561263253027,
+      "loss": 1.2455,
+      "step": 4425
+    },
+    {
+      "epoch": 0.3884681739263997,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0022906828466435895,
+      "loss": 1.1948,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3885559435093022,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0022903095037290133,
+      "loss": 1.2912,
+      "step": 4427
+    },
+    {
+      "epoch": 0.3886437130922047,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0022899360976184236,
+      "loss": 1.2281,
+      "step": 4428
+    },
+    {
+      "epoch": 0.38873148267510715,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002289562628348673,
+      "loss": 1.2357,
+      "step": 4429
+    },
+    {
+      "epoch": 0.3888192522580097,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0022891890959566216,
+      "loss": 1.2234,
+      "step": 4430
+    },
+    {
+      "epoch": 0.38890702184091214,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.002288815500479135,
+      "loss": 1.2021,
+      "step": 4431
+    },
+    {
+      "epoch": 0.38899479142381466,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002288441841953087,
+      "loss": 1.2222,
+      "step": 4432
+    },
+    {
+      "epoch": 0.3890825610067171,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002288068120415355,
+      "loss": 1.2676,
+      "step": 4433
+    },
+    {
+      "epoch": 0.3891703305896196,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0022876943359028245,
+      "loss": 1.2517,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3892581001725221,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0022873204884523855,
+      "loss": 1.2502,
+      "step": 4435
+    },
+    {
+      "epoch": 0.38934586975542457,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0022869465781009367,
+      "loss": 1.2793,
+      "step": 4436
+    },
+    {
+      "epoch": 0.38943363933832703,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0022865726048853807,
+      "loss": 1.19,
+      "step": 4437
+    },
+    {
+      "epoch": 0.38952140892122955,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0022861985688426277,
+      "loss": 1.1752,
+      "step": 4438
+    },
+    {
+      "epoch": 0.389609178504132,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.002285824470009593,
+      "loss": 1.2646,
+      "step": 4439
+    },
+    {
+      "epoch": 0.38969694808703453,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0022854503084231994,
+      "loss": 1.293,
+      "step": 4440
+    },
+    {
+      "epoch": 0.389784717669937,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.002285076084120374,
+      "loss": 1.2106,
+      "step": 4441
+    },
+    {
+      "epoch": 0.38987248725283946,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0022847017971380523,
+      "loss": 1.2581,
+      "step": 4442
+    },
+    {
+      "epoch": 0.389960256835742,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.002284327447513174,
+      "loss": 1.2417,
+      "step": 4443
+    },
+    {
+      "epoch": 0.39004802641864444,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0022839530352826874,
+      "loss": 1.261,
+      "step": 4444
+    },
+    {
+      "epoch": 0.39013579600154696,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0022835785604835446,
+      "loss": 1.1875,
+      "step": 4445
+    },
+    {
+      "epoch": 0.39022356558444943,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002283204023152704,
+      "loss": 1.1846,
+      "step": 4446
+    },
+    {
+      "epoch": 0.3903113351673519,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0022828294233271326,
+      "loss": 1.2762,
+      "step": 4447
+    },
+    {
+      "epoch": 0.3903991047502544,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.002282454761043801,
+      "loss": 1.2679,
+      "step": 4448
+    },
+    {
+      "epoch": 0.3904868743331569,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0022820800363396866,
+      "loss": 1.3145,
+      "step": 4449
+    },
+    {
+      "epoch": 0.39057464391605934,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.002281705249251774,
+      "loss": 1.259,
+      "step": 4450
+    },
+    {
+      "epoch": 0.39066241349896186,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002281330399817053,
+      "loss": 1.2508,
+      "step": 4451
+    },
+    {
+      "epoch": 0.3907501830818643,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0022809554880725193,
+      "loss": 1.2042,
+      "step": 4452
+    },
+    {
+      "epoch": 0.39083795266476684,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.002280580514055175,
+      "loss": 1.249,
+      "step": 4453
+    },
+    {
+      "epoch": 0.3909257222476693,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00228020547780203,
+      "loss": 1.213,
+      "step": 4454
+    },
+    {
+      "epoch": 0.39101349183057177,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0022798303793500974,
+      "loss": 1.2032,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3911012614134743,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0022794552187363985,
+      "loss": 1.2177,
+      "step": 4456
+    },
+    {
+      "epoch": 0.39118903099637675,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00227907999599796,
+      "loss": 1.2172,
+      "step": 4457
+    },
+    {
+      "epoch": 0.39127680057927927,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0022787047111718153,
+      "loss": 1.268,
+      "step": 4458
+    },
+    {
+      "epoch": 0.39136457016218174,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002278329364295003,
+      "loss": 1.2312,
+      "step": 4459
+    },
+    {
+      "epoch": 0.3914523397450842,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.002277953955404569,
+      "loss": 1.2314,
+      "step": 4460
+    },
+    {
+      "epoch": 0.3915401093279867,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002277578484537564,
+      "loss": 1.1943,
+      "step": 4461
+    },
+    {
+      "epoch": 0.3916278789108892,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002277202951731046,
+      "loss": 1.2327,
+      "step": 4462
+    },
+    {
+      "epoch": 0.39171564849379165,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.002276827357022078,
+      "loss": 1.2266,
+      "step": 4463
+    },
+    {
+      "epoch": 0.39180341807669417,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0022764517004477303,
+      "loss": 1.197,
+      "step": 4464
+    },
+    {
+      "epoch": 0.39189118765959663,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0022760759820450787,
+      "loss": 1.2324,
+      "step": 4465
+    },
+    {
+      "epoch": 0.39197895724249915,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0022757002018512044,
+      "loss": 1.2474,
+      "step": 4466
+    },
+    {
+      "epoch": 0.3920667268254016,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0022753243599031967,
+      "loss": 1.2596,
+      "step": 4467
+    },
+    {
+      "epoch": 0.3921544964083041,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002274948456238148,
+      "loss": 1.2443,
+      "step": 4468
+    },
+    {
+      "epoch": 0.3922422659912066,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0022745724908931606,
+      "loss": 1.2971,
+      "step": 4469
+    },
+    {
+      "epoch": 0.39233003557410906,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002274196463905339,
+      "loss": 1.244,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3924178051570116,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0022738203753117966,
+      "loss": 1.2328,
+      "step": 4471
+    },
+    {
+      "epoch": 0.39250557473991404,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002273444225149651,
+      "loss": 1.2785,
+      "step": 4472
+    },
+    {
+      "epoch": 0.3925933443228165,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0022730680134560276,
+      "loss": 1.2153,
+      "step": 4473
+    },
+    {
+      "epoch": 0.392681113905719,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0022726917402680563,
+      "loss": 1.2289,
+      "step": 4474
+    },
+    {
+      "epoch": 0.3927688834886215,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0022723154056228743,
+      "loss": 1.2052,
+      "step": 4475
+    },
+    {
+      "epoch": 0.392856653071524,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002271939009557624,
+      "loss": 1.2234,
+      "step": 4476
+    },
+    {
+      "epoch": 0.3929444226544265,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0022715625521094545,
+      "loss": 1.1727,
+      "step": 4477
+    },
+    {
+      "epoch": 0.39303219223732894,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0022711860333155206,
+      "loss": 1.2112,
+      "step": 4478
+    },
+    {
+      "epoch": 0.39311996182023146,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0022708094532129827,
+      "loss": 1.2185,
+      "step": 4479
+    },
+    {
+      "epoch": 0.3932077314031339,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002270432811839009,
+      "loss": 1.2277,
+      "step": 4480
+    },
+    {
+      "epoch": 0.3932955009860364,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0022700561092307706,
+      "loss": 1.209,
+      "step": 4481
+    },
+    {
+      "epoch": 0.3933832705689389,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0022696793454254484,
+      "loss": 1.2295,
+      "step": 4482
+    },
+    {
+      "epoch": 0.39347104015184137,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002269302520460227,
+      "loss": 1.2747,
+      "step": 4483
+    },
+    {
+      "epoch": 0.3935588097347439,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0022689256343722966,
+      "loss": 1.2553,
+      "step": 4484
+    },
+    {
+      "epoch": 0.39364657931764635,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0022685486871988556,
+      "loss": 1.2908,
+      "step": 4485
+    },
+    {
+      "epoch": 0.3937343489005488,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0022681716789771057,
+      "loss": 1.2907,
+      "step": 4486
+    },
+    {
+      "epoch": 0.39382211848345133,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002267794609744258,
+      "loss": 1.2084,
+      "step": 4487
+    },
+    {
+      "epoch": 0.3939098880663538,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0022674174795375267,
+      "loss": 1.2433,
+      "step": 4488
+    },
+    {
+      "epoch": 0.3939976576492563,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002267040288394133,
+      "loss": 1.135,
+      "step": 4489
+    },
+    {
+      "epoch": 0.3940854272321588,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0022666630363513043,
+      "loss": 1.2265,
+      "step": 4490
+    },
+    {
+      "epoch": 0.39417319681506124,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002266285723446274,
+      "loss": 1.219,
+      "step": 4491
+    },
+    {
+      "epoch": 0.39426096639796376,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002265908349716281,
+      "loss": 1.2183,
+      "step": 4492
+    },
+    {
+      "epoch": 0.3943487359808662,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002265530915198572,
+      "loss": 1.2356,
+      "step": 4493
+    },
+    {
+      "epoch": 0.3944365055637687,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.002265153419930396,
+      "loss": 1.2592,
+      "step": 4494
+    },
+    {
+      "epoch": 0.3945242751466712,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0022647758639490123,
+      "loss": 1.2597,
+      "step": 4495
+    },
+    {
+      "epoch": 0.3946120447295737,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0022643982472916827,
+      "loss": 1.2391,
+      "step": 4496
+    },
+    {
+      "epoch": 0.3946998143124762,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0022640205699956772,
+      "loss": 1.225,
+      "step": 4497
+    },
+    {
+      "epoch": 0.39478758389537866,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0022636428320982715,
+      "loss": 1.196,
+      "step": 4498
+    },
+    {
+      "epoch": 0.3948753534782811,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0022632650336367457,
+      "loss": 1.2195,
+      "step": 4499
+    },
+    {
+      "epoch": 0.39496312306118364,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.002262887174648388,
+      "loss": 1.2091,
+      "step": 4500
+    },
+    {
+      "epoch": 0.39496312306118364,
+      "eval_loss": 1.2186098098754883,
+      "eval_runtime": 437.68,
+      "eval_samples_per_second": 33.678,
+      "eval_steps_per_second": 8.419,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3950508926440861,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0022625092551704913,
+      "loss": 1.2265,
+      "step": 4501
+    },
+    {
+      "epoch": 0.3951386622269886,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0022621312752403545,
+      "loss": 1.3061,
+      "step": 4502
+    },
+    {
+      "epoch": 0.3952264318098911,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0022617532348952834,
+      "loss": 1.1992,
+      "step": 4503
+    },
+    {
+      "epoch": 0.39531420139279355,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0022613751341725883,
+      "loss": 1.2589,
+      "step": 4504
+    },
+    {
+      "epoch": 0.39540197097569607,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0022609969731095863,
+      "loss": 1.1967,
+      "step": 4505
+    },
+    {
+      "epoch": 0.39548974055859853,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.002260618751743601,
+      "loss": 1.2126,
+      "step": 4506
+    },
+    {
+      "epoch": 0.395577510141501,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0022602404701119604,
+      "loss": 1.2631,
+      "step": 4507
+    },
+    {
+      "epoch": 0.3956652797244035,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0022598621282520007,
+      "loss": 1.2702,
+      "step": 4508
+    },
+    {
+      "epoch": 0.395753049307306,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002259483726201062,
+      "loss": 1.2373,
+      "step": 4509
+    },
+    {
+      "epoch": 0.3958408188902085,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002259105263996491,
+      "loss": 1.2351,
+      "step": 4510
+    },
+    {
+      "epoch": 0.39592858847311097,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.00225872674167564,
+      "loss": 1.205,
+      "step": 4511
+    },
+    {
+      "epoch": 0.39601635805601343,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002258348159275869,
+      "loss": 1.2386,
+      "step": 4512
+    },
+    {
+      "epoch": 0.39610412763891595,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0022579695168345416,
+      "loss": 1.2687,
+      "step": 4513
+    },
+    {
+      "epoch": 0.3961918972218184,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0022575908143890283,
+      "loss": 1.2289,
+      "step": 4514
+    },
+    {
+      "epoch": 0.39627966680472093,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0022572120519767057,
+      "loss": 1.2376,
+      "step": 4515
+    },
+    {
+      "epoch": 0.3963674363876234,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002256833229634956,
+      "loss": 1.1951,
+      "step": 4516
+    },
+    {
+      "epoch": 0.39645520597052586,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0022564543474011683,
+      "loss": 1.282,
+      "step": 4517
+    },
+    {
+      "epoch": 0.3965429755534284,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0022560754053127362,
+      "loss": 1.2294,
+      "step": 4518
+    },
+    {
+      "epoch": 0.39663074513633084,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002255696403407059,
+      "loss": 1.1793,
+      "step": 4519
+    },
+    {
+      "epoch": 0.3967185147192333,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.002255317341721544,
+      "loss": 1.1754,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3968062843021358,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0022549382202936017,
+      "loss": 1.2479,
+      "step": 4521
+    },
+    {
+      "epoch": 0.3968940538850383,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0022545590391606514,
+      "loss": 1.205,
+      "step": 4522
+    },
+    {
+      "epoch": 0.3969818234679408,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0022541797983601157,
+      "loss": 1.255,
+      "step": 4523
+    },
+    {
+      "epoch": 0.3970695930508433,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002253800497929425,
+      "loss": 1.2207,
+      "step": 4524
+    },
+    {
+      "epoch": 0.39715736263374574,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0022534211379060137,
+      "loss": 1.1946,
+      "step": 4525
+    },
+    {
+      "epoch": 0.39724513221664826,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0022530417183273237,
+      "loss": 1.2509,
+      "step": 4526
+    },
+    {
+      "epoch": 0.3973329017995507,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0022526622392308026,
+      "loss": 1.2738,
+      "step": 4527
+    },
+    {
+      "epoch": 0.39742067138245324,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.002252282700653903,
+      "loss": 1.1748,
+      "step": 4528
+    },
+    {
+      "epoch": 0.3975084409653557,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0022519031026340835,
+      "loss": 1.2356,
+      "step": 4529
+    },
+    {
+      "epoch": 0.39759621054825817,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0022515234452088104,
+      "loss": 1.1747,
+      "step": 4530
+    },
+    {
+      "epoch": 0.3976839801311607,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002251143728415552,
+      "loss": 1.2162,
+      "step": 4531
+    },
+    {
+      "epoch": 0.39777174971406315,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0022507639522917873,
+      "loss": 1.2556,
+      "step": 4532
+    },
+    {
+      "epoch": 0.3978595192969656,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002250384116874997,
+      "loss": 1.2334,
+      "step": 4533
+    },
+    {
+      "epoch": 0.39794728887986813,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0022500042222026707,
+      "loss": 1.2725,
+      "step": 4534
+    },
+    {
+      "epoch": 0.3980350584627706,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002249624268312301,
+      "loss": 1.2755,
+      "step": 4535
+    },
+    {
+      "epoch": 0.3981228280456731,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002249244255241389,
+      "loss": 1.2529,
+      "step": 4536
+    },
+    {
+      "epoch": 0.3982105976285756,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00224886418302744,
+      "loss": 1.182,
+      "step": 4537
+    },
+    {
+      "epoch": 0.39829836721147804,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002248484051707966,
+      "loss": 1.1488,
+      "step": 4538
+    },
+    {
+      "epoch": 0.39838613679438056,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0022481038613204837,
+      "loss": 1.2066,
+      "step": 4539
+    },
+    {
+      "epoch": 0.398473906377283,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0022477236119025175,
+      "loss": 1.2572,
+      "step": 4540
+    },
+    {
+      "epoch": 0.39856167596018555,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002247343303491595,
+      "loss": 1.2349,
+      "step": 4541
+    },
+    {
+      "epoch": 0.398649445543088,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0022469629361252528,
+      "loss": 1.2365,
+      "step": 4542
+    },
+    {
+      "epoch": 0.3987372151259905,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0022465825098410305,
+      "loss": 1.1759,
+      "step": 4543
+    },
+    {
+      "epoch": 0.398824984708893,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.002246202024676475,
+      "loss": 1.2416,
+      "step": 4544
+    },
+    {
+      "epoch": 0.39891275429179546,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0022458214806691384,
+      "loss": 1.1787,
+      "step": 4545
+    },
+    {
+      "epoch": 0.399000523874698,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0022454408778565796,
+      "loss": 1.2258,
+      "step": 4546
+    },
+    {
+      "epoch": 0.39908829345760044,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0022450602162763624,
+      "loss": 1.2491,
+      "step": 4547
+    },
+    {
+      "epoch": 0.3991760630405029,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0022446794959660563,
+      "loss": 1.2196,
+      "step": 4548
+    },
+    {
+      "epoch": 0.3992638326234054,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0022442987169632366,
+      "loss": 1.2421,
+      "step": 4549
+    },
+    {
+      "epoch": 0.3993516022063079,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0022439178793054854,
+      "loss": 1.2328,
+      "step": 4550
+    },
+    {
+      "epoch": 0.39943937178921035,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0022435369830303897,
+      "loss": 1.1851,
+      "step": 4551
+    },
+    {
+      "epoch": 0.39952714137211287,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002243156028175542,
+      "loss": 1.3398,
+      "step": 4552
+    },
+    {
+      "epoch": 0.39961491095501533,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.002242775014778541,
+      "loss": 1.2284,
+      "step": 4553
+    },
+    {
+      "epoch": 0.39970268053791785,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002242393942876992,
+      "loss": 1.2492,
+      "step": 4554
+    },
+    {
+      "epoch": 0.3997904501208203,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0022420128125085047,
+      "loss": 1.2483,
+      "step": 4555
+    },
+    {
+      "epoch": 0.3998782197037228,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002241631623710695,
+      "loss": 1.2433,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3999659892866253,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.002241250376521186,
+      "loss": 1.2344,
+      "step": 4557
+    },
+    {
+      "epoch": 0.40005375886952776,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0022408690709776035,
+      "loss": 1.2559,
+      "step": 4558
+    },
+    {
+      "epoch": 0.4001415284524303,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0022404877071175813,
+      "loss": 1.2125,
+      "step": 4559
+    },
+    {
+      "epoch": 0.40022929803533275,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0022401062849787597,
+      "loss": 1.2232,
+      "step": 4560
+    },
+    {
+      "epoch": 0.4003170676182352,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002239724804598782,
+      "loss": 1.2485,
+      "step": 4561
+    },
+    {
+      "epoch": 0.40040483720113773,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0022393432660153,
+      "loss": 1.2028,
+      "step": 4562
+    },
+    {
+      "epoch": 0.4004926067840402,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0022389616692659693,
+      "loss": 1.2068,
+      "step": 4563
+    },
+    {
+      "epoch": 0.40058037636694266,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0022385800143884527,
+      "loss": 1.2435,
+      "step": 4564
+    },
+    {
+      "epoch": 0.4006681459498452,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002238198301420417,
+      "loss": 1.206,
+      "step": 4565
+    },
+    {
+      "epoch": 0.40075591553274764,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0022378165303995365,
+      "loss": 1.2264,
+      "step": 4566
+    },
+    {
+      "epoch": 0.40084368511565016,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0022374347013634902,
+      "loss": 1.2607,
+      "step": 4567
+    },
+    {
+      "epoch": 0.4009314546985526,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002237052814349963,
+      "loss": 1.2098,
+      "step": 4568
+    },
+    {
+      "epoch": 0.4010192242814551,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0022366708693966468,
+      "loss": 1.2563,
+      "step": 4569
+    },
+    {
+      "epoch": 0.4011069938643576,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0022362888665412362,
+      "loss": 1.2339,
+      "step": 4570
+    },
+    {
+      "epoch": 0.40119476344726007,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002235906805821435,
+      "loss": 1.2235,
+      "step": 4571
+    },
+    {
+      "epoch": 0.4012825330301626,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0022355246872749496,
+      "loss": 1.2523,
+      "step": 4572
+    },
+    {
+      "epoch": 0.40137030261306506,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002235142510939495,
+      "loss": 1.2645,
+      "step": 4573
+    },
+    {
+      "epoch": 0.4014580721959675,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00223476027685279,
+      "loss": 1.2357,
+      "step": 4574
+    },
+    {
+      "epoch": 0.40154584177887004,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002234377985052559,
+      "loss": 1.2698,
+      "step": 4575
+    },
+    {
+      "epoch": 0.4016336113617725,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002233995635576533,
+      "loss": 1.2074,
+      "step": 4576
+    },
+    {
+      "epoch": 0.40172138094467497,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.002233613228462449,
+      "loss": 1.2693,
+      "step": 4577
+    },
+    {
+      "epoch": 0.4018091505275775,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0022332307637480484,
+      "loss": 1.2476,
+      "step": 4578
+    },
+    {
+      "epoch": 0.40189692011047995,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0022328482414710796,
+      "loss": 1.1465,
+      "step": 4579
+    },
+    {
+      "epoch": 0.40198468969338247,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0022324656616692948,
+      "loss": 1.198,
+      "step": 4580
+    },
+    {
+      "epoch": 0.40207245927628493,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.002232083024380455,
+      "loss": 1.2371,
+      "step": 4581
+    },
+    {
+      "epoch": 0.4021602288591874,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0022317003296423218,
+      "loss": 1.2475,
+      "step": 4582
+    },
+    {
+      "epoch": 0.4022479984420899,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00223131757749267,
+      "loss": 1.2169,
+      "step": 4583
+    },
+    {
+      "epoch": 0.4023357680249924,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002230934767969272,
+      "loss": 1.1986,
+      "step": 4584
+    },
+    {
+      "epoch": 0.4024235376078949,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.002230551901109912,
+      "loss": 1.214,
+      "step": 4585
+    },
+    {
+      "epoch": 0.40251130719079736,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0022301689769523756,
+      "loss": 1.281,
+      "step": 4586
+    },
+    {
+      "epoch": 0.4025990767736998,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.002229785995534457,
+      "loss": 1.2614,
+      "step": 4587
+    },
+    {
+      "epoch": 0.40268684635660235,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002229402956893955,
+      "loss": 1.2541,
+      "step": 4588
+    },
+    {
+      "epoch": 0.4027746159395048,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002229019861068674,
+      "loss": 1.2198,
+      "step": 4589
+    },
+    {
+      "epoch": 0.4028623855224073,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.002228636708096423,
+      "loss": 1.2111,
+      "step": 4590
+    },
+    {
+      "epoch": 0.4029501551053098,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0022282534980150193,
+      "loss": 1.1959,
+      "step": 4591
+    },
+    {
+      "epoch": 0.40303792468821226,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0022278702308622827,
+      "loss": 1.2903,
+      "step": 4592
+    },
+    {
+      "epoch": 0.4031256942711148,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0022274869066760414,
+      "loss": 1.204,
+      "step": 4593
+    },
+    {
+      "epoch": 0.40321346385401724,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0022271035254941268,
+      "loss": 1.2469,
+      "step": 4594
+    },
+    {
+      "epoch": 0.4033012334369197,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0022267200873543787,
+      "loss": 1.2176,
+      "step": 4595
+    },
+    {
+      "epoch": 0.4033890030198222,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002226336592294639,
+      "loss": 1.2289,
+      "step": 4596
+    },
+    {
+      "epoch": 0.4034767726027247,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0022259530403527586,
+      "loss": 1.2417,
+      "step": 4597
+    },
+    {
+      "epoch": 0.4035645421856272,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0022255694315665916,
+      "loss": 1.2632,
+      "step": 4598
+    },
+    {
+      "epoch": 0.40365231176852967,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0022251857659739997,
+      "loss": 1.1894,
+      "step": 4599
+    },
+    {
+      "epoch": 0.40374008135143213,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0022248020436128483,
+      "loss": 1.2498,
+      "step": 4600
+    },
+    {
+      "epoch": 0.40382785093433465,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0022244182645210094,
+      "loss": 1.1926,
+      "step": 4601
+    },
+    {
+      "epoch": 0.4039156205172371,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002224034428736361,
+      "loss": 1.2626,
+      "step": 4602
+    },
+    {
+      "epoch": 0.4040033901001396,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0022236505362967854,
+      "loss": 1.2359,
+      "step": 4603
+    },
+    {
+      "epoch": 0.4040911596830421,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0022232665872401716,
+      "loss": 1.2135,
+      "step": 4604
+    },
+    {
+      "epoch": 0.40417892926594456,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0022228825816044146,
+      "loss": 1.2268,
+      "step": 4605
+    },
+    {
+      "epoch": 0.4042666988488471,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0022224985194274132,
+      "loss": 1.3129,
+      "step": 4606
+    },
+    {
+      "epoch": 0.40435446843174955,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0022221144007470734,
+      "loss": 1.2514,
+      "step": 4607
+    },
+    {
+      "epoch": 0.404442238014652,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0022217302256013058,
+      "loss": 1.2257,
+      "step": 4608
+    },
+    {
+      "epoch": 0.40453000759755453,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002221345994028027,
+      "loss": 1.3206,
+      "step": 4609
+    },
+    {
+      "epoch": 0.404617777180457,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002220961706065159,
+      "loss": 1.191,
+      "step": 4610
+    },
+    {
+      "epoch": 0.4047055467633595,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002220577361750631,
+      "loss": 1.2191,
+      "step": 4611
+    },
+    {
+      "epoch": 0.404793316346262,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0022201929611223737,
+      "loss": 1.2534,
+      "step": 4612
+    },
+    {
+      "epoch": 0.40488108592916444,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002219808504218328,
+      "loss": 1.1816,
+      "step": 4613
+    },
+    {
+      "epoch": 0.40496885551206696,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.002219423991076438,
+      "loss": 1.2388,
+      "step": 4614
+    },
+    {
+      "epoch": 0.4050566250949694,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0022190394217346523,
+      "loss": 1.2437,
+      "step": 4615
+    },
+    {
+      "epoch": 0.40514439467787194,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.002218654796230928,
+      "loss": 1.2384,
+      "step": 4616
+    },
+    {
+      "epoch": 0.4052321642607744,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002218270114603224,
+      "loss": 1.2531,
+      "step": 4617
+    },
+    {
+      "epoch": 0.40531993384367687,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00221788537688951,
+      "loss": 1.2293,
+      "step": 4618
+    },
+    {
+      "epoch": 0.4054077034265794,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002217500583127755,
+      "loss": 1.2216,
+      "step": 4619
+    },
+    {
+      "epoch": 0.40549547300948185,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002217115733355939,
+      "loss": 1.2342,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4055832425923843,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002216730827612043,
+      "loss": 1.2192,
+      "step": 4621
+    },
+    {
+      "epoch": 0.40567101217528684,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0022163458659340573,
+      "loss": 1.2177,
+      "step": 4622
+    },
+    {
+      "epoch": 0.4057587817581893,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.002215960848359975,
+      "loss": 1.2334,
+      "step": 4623
+    },
+    {
+      "epoch": 0.4058465513410918,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0022155757749277977,
+      "loss": 1.2084,
+      "step": 4624
+    },
+    {
+      "epoch": 0.4059343209239943,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002215190645675528,
+      "loss": 1.209,
+      "step": 4625
+    },
+    {
+      "epoch": 0.40602209050689675,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0022148054606411787,
+      "loss": 1.1929,
+      "step": 4626
+    },
+    {
+      "epoch": 0.40610986008979927,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0022144202198627646,
+      "loss": 1.1999,
+      "step": 4627
+    },
+    {
+      "epoch": 0.40619762967270173,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0022140349233783084,
+      "loss": 1.268,
+      "step": 4628
+    },
+    {
+      "epoch": 0.40628539925560425,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002213649571225837,
+      "loss": 1.2848,
+      "step": 4629
+    },
+    {
+      "epoch": 0.4063731688385067,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0022132641634433833,
+      "loss": 1.2309,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4064609384214092,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002212878700068985,
+      "loss": 1.2424,
+      "step": 4631
+    },
+    {
+      "epoch": 0.4065487080043117,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002212493181140687,
+      "loss": 1.2071,
+      "step": 4632
+    },
+    {
+      "epoch": 0.40663647758721416,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0022121076066965364,
+      "loss": 1.1878,
+      "step": 4633
+    },
+    {
+      "epoch": 0.4067242471701166,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0022117219767745903,
+      "loss": 1.177,
+      "step": 4634
+    },
+    {
+      "epoch": 0.40681201675301915,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0022113362914129076,
+      "loss": 1.2387,
+      "step": 4635
+    },
+    {
+      "epoch": 0.4068997863359216,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002210950550649554,
+      "loss": 1.2387,
+      "step": 4636
+    },
+    {
+      "epoch": 0.40698755591882413,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0022105647545226,
+      "loss": 1.2116,
+      "step": 4637
+    },
+    {
+      "epoch": 0.4070753255017266,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.002210178903070123,
+      "loss": 1.2735,
+      "step": 4638
+    },
+    {
+      "epoch": 0.40716309508462906,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0022097929963302053,
+      "loss": 1.2082,
+      "step": 4639
+    },
+    {
+      "epoch": 0.4072508646675316,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0022094070343409338,
+      "loss": 1.3171,
+      "step": 4640
+    },
+    {
+      "epoch": 0.40733863425043404,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0022090210171404013,
+      "loss": 1.1991,
+      "step": 4641
+    },
+    {
+      "epoch": 0.40742640383333656,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.002208634944766706,
+      "loss": 1.2177,
+      "step": 4642
+    },
+    {
+      "epoch": 0.407514173416239,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.002208248817257952,
+      "loss": 1.1877,
+      "step": 4643
+    },
+    {
+      "epoch": 0.4076019429991415,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0022078626346522493,
+      "loss": 1.1938,
+      "step": 4644
+    },
+    {
+      "epoch": 0.407689712582044,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0022074763969877113,
+      "loss": 1.2259,
+      "step": 4645
+    },
+    {
+      "epoch": 0.40777748216494647,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0022070901043024596,
+      "loss": 1.2457,
+      "step": 4646
+    },
+    {
+      "epoch": 0.40786525174784893,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0022067037566346178,
+      "loss": 1.2244,
+      "step": 4647
+    },
+    {
+      "epoch": 0.40795302133075145,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0022063173540223186,
+      "loss": 1.2694,
+      "step": 4648
+    },
+    {
+      "epoch": 0.4080407909136539,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0022059308965036977,
+      "loss": 1.2373,
+      "step": 4649
+    },
+    {
+      "epoch": 0.40812856049655644,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002205544384116897,
+      "loss": 1.282,
+      "step": 4650
+    },
+    {
+      "epoch": 0.4082163300794589,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0022051578169000638,
+      "loss": 1.2021,
+      "step": 4651
+    },
+    {
+      "epoch": 0.40830409966236136,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0022047711948913504,
+      "loss": 1.2458,
+      "step": 4652
+    },
+    {
+      "epoch": 0.4083918692452639,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002204384518128916,
+      "loss": 1.1999,
+      "step": 4653
+    },
+    {
+      "epoch": 0.40847963882816635,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0022039977866509224,
+      "loss": 1.1851,
+      "step": 4654
+    },
+    {
+      "epoch": 0.40856740841106887,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0022036110004955392,
+      "loss": 1.2402,
+      "step": 4655
+    },
+    {
+      "epoch": 0.40865517799397133,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0022032241597009414,
+      "loss": 1.1925,
+      "step": 4656
+    },
+    {
+      "epoch": 0.4087429475768738,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002202837264305308,
+      "loss": 1.2129,
+      "step": 4657
+    },
+    {
+      "epoch": 0.4088307171597763,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002202450314346824,
+      "loss": 1.2575,
+      "step": 4658
+    },
+    {
+      "epoch": 0.4089184867426788,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00220206330986368,
+      "loss": 1.2332,
+      "step": 4659
+    },
+    {
+      "epoch": 0.40900625632558124,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0022016762508940716,
+      "loss": 1.2346,
+      "step": 4660
+    },
+    {
+      "epoch": 0.40909402590848376,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0022012891374762,
+      "loss": 1.2313,
+      "step": 4661
+    },
+    {
+      "epoch": 0.4091817954913862,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0022009019696482717,
+      "loss": 1.2965,
+      "step": 4662
+    },
+    {
+      "epoch": 0.40926956507428874,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0022005147474484987,
+      "loss": 1.1983,
+      "step": 4663
+    },
+    {
+      "epoch": 0.4093573346571912,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002200127470915099,
+      "loss": 1.2684,
+      "step": 4664
+    },
+    {
+      "epoch": 0.40944510424009367,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002199740140086295,
+      "loss": 1.1919,
+      "step": 4665
+    },
+    {
+      "epoch": 0.4095328738229962,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0021993527550003135,
+      "loss": 1.2368,
+      "step": 4666
+    },
+    {
+      "epoch": 0.40962064340589865,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0021989653156953884,
+      "loss": 1.2697,
+      "step": 4667
+    },
+    {
+      "epoch": 0.4097084129888012,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0021985778222097596,
+      "loss": 1.2202,
+      "step": 4668
+    },
+    {
+      "epoch": 0.40979618257170364,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0021981902745816704,
+      "loss": 1.2344,
+      "step": 4669
+    },
+    {
+      "epoch": 0.4098839521546061,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00219780267284937,
+      "loss": 1.2775,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4099717217375086,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0021974150170511134,
+      "loss": 1.1451,
+      "step": 4671
+    },
+    {
+      "epoch": 0.4100594913204111,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0021970273072251605,
+      "loss": 1.2525,
+      "step": 4672
+    },
+    {
+      "epoch": 0.41014726090331355,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002196639543409777,
+      "loss": 1.206,
+      "step": 4673
+    },
+    {
+      "epoch": 0.41023503048621607,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.002196251725643233,
+      "loss": 1.1734,
+      "step": 4674
+    },
+    {
+      "epoch": 0.41032280006911853,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0021958638539638056,
+      "loss": 1.1711,
+      "step": 4675
+    },
+    {
+      "epoch": 0.41041056965202105,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.002195475928409776,
+      "loss": 1.2429,
+      "step": 4676
+    },
+    {
+      "epoch": 0.4104983392349235,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00219508794901943,
+      "loss": 1.2194,
+      "step": 4677
+    },
+    {
+      "epoch": 0.410586108817826,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0021946999158310603,
+      "loss": 1.237,
+      "step": 4678
+    },
+    {
+      "epoch": 0.4106738784007285,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0021943118288829644,
+      "loss": 1.2404,
+      "step": 4679
+    },
+    {
+      "epoch": 0.41076164798363096,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0021939236882134445,
+      "loss": 1.1934,
+      "step": 4680
+    },
+    {
+      "epoch": 0.4108494175665335,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002193535493860809,
+      "loss": 1.2583,
+      "step": 4681
+    },
+    {
+      "epoch": 0.41093718714943595,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0021931472458633704,
+      "loss": 1.2277,
+      "step": 4682
+    },
+    {
+      "epoch": 0.4110249567323384,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002192758944259449,
+      "loss": 1.213,
+      "step": 4683
+    },
+    {
+      "epoch": 0.41111272631524093,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002192370589087366,
+      "loss": 1.2184,
+      "step": 4684
+    },
+    {
+      "epoch": 0.4112004958981434,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002191982180385453,
+      "loss": 1.1809,
+      "step": 4685
+    },
+    {
+      "epoch": 0.4112882654810459,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0021915937181920422,
+      "loss": 1.238,
+      "step": 4686
+    },
+    {
+      "epoch": 0.4113760350639484,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.002191205202545475,
+      "loss": 1.2676,
+      "step": 4687
+    },
+    {
+      "epoch": 0.41146380464685084,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002190816633484096,
+      "loss": 1.2123,
+      "step": 4688
+    },
+    {
+      "epoch": 0.41155157422975336,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.002190428011046255,
+      "loss": 1.2053,
+      "step": 4689
+    },
+    {
+      "epoch": 0.4116393438126558,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002190039335270307,
+      "loss": 1.2155,
+      "step": 4690
+    },
+    {
+      "epoch": 0.4117271133955583,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0021896506061946145,
+      "loss": 1.2653,
+      "step": 4691
+    },
+    {
+      "epoch": 0.4118148829784608,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0021892618238575417,
+      "loss": 1.2147,
+      "step": 4692
+    },
+    {
+      "epoch": 0.41190265256136327,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002188872988297461,
+      "loss": 1.2205,
+      "step": 4693
+    },
+    {
+      "epoch": 0.4119904221442658,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0021884840995527485,
+      "loss": 1.223,
+      "step": 4694
+    },
+    {
+      "epoch": 0.41207819172716825,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0021880951576617865,
+      "loss": 1.2465,
+      "step": 4695
+    },
+    {
+      "epoch": 0.4121659613100707,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.002187706162662961,
+      "loss": 1.1954,
+      "step": 4696
+    },
+    {
+      "epoch": 0.41225373089297324,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002187317114594665,
+      "loss": 1.155,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4123415004758757,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002186928013495296,
+      "loss": 1.1648,
+      "step": 4698
+    },
+    {
+      "epoch": 0.4124292700587782,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.002186538859403256,
+      "loss": 1.196,
+      "step": 4699
+    },
+    {
+      "epoch": 0.4125170396416807,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0021861496523569537,
+      "loss": 1.2401,
+      "step": 4700
+    },
+    {
+      "epoch": 0.41260480922458315,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0021857603923948025,
+      "loss": 1.2838,
+      "step": 4701
+    },
+    {
+      "epoch": 0.41269257880748567,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0021853710795552207,
+      "loss": 1.2658,
+      "step": 4702
+    },
+    {
+      "epoch": 0.41278034839038813,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002184981713876631,
+      "loss": 1.1432,
+      "step": 4703
+    },
+    {
+      "epoch": 0.4128681179732906,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0021845922953974634,
+      "loss": 1.2231,
+      "step": 4704
+    },
+    {
+      "epoch": 0.4129558875561931,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0021842028241561514,
+      "loss": 1.2206,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4130436571390956,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0021838133001911343,
+      "loss": 1.2641,
+      "step": 4706
+    },
+    {
+      "epoch": 0.4131314267219981,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0021834237235408573,
+      "loss": 1.2306,
+      "step": 4707
+    },
+    {
+      "epoch": 0.41321919630490056,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0021830340942437685,
+      "loss": 1.2655,
+      "step": 4708
+    },
+    {
+      "epoch": 0.413306965887803,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.002182644412338324,
+      "loss": 1.2207,
+      "step": 4709
+    },
+    {
+      "epoch": 0.41339473547070554,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002182254677862984,
+      "loss": 1.2055,
+      "step": 4710
+    },
+    {
+      "epoch": 0.413482505053608,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.002181864890856213,
+      "loss": 1.2459,
+      "step": 4711
+    },
+    {
+      "epoch": 0.4135702746365105,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002181475051356482,
+      "loss": 1.2187,
+      "step": 4712
+    },
+    {
+      "epoch": 0.413658044219413,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002181085159402266,
+      "loss": 1.2948,
+      "step": 4713
+    },
+    {
+      "epoch": 0.41374581380231545,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0021806952150320463,
+      "loss": 1.2339,
+      "step": 4714
+    },
+    {
+      "epoch": 0.413833583385218,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0021803052182843086,
+      "loss": 1.2105,
+      "step": 4715
+    },
+    {
+      "epoch": 0.41392135296812044,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0021799151691975446,
+      "loss": 1.2213,
+      "step": 4716
+    },
+    {
+      "epoch": 0.4140091225510229,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00217952506781025,
+      "loss": 1.2304,
+      "step": 4717
+    },
+    {
+      "epoch": 0.4140968921339254,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002179134914160926,
+      "loss": 1.2412,
+      "step": 4718
+    },
+    {
+      "epoch": 0.4141846617168279,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.00217874470828808,
+      "loss": 1.2214,
+      "step": 4719
+    },
+    {
+      "epoch": 0.4142724312997304,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0021783544502302237,
+      "loss": 1.2616,
+      "step": 4720
+    },
+    {
+      "epoch": 0.41436020088263287,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0021779641400258734,
+      "loss": 1.234,
+      "step": 4721
+    },
+    {
+      "epoch": 0.41444797046553533,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0021775737777135515,
+      "loss": 1.2485,
+      "step": 4722
+    },
+    {
+      "epoch": 0.41453574004843785,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.002177183363331785,
+      "loss": 1.2255,
+      "step": 4723
+    },
+    {
+      "epoch": 0.4146235096313403,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002176792896919107,
+      "loss": 1.2584,
+      "step": 4724
+    },
+    {
+      "epoch": 0.41471127921424283,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0021764023785140543,
+      "loss": 1.2259,
+      "step": 4725
+    },
+    {
+      "epoch": 0.4147990487971453,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.00217601180815517,
+      "loss": 1.1945,
+      "step": 4726
+    },
+    {
+      "epoch": 0.41488681838004776,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002175621185881001,
+      "loss": 1.232,
+      "step": 4727
+    },
+    {
+      "epoch": 0.4149745879629503,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0021752305117301012,
+      "loss": 1.174,
+      "step": 4728
+    },
+    {
+      "epoch": 0.41506235754585274,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002174839785741028,
+      "loss": 1.2413,
+      "step": 4729
+    },
+    {
+      "epoch": 0.4151501271287552,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0021744490079523443,
+      "loss": 1.2116,
+      "step": 4730
+    },
+    {
+      "epoch": 0.41523789671165773,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.002174058178402619,
+      "loss": 1.1809,
+      "step": 4731
+    },
+    {
+      "epoch": 0.4153256662945602,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0021736672971304255,
+      "loss": 1.1998,
+      "step": 4732
+    },
+    {
+      "epoch": 0.4154134358774627,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002173276364174341,
+      "loss": 1.2295,
+      "step": 4733
+    },
+    {
+      "epoch": 0.4155012054603652,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00217288537957295,
+      "loss": 1.19,
+      "step": 4734
+    },
+    {
+      "epoch": 0.41558897504326764,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.002172494343364841,
+      "loss": 1.2175,
+      "step": 4735
+    },
+    {
+      "epoch": 0.41567674462617016,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0021721032555886083,
+      "loss": 1.1389,
+      "step": 4736
+    },
+    {
+      "epoch": 0.4157645142090726,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00217171211628285,
+      "loss": 1.2296,
+      "step": 4737
+    },
+    {
+      "epoch": 0.41585228379197514,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.002171320925486169,
+      "loss": 1.2218,
+      "step": 4738
+    },
+    {
+      "epoch": 0.4159400533748776,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0021709296832371763,
+      "loss": 1.189,
+      "step": 4739
+    },
+    {
+      "epoch": 0.41602782295778007,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0021705383895744856,
+      "loss": 1.259,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4161155925406826,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002170147044536715,
+      "loss": 1.154,
+      "step": 4741
+    },
+    {
+      "epoch": 0.41620336212358505,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002169755648162489,
+      "loss": 1.204,
+      "step": 4742
+    },
+    {
+      "epoch": 0.4162911317064875,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0021693642004904376,
+      "loss": 1.2167,
+      "step": 4743
+    },
+    {
+      "epoch": 0.41637890128939004,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0021689727015591943,
+      "loss": 1.2386,
+      "step": 4744
+    },
+    {
+      "epoch": 0.4164666708722925,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0021685811514073985,
+      "loss": 1.2215,
+      "step": 4745
+    },
+    {
+      "epoch": 0.416554440455195,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0021681895500736954,
+      "loss": 1.1829,
+      "step": 4746
+    },
+    {
+      "epoch": 0.4166422100380975,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.002167797897596734,
+      "loss": 1.219,
+      "step": 4747
+    },
+    {
+      "epoch": 0.41672997962099995,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0021674061940151687,
+      "loss": 1.306,
+      "step": 4748
+    },
+    {
+      "epoch": 0.41681774920390247,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00216701443936766,
+      "loss": 1.2115,
+      "step": 4749
+    },
+    {
+      "epoch": 0.41690551878680493,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.002166622633692871,
+      "loss": 1.2309,
+      "step": 4750
+    },
+    {
+      "epoch": 0.41699328836970745,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0021662307770294728,
+      "loss": 1.1403,
+      "step": 4751
+    },
+    {
+      "epoch": 0.4170810579526099,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0021658388694161396,
+      "loss": 1.2352,
+      "step": 4752
+    },
+    {
+      "epoch": 0.4171688275355124,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0021654469108915507,
+      "loss": 1.2532,
+      "step": 4753
+    },
+    {
+      "epoch": 0.4172565971184149,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0021650549014943913,
+      "loss": 1.136,
+      "step": 4754
+    },
+    {
+      "epoch": 0.41734436670131736,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0021646628412633514,
+      "loss": 1.2391,
+      "step": 4755
+    },
+    {
+      "epoch": 0.4174321362842199,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.002164270730237125,
+      "loss": 1.2304,
+      "step": 4756
+    },
+    {
+      "epoch": 0.41751990586712234,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0021638785684544126,
+      "loss": 1.2285,
+      "step": 4757
+    },
+    {
+      "epoch": 0.4176076754500248,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002163486355953919,
+      "loss": 1.267,
+      "step": 4758
+    },
+    {
+      "epoch": 0.4176954450329273,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0021630940927743537,
+      "loss": 1.2449,
+      "step": 4759
+    },
+    {
+      "epoch": 0.4177832146158298,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0021627017789544317,
+      "loss": 1.2094,
+      "step": 4760
+    },
+    {
+      "epoch": 0.41787098419873225,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0021623094145328727,
+      "loss": 1.2583,
+      "step": 4761
+    },
+    {
+      "epoch": 0.4179587537816348,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002161916999548402,
+      "loss": 1.2617,
+      "step": 4762
+    },
+    {
+      "epoch": 0.41804652336453724,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0021615245340397484,
+      "loss": 1.2177,
+      "step": 4763
+    },
+    {
+      "epoch": 0.41813429294743976,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.002161132018045647,
+      "loss": 1.1464,
+      "step": 4764
+    },
+    {
+      "epoch": 0.4182220625303422,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0021607394516048387,
+      "loss": 1.1842,
+      "step": 4765
+    },
+    {
+      "epoch": 0.4183098321132447,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.002160346834756067,
+      "loss": 1.2119,
+      "step": 4766
+    },
+    {
+      "epoch": 0.4183976016961472,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0021599541675380823,
+      "loss": 1.2508,
+      "step": 4767
+    },
+    {
+      "epoch": 0.41848537127904967,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0021595614499896384,
+      "loss": 1.2805,
+      "step": 4768
+    },
+    {
+      "epoch": 0.4185731408619522,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.002159168682149496,
+      "loss": 1.2457,
+      "step": 4769
+    },
+    {
+      "epoch": 0.41866091044485465,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0021587758640564197,
+      "loss": 1.2406,
+      "step": 4770
+    },
+    {
+      "epoch": 0.4187486800277571,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0021583829957491783,
+      "loss": 1.2481,
+      "step": 4771
+    },
+    {
+      "epoch": 0.41883644961065963,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002157990077266547,
+      "loss": 1.2297,
+      "step": 4772
+    },
+    {
+      "epoch": 0.4189242191935621,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.002157597108647305,
+      "loss": 1.2477,
+      "step": 4773
+    },
+    {
+      "epoch": 0.41901198877646456,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002157204089930237,
+      "loss": 1.197,
+      "step": 4774
+    },
+    {
+      "epoch": 0.4190997583593671,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0021568110211541325,
+      "loss": 1.2114,
+      "step": 4775
+    },
+    {
+      "epoch": 0.41918752794226954,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0021564179023577847,
+      "loss": 1.189,
+      "step": 4776
+    },
+    {
+      "epoch": 0.41927529752517206,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0021560247335799944,
+      "loss": 1.2591,
+      "step": 4777
+    },
+    {
+      "epoch": 0.4193630671080745,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.002155631514859565,
+      "loss": 1.2123,
+      "step": 4778
+    },
+    {
+      "epoch": 0.419450836690977,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0021552382462353054,
+      "loss": 1.2137,
+      "step": 4779
+    },
+    {
+      "epoch": 0.4195386062738795,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0021548449277460306,
+      "loss": 1.2143,
+      "step": 4780
+    },
+    {
+      "epoch": 0.419626375856782,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.002154451559430559,
+      "loss": 1.2746,
+      "step": 4781
+    },
+    {
+      "epoch": 0.4197141454396845,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0021540581413277138,
+      "loss": 1.1956,
+      "step": 4782
+    },
+    {
+      "epoch": 0.41980191502258696,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002153664673476325,
+      "loss": 1.2191,
+      "step": 4783
+    },
+    {
+      "epoch": 0.4198896846054894,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002153271155915226,
+      "loss": 1.3266,
+      "step": 4784
+    },
+    {
+      "epoch": 0.41997745418839194,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.002152877588683255,
+      "loss": 1.2224,
+      "step": 4785
+    },
+    {
+      "epoch": 0.4200652237712944,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0021524839718192563,
+      "loss": 1.1673,
+      "step": 4786
+    },
+    {
+      "epoch": 0.42015299335419687,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.002152090305362077,
+      "loss": 1.2373,
+      "step": 4787
+    },
+    {
+      "epoch": 0.4202407629370994,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002151696589350571,
+      "loss": 1.1805,
+      "step": 4788
+    },
+    {
+      "epoch": 0.42032853252000185,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002151302823823598,
+      "loss": 1.2032,
+      "step": 4789
+    },
+    {
+      "epoch": 0.42041630210290437,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.002150909008820019,
+      "loss": 1.2398,
+      "step": 4790
+    },
+    {
+      "epoch": 0.42050407168580684,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.002150515144378703,
+      "loss": 1.2101,
+      "step": 4791
+    },
+    {
+      "epoch": 0.4205918412687093,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002150121230538523,
+      "loss": 1.2529,
+      "step": 4792
+    },
+    {
+      "epoch": 0.4206796108516118,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0021497272673383556,
+      "loss": 1.2364,
+      "step": 4793
+    },
+    {
+      "epoch": 0.4207673804345143,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0021493332548170848,
+      "loss": 1.2102,
+      "step": 4794
+    },
+    {
+      "epoch": 0.4208551500174168,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0021489391930135976,
+      "loss": 1.2395,
+      "step": 4795
+    },
+    {
+      "epoch": 0.42094291960031927,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.002148545081966786,
+      "loss": 1.2086,
+      "step": 4796
+    },
+    {
+      "epoch": 0.42103068918322173,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0021481509217155478,
+      "loss": 1.1661,
+      "step": 4797
+    },
+    {
+      "epoch": 0.42111845876612425,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.002147756712298784,
+      "loss": 1.1985,
+      "step": 4798
+    },
+    {
+      "epoch": 0.4212062283490267,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0021473624537554025,
+      "loss": 1.1822,
+      "step": 4799
+    },
+    {
+      "epoch": 0.4212939979319292,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0021469681461243157,
+      "loss": 1.2171,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4213817675148317,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0021465737894444384,
+      "loss": 1.2201,
+      "step": 4801
+    },
+    {
+      "epoch": 0.42146953709773416,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0021461793837546927,
+      "loss": 1.2769,
+      "step": 4802
+    },
+    {
+      "epoch": 0.4215573066806367,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0021457849290940053,
+      "loss": 1.2378,
+      "step": 4803
+    },
+    {
+      "epoch": 0.42164507626353914,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.002145390425501308,
+      "loss": 1.2248,
+      "step": 4804
+    },
+    {
+      "epoch": 0.4217328458464416,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0021449958730155347,
+      "loss": 1.2129,
+      "step": 4805
+    },
+    {
+      "epoch": 0.4218206154293441,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0021446012716756276,
+      "loss": 1.201,
+      "step": 4806
+    },
+    {
+      "epoch": 0.4219083850122466,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0021442066215205326,
+      "loss": 1.2552,
+      "step": 4807
+    },
+    {
+      "epoch": 0.4219961545951491,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002143811922589199,
+      "loss": 1.2031,
+      "step": 4808
+    },
+    {
+      "epoch": 0.4220839241780516,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0021434171749205827,
+      "loss": 1.2108,
+      "step": 4809
+    },
+    {
+      "epoch": 0.42217169376095404,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002143022378553644,
+      "loss": 1.2226,
+      "step": 4810
+    },
+    {
+      "epoch": 0.42225946334385656,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0021426275335273462,
+      "loss": 1.232,
+      "step": 4811
+    },
+    {
+      "epoch": 0.422347232926759,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0021422326398806615,
+      "loss": 1.2116,
+      "step": 4812
+    },
+    {
+      "epoch": 0.4224350025096615,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0021418376976525618,
+      "loss": 1.2245,
+      "step": 4813
+    },
+    {
+      "epoch": 0.422522772092564,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.002141442706882028,
+      "loss": 1.203,
+      "step": 4814
+    },
+    {
+      "epoch": 0.42261054167546647,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.002141047667608043,
+      "loss": 1.2173,
+      "step": 4815
+    },
+    {
+      "epoch": 0.422698311258369,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.002140652579869597,
+      "loss": 1.1809,
+      "step": 4816
+    },
+    {
+      "epoch": 0.42278608084127145,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0021402574437056824,
+      "loss": 1.2839,
+      "step": 4817
+    },
+    {
+      "epoch": 0.4228738504241739,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002139862259155298,
+      "loss": 1.2676,
+      "step": 4818
+    },
+    {
+      "epoch": 0.42296162000707643,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0021394670262574467,
+      "loss": 1.225,
+      "step": 4819
+    },
+    {
+      "epoch": 0.4230493895899789,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002139071745051137,
+      "loss": 1.1692,
+      "step": 4820
+    },
+    {
+      "epoch": 0.4231371591728814,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.002138676415575381,
+      "loss": 1.2702,
+      "step": 4821
+    },
+    {
+      "epoch": 0.4232249287557839,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002138281037869196,
+      "loss": 1.2643,
+      "step": 4822
+    },
+    {
+      "epoch": 0.42331269833868634,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002137885611971605,
+      "loss": 1.289,
+      "step": 4823
+    },
+    {
+      "epoch": 0.42340046792158886,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0021374901379216343,
+      "loss": 1.213,
+      "step": 4824
+    },
+    {
+      "epoch": 0.4234882375044913,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0021370946157583163,
+      "loss": 1.2003,
+      "step": 4825
+    },
+    {
+      "epoch": 0.42357600708739385,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002136699045520687,
+      "loss": 1.1635,
+      "step": 4826
+    },
+    {
+      "epoch": 0.4236637766702963,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002136303427247788,
+      "loss": 1.212,
+      "step": 4827
+    },
+    {
+      "epoch": 0.4237515462531988,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0021359077609786645,
+      "loss": 1.1596,
+      "step": 4828
+    },
+    {
+      "epoch": 0.4238393158361013,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002135512046752367,
+      "loss": 1.2204,
+      "step": 4829
+    },
+    {
+      "epoch": 0.42392708541900376,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0021351162846079525,
+      "loss": 1.1735,
+      "step": 4830
+    },
+    {
+      "epoch": 0.4240148550019062,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00213472047458448,
+      "loss": 1.2176,
+      "step": 4831
+    },
+    {
+      "epoch": 0.42410262458480874,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002134324616721015,
+      "loss": 1.205,
+      "step": 4832
+    },
+    {
+      "epoch": 0.4241903941677112,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0021339287110566257,
+      "loss": 1.2208,
+      "step": 4833
+    },
+    {
+      "epoch": 0.4242781637506137,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0021335327576303877,
+      "loss": 1.2623,
+      "step": 4834
+    },
+    {
+      "epoch": 0.4243659333335162,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00213313675648138,
+      "loss": 1.2083,
+      "step": 4835
+    },
+    {
+      "epoch": 0.42445370291641865,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0021327407076486863,
+      "loss": 1.2316,
+      "step": 4836
+    },
+    {
+      "epoch": 0.42454147249932117,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0021323446111713947,
+      "loss": 1.2241,
+      "step": 4837
+    },
+    {
+      "epoch": 0.42462924208222363,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002131948467088599,
+      "loss": 1.2591,
+      "step": 4838
+    },
+    {
+      "epoch": 0.42471701166512615,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0021315522754393953,
+      "loss": 1.2407,
+      "step": 4839
+    },
+    {
+      "epoch": 0.4248047812480286,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0021311560362628885,
+      "loss": 1.1952,
+      "step": 4840
+    },
+    {
+      "epoch": 0.4248925508309311,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.002130759749598184,
+      "loss": 1.1734,
+      "step": 4841
+    },
+    {
+      "epoch": 0.4249803204138336,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0021303634154843953,
+      "loss": 1.2182,
+      "step": 4842
+    },
+    {
+      "epoch": 0.42506808999673606,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002129967033960638,
+      "loss": 1.2365,
+      "step": 4843
+    },
+    {
+      "epoch": 0.42515585957963853,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002129570605066033,
+      "loss": 1.2147,
+      "step": 4844
+    },
+    {
+      "epoch": 0.42524362916254105,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0021291741288397077,
+      "loss": 1.1792,
+      "step": 4845
+    },
+    {
+      "epoch": 0.4253313987454435,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002128777605320792,
+      "loss": 1.1898,
+      "step": 4846
+    },
+    {
+      "epoch": 0.42541916832834603,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00212838103454842,
+      "loss": 1.2829,
+      "step": 4847
+    },
+    {
+      "epoch": 0.4255069379112485,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0021279844165617333,
+      "loss": 1.2568,
+      "step": 4848
+    },
+    {
+      "epoch": 0.42559470749415096,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.002127587751399876,
+      "loss": 1.1785,
+      "step": 4849
+    },
+    {
+      "epoch": 0.4256824770770535,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0021271910391019973,
+      "loss": 1.2474,
+      "step": 4850
+    },
+    {
+      "epoch": 0.42577024665995594,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0021267942797072513,
+      "loss": 1.2818,
+      "step": 4851
+    },
+    {
+      "epoch": 0.42585801624285846,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0021263974732547964,
+      "loss": 1.21,
+      "step": 4852
+    },
+    {
+      "epoch": 0.4259457858257609,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0021260006197837958,
+      "loss": 1.2483,
+      "step": 4853
+    },
+    {
+      "epoch": 0.4260335554086634,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0021256037193334173,
+      "loss": 1.2196,
+      "step": 4854
+    },
+    {
+      "epoch": 0.4261213249915659,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0021252067719428345,
+      "loss": 1.2145,
+      "step": 4855
+    },
+    {
+      "epoch": 0.4262090945744684,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.002124809777651222,
+      "loss": 1.2299,
+      "step": 4856
+    },
+    {
+      "epoch": 0.42629686415737084,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0021244127364977647,
+      "loss": 1.2151,
+      "step": 4857
+    },
+    {
+      "epoch": 0.42638463374027336,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.002124015648521647,
+      "loss": 1.2075,
+      "step": 4858
+    },
+    {
+      "epoch": 0.4264724033231758,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0021236185137620597,
+      "loss": 1.2226,
+      "step": 4859
+    },
+    {
+      "epoch": 0.42656017290607834,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0021232213322581995,
+      "loss": 1.2399,
+      "step": 4860
+    },
+    {
+      "epoch": 0.4266479424889808,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0021228241040492668,
+      "loss": 1.2479,
+      "step": 4861
+    },
+    {
+      "epoch": 0.42673571207188327,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0021224268291744654,
+      "loss": 1.2049,
+      "step": 4862
+    },
+    {
+      "epoch": 0.4268234816547858,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0021220295076730053,
+      "loss": 1.2894,
+      "step": 4863
+    },
+    {
+      "epoch": 0.42691125123768825,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0021216321395841004,
+      "loss": 1.1894,
+      "step": 4864
+    },
+    {
+      "epoch": 0.42699902082059077,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0021212347249469697,
+      "loss": 1.279,
+      "step": 4865
+    },
+    {
+      "epoch": 0.42708679040349323,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002120837263800836,
+      "loss": 1.2459,
+      "step": 4866
+    },
+    {
+      "epoch": 0.4271745599863957,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002120439756184928,
+      "loss": 1.1781,
+      "step": 4867
+    },
+    {
+      "epoch": 0.4272623295692982,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002120042202138477,
+      "loss": 1.2345,
+      "step": 4868
+    },
+    {
+      "epoch": 0.4273500991522007,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0021196446017007205,
+      "loss": 1.2378,
+      "step": 4869
+    },
+    {
+      "epoch": 0.42743786873510314,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0021192469549109,
+      "loss": 1.2194,
+      "step": 4870
+    },
+    {
+      "epoch": 0.42752563831800566,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002118849261808262,
+      "loss": 1.1683,
+      "step": 4871
+    },
+    {
+      "epoch": 0.4276134079009081,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.002118451522432057,
+      "loss": 1.2506,
+      "step": 4872
+    },
+    {
+      "epoch": 0.42770117748381065,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0021180537368215404,
+      "loss": 1.2513,
+      "step": 4873
+    },
+    {
+      "epoch": 0.4277889470667131,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0021176559050159714,
+      "loss": 1.2245,
+      "step": 4874
+    },
+    {
+      "epoch": 0.4278767166496156,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002117258027054616,
+      "loss": 1.1522,
+      "step": 4875
+    },
+    {
+      "epoch": 0.4279644862325181,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.002116860102976741,
+      "loss": 1.1851,
+      "step": 4876
+    },
+    {
+      "epoch": 0.42805225581542056,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.002116462132821622,
+      "loss": 1.2657,
+      "step": 4877
+    },
+    {
+      "epoch": 0.4281400253983231,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.002116064116628536,
+      "loss": 1.2087,
+      "step": 4878
+    },
+    {
+      "epoch": 0.42822779498122554,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002115666054436765,
+      "loss": 1.3052,
+      "step": 4879
+    },
+    {
+      "epoch": 0.428315564564128,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0021152679462855977,
+      "loss": 1.2098,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4284033341470305,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.002114869792214326,
+      "loss": 1.2051,
+      "step": 4881
+    },
+    {
+      "epoch": 0.428491103729933,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0021144715922622433,
+      "loss": 1.2361,
+      "step": 4882
+    },
+    {
+      "epoch": 0.42857887331283545,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.002114073346468654,
+      "loss": 1.2149,
+      "step": 4883
+    },
+    {
+      "epoch": 0.42866664289573797,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0021136750548728604,
+      "loss": 1.2524,
+      "step": 4884
+    },
+    {
+      "epoch": 0.42875441247864043,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0021132767175141742,
+      "loss": 1.2889,
+      "step": 4885
+    },
+    {
+      "epoch": 0.42884218206154295,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0021128783344319087,
+      "loss": 1.2189,
+      "step": 4886
+    },
+    {
+      "epoch": 0.4289299516444454,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0021124799056653834,
+      "loss": 1.259,
+      "step": 4887
+    },
+    {
+      "epoch": 0.4290177212273479,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.002112081431253921,
+      "loss": 1.207,
+      "step": 4888
+    },
+    {
+      "epoch": 0.4291054908102504,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00211168291123685,
+      "loss": 1.2211,
+      "step": 4889
+    },
+    {
+      "epoch": 0.42919326039315286,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0021112843456535025,
+      "loss": 1.2211,
+      "step": 4890
+    },
+    {
+      "epoch": 0.4292810299760554,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.002110885734543215,
+      "loss": 1.2225,
+      "step": 4891
+    },
+    {
+      "epoch": 0.42936879955895785,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0021104870779453293,
+      "loss": 1.2516,
+      "step": 4892
+    },
+    {
+      "epoch": 0.4294565691418603,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0021100883758991915,
+      "loss": 1.192,
+      "step": 4893
+    },
+    {
+      "epoch": 0.42954433872476283,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.002109689628444151,
+      "loss": 1.2158,
+      "step": 4894
+    },
+    {
+      "epoch": 0.4296321083076653,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.002109290835619563,
+      "loss": 1.2736,
+      "step": 4895
+    },
+    {
+      "epoch": 0.4297198778905678,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0021088919974647873,
+      "loss": 1.2297,
+      "step": 4896
+    },
+    {
+      "epoch": 0.4298076474734703,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.002108493114019187,
+      "loss": 1.2263,
+      "step": 4897
+    },
+    {
+      "epoch": 0.42989541705637274,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00210809418532213,
+      "loss": 1.2462,
+      "step": 4898
+    },
+    {
+      "epoch": 0.42998318663927526,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0021076952114129897,
+      "loss": 1.2258,
+      "step": 4899
+    },
+    {
+      "epoch": 0.4300709562221777,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0021072961923311435,
+      "loss": 1.2181,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4301587258050802,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.002106897128115972,
+      "loss": 1.1706,
+      "step": 4901
+    },
+    {
+      "epoch": 0.4302464953879827,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0021064980188068624,
+      "loss": 1.2507,
+      "step": 4902
+    },
+    {
+      "epoch": 0.43033426497088517,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0021060988644432043,
+      "loss": 1.2778,
+      "step": 4903
+    },
+    {
+      "epoch": 0.4304220345537877,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.002105699665064393,
+      "loss": 1.2409,
+      "step": 4904
+    },
+    {
+      "epoch": 0.43050980413669016,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0021053004207098277,
+      "loss": 1.1897,
+      "step": 4905
+    },
+    {
+      "epoch": 0.4305975737195926,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002104901131418913,
+      "loss": 1.2217,
+      "step": 4906
+    },
+    {
+      "epoch": 0.43068534330249514,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002104501797231056,
+      "loss": 1.1943,
+      "step": 4907
+    },
+    {
+      "epoch": 0.4307731128853976,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0021041024181856697,
+      "loss": 1.2348,
+      "step": 4908
+    },
+    {
+      "epoch": 0.4308608824683001,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0021037029943221717,
+      "loss": 1.2034,
+      "step": 4909
+    },
+    {
+      "epoch": 0.4309486520512026,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.002103303525679984,
+      "loss": 1.2225,
+      "step": 4910
+    },
+    {
+      "epoch": 0.43103642163410505,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.002102904012298531,
+      "loss": 1.2216,
+      "step": 4911
+    },
+    {
+      "epoch": 0.43112419121700757,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.002102504454217245,
+      "loss": 1.1379,
+      "step": 4912
+    },
+    {
+      "epoch": 0.43121196079991003,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00210210485147556,
+      "loss": 1.1809,
+      "step": 4913
+    },
+    {
+      "epoch": 0.4312997303828125,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002101705204112914,
+      "loss": 1.2513,
+      "step": 4914
+    },
+    {
+      "epoch": 0.431387499965715,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0021013055121687518,
+      "loss": 1.1985,
+      "step": 4915
+    },
+    {
+      "epoch": 0.4314752695486175,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0021009057756825223,
+      "loss": 1.3088,
+      "step": 4916
+    },
+    {
+      "epoch": 0.43156303913152,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.002100505994693676,
+      "loss": 1.2535,
+      "step": 4917
+    },
+    {
+      "epoch": 0.43165080871442246,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0021001061692416706,
+      "loss": 1.3011,
+      "step": 4918
+    },
+    {
+      "epoch": 0.4317385782973249,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0020997062993659676,
+      "loss": 1.286,
+      "step": 4919
+    },
+    {
+      "epoch": 0.43182634788022745,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0020993063851060317,
+      "loss": 1.2998,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4319141174631299,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.002098906426501334,
+      "loss": 1.1871,
+      "step": 4921
+    },
+    {
+      "epoch": 0.43200188704603243,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002098506423591348,
+      "loss": 1.2161,
+      "step": 4922
+    },
+    {
+      "epoch": 0.4320896566289349,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0020981063764155526,
+      "loss": 1.1934,
+      "step": 4923
+    },
+    {
+      "epoch": 0.43217742621183736,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0020977062850134317,
+      "loss": 1.2146,
+      "step": 4924
+    },
+    {
+      "epoch": 0.4322651957947399,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0020973061494244704,
+      "loss": 1.221,
+      "step": 4925
+    },
+    {
+      "epoch": 0.43235296537764234,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0020969059696881636,
+      "loss": 1.2071,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4324407349605448,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002096505745844005,
+      "loss": 1.1928,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4325285045434473,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0020961054779314965,
+      "loss": 1.2045,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4326162741263498,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0020957051659901416,
+      "loss": 1.2328,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4327040437092523,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0020953048100594515,
+      "loss": 1.2063,
+      "step": 4930
+    },
+    {
+      "epoch": 0.43279181329215477,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.002094904410178938,
+      "loss": 1.23,
+      "step": 4931
+    },
+    {
+      "epoch": 0.43287958287505723,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0020945039663881203,
+      "loss": 1.1795,
+      "step": 4932
+    },
+    {
+      "epoch": 0.43296735245795975,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.002094103478726519,
+      "loss": 1.2468,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4330551220408622,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0020937029472336615,
+      "loss": 1.195,
+      "step": 4934
+    },
+    {
+      "epoch": 0.43314289162376474,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0020933023719490796,
+      "loss": 1.2679,
+      "step": 4935
+    },
+    {
+      "epoch": 0.4332306612066672,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0020929017529123075,
+      "loss": 1.23,
+      "step": 4936
+    },
+    {
+      "epoch": 0.43331843078956966,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002092501090162884,
+      "loss": 1.2175,
+      "step": 4937
+    },
+    {
+      "epoch": 0.4334062003724722,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002092100383740355,
+      "loss": 1.2027,
+      "step": 4938
+    },
+    {
+      "epoch": 0.43349396995537465,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0020916996336842668,
+      "loss": 1.2464,
+      "step": 4939
+    },
+    {
+      "epoch": 0.4335817395382771,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.002091298840034172,
+      "loss": 1.237,
+      "step": 4940
+    },
+    {
+      "epoch": 0.43366950912117963,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0020908980028296288,
+      "loss": 1.2336,
+      "step": 4941
+    },
+    {
+      "epoch": 0.4337572787040821,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0020904971221101967,
+      "loss": 1.2119,
+      "step": 4942
+    },
+    {
+      "epoch": 0.4338450482869846,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.002090096197915442,
+      "loss": 1.2757,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4339328178698871,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0020896952302849334,
+      "loss": 1.2071,
+      "step": 4944
+    },
+    {
+      "epoch": 0.43402058745278954,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0020892942192582455,
+      "loss": 1.2993,
+      "step": 4945
+    },
+    {
+      "epoch": 0.43410835703569206,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0020888931648749567,
+      "loss": 1.1966,
+      "step": 4946
+    },
+    {
+      "epoch": 0.4341961266185945,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.002088492067174649,
+      "loss": 1.221,
+      "step": 4947
+    },
+    {
+      "epoch": 0.43428389620149704,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002088090926196909,
+      "loss": 1.2194,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4343716657843995,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.002087689741981328,
+      "loss": 1.2315,
+      "step": 4949
+    },
+    {
+      "epoch": 0.43445943536730197,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002087288514567502,
+      "loss": 1.2816,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4345472049502045,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0020868872439950294,
+      "loss": 1.2106,
+      "step": 4951
+    },
+    {
+      "epoch": 0.43463497453310695,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0020864859303035147,
+      "loss": 1.2538,
+      "step": 4952
+    },
+    {
+      "epoch": 0.4347227441160094,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0020860845735325654,
+      "loss": 1.2611,
+      "step": 4953
+    },
+    {
+      "epoch": 0.43481051369891194,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0020856831737217944,
+      "loss": 1.1426,
+      "step": 4954
+    },
+    {
+      "epoch": 0.4348982832818144,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.002085281730910818,
+      "loss": 1.2113,
+      "step": 4955
+    },
+    {
+      "epoch": 0.4349860528647169,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.002084880245139257,
+      "loss": 1.2201,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4350738224476194,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0020844787164467364,
+      "loss": 1.2335,
+      "step": 4957
+    },
+    {
+      "epoch": 0.43516159203052185,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.002084077144872886,
+      "loss": 1.2024,
+      "step": 4958
+    },
+    {
+      "epoch": 0.43524936161342437,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002083675530457339,
+      "loss": 1.199,
+      "step": 4959
+    },
+    {
+      "epoch": 0.43533713119632683,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0020832738732397328,
+      "loss": 1.2214,
+      "step": 4960
+    },
+    {
+      "epoch": 0.43542490077922935,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00208287217325971,
+      "loss": 1.1816,
+      "step": 4961
+    },
+    {
+      "epoch": 0.4355126703621318,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.002082470430556916,
+      "loss": 1.2475,
+      "step": 4962
+    },
+    {
+      "epoch": 0.4356004399450343,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0020820686451710024,
+      "loss": 1.1901,
+      "step": 4963
+    },
+    {
+      "epoch": 0.4356882095279368,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.002081666817141623,
+      "loss": 1.2312,
+      "step": 4964
+    },
+    {
+      "epoch": 0.43577597911083926,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.002081264946508436,
+      "loss": 1.2,
+      "step": 4965
+    },
+    {
+      "epoch": 0.4358637486937418,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.002080863033311106,
+      "loss": 1.2454,
+      "step": 4966
+    },
+    {
+      "epoch": 0.43595151827664425,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.002080461077589299,
+      "loss": 1.2297,
+      "step": 4967
+    },
+    {
+      "epoch": 0.4360392878595467,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0020800590793826875,
+      "loss": 1.286,
+      "step": 4968
+    },
+    {
+      "epoch": 0.43612705744244923,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002079657038730946,
+      "loss": 1.2141,
+      "step": 4969
+    },
+    {
+      "epoch": 0.4362148270253517,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0020792549556737545,
+      "loss": 1.2259,
+      "step": 4970
+    },
+    {
+      "epoch": 0.43630259660825416,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.002078852830250798,
+      "loss": 1.2029,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4363903661911567,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.002078450662501764,
+      "loss": 1.2036,
+      "step": 4972
+    },
+    {
+      "epoch": 0.43647813577405914,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.002078048452466345,
+      "loss": 1.2082,
+      "step": 4973
+    },
+    {
+      "epoch": 0.43656590535696166,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.002077646200184237,
+      "loss": 1.2463,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4366536749398641,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002077243905695141,
+      "loss": 1.245,
+      "step": 4975
+    },
+    {
+      "epoch": 0.4367414445227666,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0020768415690387626,
+      "loss": 1.1937,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4368292141056691,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0020764391902548104,
+      "loss": 1.1421,
+      "step": 4977
+    },
+    {
+      "epoch": 0.43691698368857157,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.002076036769382997,
+      "loss": 1.1797,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4370047532714741,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00207563430646304,
+      "loss": 1.1915,
+      "step": 4979
+    },
+    {
+      "epoch": 0.43709252285437655,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0020752318015346614,
+      "loss": 1.2574,
+      "step": 4980
+    },
+    {
+      "epoch": 0.437180292437279,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.002074829254637587,
+      "loss": 1.2448,
+      "step": 4981
+    },
+    {
+      "epoch": 0.43726806202018154,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0020744266658115455,
+      "loss": 1.2664,
+      "step": 4982
+    },
+    {
+      "epoch": 0.437355831603084,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0020740240350962716,
+      "loss": 1.2013,
+      "step": 4983
+    },
+    {
+      "epoch": 0.43744360118598646,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0020736213625315034,
+      "loss": 1.2475,
+      "step": 4984
+    },
+    {
+      "epoch": 0.437531370768889,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0020732186481569824,
+      "loss": 1.1894,
+      "step": 4985
+    },
+    {
+      "epoch": 0.43761914035179145,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002072815892012456,
+      "loss": 1.1894,
+      "step": 4986
+    },
+    {
+      "epoch": 0.43770690993469397,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0020724130941376737,
+      "loss": 1.2059,
+      "step": 4987
+    },
+    {
+      "epoch": 0.43779467951759643,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0020720102545723905,
+      "loss": 1.2256,
+      "step": 4988
+    },
+    {
+      "epoch": 0.4378824491004989,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.002071607373356365,
+      "loss": 1.2262,
+      "step": 4989
+    },
+    {
+      "epoch": 0.4379702186834014,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00207120445052936,
+      "loss": 1.2233,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4380579882663039,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0020708014861311424,
+      "loss": 1.2469,
+      "step": 4991
+    },
+    {
+      "epoch": 0.4381457578492064,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0020703984802014834,
+      "loss": 1.2283,
+      "step": 4992
+    },
+    {
+      "epoch": 0.43823352743210886,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.002069995432780157,
+      "loss": 1.1925,
+      "step": 4993
+    },
+    {
+      "epoch": 0.4383212970150113,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.002069592343906944,
+      "loss": 1.2098,
+      "step": 4994
+    },
+    {
+      "epoch": 0.43840906659791384,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0020691892136216264,
+      "loss": 1.1862,
+      "step": 4995
+    },
+    {
+      "epoch": 0.4384968361808163,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.002068786041963993,
+      "loss": 1.1716,
+      "step": 4996
+    },
+    {
+      "epoch": 0.43858460576371877,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0020683828289738333,
+      "loss": 1.2139,
+      "step": 4997
+    },
+    {
+      "epoch": 0.4386723753466213,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0020679795746909443,
+      "loss": 1.2342,
+      "step": 4998
+    },
+    {
+      "epoch": 0.43876014492952375,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.002067576279155125,
+      "loss": 1.2222,
+      "step": 4999
+    },
+    {
+      "epoch": 0.4388479145124263,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.002067172942406179,
+      "loss": 1.2803,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4388479145124263,
+      "eval_loss": 1.2060072422027588,
+      "eval_runtime": 437.7156,
+      "eval_samples_per_second": 33.675,
+      "eval_steps_per_second": 8.419,
+      "step": 5000
+    },
+    {
+      "epoch": 0.43893568409532874,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0020667695644839145,
+      "loss": 1.2491,
+      "step": 5001
+    },
+    {
+      "epoch": 0.4390234536782312,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0020663661454281435,
+      "loss": 1.2654,
+      "step": 5002
+    },
+    {
+      "epoch": 0.4391112232611337,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0020659626852786803,
+      "loss": 1.2196,
+      "step": 5003
+    },
+    {
+      "epoch": 0.4391989928440362,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0020655591840753468,
+      "loss": 1.2424,
+      "step": 5004
+    },
+    {
+      "epoch": 0.4392867624269387,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.002065155641857966,
+      "loss": 1.2331,
+      "step": 5005
+    },
+    {
+      "epoch": 0.43937453200984117,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.002064752058666366,
+      "loss": 1.2512,
+      "step": 5006
+    },
+    {
+      "epoch": 0.43946230159274363,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.002064348434540379,
+      "loss": 1.2417,
+      "step": 5007
+    },
+    {
+      "epoch": 0.43955007117564615,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.002063944769519841,
+      "loss": 1.2213,
+      "step": 5008
+    },
+    {
+      "epoch": 0.4396378407585486,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002063541063644592,
+      "loss": 1.1822,
+      "step": 5009
+    },
+    {
+      "epoch": 0.4397256103414511,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.002063137316954476,
+      "loss": 1.212,
+      "step": 5010
+    },
+    {
+      "epoch": 0.4398133799243536,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.002062733529489342,
+      "loss": 1.2641,
+      "step": 5011
+    },
+    {
+      "epoch": 0.43990114950725606,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0020623297012890415,
+      "loss": 1.3075,
+      "step": 5012
+    },
+    {
+      "epoch": 0.4399889190901586,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.002061925832393431,
+      "loss": 1.207,
+      "step": 5013
+    },
+    {
+      "epoch": 0.44007668867306104,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0020615219228423713,
+      "loss": 1.14,
+      "step": 5014
+    },
+    {
+      "epoch": 0.4401644582559635,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.002061117972675725,
+      "loss": 1.1703,
+      "step": 5015
+    },
+    {
+      "epoch": 0.44025222783886603,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0020607139819333624,
+      "loss": 1.2169,
+      "step": 5016
+    },
+    {
+      "epoch": 0.4403399974217685,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0020603099506551546,
+      "loss": 1.2056,
+      "step": 5017
+    },
+    {
+      "epoch": 0.440427767004671,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0020599058788809783,
+      "loss": 1.1458,
+      "step": 5018
+    },
+    {
+      "epoch": 0.4405155365875735,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.002059501766650714,
+      "loss": 1.2067,
+      "step": 5019
+    },
+    {
+      "epoch": 0.44060330617047594,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.002059097614004245,
+      "loss": 1.2233,
+      "step": 5020
+    },
+    {
+      "epoch": 0.44069107575337846,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00205869342098146,
+      "loss": 1.1693,
+      "step": 5021
+    },
+    {
+      "epoch": 0.4407788453362809,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0020582891876222524,
+      "loss": 1.2176,
+      "step": 5022
+    },
+    {
+      "epoch": 0.4408666149191834,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0020578849139665173,
+      "loss": 1.1948,
+      "step": 5023
+    },
+    {
+      "epoch": 0.4409543845020859,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.002057480600054155,
+      "loss": 1.2207,
+      "step": 5024
+    },
+    {
+      "epoch": 0.44104215408498837,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.002057076245925069,
+      "loss": 1.2369,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4411299236678909,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.002056671851619169,
+      "loss": 1.2231,
+      "step": 5026
+    },
+    {
+      "epoch": 0.44121769325079335,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.002056267417176366,
+      "loss": 1.2283,
+      "step": 5027
+    },
+    {
+      "epoch": 0.4413054628336958,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.002055862942636577,
+      "loss": 1.2307,
+      "step": 5028
+    },
+    {
+      "epoch": 0.44139323241659834,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0020554584280397213,
+      "loss": 1.2194,
+      "step": 5029
+    },
+    {
+      "epoch": 0.4414810019995008,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.002055053873425723,
+      "loss": 1.2712,
+      "step": 5030
+    },
+    {
+      "epoch": 0.4415687715824033,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0020546492788345097,
+      "loss": 1.2531,
+      "step": 5031
+    },
+    {
+      "epoch": 0.4416565411653058,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0020542446443060146,
+      "loss": 1.2172,
+      "step": 5032
+    },
+    {
+      "epoch": 0.44174431074820825,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.002053839969880172,
+      "loss": 1.1538,
+      "step": 5033
+    },
+    {
+      "epoch": 0.44183208033111077,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0020534352555969227,
+      "loss": 1.228,
+      "step": 5034
+    },
+    {
+      "epoch": 0.44191984991401323,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0020530305014962095,
+      "loss": 1.2101,
+      "step": 5035
+    },
+    {
+      "epoch": 0.44200761949691575,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0020526257076179805,
+      "loss": 1.2118,
+      "step": 5036
+    },
+    {
+      "epoch": 0.4420953890798182,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0020522208740021875,
+      "loss": 1.2232,
+      "step": 5037
+    },
+    {
+      "epoch": 0.4421831586627207,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0020518160006887855,
+      "loss": 1.2031,
+      "step": 5038
+    },
+    {
+      "epoch": 0.4422709282456232,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0020514110877177345,
+      "loss": 1.1569,
+      "step": 5039
+    },
+    {
+      "epoch": 0.44235869782852566,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002051006135128997,
+      "loss": 1.2677,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4424464674114281,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002050601142962541,
+      "loss": 1.1918,
+      "step": 5041
+    },
+    {
+      "epoch": 0.44253423699433064,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0020501961112583374,
+      "loss": 1.2445,
+      "step": 5042
+    },
+    {
+      "epoch": 0.4426220065772331,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.002049791040056361,
+      "loss": 1.234,
+      "step": 5043
+    },
+    {
+      "epoch": 0.4427097761601356,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002049385929396591,
+      "loss": 1.2196,
+      "step": 5044
+    },
+    {
+      "epoch": 0.4427975457430381,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0020489807793190092,
+      "loss": 1.1715,
+      "step": 5045
+    },
+    {
+      "epoch": 0.44288531532594055,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0020485755898636045,
+      "loss": 1.208,
+      "step": 5046
+    },
+    {
+      "epoch": 0.4429730849088431,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002048170361070365,
+      "loss": 1.2128,
+      "step": 5047
+    },
+    {
+      "epoch": 0.44306085449174554,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0020477650929792866,
+      "loss": 1.2059,
+      "step": 5048
+    },
+    {
+      "epoch": 0.44314862407464806,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.002047359785630368,
+      "loss": 1.2627,
+      "step": 5049
+    },
+    {
+      "epoch": 0.4432363936575505,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0020469544390636103,
+      "loss": 1.218,
+      "step": 5050
+    },
+    {
+      "epoch": 0.443324163240453,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0020465490533190206,
+      "loss": 1.1832,
+      "step": 5051
+    },
+    {
+      "epoch": 0.4434119328233555,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0020461436284366077,
+      "loss": 1.2542,
+      "step": 5052
+    },
+    {
+      "epoch": 0.44349970240625797,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0020457381644563865,
+      "loss": 1.2751,
+      "step": 5053
+    },
+    {
+      "epoch": 0.44358747198916043,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0020453326614183745,
+      "loss": 1.2302,
+      "step": 5054
+    },
+    {
+      "epoch": 0.44367524157206295,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.002044927119362593,
+      "loss": 1.1618,
+      "step": 5055
+    },
+    {
+      "epoch": 0.4437630111549654,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.002044521538329067,
+      "loss": 1.2212,
+      "step": 5056
+    },
+    {
+      "epoch": 0.44385078073786793,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.002044115918357827,
+      "loss": 1.1432,
+      "step": 5057
+    },
+    {
+      "epoch": 0.4439385503207704,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.002043710259488905,
+      "loss": 1.2317,
+      "step": 5058
+    },
+    {
+      "epoch": 0.44402631990367286,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002043304561762338,
+      "loss": 1.2073,
+      "step": 5059
+    },
+    {
+      "epoch": 0.4441140894865754,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002042898825218167,
+      "loss": 1.2476,
+      "step": 5060
+    },
+    {
+      "epoch": 0.44420185906947784,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0020424930498964366,
+      "loss": 1.2567,
+      "step": 5061
+    },
+    {
+      "epoch": 0.44428962865238036,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0020420872358371954,
+      "loss": 1.2038,
+      "step": 5062
+    },
+    {
+      "epoch": 0.44437739823528283,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0020416813830804953,
+      "loss": 1.2564,
+      "step": 5063
+    },
+    {
+      "epoch": 0.4444651678181853,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0020412754916663924,
+      "loss": 1.274,
+      "step": 5064
+    },
+    {
+      "epoch": 0.4445529374010878,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002040869561634947,
+      "loss": 1.2053,
+      "step": 5065
+    },
+    {
+      "epoch": 0.4446407069839903,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0020404635930262216,
+      "loss": 1.1869,
+      "step": 5066
+    },
+    {
+      "epoch": 0.44472847656689274,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0020400575858802853,
+      "loss": 1.2691,
+      "step": 5067
+    },
+    {
+      "epoch": 0.44481624614979526,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.002039651540237209,
+      "loss": 1.2495,
+      "step": 5068
+    },
+    {
+      "epoch": 0.4449040157326977,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.002039245456137067,
+      "loss": 1.2146,
+      "step": 5069
+    },
+    {
+      "epoch": 0.44499178531560024,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0020388393336199383,
+      "loss": 1.2111,
+      "step": 5070
+    },
+    {
+      "epoch": 0.4450795548985027,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0020384331727259066,
+      "loss": 1.2555,
+      "step": 5071
+    },
+    {
+      "epoch": 0.44516732448140517,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.002038026973495057,
+      "loss": 1.1797,
+      "step": 5072
+    },
+    {
+      "epoch": 0.4452550940643077,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.002037620735967481,
+      "loss": 1.2466,
+      "step": 5073
+    },
+    {
+      "epoch": 0.44534286364721015,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.002037214460183272,
+      "loss": 1.1843,
+      "step": 5074
+    },
+    {
+      "epoch": 0.44543063323011267,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0020368081461825276,
+      "loss": 1.1285,
+      "step": 5075
+    },
+    {
+      "epoch": 0.44551840281301514,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0020364017940053504,
+      "loss": 1.205,
+      "step": 5076
+    },
+    {
+      "epoch": 0.4456061723959176,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0020359954036918444,
+      "loss": 1.1634,
+      "step": 5077
+    },
+    {
+      "epoch": 0.4456939419788201,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0020355889752821195,
+      "loss": 1.2351,
+      "step": 5078
+    },
+    {
+      "epoch": 0.4457817115617226,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0020351825088162885,
+      "loss": 1.24,
+      "step": 5079
+    },
+    {
+      "epoch": 0.44586948114462505,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.002034776004334468,
+      "loss": 1.1741,
+      "step": 5080
+    },
+    {
+      "epoch": 0.44595725072752757,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0020343694618767784,
+      "loss": 1.2357,
+      "step": 5081
+    },
+    {
+      "epoch": 0.44604502031043003,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.002033962881483343,
+      "loss": 1.2178,
+      "step": 5082
+    },
+    {
+      "epoch": 0.44613278989333255,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0020335562631942916,
+      "loss": 1.1827,
+      "step": 5083
+    },
+    {
+      "epoch": 0.446220559476235,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0020331496070497544,
+      "loss": 1.185,
+      "step": 5084
+    },
+    {
+      "epoch": 0.4463083290591375,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.002032742913089867,
+      "loss": 1.2288,
+      "step": 5085
+    },
+    {
+      "epoch": 0.44639609864204,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.002032336181354768,
+      "loss": 1.258,
+      "step": 5086
+    },
+    {
+      "epoch": 0.44648386822494246,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.002031929411884601,
+      "loss": 1.2355,
+      "step": 5087
+    },
+    {
+      "epoch": 0.446571637807845,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0020315226047195128,
+      "loss": 1.2364,
+      "step": 5088
+    },
+    {
+      "epoch": 0.44665940739074744,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0020311157598996523,
+      "loss": 1.2113,
+      "step": 5089
+    },
+    {
+      "epoch": 0.4467471769736499,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0020307088774651747,
+      "loss": 1.234,
+      "step": 5090
+    },
+    {
+      "epoch": 0.4468349465565524,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0020303019574562375,
+      "loss": 1.1739,
+      "step": 5091
+    },
+    {
+      "epoch": 0.4469227161394549,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0020298949999130013,
+      "loss": 1.1588,
+      "step": 5092
+    },
+    {
+      "epoch": 0.44701048572235735,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0020294880048756323,
+      "loss": 1.2142,
+      "step": 5093
+    },
+    {
+      "epoch": 0.4470982553052599,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.002029080972384299,
+      "loss": 1.2028,
+      "step": 5094
+    },
+    {
+      "epoch": 0.44718602488816234,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.002028673902479173,
+      "loss": 1.266,
+      "step": 5095
+    },
+    {
+      "epoch": 0.44727379447106486,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0020282667952004315,
+      "loss": 1.2095,
+      "step": 5096
+    },
+    {
+      "epoch": 0.4473615640539673,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0020278596505882538,
+      "loss": 1.1642,
+      "step": 5097
+    },
+    {
+      "epoch": 0.4474493336368698,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0020274524686828242,
+      "loss": 1.2864,
+      "step": 5098
+    },
+    {
+      "epoch": 0.4475371032197723,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0020270452495243286,
+      "loss": 1.1768,
+      "step": 5099
+    },
+    {
+      "epoch": 0.44762487280267477,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0020266379931529596,
+      "loss": 1.2167,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4477126423855773,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0020262306996089107,
+      "loss": 1.2445,
+      "step": 5101
+    },
+    {
+      "epoch": 0.44780041196847975,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0020258233689323807,
+      "loss": 1.2111,
+      "step": 5102
+    },
+    {
+      "epoch": 0.4478881815513822,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0020254160011635705,
+      "loss": 1.2168,
+      "step": 5103
+    },
+    {
+      "epoch": 0.44797595113428473,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.002025008596342687,
+      "loss": 1.1768,
+      "step": 5104
+    },
+    {
+      "epoch": 0.4480637207171872,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0020246011545099387,
+      "loss": 1.1868,
+      "step": 5105
+    },
+    {
+      "epoch": 0.4481514903000897,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0020241936757055385,
+      "loss": 1.1819,
+      "step": 5106
+    },
+    {
+      "epoch": 0.4482392598829922,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0020237861599697034,
+      "loss": 1.2116,
+      "step": 5107
+    },
+    {
+      "epoch": 0.44832702946589464,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0020233786073426533,
+      "loss": 1.1828,
+      "step": 5108
+    },
+    {
+      "epoch": 0.44841479904879716,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.002022971017864612,
+      "loss": 1.2398,
+      "step": 5109
+    },
+    {
+      "epoch": 0.4485025686316996,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.002022563391575807,
+      "loss": 1.1566,
+      "step": 5110
+    },
+    {
+      "epoch": 0.4485903382146021,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.002022155728516469,
+      "loss": 1.2004,
+      "step": 5111
+    },
+    {
+      "epoch": 0.4486781077975046,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0020217480287268337,
+      "loss": 1.2625,
+      "step": 5112
+    },
+    {
+      "epoch": 0.4487658773804071,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0020213402922471388,
+      "loss": 1.2128,
+      "step": 5113
+    },
+    {
+      "epoch": 0.4488536469633096,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0020209325191176265,
+      "loss": 1.209,
+      "step": 5114
+    },
+    {
+      "epoch": 0.44894141654621206,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.002020524709378542,
+      "loss": 1.217,
+      "step": 5115
+    },
+    {
+      "epoch": 0.4490291861291145,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0020201168630701346,
+      "loss": 1.205,
+      "step": 5116
+    },
+    {
+      "epoch": 0.44911695571201704,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0020197089802326575,
+      "loss": 1.2094,
+      "step": 5117
+    },
+    {
+      "epoch": 0.4492047252949195,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.002019301060906367,
+      "loss": 1.2495,
+      "step": 5118
+    },
+    {
+      "epoch": 0.449292494877822,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.002018893105131523,
+      "loss": 1.1558,
+      "step": 5119
+    },
+    {
+      "epoch": 0.4493802644607245,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.002018485112948389,
+      "loss": 1.2127,
+      "step": 5120
+    },
+    {
+      "epoch": 0.44946803404362695,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0020180770843972328,
+      "loss": 1.2587,
+      "step": 5121
+    },
+    {
+      "epoch": 0.44955580362652947,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.002017669019518324,
+      "loss": 1.2013,
+      "step": 5122
+    },
+    {
+      "epoch": 0.44964357320943193,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0020172609183519387,
+      "loss": 1.1907,
+      "step": 5123
+    },
+    {
+      "epoch": 0.4497313427923344,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0020168527809383534,
+      "loss": 1.2245,
+      "step": 5124
+    },
+    {
+      "epoch": 0.4498191123752369,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00201644460731785,
+      "loss": 1.1919,
+      "step": 5125
+    },
+    {
+      "epoch": 0.4499068819581394,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0020160363975307146,
+      "loss": 1.1589,
+      "step": 5126
+    },
+    {
+      "epoch": 0.4499946515410419,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.002015628151617234,
+      "loss": 1.1259,
+      "step": 5127
+    },
+    {
+      "epoch": 0.45008242112394437,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.002015219869617703,
+      "loss": 1.1923,
+      "step": 5128
+    },
+    {
+      "epoch": 0.45017019070684683,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0020148115515724145,
+      "loss": 1.1882,
+      "step": 5129
+    },
+    {
+      "epoch": 0.45025796028974935,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0020144031975216706,
+      "loss": 1.2328,
+      "step": 5130
+    },
+    {
+      "epoch": 0.4503457298726518,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0020139948075057717,
+      "loss": 1.1982,
+      "step": 5131
+    },
+    {
+      "epoch": 0.45043349945555433,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0020135863815650265,
+      "loss": 1.252,
+      "step": 5132
+    },
+    {
+      "epoch": 0.4505212690384568,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0020131779197397432,
+      "loss": 1.241,
+      "step": 5133
+    },
+    {
+      "epoch": 0.45060903862135926,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0020127694220702375,
+      "loss": 1.2009,
+      "step": 5134
+    },
+    {
+      "epoch": 0.4506968082042618,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.002012360888596824,
+      "loss": 1.2057,
+      "step": 5135
+    },
+    {
+      "epoch": 0.45078457778716424,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0020119523193598255,
+      "loss": 1.1685,
+      "step": 5136
+    },
+    {
+      "epoch": 0.4508723473700667,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.002011543714399565,
+      "loss": 1.2131,
+      "step": 5137
+    },
+    {
+      "epoch": 0.4509601169529692,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00201113507375637,
+      "loss": 1.2451,
+      "step": 5138
+    },
+    {
+      "epoch": 0.4510478865358717,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.002010726397470573,
+      "loss": 1.2106,
+      "step": 5139
+    },
+    {
+      "epoch": 0.4511356561187742,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.002010317685582507,
+      "loss": 1.1863,
+      "step": 5140
+    },
+    {
+      "epoch": 0.4512234257016767,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0020099089381325117,
+      "loss": 1.2255,
+      "step": 5141
+    },
+    {
+      "epoch": 0.45131119528457914,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.002009500155160928,
+      "loss": 1.2686,
+      "step": 5142
+    },
+    {
+      "epoch": 0.45139896486748166,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0020090913367081017,
+      "loss": 1.2641,
+      "step": 5143
+    },
+    {
+      "epoch": 0.4514867344503841,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002008682482814381,
+      "loss": 1.2082,
+      "step": 5144
+    },
+    {
+      "epoch": 0.45157450403328664,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002008273593520119,
+      "loss": 1.2125,
+      "step": 5145
+    },
+    {
+      "epoch": 0.4516622736161891,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00200786466886567,
+      "loss": 1.2099,
+      "step": 5146
+    },
+    {
+      "epoch": 0.45175004319909157,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0020074557088913952,
+      "loss": 1.1761,
+      "step": 5147
+    },
+    {
+      "epoch": 0.4518378127819941,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.002007046713637656,
+      "loss": 1.1974,
+      "step": 5148
+    },
+    {
+      "epoch": 0.45192558236489655,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.002006637683144819,
+      "loss": 1.1951,
+      "step": 5149
+    },
+    {
+      "epoch": 0.452013351947799,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0020062286174532535,
+      "loss": 1.1793,
+      "step": 5150
+    },
+    {
+      "epoch": 0.45210112153070153,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.002005819516603333,
+      "loss": 1.2228,
+      "step": 5151
+    },
+    {
+      "epoch": 0.452188891113604,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.002005410380635434,
+      "loss": 1.1206,
+      "step": 5152
+    },
+    {
+      "epoch": 0.4522766606965065,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0020050012095899376,
+      "loss": 1.144,
+      "step": 5153
+    },
+    {
+      "epoch": 0.452364430279409,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0020045920035072257,
+      "loss": 1.2346,
+      "step": 5154
+    },
+    {
+      "epoch": 0.45245219986231144,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0020041827624276867,
+      "loss": 1.2127,
+      "step": 5155
+    },
+    {
+      "epoch": 0.45253996944521396,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0020037734863917094,
+      "loss": 1.1855,
+      "step": 5156
+    },
+    {
+      "epoch": 0.4526277390281164,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00200336417543969,
+      "loss": 1.2161,
+      "step": 5157
+    },
+    {
+      "epoch": 0.45271550861101895,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0020029548296120235,
+      "loss": 1.1987,
+      "step": 5158
+    },
+    {
+      "epoch": 0.4528032781939214,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.002002545448949113,
+      "loss": 1.1921,
+      "step": 5159
+    },
+    {
+      "epoch": 0.4528910477768239,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.002002136033491361,
+      "loss": 1.2121,
+      "step": 5160
+    },
+    {
+      "epoch": 0.4529788173597264,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.002001726583279176,
+      "loss": 1.1847,
+      "step": 5161
+    },
+    {
+      "epoch": 0.45306658694262886,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0020013170983529684,
+      "loss": 1.1796,
+      "step": 5162
+    },
+    {
+      "epoch": 0.4531543565255314,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.002000907578753154,
+      "loss": 1.2229,
+      "step": 5163
+    },
+    {
+      "epoch": 0.45324212610843384,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00200049802452015,
+      "loss": 1.2243,
+      "step": 5164
+    },
+    {
+      "epoch": 0.4533298956913363,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0020000884356943783,
+      "loss": 1.1839,
+      "step": 5165
+    },
+    {
+      "epoch": 0.4534176652742388,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.001999678812316262,
+      "loss": 1.2343,
+      "step": 5166
+    },
+    {
+      "epoch": 0.4535054348571413,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0019992691544262315,
+      "loss": 1.2151,
+      "step": 5167
+    },
+    {
+      "epoch": 0.45359320444004375,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0019988594620647167,
+      "loss": 1.1935,
+      "step": 5168
+    },
+    {
+      "epoch": 0.45368097402294627,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001998449735272154,
+      "loss": 1.2349,
+      "step": 5169
+    },
+    {
+      "epoch": 0.45376874360584873,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0019980399740889805,
+      "loss": 1.2087,
+      "step": 5170
+    },
+    {
+      "epoch": 0.45385651318875125,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0019976301785556393,
+      "loss": 1.1648,
+      "step": 5171
+    },
+    {
+      "epoch": 0.4539442827716537,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.001997220348712575,
+      "loss": 1.2344,
+      "step": 5172
+    },
+    {
+      "epoch": 0.4540320523545562,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001996810484600236,
+      "loss": 1.154,
+      "step": 5173
+    },
+    {
+      "epoch": 0.4541198219374587,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0019964005862590744,
+      "loss": 1.2264,
+      "step": 5174
+    },
+    {
+      "epoch": 0.45420759152036116,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0019959906537295453,
+      "loss": 1.2002,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4542953611032637,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.001995580687052108,
+      "loss": 1.1964,
+      "step": 5176
+    },
+    {
+      "epoch": 0.45438313068616615,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0019951706862672246,
+      "loss": 1.262,
+      "step": 5177
+    },
+    {
+      "epoch": 0.4544709002690686,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0019947606514153595,
+      "loss": 1.2685,
+      "step": 5178
+    },
+    {
+      "epoch": 0.45455866985197113,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.001994350582536983,
+      "loss": 1.2152,
+      "step": 5179
+    },
+    {
+      "epoch": 0.4546464394348736,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.001993940479672566,
+      "loss": 1.195,
+      "step": 5180
+    },
+    {
+      "epoch": 0.45473420901777606,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.001993530342862585,
+      "loss": 1.2073,
+      "step": 5181
+    },
+    {
+      "epoch": 0.4548219786006786,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.001993120172147518,
+      "loss": 1.2134,
+      "step": 5182
+    },
+    {
+      "epoch": 0.45490974818358104,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.001992709967567848,
+      "loss": 1.1748,
+      "step": 5183
+    },
+    {
+      "epoch": 0.45499751776648356,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0019922997291640598,
+      "loss": 1.168,
+      "step": 5184
+    },
+    {
+      "epoch": 0.455085287349386,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0019918894569766433,
+      "loss": 1.2917,
+      "step": 5185
+    },
+    {
+      "epoch": 0.4551730569322885,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0019914791510460897,
+      "loss": 1.2036,
+      "step": 5186
+    },
+    {
+      "epoch": 0.455260826515191,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.001991068811412895,
+      "loss": 1.236,
+      "step": 5187
+    },
+    {
+      "epoch": 0.45534859609809347,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0019906584381175585,
+      "loss": 1.1489,
+      "step": 5188
+    },
+    {
+      "epoch": 0.455436365680996,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.001990248031200582,
+      "loss": 1.2006,
+      "step": 5189
+    },
+    {
+      "epoch": 0.45552413526389846,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0019898375907024713,
+      "loss": 1.2246,
+      "step": 5190
+    },
+    {
+      "epoch": 0.4556119048468009,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001989427116663735,
+      "loss": 1.2291,
+      "step": 5191
+    },
+    {
+      "epoch": 0.45569967442970344,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0019890166091248848,
+      "loss": 1.1688,
+      "step": 5192
+    },
+    {
+      "epoch": 0.4557874440126059,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.001988606068126437,
+      "loss": 1.2742,
+      "step": 5193
+    },
+    {
+      "epoch": 0.45587521359550837,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0019881954937089105,
+      "loss": 1.2066,
+      "step": 5194
+    },
+    {
+      "epoch": 0.4559629831784109,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0019877848859128264,
+      "loss": 1.2447,
+      "step": 5195
+    },
+    {
+      "epoch": 0.45605075276131335,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0019873742447787113,
+      "loss": 1.2446,
+      "step": 5196
+    },
+    {
+      "epoch": 0.45613852234421587,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0019869635703470925,
+      "loss": 1.241,
+      "step": 5197
+    },
+    {
+      "epoch": 0.45622629192711833,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0019865528626585027,
+      "loss": 1.2343,
+      "step": 5198
+    },
+    {
+      "epoch": 0.4563140615100208,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.001986142121753477,
+      "loss": 1.1629,
+      "step": 5199
+    },
+    {
+      "epoch": 0.4564018310929233,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.001985731347672554,
+      "loss": 1.2155,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4564896006758258,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0019853205404562757,
+      "loss": 1.2564,
+      "step": 5201
+    },
+    {
+      "epoch": 0.4565773702587283,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.001984909700145186,
+      "loss": 1.241,
+      "step": 5202
+    },
+    {
+      "epoch": 0.45666513984163076,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001984498826779835,
+      "loss": 1.2682,
+      "step": 5203
+    },
+    {
+      "epoch": 0.4567529094245332,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0019840879204007732,
+      "loss": 1.1949,
+      "step": 5204
+    },
+    {
+      "epoch": 0.45684067900743575,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.001983676981048555,
+      "loss": 1.2734,
+      "step": 5205
+    },
+    {
+      "epoch": 0.4569284485903382,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0019832660087637398,
+      "loss": 1.2198,
+      "step": 5206
+    },
+    {
+      "epoch": 0.4570162181732407,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0019828550035868877,
+      "loss": 1.1491,
+      "step": 5207
+    },
+    {
+      "epoch": 0.4571039877561432,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0019824439655585645,
+      "loss": 1.2804,
+      "step": 5208
+    },
+    {
+      "epoch": 0.45719175733904566,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0019820328947193363,
+      "loss": 1.2591,
+      "step": 5209
+    },
+    {
+      "epoch": 0.4572795269219482,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.001981621791109776,
+      "loss": 1.2192,
+      "step": 5210
+    },
+    {
+      "epoch": 0.45736729650485064,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0019812106547704563,
+      "loss": 1.2008,
+      "step": 5211
+    },
+    {
+      "epoch": 0.4574550660877531,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.001980799485741956,
+      "loss": 1.2236,
+      "step": 5212
+    },
+    {
+      "epoch": 0.4575428356706556,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0019803882840648554,
+      "loss": 1.2817,
+      "step": 5213
+    },
+    {
+      "epoch": 0.4576306052535581,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0019799770497797384,
+      "loss": 1.1848,
+      "step": 5214
+    },
+    {
+      "epoch": 0.4577183748364606,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0019795657829271922,
+      "loss": 1.2408,
+      "step": 5215
+    },
+    {
+      "epoch": 0.45780614441936307,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001979154483547807,
+      "loss": 1.2437,
+      "step": 5216
+    },
+    {
+      "epoch": 0.45789391400226553,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0019787431516821766,
+      "loss": 1.1442,
+      "step": 5217
+    },
+    {
+      "epoch": 0.45798168358516805,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0019783317873708984,
+      "loss": 1.2639,
+      "step": 5218
+    },
+    {
+      "epoch": 0.4580694531680705,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.001977920390654572,
+      "loss": 1.2405,
+      "step": 5219
+    },
+    {
+      "epoch": 0.458157222750973,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0019775089615738,
+      "loss": 1.2391,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4582449923338755,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0019770975001691896,
+      "loss": 1.2347,
+      "step": 5221
+    },
+    {
+      "epoch": 0.45833276191677796,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0019766860064813502,
+      "loss": 1.1919,
+      "step": 5222
+    },
+    {
+      "epoch": 0.4584205314996805,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0019762744805508947,
+      "loss": 1.1622,
+      "step": 5223
+    },
+    {
+      "epoch": 0.45850830108258295,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.001975862922418439,
+      "loss": 1.1976,
+      "step": 5224
+    },
+    {
+      "epoch": 0.4585960706654854,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.001975451332124602,
+      "loss": 1.2503,
+      "step": 5225
+    },
+    {
+      "epoch": 0.45868384024838793,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.001975039709710007,
+      "loss": 1.1962,
+      "step": 5226
+    },
+    {
+      "epoch": 0.4587716098312904,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.001974628055215278,
+      "loss": 1.1837,
+      "step": 5227
+    },
+    {
+      "epoch": 0.4588593794141929,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0019742163686810445,
+      "loss": 1.1658,
+      "step": 5228
+    },
+    {
+      "epoch": 0.4589471489970954,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.001973804650147938,
+      "loss": 1.1734,
+      "step": 5229
+    },
+    {
+      "epoch": 0.45903491857999784,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0019733928996565945,
+      "loss": 1.208,
+      "step": 5230
+    },
+    {
+      "epoch": 0.45912268816290036,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001972981117247651,
+      "loss": 1.2184,
+      "step": 5231
+    },
+    {
+      "epoch": 0.4592104577458028,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0019725693029617494,
+      "loss": 1.2016,
+      "step": 5232
+    },
+    {
+      "epoch": 0.45929822732870534,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0019721574568395336,
+      "loss": 1.2306,
+      "step": 5233
+    },
+    {
+      "epoch": 0.4593859969116078,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001971745578921652,
+      "loss": 1.1815,
+      "step": 5234
+    },
+    {
+      "epoch": 0.45947376649451027,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0019713336692487545,
+      "loss": 1.2226,
+      "step": 5235
+    },
+    {
+      "epoch": 0.4595615360774128,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0019709217278614955,
+      "loss": 1.1379,
+      "step": 5236
+    },
+    {
+      "epoch": 0.45964930566031525,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0019705097548005316,
+      "loss": 1.233,
+      "step": 5237
+    },
+    {
+      "epoch": 0.4597370752432177,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001970097750106523,
+      "loss": 1.1768,
+      "step": 5238
+    },
+    {
+      "epoch": 0.45982484482612024,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0019696857138201333,
+      "loss": 1.2292,
+      "step": 5239
+    },
+    {
+      "epoch": 0.4599126144090227,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0019692736459820284,
+      "loss": 1.2226,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4600003839919252,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.001968861546632878,
+      "loss": 1.2158,
+      "step": 5241
+    },
+    {
+      "epoch": 0.4600881535748277,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0019684494158133544,
+      "loss": 1.2313,
+      "step": 5242
+    },
+    {
+      "epoch": 0.46017592315773015,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.001968037253564134,
+      "loss": 1.2212,
+      "step": 5243
+    },
+    {
+      "epoch": 0.46026369274063267,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0019676250599258947,
+      "loss": 1.1952,
+      "step": 5244
+    },
+    {
+      "epoch": 0.46035146232353513,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.001967212834939319,
+      "loss": 1.2992,
+      "step": 5245
+    },
+    {
+      "epoch": 0.46043923190643765,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0019668005786450914,
+      "loss": 1.2448,
+      "step": 5246
+    },
+    {
+      "epoch": 0.4605270014893401,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0019663882910839004,
+      "loss": 1.182,
+      "step": 5247
+    },
+    {
+      "epoch": 0.4606147710722426,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001965975972296437,
+      "loss": 1.1869,
+      "step": 5248
+    },
+    {
+      "epoch": 0.4607025406551451,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0019655636223233955,
+      "loss": 1.159,
+      "step": 5249
+    },
+    {
+      "epoch": 0.46079031023804756,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0019651512412054723,
+      "loss": 1.2175,
+      "step": 5250
+    },
+    {
+      "epoch": 0.46087807982095,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0019647388289833695,
+      "loss": 1.194,
+      "step": 5251
+    },
+    {
+      "epoch": 0.46096584940385255,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.001964326385697789,
+      "loss": 1.2138,
+      "step": 5252
+    },
+    {
+      "epoch": 0.461053618986755,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.001963913911389438,
+      "loss": 1.2219,
+      "step": 5253
+    },
+    {
+      "epoch": 0.46114138856965753,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0019635014060990262,
+      "loss": 1.1674,
+      "step": 5254
+    },
+    {
+      "epoch": 0.46122915815256,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0019630888698672667,
+      "loss": 1.2108,
+      "step": 5255
+    },
+    {
+      "epoch": 0.46131692773546246,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001962676302734874,
+      "loss": 1.2211,
+      "step": 5256
+    },
+    {
+      "epoch": 0.461404697318365,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0019622637047425674,
+      "loss": 1.228,
+      "step": 5257
+    },
+    {
+      "epoch": 0.46149246690126744,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0019618510759310684,
+      "loss": 1.2751,
+      "step": 5258
+    },
+    {
+      "epoch": 0.46158023648416996,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.001961438416341103,
+      "loss": 1.2846,
+      "step": 5259
+    },
+    {
+      "epoch": 0.4616680060670724,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0019610257260133973,
+      "loss": 1.2056,
+      "step": 5260
+    },
+    {
+      "epoch": 0.4617557756499749,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0019606130049886835,
+      "loss": 1.2114,
+      "step": 5261
+    },
+    {
+      "epoch": 0.4618435452328774,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0019602002533076947,
+      "loss": 1.1723,
+      "step": 5262
+    },
+    {
+      "epoch": 0.46193131481577987,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0019597874710111695,
+      "loss": 1.1502,
+      "step": 5263
+    },
+    {
+      "epoch": 0.46201908439868233,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.001959374658139846,
+      "loss": 1.1504,
+      "step": 5264
+    },
+    {
+      "epoch": 0.46210685398158485,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0019589618147344673,
+      "loss": 1.1947,
+      "step": 5265
+    },
+    {
+      "epoch": 0.4621946235644873,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0019585489408357804,
+      "loss": 1.1834,
+      "step": 5266
+    },
+    {
+      "epoch": 0.46228239314738984,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0019581360364845342,
+      "loss": 1.2166,
+      "step": 5267
+    },
+    {
+      "epoch": 0.4623701627302923,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00195772310172148,
+      "loss": 1.2282,
+      "step": 5268
+    },
+    {
+      "epoch": 0.46245793231319476,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0019573101365873737,
+      "loss": 1.2422,
+      "step": 5269
+    },
+    {
+      "epoch": 0.4625457018960973,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001956897141122973,
+      "loss": 1.2101,
+      "step": 5270
+    },
+    {
+      "epoch": 0.46263347147899975,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0019564841153690386,
+      "loss": 1.2027,
+      "step": 5271
+    },
+    {
+      "epoch": 0.46272124106190227,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0019560710593663345,
+      "loss": 1.1852,
+      "step": 5272
+    },
+    {
+      "epoch": 0.46280901064480473,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0019556579731556277,
+      "loss": 1.1748,
+      "step": 5273
+    },
+    {
+      "epoch": 0.4628967802277072,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0019552448567776884,
+      "loss": 1.1634,
+      "step": 5274
+    },
+    {
+      "epoch": 0.4629845498106097,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00195483171027329,
+      "loss": 1.1983,
+      "step": 5275
+    },
+    {
+      "epoch": 0.4630723193935122,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0019544185336832077,
+      "loss": 1.1762,
+      "step": 5276
+    },
+    {
+      "epoch": 0.46316008897641464,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.001954005327048221,
+      "loss": 1.2566,
+      "step": 5277
+    },
+    {
+      "epoch": 0.46324785855931716,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.001953592090409111,
+      "loss": 1.2028,
+      "step": 5278
+    },
+    {
+      "epoch": 0.4633356281422196,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001953178823806663,
+      "loss": 1.1755,
+      "step": 5279
+    },
+    {
+      "epoch": 0.46342339772512214,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0019527655272816645,
+      "loss": 1.2327,
+      "step": 5280
+    },
+    {
+      "epoch": 0.4635111673080246,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0019523522008749068,
+      "loss": 1.1879,
+      "step": 5281
+    },
+    {
+      "epoch": 0.46359893689092707,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0019519388446271831,
+      "loss": 1.2911,
+      "step": 5282
+    },
+    {
+      "epoch": 0.4636867064738296,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0019515254585792904,
+      "loss": 1.1939,
+      "step": 5283
+    },
+    {
+      "epoch": 0.46377447605673205,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001951112042772028,
+      "loss": 1.2266,
+      "step": 5284
+    },
+    {
+      "epoch": 0.4638622456396346,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0019506985972461977,
+      "loss": 1.1927,
+      "step": 5285
+    },
+    {
+      "epoch": 0.46395001522253704,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0019502851220426067,
+      "loss": 1.2419,
+      "step": 5286
+    },
+    {
+      "epoch": 0.4640377848054395,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.001949871617202061,
+      "loss": 1.1644,
+      "step": 5287
+    },
+    {
+      "epoch": 0.464125554388342,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0019494580827653742,
+      "loss": 1.1231,
+      "step": 5288
+    },
+    {
+      "epoch": 0.4642133239712445,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0019490445187733594,
+      "loss": 1.191,
+      "step": 5289
+    },
+    {
+      "epoch": 0.46430109355414695,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0019486309252668339,
+      "loss": 1.1483,
+      "step": 5290
+    },
+    {
+      "epoch": 0.46438886313704947,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0019482173022866173,
+      "loss": 1.219,
+      "step": 5291
+    },
+    {
+      "epoch": 0.46447663271995193,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0019478036498735331,
+      "loss": 1.242,
+      "step": 5292
+    },
+    {
+      "epoch": 0.46456440230285445,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0019473899680684072,
+      "loss": 1.2145,
+      "step": 5293
+    },
+    {
+      "epoch": 0.4646521718857569,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001946976256912068,
+      "loss": 1.1927,
+      "step": 5294
+    },
+    {
+      "epoch": 0.4647399414686594,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0019465625164453476,
+      "loss": 1.2113,
+      "step": 5295
+    },
+    {
+      "epoch": 0.4648277110515619,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0019461487467090802,
+      "loss": 1.2007,
+      "step": 5296
+    },
+    {
+      "epoch": 0.46491548063446436,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0019457349477441028,
+      "loss": 1.1845,
+      "step": 5297
+    },
+    {
+      "epoch": 0.4650032502173669,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0019453211195912571,
+      "loss": 1.2784,
+      "step": 5298
+    },
+    {
+      "epoch": 0.46509101980026935,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0019449072622913846,
+      "loss": 1.1871,
+      "step": 5299
+    },
+    {
+      "epoch": 0.4651787893831718,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0019444933758853327,
+      "loss": 1.1931,
+      "step": 5300
+    },
+    {
+      "epoch": 0.46526655896607433,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0019440794604139494,
+      "loss": 1.2036,
+      "step": 5301
+    },
+    {
+      "epoch": 0.4653543285489768,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0019436655159180876,
+      "loss": 1.1831,
+      "step": 5302
+    },
+    {
+      "epoch": 0.4654420981318793,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0019432515424386008,
+      "loss": 1.1998,
+      "step": 5303
+    },
+    {
+      "epoch": 0.4655298677147818,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0019428375400163474,
+      "loss": 1.1823,
+      "step": 5304
+    },
+    {
+      "epoch": 0.46561763729768424,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001942423508692187,
+      "loss": 1.1731,
+      "step": 5305
+    },
+    {
+      "epoch": 0.46570540688058676,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0019420094485069841,
+      "loss": 1.2139,
+      "step": 5306
+    },
+    {
+      "epoch": 0.4657931764634892,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0019415953595016036,
+      "loss": 1.1909,
+      "step": 5307
+    },
+    {
+      "epoch": 0.4658809460463917,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0019411812417169149,
+      "loss": 1.2421,
+      "step": 5308
+    },
+    {
+      "epoch": 0.4659687156292942,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00194076709519379,
+      "loss": 1.1763,
+      "step": 5309
+    },
+    {
+      "epoch": 0.46605648521219667,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0019403529199731031,
+      "loss": 1.195,
+      "step": 5310
+    },
+    {
+      "epoch": 0.4661442547950992,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0019399387160957322,
+      "loss": 1.1828,
+      "step": 5311
+    },
+    {
+      "epoch": 0.46623202437800165,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0019395244836025566,
+      "loss": 1.2152,
+      "step": 5312
+    },
+    {
+      "epoch": 0.4663197939609041,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0019391102225344607,
+      "loss": 1.2397,
+      "step": 5313
+    },
+    {
+      "epoch": 0.46640756354380664,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0019386959329323297,
+      "loss": 1.1872,
+      "step": 5314
+    },
+    {
+      "epoch": 0.4664953331267091,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0019382816148370524,
+      "loss": 1.2148,
+      "step": 5315
+    },
+    {
+      "epoch": 0.4665831027096116,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0019378672682895202,
+      "loss": 1.189,
+      "step": 5316
+    },
+    {
+      "epoch": 0.4666708722925141,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0019374528933306277,
+      "loss": 1.2509,
+      "step": 5317
+    },
+    {
+      "epoch": 0.46675864187541655,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001937038490001272,
+      "loss": 1.1565,
+      "step": 5318
+    },
+    {
+      "epoch": 0.46684641145831907,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001936624058342353,
+      "loss": 1.2591,
+      "step": 5319
+    },
+    {
+      "epoch": 0.46693418104122153,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.001936209598394774,
+      "loss": 1.1724,
+      "step": 5320
+    },
+    {
+      "epoch": 0.467021950624124,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0019357951101994396,
+      "loss": 1.2657,
+      "step": 5321
+    },
+    {
+      "epoch": 0.4671097202070265,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0019353805937972592,
+      "loss": 1.158,
+      "step": 5322
+    },
+    {
+      "epoch": 0.467197489789929,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0019349660492291427,
+      "loss": 1.1995,
+      "step": 5323
+    },
+    {
+      "epoch": 0.4672852593728315,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0019345514765360052,
+      "loss": 1.251,
+      "step": 5324
+    },
+    {
+      "epoch": 0.46737302895573396,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0019341368757587623,
+      "loss": 1.2142,
+      "step": 5325
+    },
+    {
+      "epoch": 0.4674607985386364,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0019337222469383351,
+      "loss": 1.2277,
+      "step": 5326
+    },
+    {
+      "epoch": 0.46754856812153894,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0019333075901156438,
+      "loss": 1.1799,
+      "step": 5327
+    },
+    {
+      "epoch": 0.4676363377044414,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0019328929053316147,
+      "loss": 1.246,
+      "step": 5328
+    },
+    {
+      "epoch": 0.4677241072873439,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0019324781926271751,
+      "loss": 1.2057,
+      "step": 5329
+    },
+    {
+      "epoch": 0.4678118768702464,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0019320634520432558,
+      "loss": 1.1849,
+      "step": 5330
+    },
+    {
+      "epoch": 0.46789964645314885,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0019316486836207895,
+      "loss": 1.1744,
+      "step": 5331
+    },
+    {
+      "epoch": 0.4679874160360514,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0019312338874007133,
+      "loss": 1.2112,
+      "step": 5332
+    },
+    {
+      "epoch": 0.46807518561895384,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0019308190634239647,
+      "loss": 1.189,
+      "step": 5333
+    },
+    {
+      "epoch": 0.4681629552018563,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0019304042117314856,
+      "loss": 1.1498,
+      "step": 5334
+    },
+    {
+      "epoch": 0.4682507247847588,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0019299893323642204,
+      "loss": 1.1582,
+      "step": 5335
+    },
+    {
+      "epoch": 0.4683384943676613,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0019295744253631157,
+      "loss": 1.2033,
+      "step": 5336
+    },
+    {
+      "epoch": 0.4684262639505638,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.001929159490769122,
+      "loss": 1.1697,
+      "step": 5337
+    },
+    {
+      "epoch": 0.46851403353346627,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0019287445286231902,
+      "loss": 1.2606,
+      "step": 5338
+    },
+    {
+      "epoch": 0.46860180311636873,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0019283295389662775,
+      "loss": 1.2327,
+      "step": 5339
+    },
+    {
+      "epoch": 0.46868957269927125,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00192791452183934,
+      "loss": 1.228,
+      "step": 5340
+    },
+    {
+      "epoch": 0.4687773422821737,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.001927499477283339,
+      "loss": 1.1574,
+      "step": 5341
+    },
+    {
+      "epoch": 0.46886511186507623,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0019270844053392371,
+      "loss": 1.2046,
+      "step": 5342
+    },
+    {
+      "epoch": 0.4689528814479787,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0019266693060480016,
+      "loss": 1.1962,
+      "step": 5343
+    },
+    {
+      "epoch": 0.46904065103088116,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0019262541794505996,
+      "loss": 1.1862,
+      "step": 5344
+    },
+    {
+      "epoch": 0.4691284206137837,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0019258390255880037,
+      "loss": 1.2534,
+      "step": 5345
+    },
+    {
+      "epoch": 0.46921619019668614,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0019254238445011874,
+      "loss": 1.1938,
+      "step": 5346
+    },
+    {
+      "epoch": 0.4693039597795886,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.001925008636231127,
+      "loss": 1.3485,
+      "step": 5347
+    },
+    {
+      "epoch": 0.46939172936249113,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.001924593400818802,
+      "loss": 1.2057,
+      "step": 5348
+    },
+    {
+      "epoch": 0.4694794989453936,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0019241781383051958,
+      "loss": 1.1769,
+      "step": 5349
+    },
+    {
+      "epoch": 0.4695672685282961,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.001923762848731292,
+      "loss": 1.2,
+      "step": 5350
+    },
+    {
+      "epoch": 0.4696550381111986,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0019233475321380781,
+      "loss": 1.2359,
+      "step": 5351
+    },
+    {
+      "epoch": 0.46974280769410104,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0019229321885665444,
+      "loss": 1.2209,
+      "step": 5352
+    },
+    {
+      "epoch": 0.46983057727700356,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0019225168180576832,
+      "loss": 1.187,
+      "step": 5353
+    },
+    {
+      "epoch": 0.469918346859906,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0019221014206524908,
+      "loss": 1.1831,
+      "step": 5354
+    },
+    {
+      "epoch": 0.47000611644280854,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.001921685996391965,
+      "loss": 1.1781,
+      "step": 5355
+    },
+    {
+      "epoch": 0.470093886025711,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0019212705453171062,
+      "loss": 1.1746,
+      "step": 5356
+    },
+    {
+      "epoch": 0.47018165560861347,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0019208550674689178,
+      "loss": 1.1832,
+      "step": 5357
+    },
+    {
+      "epoch": 0.470269425191516,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0019204395628884057,
+      "loss": 1.1755,
+      "step": 5358
+    },
+    {
+      "epoch": 0.47035719477441845,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0019200240316165794,
+      "loss": 1.2429,
+      "step": 5359
+    },
+    {
+      "epoch": 0.4704449643573209,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0019196084736944493,
+      "loss": 1.1878,
+      "step": 5360
+    },
+    {
+      "epoch": 0.47053273394022344,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0019191928891630302,
+      "loss": 1.2124,
+      "step": 5361
+    },
+    {
+      "epoch": 0.4706205035231259,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0019187772780633373,
+      "loss": 1.2221,
+      "step": 5362
+    },
+    {
+      "epoch": 0.4707082731060284,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0019183616404363907,
+      "loss": 1.2346,
+      "step": 5363
+    },
+    {
+      "epoch": 0.4707960426889309,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0019179459763232123,
+      "loss": 1.1943,
+      "step": 5364
+    },
+    {
+      "epoch": 0.47088381227183335,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0019175302857648263,
+      "loss": 1.2541,
+      "step": 5365
+    },
+    {
+      "epoch": 0.47097158185473587,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0019171145688022593,
+      "loss": 1.1844,
+      "step": 5366
+    },
+    {
+      "epoch": 0.47105935143763833,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.001916698825476542,
+      "loss": 1.2668,
+      "step": 5367
+    },
+    {
+      "epoch": 0.47114712102054085,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0019162830558287051,
+      "loss": 1.2319,
+      "step": 5368
+    },
+    {
+      "epoch": 0.4712348906034433,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0019158672598997843,
+      "loss": 1.1886,
+      "step": 5369
+    },
+    {
+      "epoch": 0.4713226601863458,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001915451437730817,
+      "loss": 1.219,
+      "step": 5370
+    },
+    {
+      "epoch": 0.4714104297692483,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0019150355893628433,
+      "loss": 1.193,
+      "step": 5371
+    },
+    {
+      "epoch": 0.47149819935215076,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0019146197148369056,
+      "loss": 1.16,
+      "step": 5372
+    },
+    {
+      "epoch": 0.4715859689350533,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.001914203814194049,
+      "loss": 1.2006,
+      "step": 5373
+    },
+    {
+      "epoch": 0.47167373851795574,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0019137878874753214,
+      "loss": 1.2333,
+      "step": 5374
+    },
+    {
+      "epoch": 0.4717615081008582,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0019133719347217734,
+      "loss": 1.2414,
+      "step": 5375
+    },
+    {
+      "epoch": 0.4718492776837607,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0019129559559744572,
+      "loss": 1.1961,
+      "step": 5376
+    },
+    {
+      "epoch": 0.4719370472666632,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.001912539951274429,
+      "loss": 1.1999,
+      "step": 5377
+    },
+    {
+      "epoch": 0.47202481684956565,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0019121239206627456,
+      "loss": 1.2018,
+      "step": 5378
+    },
+    {
+      "epoch": 0.4721125864324682,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0019117078641804688,
+      "loss": 1.1275,
+      "step": 5379
+    },
+    {
+      "epoch": 0.47220035601537064,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0019112917818686617,
+      "loss": 1.2018,
+      "step": 5380
+    },
+    {
+      "epoch": 0.47228812559827316,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0019108756737683894,
+      "loss": 1.1821,
+      "step": 5381
+    },
+    {
+      "epoch": 0.4723758951811756,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0019104595399207199,
+      "loss": 1.1848,
+      "step": 5382
+    },
+    {
+      "epoch": 0.4724636647640781,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0019100433803667246,
+      "loss": 1.274,
+      "step": 5383
+    },
+    {
+      "epoch": 0.4725514343469806,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0019096271951474768,
+      "loss": 1.2939,
+      "step": 5384
+    },
+    {
+      "epoch": 0.47263920392988307,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0019092109843040518,
+      "loss": 1.2287,
+      "step": 5385
+    },
+    {
+      "epoch": 0.4727269735127856,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0019087947478775291,
+      "loss": 1.2127,
+      "step": 5386
+    },
+    {
+      "epoch": 0.47281474309568805,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001908378485908988,
+      "loss": 1.2372,
+      "step": 5387
+    },
+    {
+      "epoch": 0.4729025126785905,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001907962198439513,
+      "loss": 1.2388,
+      "step": 5388
+    },
+    {
+      "epoch": 0.47299028226149303,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0019075458855101896,
+      "loss": 1.2305,
+      "step": 5389
+    },
+    {
+      "epoch": 0.4730780518443955,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0019071295471621065,
+      "loss": 1.2861,
+      "step": 5390
+    },
+    {
+      "epoch": 0.47316582142729796,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0019067131834363544,
+      "loss": 1.1988,
+      "step": 5391
+    },
+    {
+      "epoch": 0.4732535910102005,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0019062967943740273,
+      "loss": 1.2323,
+      "step": 5392
+    },
+    {
+      "epoch": 0.47334136059310294,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0019058803800162197,
+      "loss": 1.1763,
+      "step": 5393
+    },
+    {
+      "epoch": 0.47342913017600546,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0019054639404040318,
+      "loss": 1.24,
+      "step": 5394
+    },
+    {
+      "epoch": 0.4735168997589079,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0019050474755785635,
+      "loss": 1.2037,
+      "step": 5395
+    },
+    {
+      "epoch": 0.4736046693418104,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0019046309855809192,
+      "loss": 1.1655,
+      "step": 5396
+    },
+    {
+      "epoch": 0.4736924389247129,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0019042144704522032,
+      "loss": 1.2159,
+      "step": 5397
+    },
+    {
+      "epoch": 0.4737802085076154,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0019037979302335252,
+      "loss": 1.1991,
+      "step": 5398
+    },
+    {
+      "epoch": 0.4738679780905179,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.001903381364965996,
+      "loss": 1.2288,
+      "step": 5399
+    },
+    {
+      "epoch": 0.47395574767342036,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0019029647746907285,
+      "loss": 1.225,
+      "step": 5400
+    },
+    {
+      "epoch": 0.4740435172563228,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0019025481594488385,
+      "loss": 1.2515,
+      "step": 5401
+    },
+    {
+      "epoch": 0.47413128683922534,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.001902131519281445,
+      "loss": 1.1419,
+      "step": 5402
+    },
+    {
+      "epoch": 0.4742190564221278,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0019017148542296671,
+      "loss": 1.2398,
+      "step": 5403
+    },
+    {
+      "epoch": 0.47430682600503027,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0019012981643346302,
+      "loss": 1.2144,
+      "step": 5404
+    },
+    {
+      "epoch": 0.4743945955879328,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0019008814496374578,
+      "loss": 1.251,
+      "step": 5405
+    },
+    {
+      "epoch": 0.47448236517083525,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00190046471017928,
+      "loss": 1.2039,
+      "step": 5406
+    },
+    {
+      "epoch": 0.47457013475373777,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0019000479460012255,
+      "loss": 1.2528,
+      "step": 5407
+    },
+    {
+      "epoch": 0.47465790433664024,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0018996311571444285,
+      "loss": 1.1995,
+      "step": 5408
+    },
+    {
+      "epoch": 0.4747456739195427,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0018992143436500238,
+      "loss": 1.1862,
+      "step": 5409
+    },
+    {
+      "epoch": 0.4748334435024452,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0018987975055591501,
+      "loss": 1.1682,
+      "step": 5410
+    },
+    {
+      "epoch": 0.4749212130853477,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0018983806429129465,
+      "loss": 1.2197,
+      "step": 5411
+    },
+    {
+      "epoch": 0.4750089826682502,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0018979637557525566,
+      "loss": 1.2137,
+      "step": 5412
+    },
+    {
+      "epoch": 0.47509675225115267,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0018975468441191246,
+      "loss": 1.2378,
+      "step": 5413
+    },
+    {
+      "epoch": 0.47518452183405513,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001897129908053799,
+      "loss": 1.1529,
+      "step": 5414
+    },
+    {
+      "epoch": 0.47527229141695765,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0018967129475977293,
+      "loss": 1.2306,
+      "step": 5415
+    },
+    {
+      "epoch": 0.4753600609998601,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001896295962792068,
+      "loss": 1.2262,
+      "step": 5416
+    },
+    {
+      "epoch": 0.4754478305827626,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0018958789536779692,
+      "loss": 1.1877,
+      "step": 5417
+    },
+    {
+      "epoch": 0.4755356001656651,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0018954619202965909,
+      "loss": 1.256,
+      "step": 5418
+    },
+    {
+      "epoch": 0.47562336974856756,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0018950448626890925,
+      "loss": 1.1814,
+      "step": 5419
+    },
+    {
+      "epoch": 0.4757111393314701,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0018946277808966354,
+      "loss": 1.187,
+      "step": 5420
+    },
+    {
+      "epoch": 0.47579890891437254,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0018942106749603846,
+      "loss": 1.2209,
+      "step": 5421
+    },
+    {
+      "epoch": 0.475886678497275,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0018937935449215063,
+      "loss": 1.2644,
+      "step": 5422
+    },
+    {
+      "epoch": 0.4759744480801775,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0018933763908211697,
+      "loss": 1.1862,
+      "step": 5423
+    },
+    {
+      "epoch": 0.47606221766308,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0018929592127005465,
+      "loss": 1.2334,
+      "step": 5424
+    },
+    {
+      "epoch": 0.4761499872459825,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0018925420106008103,
+      "loss": 1.2364,
+      "step": 5425
+    },
+    {
+      "epoch": 0.476237756828885,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.001892124784563137,
+      "loss": 1.2192,
+      "step": 5426
+    },
+    {
+      "epoch": 0.47632552641178744,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0018917075346287057,
+      "loss": 1.2153,
+      "step": 5427
+    },
+    {
+      "epoch": 0.47641329599468996,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0018912902608386971,
+      "loss": 1.2239,
+      "step": 5428
+    },
+    {
+      "epoch": 0.4765010655775924,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0018908729632342947,
+      "loss": 1.1532,
+      "step": 5429
+    },
+    {
+      "epoch": 0.4765888351604949,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0018904556418566836,
+      "loss": 1.1592,
+      "step": 5430
+    },
+    {
+      "epoch": 0.4766766047433974,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0018900382967470526,
+      "loss": 1.2249,
+      "step": 5431
+    },
+    {
+      "epoch": 0.47676437432629987,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0018896209279465912,
+      "loss": 1.2173,
+      "step": 5432
+    },
+    {
+      "epoch": 0.4768521439092024,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001889203535496492,
+      "loss": 1.1793,
+      "step": 5433
+    },
+    {
+      "epoch": 0.47693991349210485,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0018887861194379508,
+      "loss": 1.2489,
+      "step": 5434
+    },
+    {
+      "epoch": 0.4770276830750073,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0018883686798121643,
+      "loss": 1.1835,
+      "step": 5435
+    },
+    {
+      "epoch": 0.47711545265790983,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0018879512166603322,
+      "loss": 1.2847,
+      "step": 5436
+    },
+    {
+      "epoch": 0.4772032222408123,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0018875337300236568,
+      "loss": 1.1481,
+      "step": 5437
+    },
+    {
+      "epoch": 0.4772909918237148,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0018871162199433418,
+      "loss": 1.2312,
+      "step": 5438
+    },
+    {
+      "epoch": 0.4773787614066173,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0018866986864605941,
+      "loss": 1.1675,
+      "step": 5439
+    },
+    {
+      "epoch": 0.47746653098951974,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0018862811296166228,
+      "loss": 1.1945,
+      "step": 5440
+    },
+    {
+      "epoch": 0.47755430057242226,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001885863549452639,
+      "loss": 1.2894,
+      "step": 5441
+    },
+    {
+      "epoch": 0.4776420701553247,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0018854459460098557,
+      "loss": 1.1475,
+      "step": 5442
+    },
+    {
+      "epoch": 0.47772983973822725,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0018850283193294897,
+      "loss": 1.243,
+      "step": 5443
+    },
+    {
+      "epoch": 0.4778176093211297,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0018846106694527579,
+      "loss": 1.1649,
+      "step": 5444
+    },
+    {
+      "epoch": 0.4779053789040322,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.001884192996420882,
+      "loss": 1.2046,
+      "step": 5445
+    },
+    {
+      "epoch": 0.4779931484869347,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0018837753002750837,
+      "loss": 1.233,
+      "step": 5446
+    },
+    {
+      "epoch": 0.47808091806983716,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0018833575810565887,
+      "loss": 1.1433,
+      "step": 5447
+    },
+    {
+      "epoch": 0.4781686876527396,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.001882939838806623,
+      "loss": 1.2221,
+      "step": 5448
+    },
+    {
+      "epoch": 0.47825645723564214,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0018825220735664173,
+      "loss": 1.2185,
+      "step": 5449
+    },
+    {
+      "epoch": 0.4783442268185446,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0018821042853772027,
+      "loss": 1.2177,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4784319964014471,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001881686474280214,
+      "loss": 1.2403,
+      "step": 5451
+    },
+    {
+      "epoch": 0.4785197659843496,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0018812686403166868,
+      "loss": 1.2125,
+      "step": 5452
+    },
+    {
+      "epoch": 0.47860753556725205,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0018808507835278603,
+      "loss": 1.1971,
+      "step": 5453
+    },
+    {
+      "epoch": 0.47869530515015457,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.001880432903954974,
+      "loss": 1.2094,
+      "step": 5454
+    },
+    {
+      "epoch": 0.47878307473305703,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0018800150016392726,
+      "loss": 1.2247,
+      "step": 5455
+    },
+    {
+      "epoch": 0.47887084431595955,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0018795970766220002,
+      "loss": 1.2075,
+      "step": 5456
+    },
+    {
+      "epoch": 0.478958613898862,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0018791791289444054,
+      "loss": 1.2449,
+      "step": 5457
+    },
+    {
+      "epoch": 0.4790463834817645,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0018787611586477367,
+      "loss": 1.185,
+      "step": 5458
+    },
+    {
+      "epoch": 0.479134153064667,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0018783431657732475,
+      "loss": 1.1308,
+      "step": 5459
+    },
+    {
+      "epoch": 0.47922192264756946,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0018779251503621905,
+      "loss": 1.2188,
+      "step": 5460
+    },
+    {
+      "epoch": 0.47930969223047193,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0018775071124558238,
+      "loss": 1.2285,
+      "step": 5461
+    },
+    {
+      "epoch": 0.47939746181337445,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0018770890520954052,
+      "loss": 1.2429,
+      "step": 5462
+    },
+    {
+      "epoch": 0.4794852313962769,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0018766709693221954,
+      "loss": 1.1764,
+      "step": 5463
+    },
+    {
+      "epoch": 0.47957300097917943,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0018762528641774574,
+      "loss": 1.2362,
+      "step": 5464
+    },
+    {
+      "epoch": 0.4796607705620819,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0018758347367024578,
+      "loss": 1.1588,
+      "step": 5465
+    },
+    {
+      "epoch": 0.47974854014498436,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001875416586938463,
+      "loss": 1.1684,
+      "step": 5466
+    },
+    {
+      "epoch": 0.4798363097278869,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.001874998414926743,
+      "loss": 1.2048,
+      "step": 5467
+    },
+    {
+      "epoch": 0.47992407931078934,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0018745802207085697,
+      "loss": 1.2096,
+      "step": 5468
+    },
+    {
+      "epoch": 0.48001184889369186,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001874162004325217,
+      "loss": 1.1891,
+      "step": 5469
+    },
+    {
+      "epoch": 0.4800996184765943,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0018737437658179615,
+      "loss": 1.2167,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4801873880594968,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0018733255052280817,
+      "loss": 1.2415,
+      "step": 5471
+    },
+    {
+      "epoch": 0.4802751576423993,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.001872907222596858,
+      "loss": 1.2448,
+      "step": 5472
+    },
+    {
+      "epoch": 0.4803629272253018,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0018724889179655743,
+      "loss": 1.2422,
+      "step": 5473
+    },
+    {
+      "epoch": 0.48045069680820424,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.001872070591375514,
+      "loss": 1.2266,
+      "step": 5474
+    },
+    {
+      "epoch": 0.48053846639110676,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001871652242867965,
+      "loss": 1.2296,
+      "step": 5475
+    },
+    {
+      "epoch": 0.4806262359740092,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001871233872484217,
+      "loss": 1.2309,
+      "step": 5476
+    },
+    {
+      "epoch": 0.48071400555691174,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0018708154802655607,
+      "loss": 1.1969,
+      "step": 5477
+    },
+    {
+      "epoch": 0.4808017751398142,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0018703970662532906,
+      "loss": 1.2354,
+      "step": 5478
+    },
+    {
+      "epoch": 0.48088954472271667,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001869978630488702,
+      "loss": 1.2126,
+      "step": 5479
+    },
+    {
+      "epoch": 0.4809773143056192,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001869560173013093,
+      "loss": 1.2426,
+      "step": 5480
+    },
+    {
+      "epoch": 0.48106508388852165,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0018691416938677633,
+      "loss": 1.2718,
+      "step": 5481
+    },
+    {
+      "epoch": 0.48115285347142417,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0018687231930940163,
+      "loss": 1.2448,
+      "step": 5482
+    },
+    {
+      "epoch": 0.48124062305432663,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.001868304670733155,
+      "loss": 1.2268,
+      "step": 5483
+    },
+    {
+      "epoch": 0.4813283926372291,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0018678861268264868,
+      "loss": 1.2446,
+      "step": 5484
+    },
+    {
+      "epoch": 0.4814161622201316,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0018674675614153199,
+      "loss": 1.2098,
+      "step": 5485
+    },
+    {
+      "epoch": 0.4815039318030341,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0018670489745409651,
+      "loss": 1.2223,
+      "step": 5486
+    },
+    {
+      "epoch": 0.48159170138593654,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.001866630366244735,
+      "loss": 1.1832,
+      "step": 5487
+    },
+    {
+      "epoch": 0.48167947096883906,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0018662117365679455,
+      "loss": 1.2633,
+      "step": 5488
+    },
+    {
+      "epoch": 0.4817672405517415,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0018657930855519127,
+      "loss": 1.1947,
+      "step": 5489
+    },
+    {
+      "epoch": 0.48185501013464405,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0018653744132379566,
+      "loss": 1.2511,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4819427797175465,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.001864955719667398,
+      "loss": 1.1978,
+      "step": 5491
+    },
+    {
+      "epoch": 0.482030549300449,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00186453700488156,
+      "loss": 1.2001,
+      "step": 5492
+    },
+    {
+      "epoch": 0.4821183188833515,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0018641182689217693,
+      "loss": 1.1614,
+      "step": 5493
+    },
+    {
+      "epoch": 0.48220608846625396,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0018636995118293518,
+      "loss": 1.2405,
+      "step": 5494
+    },
+    {
+      "epoch": 0.4822938580491565,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0018632807336456385,
+      "loss": 1.1465,
+      "step": 5495
+    },
+    {
+      "epoch": 0.48238162763205894,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0018628619344119607,
+      "loss": 1.2387,
+      "step": 5496
+    },
+    {
+      "epoch": 0.4824693972149614,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0018624431141696521,
+      "loss": 1.1818,
+      "step": 5497
+    },
+    {
+      "epoch": 0.4825571667978639,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.001862024272960049,
+      "loss": 1.1972,
+      "step": 5498
+    },
+    {
+      "epoch": 0.4826449363807664,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0018616054108244883,
+      "loss": 1.2262,
+      "step": 5499
+    },
+    {
+      "epoch": 0.48273270596366885,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.001861186527804312,
+      "loss": 1.222,
+      "step": 5500
+    },
+    {
+      "epoch": 0.48273270596366885,
+      "eval_loss": 1.1942256689071655,
+      "eval_runtime": 437.7379,
+      "eval_samples_per_second": 33.673,
+      "eval_steps_per_second": 8.418,
+      "step": 5500
+    },
+    {
+      "epoch": 0.48282047554657137,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0018607676239408603,
+      "loss": 1.1929,
+      "step": 5501
+    },
+    {
+      "epoch": 0.48290824512947383,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0018603486992754782,
+      "loss": 1.1682,
+      "step": 5502
+    },
+    {
+      "epoch": 0.48299601471237635,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0018599297538495116,
+      "loss": 1.1759,
+      "step": 5503
+    },
+    {
+      "epoch": 0.4830837842952788,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0018595107877043093,
+      "loss": 1.1654,
+      "step": 5504
+    },
+    {
+      "epoch": 0.4831715538781813,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.001859091800881221,
+      "loss": 1.203,
+      "step": 5505
+    },
+    {
+      "epoch": 0.4832593234610838,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0018586727934215993,
+      "loss": 1.2621,
+      "step": 5506
+    },
+    {
+      "epoch": 0.48334709304398626,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0018582537653667988,
+      "loss": 1.1855,
+      "step": 5507
+    },
+    {
+      "epoch": 0.4834348626268888,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0018578347167581759,
+      "loss": 1.2118,
+      "step": 5508
+    },
+    {
+      "epoch": 0.48352263220979125,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0018574156476370877,
+      "loss": 1.2565,
+      "step": 5509
+    },
+    {
+      "epoch": 0.4836104017926937,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0018569965580448968,
+      "loss": 1.1492,
+      "step": 5510
+    },
+    {
+      "epoch": 0.48369817137559623,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0018565774480229645,
+      "loss": 1.2336,
+      "step": 5511
+    },
+    {
+      "epoch": 0.4837859409584987,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0018561583176126555,
+      "loss": 1.1388,
+      "step": 5512
+    },
+    {
+      "epoch": 0.4838737105414012,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0018557391668553359,
+      "loss": 1.2848,
+      "step": 5513
+    },
+    {
+      "epoch": 0.4839614801243037,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0018553199957923748,
+      "loss": 1.2172,
+      "step": 5514
+    },
+    {
+      "epoch": 0.48404924970720614,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0018549008044651428,
+      "loss": 1.1855,
+      "step": 5515
+    },
+    {
+      "epoch": 0.48413701929010866,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0018544815929150116,
+      "loss": 1.2761,
+      "step": 5516
+    },
+    {
+      "epoch": 0.4842247888730111,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001854062361183357,
+      "loss": 1.2249,
+      "step": 5517
+    },
+    {
+      "epoch": 0.4843125584559136,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0018536431093115543,
+      "loss": 1.2337,
+      "step": 5518
+    },
+    {
+      "epoch": 0.4844003280388161,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0018532238373409826,
+      "loss": 1.1951,
+      "step": 5519
+    },
+    {
+      "epoch": 0.48448809762171857,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0018528045453130222,
+      "loss": 1.2532,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4845758672046211,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0018523852332690556,
+      "loss": 1.2508,
+      "step": 5521
+    },
+    {
+      "epoch": 0.48466363678752356,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0018519659012504672,
+      "loss": 1.1902,
+      "step": 5522
+    },
+    {
+      "epoch": 0.484751406370426,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0018515465492986436,
+      "loss": 1.2847,
+      "step": 5523
+    },
+    {
+      "epoch": 0.48483917595332854,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0018511271774549732,
+      "loss": 1.1904,
+      "step": 5524
+    },
+    {
+      "epoch": 0.484926945536231,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0018507077857608464,
+      "loss": 1.1808,
+      "step": 5525
+    },
+    {
+      "epoch": 0.4850147151191335,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0018502883742576549,
+      "loss": 1.1845,
+      "step": 5526
+    },
+    {
+      "epoch": 0.485102484702036,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0018498689429867936,
+      "loss": 1.199,
+      "step": 5527
+    },
+    {
+      "epoch": 0.48519025428493845,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001849449491989658,
+      "loss": 1.1831,
+      "step": 5528
+    },
+    {
+      "epoch": 0.48527802386784097,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.001849030021307647,
+      "loss": 1.1872,
+      "step": 5529
+    },
+    {
+      "epoch": 0.48536579345074343,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0018486105309821607,
+      "loss": 1.1897,
+      "step": 5530
+    },
+    {
+      "epoch": 0.4854535630336459,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0018481910210546005,
+      "loss": 1.1876,
+      "step": 5531
+    },
+    {
+      "epoch": 0.4855413326165484,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001847771491566371,
+      "loss": 1.2325,
+      "step": 5532
+    },
+    {
+      "epoch": 0.4856291021994509,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0018473519425588779,
+      "loss": 1.2243,
+      "step": 5533
+    },
+    {
+      "epoch": 0.4857168717823534,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0018469323740735281,
+      "loss": 1.2145,
+      "step": 5534
+    },
+    {
+      "epoch": 0.48580464136525586,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0018465127861517331,
+      "loss": 1.2092,
+      "step": 5535
+    },
+    {
+      "epoch": 0.4858924109481583,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.001846093178834903,
+      "loss": 1.2302,
+      "step": 5536
+    },
+    {
+      "epoch": 0.48598018053106085,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.001845673552164453,
+      "loss": 1.1447,
+      "step": 5537
+    },
+    {
+      "epoch": 0.4860679501139633,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001845253906181797,
+      "loss": 1.1813,
+      "step": 5538
+    },
+    {
+      "epoch": 0.48615571969686583,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.001844834240928353,
+      "loss": 1.2512,
+      "step": 5539
+    },
+    {
+      "epoch": 0.4862434892797683,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0018444145564455404,
+      "loss": 1.2761,
+      "step": 5540
+    },
+    {
+      "epoch": 0.48633125886267076,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0018439948527747808,
+      "loss": 1.1943,
+      "step": 5541
+    },
+    {
+      "epoch": 0.4864190284455733,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001843575129957496,
+      "loss": 1.211,
+      "step": 5542
+    },
+    {
+      "epoch": 0.48650679802847574,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0018431553880351127,
+      "loss": 1.2018,
+      "step": 5543
+    },
+    {
+      "epoch": 0.4865945676113782,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0018427356270490558,
+      "loss": 1.1885,
+      "step": 5544
+    },
+    {
+      "epoch": 0.4866823371942807,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0018423158470407563,
+      "loss": 1.2539,
+      "step": 5545
+    },
+    {
+      "epoch": 0.4867701067771832,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0018418960480516427,
+      "loss": 1.2094,
+      "step": 5546
+    },
+    {
+      "epoch": 0.4868578763600857,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001841476230123149,
+      "loss": 1.1734,
+      "step": 5547
+    },
+    {
+      "epoch": 0.48694564594298817,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0018410563932967088,
+      "loss": 1.254,
+      "step": 5548
+    },
+    {
+      "epoch": 0.48703341552589063,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0018406365376137587,
+      "loss": 1.1996,
+      "step": 5549
+    },
+    {
+      "epoch": 0.48712118510879315,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0018402166631157364,
+      "loss": 1.2518,
+      "step": 5550
+    },
+    {
+      "epoch": 0.4872089546916956,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0018397967698440825,
+      "loss": 1.2485,
+      "step": 5551
+    },
+    {
+      "epoch": 0.48729672427459814,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.001839376857840238,
+      "loss": 1.1921,
+      "step": 5552
+    },
+    {
+      "epoch": 0.4873844938575006,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0018389569271456477,
+      "loss": 1.2644,
+      "step": 5553
+    },
+    {
+      "epoch": 0.48747226344040306,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0018385369778017554,
+      "loss": 1.2777,
+      "step": 5554
+    },
+    {
+      "epoch": 0.4875600330233056,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00183811700985001,
+      "loss": 1.2682,
+      "step": 5555
+    },
+    {
+      "epoch": 0.48764780260620805,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0018376970233318596,
+      "loss": 1.1651,
+      "step": 5556
+    },
+    {
+      "epoch": 0.4877355721891105,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0018372770182887562,
+      "loss": 1.1804,
+      "step": 5557
+    },
+    {
+      "epoch": 0.48782334177201303,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0018368569947621518,
+      "loss": 1.2092,
+      "step": 5558
+    },
+    {
+      "epoch": 0.4879111113549155,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0018364369527935015,
+      "loss": 1.1668,
+      "step": 5559
+    },
+    {
+      "epoch": 0.487998880937818,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0018360168924242614,
+      "loss": 1.2076,
+      "step": 5560
+    },
+    {
+      "epoch": 0.4880866505207205,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00183559681369589,
+      "loss": 1.2162,
+      "step": 5561
+    },
+    {
+      "epoch": 0.48817442010362294,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001835176716649848,
+      "loss": 1.1691,
+      "step": 5562
+    },
+    {
+      "epoch": 0.48826218968652546,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0018347566013275963,
+      "loss": 1.1655,
+      "step": 5563
+    },
+    {
+      "epoch": 0.4883499592694279,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0018343364677705994,
+      "loss": 1.2172,
+      "step": 5564
+    },
+    {
+      "epoch": 0.48843772885233044,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.001833916316020322,
+      "loss": 1.1728,
+      "step": 5565
+    },
+    {
+      "epoch": 0.4885254984352329,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001833496146118233,
+      "loss": 1.2448,
+      "step": 5566
+    },
+    {
+      "epoch": 0.48861326801813537,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001833075958105799,
+      "loss": 1.2022,
+      "step": 5567
+    },
+    {
+      "epoch": 0.4887010376010379,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0018326557520244933,
+      "loss": 1.2418,
+      "step": 5568
+    },
+    {
+      "epoch": 0.48878880718394035,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0018322355279157877,
+      "loss": 1.1884,
+      "step": 5569
+    },
+    {
+      "epoch": 0.4888765767668428,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.001831815285821156,
+      "loss": 1.1542,
+      "step": 5570
+    },
+    {
+      "epoch": 0.48896434634974534,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0018313950257820752,
+      "loss": 1.2073,
+      "step": 5571
+    },
+    {
+      "epoch": 0.4890521159326478,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0018309747478400232,
+      "loss": 1.2873,
+      "step": 5572
+    },
+    {
+      "epoch": 0.4891398855155503,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0018305544520364797,
+      "loss": 1.1929,
+      "step": 5573
+    },
+    {
+      "epoch": 0.4892276550984528,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0018301341384129258,
+      "loss": 1.2649,
+      "step": 5574
+    },
+    {
+      "epoch": 0.48931542468135525,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0018297138070108457,
+      "loss": 1.2054,
+      "step": 5575
+    },
+    {
+      "epoch": 0.48940319426425777,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0018292934578717239,
+      "loss": 1.2105,
+      "step": 5576
+    },
+    {
+      "epoch": 0.48949096384716023,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.001828873091037047,
+      "loss": 1.1405,
+      "step": 5577
+    },
+    {
+      "epoch": 0.48957873343006275,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001828452706548304,
+      "loss": 1.1334,
+      "step": 5578
+    },
+    {
+      "epoch": 0.4896665030129652,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0018280323044469846,
+      "loss": 1.242,
+      "step": 5579
+    },
+    {
+      "epoch": 0.4897542725958677,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0018276118847745813,
+      "loss": 1.1789,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4898420421787702,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0018271914475725876,
+      "loss": 1.1727,
+      "step": 5581
+    },
+    {
+      "epoch": 0.48992981176167266,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0018267709928824997,
+      "loss": 1.2666,
+      "step": 5582
+    },
+    {
+      "epoch": 0.4900175813445752,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0018263505207458141,
+      "loss": 1.2665,
+      "step": 5583
+    },
+    {
+      "epoch": 0.49010535092747765,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00182593003120403,
+      "loss": 1.2434,
+      "step": 5584
+    },
+    {
+      "epoch": 0.4901931205103801,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0018255095242986471,
+      "loss": 1.3232,
+      "step": 5585
+    },
+    {
+      "epoch": 0.49028089009328263,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0018250890000711694,
+      "loss": 1.2335,
+      "step": 5586
+    },
+    {
+      "epoch": 0.4903686596761851,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0018246684585631002,
+      "loss": 1.2365,
+      "step": 5587
+    },
+    {
+      "epoch": 0.49045642925908756,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001824247899815945,
+      "loss": 1.2217,
+      "step": 5588
+    },
+    {
+      "epoch": 0.4905441988419901,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0018238273238712114,
+      "loss": 1.1816,
+      "step": 5589
+    },
+    {
+      "epoch": 0.49063196842489254,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.001823406730770409,
+      "loss": 1.2392,
+      "step": 5590
+    },
+    {
+      "epoch": 0.49071973800779506,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0018229861205550481,
+      "loss": 1.205,
+      "step": 5591
+    },
+    {
+      "epoch": 0.4908075075906975,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.001822565493266642,
+      "loss": 1.2635,
+      "step": 5592
+    },
+    {
+      "epoch": 0.4908952771736,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0018221448489467043,
+      "loss": 1.2,
+      "step": 5593
+    },
+    {
+      "epoch": 0.4909830467565025,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0018217241876367514,
+      "loss": 1.179,
+      "step": 5594
+    },
+    {
+      "epoch": 0.49107081633940497,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0018213035093783,
+      "loss": 1.1897,
+      "step": 5595
+    },
+    {
+      "epoch": 0.4911585859223075,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0018208828142128705,
+      "loss": 1.2188,
+      "step": 5596
+    },
+    {
+      "epoch": 0.49124635550520995,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0018204621021819833,
+      "loss": 1.2007,
+      "step": 5597
+    },
+    {
+      "epoch": 0.4913341250881124,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0018200413733271614,
+      "loss": 1.219,
+      "step": 5598
+    },
+    {
+      "epoch": 0.49142189467101494,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0018196206276899283,
+      "loss": 1.1905,
+      "step": 5599
+    },
+    {
+      "epoch": 0.4915096642539174,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001819199865311811,
+      "loss": 1.1581,
+      "step": 5600
+    },
+    {
+      "epoch": 0.49159743383681986,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001818779086234336,
+      "loss": 1.1665,
+      "step": 5601
+    },
+    {
+      "epoch": 0.4916852034197224,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0018183582904990332,
+      "loss": 1.186,
+      "step": 5602
+    },
+    {
+      "epoch": 0.49177297300262485,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0018179374781474328,
+      "loss": 1.227,
+      "step": 5603
+    },
+    {
+      "epoch": 0.49186074258552737,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0018175166492210688,
+      "loss": 1.2059,
+      "step": 5604
+    },
+    {
+      "epoch": 0.49194851216842983,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.001817095803761473,
+      "loss": 1.1909,
+      "step": 5605
+    },
+    {
+      "epoch": 0.4920362817513323,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0018166749418101838,
+      "loss": 1.1545,
+      "step": 5606
+    },
+    {
+      "epoch": 0.4921240513342348,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0018162540634087366,
+      "loss": 1.1572,
+      "step": 5607
+    },
+    {
+      "epoch": 0.4922118209171373,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001815833168598672,
+      "loss": 1.2136,
+      "step": 5608
+    },
+    {
+      "epoch": 0.4922995905000398,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0018154122574215287,
+      "loss": 1.2015,
+      "step": 5609
+    },
+    {
+      "epoch": 0.49238736008294226,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0018149913299188507,
+      "loss": 1.2142,
+      "step": 5610
+    },
+    {
+      "epoch": 0.4924751296658447,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.001814570386132181,
+      "loss": 1.2344,
+      "step": 5611
+    },
+    {
+      "epoch": 0.49256289924874724,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0018141494261030662,
+      "loss": 1.1704,
+      "step": 5612
+    },
+    {
+      "epoch": 0.4926506688316497,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001813728449873052,
+      "loss": 1.2038,
+      "step": 5613
+    },
+    {
+      "epoch": 0.49273843841455217,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0018133074574836873,
+      "loss": 1.2406,
+      "step": 5614
+    },
+    {
+      "epoch": 0.4928262079974547,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.001812886448976523,
+      "loss": 1.2321,
+      "step": 5615
+    },
+    {
+      "epoch": 0.49291397758035715,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0018124654243931106,
+      "loss": 1.1943,
+      "step": 5616
+    },
+    {
+      "epoch": 0.4930017471632597,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0018120443837750038,
+      "loss": 1.2525,
+      "step": 5617
+    },
+    {
+      "epoch": 0.49308951674616214,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0018116233271637573,
+      "loss": 1.2237,
+      "step": 5618
+    },
+    {
+      "epoch": 0.4931772863290646,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.001811202254600928,
+      "loss": 1.2018,
+      "step": 5619
+    },
+    {
+      "epoch": 0.4932650559119671,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0018107811661280744,
+      "loss": 1.203,
+      "step": 5620
+    },
+    {
+      "epoch": 0.4933528254948696,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0018103600617867552,
+      "loss": 1.298,
+      "step": 5621
+    },
+    {
+      "epoch": 0.4934405950777721,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.001809938941618533,
+      "loss": 1.2694,
+      "step": 5622
+    },
+    {
+      "epoch": 0.49352836466067457,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0018095178056649701,
+      "loss": 1.1948,
+      "step": 5623
+    },
+    {
+      "epoch": 0.49361613424357703,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0018090966539676305,
+      "loss": 1.2694,
+      "step": 5624
+    },
+    {
+      "epoch": 0.49370390382647955,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.001808675486568081,
+      "loss": 1.1812,
+      "step": 5625
+    },
+    {
+      "epoch": 0.493791673409382,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0018082543035078888,
+      "loss": 1.1932,
+      "step": 5626
+    },
+    {
+      "epoch": 0.4938794429922845,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0018078331048286232,
+      "loss": 1.1834,
+      "step": 5627
+    },
+    {
+      "epoch": 0.493967212575187,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001807411890571854,
+      "loss": 1.1792,
+      "step": 5628
+    },
+    {
+      "epoch": 0.49405498215808946,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0018069906607791548,
+      "loss": 1.2386,
+      "step": 5629
+    },
+    {
+      "epoch": 0.494142751740992,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.001806569415492098,
+      "loss": 1.143,
+      "step": 5630
+    },
+    {
+      "epoch": 0.49423052132389445,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0018061481547522597,
+      "loss": 1.1354,
+      "step": 5631
+    },
+    {
+      "epoch": 0.4943182909067969,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001805726878601216,
+      "loss": 1.1726,
+      "step": 5632
+    },
+    {
+      "epoch": 0.49440606048969943,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0018053055870805462,
+      "loss": 1.1775,
+      "step": 5633
+    },
+    {
+      "epoch": 0.4944938300726019,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.001804884280231829,
+      "loss": 1.2527,
+      "step": 5634
+    },
+    {
+      "epoch": 0.4945815996555044,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0018044629580966465,
+      "loss": 1.2377,
+      "step": 5635
+    },
+    {
+      "epoch": 0.4946693692384069,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0018040416207165812,
+      "loss": 1.2285,
+      "step": 5636
+    },
+    {
+      "epoch": 0.49475713882130934,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.001803620268133217,
+      "loss": 1.1732,
+      "step": 5637
+    },
+    {
+      "epoch": 0.49484490840421186,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0018031989003881403,
+      "loss": 1.2191,
+      "step": 5638
+    },
+    {
+      "epoch": 0.4949326779871143,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.001802777517522939,
+      "loss": 1.1822,
+      "step": 5639
+    },
+    {
+      "epoch": 0.4950204475700168,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0018023561195792,
+      "loss": 1.2544,
+      "step": 5640
+    },
+    {
+      "epoch": 0.4951082171529193,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0018019347065985163,
+      "loss": 1.1761,
+      "step": 5641
+    },
+    {
+      "epoch": 0.49519598673582177,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0018015132786224771,
+      "loss": 1.21,
+      "step": 5642
+    },
+    {
+      "epoch": 0.4952837563187243,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0018010918356926773,
+      "loss": 1.2362,
+      "step": 5643
+    },
+    {
+      "epoch": 0.49537152590162675,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001800670377850711,
+      "loss": 1.1938,
+      "step": 5644
+    },
+    {
+      "epoch": 0.4954592954845292,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0018002489051381747,
+      "loss": 1.1805,
+      "step": 5645
+    },
+    {
+      "epoch": 0.49554706506743174,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0017998274175966657,
+      "loss": 1.1504,
+      "step": 5646
+    },
+    {
+      "epoch": 0.4956348346503342,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0017994059152677839,
+      "loss": 1.2099,
+      "step": 5647
+    },
+    {
+      "epoch": 0.4957226042332367,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001798984398193129,
+      "loss": 1.1844,
+      "step": 5648
+    },
+    {
+      "epoch": 0.4958103738161392,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0017985628664143039,
+      "loss": 1.1532,
+      "step": 5649
+    },
+    {
+      "epoch": 0.49589814339904165,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.001798141319972911,
+      "loss": 1.1587,
+      "step": 5650
+    },
+    {
+      "epoch": 0.49598591298194417,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0017977197589105567,
+      "loss": 1.2439,
+      "step": 5651
+    },
+    {
+      "epoch": 0.49607368256484663,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0017972981832688462,
+      "loss": 1.2248,
+      "step": 5652
+    },
+    {
+      "epoch": 0.49616145214774915,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0017968765930893886,
+      "loss": 1.203,
+      "step": 5653
+    },
+    {
+      "epoch": 0.4962492217306516,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0017964549884137916,
+      "loss": 1.2302,
+      "step": 5654
+    },
+    {
+      "epoch": 0.4963369913135541,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0017960333692836673,
+      "loss": 1.164,
+      "step": 5655
+    },
+    {
+      "epoch": 0.4964247608964566,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0017956117357406268,
+      "loss": 1.2269,
+      "step": 5656
+    },
+    {
+      "epoch": 0.49651253047935906,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0017951900878262843,
+      "loss": 1.187,
+      "step": 5657
+    },
+    {
+      "epoch": 0.4966003000622615,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0017947684255822547,
+      "loss": 1.2321,
+      "step": 5658
+    },
+    {
+      "epoch": 0.49668806964516404,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0017943467490501546,
+      "loss": 1.1806,
+      "step": 5659
+    },
+    {
+      "epoch": 0.4967758392280665,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0017939250582716011,
+      "loss": 1.1994,
+      "step": 5660
+    },
+    {
+      "epoch": 0.496863608810969,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001793503353288214,
+      "loss": 1.1956,
+      "step": 5661
+    },
+    {
+      "epoch": 0.4969513783938715,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0017930816341416133,
+      "loss": 1.1875,
+      "step": 5662
+    },
+    {
+      "epoch": 0.49703914797677395,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0017926599008734221,
+      "loss": 1.2347,
+      "step": 5663
+    },
+    {
+      "epoch": 0.4971269175596765,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.001792238153525263,
+      "loss": 1.1824,
+      "step": 5664
+    },
+    {
+      "epoch": 0.49721468714257894,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0017918163921387608,
+      "loss": 1.1112,
+      "step": 5665
+    },
+    {
+      "epoch": 0.49730245672548146,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0017913946167555418,
+      "loss": 1.2064,
+      "step": 5666
+    },
+    {
+      "epoch": 0.4973902263083839,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0017909728274172337,
+      "loss": 1.1959,
+      "step": 5667
+    },
+    {
+      "epoch": 0.4974779958912864,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0017905510241654659,
+      "loss": 1.1864,
+      "step": 5668
+    },
+    {
+      "epoch": 0.4975657654741889,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.001790129207041867,
+      "loss": 1.1571,
+      "step": 5669
+    },
+    {
+      "epoch": 0.49765353505709137,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0017897073760880704,
+      "loss": 1.195,
+      "step": 5670
+    },
+    {
+      "epoch": 0.49774130463999383,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0017892855313457086,
+      "loss": 1.1986,
+      "step": 5671
+    },
+    {
+      "epoch": 0.49782907422289635,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0017888636728564163,
+      "loss": 1.2116,
+      "step": 5672
+    },
+    {
+      "epoch": 0.4979168438057988,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0017884418006618285,
+      "loss": 1.1817,
+      "step": 5673
+    },
+    {
+      "epoch": 0.49800461338870133,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0017880199148035832,
+      "loss": 1.1766,
+      "step": 5674
+    },
+    {
+      "epoch": 0.4980923829716038,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0017875980153233178,
+      "loss": 1.2124,
+      "step": 5675
+    },
+    {
+      "epoch": 0.49818015255450626,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001787176102262673,
+      "loss": 1.2151,
+      "step": 5676
+    },
+    {
+      "epoch": 0.4982679221374088,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0017867541756632897,
+      "loss": 1.2607,
+      "step": 5677
+    },
+    {
+      "epoch": 0.49835569172031124,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0017863322355668103,
+      "loss": 1.2518,
+      "step": 5678
+    },
+    {
+      "epoch": 0.49844346130321376,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.001785910282014879,
+      "loss": 1.2316,
+      "step": 5679
+    },
+    {
+      "epoch": 0.49853123088611623,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00178548831504914,
+      "loss": 1.1818,
+      "step": 5680
+    },
+    {
+      "epoch": 0.4986190004690187,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001785066334711241,
+      "loss": 1.2757,
+      "step": 5681
+    },
+    {
+      "epoch": 0.4987067700519212,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0017846443410428291,
+      "loss": 1.2275,
+      "step": 5682
+    },
+    {
+      "epoch": 0.4987945396348237,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0017842223340855533,
+      "loss": 1.2493,
+      "step": 5683
+    },
+    {
+      "epoch": 0.49888230921772614,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0017838003138810645,
+      "loss": 1.1619,
+      "step": 5684
+    },
+    {
+      "epoch": 0.49897007880062866,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0017833782804710138,
+      "loss": 1.2084,
+      "step": 5685
+    },
+    {
+      "epoch": 0.4990578483835311,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0017829562338970548,
+      "loss": 1.1847,
+      "step": 5686
+    },
+    {
+      "epoch": 0.49914561796643364,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0017825341742008414,
+      "loss": 1.2249,
+      "step": 5687
+    },
+    {
+      "epoch": 0.4992333875493361,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0017821121014240297,
+      "loss": 1.1999,
+      "step": 5688
+    },
+    {
+      "epoch": 0.49932115713223857,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0017816900156082762,
+      "loss": 1.1534,
+      "step": 5689
+    },
+    {
+      "epoch": 0.4994089267151411,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0017812679167952396,
+      "loss": 1.2024,
+      "step": 5690
+    },
+    {
+      "epoch": 0.49949669629804355,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.001780845805026578,
+      "loss": 1.262,
+      "step": 5691
+    },
+    {
+      "epoch": 0.49958446588094607,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0017804236803439539,
+      "loss": 1.2104,
+      "step": 5692
+    },
+    {
+      "epoch": 0.49967223546384854,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0017800015427890288,
+      "loss": 1.2089,
+      "step": 5693
+    },
+    {
+      "epoch": 0.499760005046751,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0017795793924034657,
+      "loss": 1.2167,
+      "step": 5694
+    },
+    {
+      "epoch": 0.4998477746296535,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.001779157229228929,
+      "loss": 1.2103,
+      "step": 5695
+    },
+    {
+      "epoch": 0.499935544212556,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0017787350533070847,
+      "loss": 1.199,
+      "step": 5696
+    },
+    {
+      "epoch": 0.5000233137954585,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0017783128646796003,
+      "loss": 1.2887,
+      "step": 5697
+    },
+    {
+      "epoch": 0.500111083378361,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0017778906633881442,
+      "loss": 1.1894,
+      "step": 5698
+    },
+    {
+      "epoch": 0.5001988529612634,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.001777468449474385,
+      "loss": 1.2447,
+      "step": 5699
+    },
+    {
+      "epoch": 0.5002866225441659,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0017770462229799947,
+      "loss": 1.2667,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5003743921270685,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0017766239839466442,
+      "loss": 1.1714,
+      "step": 5701
+    },
+    {
+      "epoch": 0.5004621617099709,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0017762017324160083,
+      "loss": 1.1284,
+      "step": 5702
+    },
+    {
+      "epoch": 0.5005499312928734,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0017757794684297604,
+      "loss": 1.2012,
+      "step": 5703
+    },
+    {
+      "epoch": 0.5006377008757759,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0017753571920295767,
+      "loss": 1.1714,
+      "step": 5704
+    },
+    {
+      "epoch": 0.5007254704586783,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0017749349032571342,
+      "loss": 1.1888,
+      "step": 5705
+    },
+    {
+      "epoch": 0.5008132400415808,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001774512602154111,
+      "loss": 1.2332,
+      "step": 5706
+    },
+    {
+      "epoch": 0.5009010096244834,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0017740902887621866,
+      "loss": 1.1619,
+      "step": 5707
+    },
+    {
+      "epoch": 0.5009887792073858,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0017736679631230419,
+      "loss": 1.1973,
+      "step": 5708
+    },
+    {
+      "epoch": 0.5010765487902883,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0017732456252783587,
+      "loss": 1.2145,
+      "step": 5709
+    },
+    {
+      "epoch": 0.5011643183731908,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0017728232752698197,
+      "loss": 1.2164,
+      "step": 5710
+    },
+    {
+      "epoch": 0.5012520879560932,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0017724009131391094,
+      "loss": 1.1637,
+      "step": 5711
+    },
+    {
+      "epoch": 0.5013398575389958,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.001771978538927913,
+      "loss": 1.2159,
+      "step": 5712
+    },
+    {
+      "epoch": 0.5014276271218983,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0017715561526779184,
+      "loss": 1.2481,
+      "step": 5713
+    },
+    {
+      "epoch": 0.5015153967048007,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0017711337544308113,
+      "loss": 1.1586,
+      "step": 5714
+    },
+    {
+      "epoch": 0.5016031662877032,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0017707113442282827,
+      "loss": 1.1608,
+      "step": 5715
+    },
+    {
+      "epoch": 0.5016909358706056,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0017702889221120216,
+      "loss": 1.2293,
+      "step": 5716
+    },
+    {
+      "epoch": 0.5017787054535082,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0017698664881237206,
+      "loss": 1.1586,
+      "step": 5717
+    },
+    {
+      "epoch": 0.5018664750364107,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.001769444042305071,
+      "loss": 1.2277,
+      "step": 5718
+    },
+    {
+      "epoch": 0.5019542446193132,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0017690215846977672,
+      "loss": 1.1611,
+      "step": 5719
+    },
+    {
+      "epoch": 0.5020420142022156,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0017685991153435037,
+      "loss": 1.1954,
+      "step": 5720
+    },
+    {
+      "epoch": 0.5021297837851181,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.001768176634283977,
+      "loss": 1.1809,
+      "step": 5721
+    },
+    {
+      "epoch": 0.5022175533680207,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0017677541415608837,
+      "loss": 1.1718,
+      "step": 5722
+    },
+    {
+      "epoch": 0.5023053229509231,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0017673316372159228,
+      "loss": 1.2601,
+      "step": 5723
+    },
+    {
+      "epoch": 0.5023930925338256,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0017669091212907937,
+      "loss": 1.2421,
+      "step": 5724
+    },
+    {
+      "epoch": 0.502480862116728,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0017664865938271969,
+      "loss": 1.1621,
+      "step": 5725
+    },
+    {
+      "epoch": 0.5025686316996305,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.001766064054866834,
+      "loss": 1.219,
+      "step": 5726
+    },
+    {
+      "epoch": 0.5026564012825331,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0017656415044514082,
+      "loss": 1.2024,
+      "step": 5727
+    },
+    {
+      "epoch": 0.5027441708654355,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0017652189426226237,
+      "loss": 1.1683,
+      "step": 5728
+    },
+    {
+      "epoch": 0.502831940448338,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0017647963694221852,
+      "loss": 1.1589,
+      "step": 5729
+    },
+    {
+      "epoch": 0.5029197100312405,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0017643737848917994,
+      "loss": 1.1823,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5030074796141429,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0017639511890731735,
+      "loss": 1.2139,
+      "step": 5731
+    },
+    {
+      "epoch": 0.5030952491970454,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001763528582008016,
+      "loss": 1.2255,
+      "step": 5732
+    },
+    {
+      "epoch": 0.503183018779948,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0017631059637380367,
+      "loss": 1.1533,
+      "step": 5733
+    },
+    {
+      "epoch": 0.5032707883628504,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0017626833343049467,
+      "loss": 1.18,
+      "step": 5734
+    },
+    {
+      "epoch": 0.5033585579457529,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0017622606937504576,
+      "loss": 1.2632,
+      "step": 5735
+    },
+    {
+      "epoch": 0.5034463275286554,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0017618380421162813,
+      "loss": 1.1955,
+      "step": 5736
+    },
+    {
+      "epoch": 0.5035340971115578,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0017614153794441333,
+      "loss": 1.1604,
+      "step": 5737
+    },
+    {
+      "epoch": 0.5036218666944604,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0017609927057757283,
+      "loss": 1.2527,
+      "step": 5738
+    },
+    {
+      "epoch": 0.5037096362773629,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0017605700211527826,
+      "loss": 1.202,
+      "step": 5739
+    },
+    {
+      "epoch": 0.5037974058602653,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001760147325617013,
+      "loss": 1.1991,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5038851754431678,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0017597246192101379,
+      "loss": 1.1496,
+      "step": 5741
+    },
+    {
+      "epoch": 0.5039729450260703,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0017593019019738775,
+      "loss": 1.2368,
+      "step": 5742
+    },
+    {
+      "epoch": 0.5040607146089728,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.001758879173949952,
+      "loss": 1.2158,
+      "step": 5743
+    },
+    {
+      "epoch": 0.5041484841918753,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0017584564351800823,
+      "loss": 1.2131,
+      "step": 5744
+    },
+    {
+      "epoch": 0.5042362537747778,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0017580336857059922,
+      "loss": 1.1716,
+      "step": 5745
+    },
+    {
+      "epoch": 0.5043240233576802,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0017576109255694038,
+      "loss": 1.218,
+      "step": 5746
+    },
+    {
+      "epoch": 0.5044117929405827,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0017571881548120438,
+      "loss": 1.1704,
+      "step": 5747
+    },
+    {
+      "epoch": 0.5044995625234853,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0017567653734756367,
+      "loss": 1.1634,
+      "step": 5748
+    },
+    {
+      "epoch": 0.5045873321063877,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00175634258160191,
+      "loss": 1.2068,
+      "step": 5749
+    },
+    {
+      "epoch": 0.5046751016892902,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0017559197792325909,
+      "loss": 1.2329,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5047628712721927,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0017554969664094087,
+      "loss": 1.2032,
+      "step": 5751
+    },
+    {
+      "epoch": 0.5048506408550951,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0017550741431740932,
+      "loss": 1.1963,
+      "step": 5752
+    },
+    {
+      "epoch": 0.5049384104379977,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.001754651309568376,
+      "loss": 1.2047,
+      "step": 5753
+    },
+    {
+      "epoch": 0.5050261800209002,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0017542284656339882,
+      "loss": 1.2416,
+      "step": 5754
+    },
+    {
+      "epoch": 0.5051139496038026,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.001753805611412664,
+      "loss": 1.1802,
+      "step": 5755
+    },
+    {
+      "epoch": 0.5052017191867051,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.001753382746946136,
+      "loss": 1.1666,
+      "step": 5756
+    },
+    {
+      "epoch": 0.5052894887696076,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0017529598722761397,
+      "loss": 1.1399,
+      "step": 5757
+    },
+    {
+      "epoch": 0.5053772583525101,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.001752536987444412,
+      "loss": 1.1521,
+      "step": 5758
+    },
+    {
+      "epoch": 0.5054650279354126,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0017521140924926894,
+      "loss": 1.2258,
+      "step": 5759
+    },
+    {
+      "epoch": 0.5055527975183151,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0017516911874627093,
+      "loss": 1.238,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5056405671012175,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0017512682723962115,
+      "loss": 1.1735,
+      "step": 5761
+    },
+    {
+      "epoch": 0.50572833668412,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0017508453473349364,
+      "loss": 1.2454,
+      "step": 5762
+    },
+    {
+      "epoch": 0.5058161062670224,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0017504224123206244,
+      "loss": 1.1982,
+      "step": 5763
+    },
+    {
+      "epoch": 0.505903875849925,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0017499994673950178,
+      "loss": 1.1627,
+      "step": 5764
+    },
+    {
+      "epoch": 0.5059916454328275,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0017495765125998587,
+      "loss": 1.1383,
+      "step": 5765
+    },
+    {
+      "epoch": 0.50607941501573,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0017491535479768922,
+      "loss": 1.1676,
+      "step": 5766
+    },
+    {
+      "epoch": 0.5061671845986324,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0017487305735678631,
+      "loss": 1.1698,
+      "step": 5767
+    },
+    {
+      "epoch": 0.5062549541815349,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.001748307589414517,
+      "loss": 1.2544,
+      "step": 5768
+    },
+    {
+      "epoch": 0.5063427237644375,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0017478845955586006,
+      "loss": 1.1818,
+      "step": 5769
+    },
+    {
+      "epoch": 0.5064304933473399,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0017474615920418622,
+      "loss": 1.2482,
+      "step": 5770
+    },
+    {
+      "epoch": 0.5065182629302424,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0017470385789060496,
+      "loss": 1.2053,
+      "step": 5771
+    },
+    {
+      "epoch": 0.5066060325131448,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001746615556192914,
+      "loss": 1.1992,
+      "step": 5772
+    },
+    {
+      "epoch": 0.5066938020960473,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0017461925239442052,
+      "loss": 1.2513,
+      "step": 5773
+    },
+    {
+      "epoch": 0.5067815716789499,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.001745769482201675,
+      "loss": 1.193,
+      "step": 5774
+    },
+    {
+      "epoch": 0.5068693412618523,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0017453464310070751,
+      "loss": 1.2152,
+      "step": 5775
+    },
+    {
+      "epoch": 0.5069571108447548,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0017449233704021606,
+      "loss": 1.2397,
+      "step": 5776
+    },
+    {
+      "epoch": 0.5070448804276573,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0017445003004286843,
+      "loss": 1.2336,
+      "step": 5777
+    },
+    {
+      "epoch": 0.5071326500105597,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0017440772211284028,
+      "loss": 1.2288,
+      "step": 5778
+    },
+    {
+      "epoch": 0.5072204195934623,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0017436541325430717,
+      "loss": 1.2062,
+      "step": 5779
+    },
+    {
+      "epoch": 0.5073081891763648,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0017432310347144481,
+      "loss": 1.2069,
+      "step": 5780
+    },
+    {
+      "epoch": 0.5073959587592672,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00174280792768429,
+      "loss": 1.2542,
+      "step": 5781
+    },
+    {
+      "epoch": 0.5074837283421697,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0017423848114943574,
+      "loss": 1.2265,
+      "step": 5782
+    },
+    {
+      "epoch": 0.5075714979250722,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.001741961686186409,
+      "loss": 1.2078,
+      "step": 5783
+    },
+    {
+      "epoch": 0.5076592675079747,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0017415385518022063,
+      "loss": 1.1871,
+      "step": 5784
+    },
+    {
+      "epoch": 0.5077470370908772,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0017411154083835105,
+      "loss": 1.2536,
+      "step": 5785
+    },
+    {
+      "epoch": 0.5078348066737797,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0017406922559720845,
+      "loss": 1.2,
+      "step": 5786
+    },
+    {
+      "epoch": 0.5079225762566821,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0017402690946096916,
+      "loss": 1.1885,
+      "step": 5787
+    },
+    {
+      "epoch": 0.5080103458395846,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0017398459243380966,
+      "loss": 1.1883,
+      "step": 5788
+    },
+    {
+      "epoch": 0.5080981154224871,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001739422745199064,
+      "loss": 1.1563,
+      "step": 5789
+    },
+    {
+      "epoch": 0.5081858850053896,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0017389995572343602,
+      "loss": 1.1719,
+      "step": 5790
+    },
+    {
+      "epoch": 0.5082736545882921,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0017385763604857523,
+      "loss": 1.2017,
+      "step": 5791
+    },
+    {
+      "epoch": 0.5083614241711946,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0017381531549950085,
+      "loss": 1.1683,
+      "step": 5792
+    },
+    {
+      "epoch": 0.508449193754097,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0017377299408038967,
+      "loss": 1.189,
+      "step": 5793
+    },
+    {
+      "epoch": 0.5085369633369995,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.001737306717954187,
+      "loss": 1.2054,
+      "step": 5794
+    },
+    {
+      "epoch": 0.5086247329199021,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0017368834864876497,
+      "loss": 1.2339,
+      "step": 5795
+    },
+    {
+      "epoch": 0.5087125025028045,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0017364602464460558,
+      "loss": 1.2185,
+      "step": 5796
+    },
+    {
+      "epoch": 0.508800272085707,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001736036997871178,
+      "loss": 1.2116,
+      "step": 5797
+    },
+    {
+      "epoch": 0.5088880416686095,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001735613740804789,
+      "loss": 1.2092,
+      "step": 5798
+    },
+    {
+      "epoch": 0.5089758112515119,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0017351904752886625,
+      "loss": 1.1909,
+      "step": 5799
+    },
+    {
+      "epoch": 0.5090635808344145,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0017347672013645731,
+      "loss": 1.1875,
+      "step": 5800
+    },
+    {
+      "epoch": 0.509151350417317,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0017343439190742967,
+      "loss": 1.2468,
+      "step": 5801
+    },
+    {
+      "epoch": 0.5092391200002194,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0017339206284596085,
+      "loss": 1.2145,
+      "step": 5802
+    },
+    {
+      "epoch": 0.5093268895831219,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0017334973295622866,
+      "loss": 1.2162,
+      "step": 5803
+    },
+    {
+      "epoch": 0.5094146591660244,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.001733074022424109,
+      "loss": 1.2398,
+      "step": 5804
+    },
+    {
+      "epoch": 0.5095024287489269,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0017326507070868538,
+      "loss": 1.159,
+      "step": 5805
+    },
+    {
+      "epoch": 0.5095901983318294,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0017322273835923007,
+      "loss": 1.2283,
+      "step": 5806
+    },
+    {
+      "epoch": 0.5096779679147319,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0017318040519822306,
+      "loss": 1.1445,
+      "step": 5807
+    },
+    {
+      "epoch": 0.5097657374976343,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001731380712298424,
+      "loss": 1.2191,
+      "step": 5808
+    },
+    {
+      "epoch": 0.5098535070805368,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0017309573645826629,
+      "loss": 1.1801,
+      "step": 5809
+    },
+    {
+      "epoch": 0.5099412766634394,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0017305340088767307,
+      "loss": 1.2172,
+      "step": 5810
+    },
+    {
+      "epoch": 0.5100290462463418,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0017301106452224102,
+      "loss": 1.2446,
+      "step": 5811
+    },
+    {
+      "epoch": 0.5101168158292443,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001729687273661486,
+      "loss": 1.1777,
+      "step": 5812
+    },
+    {
+      "epoch": 0.5102045854121467,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0017292638942357434,
+      "loss": 1.1731,
+      "step": 5813
+    },
+    {
+      "epoch": 0.5102923549950492,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0017288405069869677,
+      "loss": 1.2281,
+      "step": 5814
+    },
+    {
+      "epoch": 0.5103801245779517,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0017284171119569462,
+      "loss": 1.249,
+      "step": 5815
+    },
+    {
+      "epoch": 0.5104678941608543,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001727993709187466,
+      "loss": 1.1903,
+      "step": 5816
+    },
+    {
+      "epoch": 0.5105556637437567,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0017275702987203153,
+      "loss": 1.1755,
+      "step": 5817
+    },
+    {
+      "epoch": 0.5106434333266592,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001727146880597283,
+      "loss": 1.1874,
+      "step": 5818
+    },
+    {
+      "epoch": 0.5107312029095616,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0017267234548601592,
+      "loss": 1.1159,
+      "step": 5819
+    },
+    {
+      "epoch": 0.5108189724924641,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0017263000215507333,
+      "loss": 1.1645,
+      "step": 5820
+    },
+    {
+      "epoch": 0.5109067420753667,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0017258765807107975,
+      "loss": 1.2263,
+      "step": 5821
+    },
+    {
+      "epoch": 0.5109945116582691,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0017254531323821433,
+      "loss": 1.2312,
+      "step": 5822
+    },
+    {
+      "epoch": 0.5110822812411716,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0017250296766065636,
+      "loss": 1.1911,
+      "step": 5823
+    },
+    {
+      "epoch": 0.5111700508240741,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0017246062134258517,
+      "loss": 1.1471,
+      "step": 5824
+    },
+    {
+      "epoch": 0.5112578204069765,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.001724182742881802,
+      "loss": 1.1924,
+      "step": 5825
+    },
+    {
+      "epoch": 0.5113455899898791,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0017237592650162087,
+      "loss": 1.1897,
+      "step": 5826
+    },
+    {
+      "epoch": 0.5114333595727816,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0017233357798708679,
+      "loss": 1.1976,
+      "step": 5827
+    },
+    {
+      "epoch": 0.511521129155684,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.001722912287487576,
+      "loss": 1.1992,
+      "step": 5828
+    },
+    {
+      "epoch": 0.5116088987385865,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00172248878790813,
+      "loss": 1.2,
+      "step": 5829
+    },
+    {
+      "epoch": 0.511696668321489,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0017220652811743267,
+      "loss": 1.2266,
+      "step": 5830
+    },
+    {
+      "epoch": 0.5117844379043915,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0017216417673279662,
+      "loss": 1.1915,
+      "step": 5831
+    },
+    {
+      "epoch": 0.511872207487294,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.001721218246410846,
+      "loss": 1.2516,
+      "step": 5832
+    },
+    {
+      "epoch": 0.5119599770701965,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.001720794718464767,
+      "loss": 1.1898,
+      "step": 5833
+    },
+    {
+      "epoch": 0.5120477466530989,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001720371183531529,
+      "loss": 1.1375,
+      "step": 5834
+    },
+    {
+      "epoch": 0.5121355162360014,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0017199476416529344,
+      "loss": 1.1382,
+      "step": 5835
+    },
+    {
+      "epoch": 0.512223285818904,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0017195240928707836,
+      "loss": 1.246,
+      "step": 5836
+    },
+    {
+      "epoch": 0.5123110554018064,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0017191005372268803,
+      "loss": 1.2246,
+      "step": 5837
+    },
+    {
+      "epoch": 0.5123988249847089,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0017186769747630272,
+      "loss": 1.172,
+      "step": 5838
+    },
+    {
+      "epoch": 0.5124865945676114,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0017182534055210287,
+      "loss": 1.1583,
+      "step": 5839
+    },
+    {
+      "epoch": 0.5125743641505138,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001717829829542689,
+      "loss": 1.16,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5126621337334164,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0017174062468698138,
+      "loss": 1.183,
+      "step": 5841
+    },
+    {
+      "epoch": 0.5127499033163189,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.001716982657544208,
+      "loss": 1.2016,
+      "step": 5842
+    },
+    {
+      "epoch": 0.5128376728992213,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0017165590616076793,
+      "loss": 1.1783,
+      "step": 5843
+    },
+    {
+      "epoch": 0.5129254424821238,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0017161354591020347,
+      "loss": 1.1923,
+      "step": 5844
+    },
+    {
+      "epoch": 0.5130132120650263,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0017157118500690826,
+      "loss": 1.2211,
+      "step": 5845
+    },
+    {
+      "epoch": 0.5131009816479287,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0017152882345506301,
+      "loss": 1.1938,
+      "step": 5846
+    },
+    {
+      "epoch": 0.5131887512308313,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0017148646125884873,
+      "loss": 1.1713,
+      "step": 5847
+    },
+    {
+      "epoch": 0.5132765208137338,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0017144409842244639,
+      "loss": 1.2203,
+      "step": 5848
+    },
+    {
+      "epoch": 0.5133642903966362,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0017140173495003705,
+      "loss": 1.1886,
+      "step": 5849
+    },
+    {
+      "epoch": 0.5134520599795387,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0017135937084580175,
+      "loss": 1.1353,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5135398295624412,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0017131700611392178,
+      "loss": 1.288,
+      "step": 5851
+    },
+    {
+      "epoch": 0.5136275991453437,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0017127464075857828,
+      "loss": 1.1695,
+      "step": 5852
+    },
+    {
+      "epoch": 0.5137153687282462,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0017123227478395254,
+      "loss": 1.2113,
+      "step": 5853
+    },
+    {
+      "epoch": 0.5138031383111487,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0017118990819422594,
+      "loss": 1.1542,
+      "step": 5854
+    },
+    {
+      "epoch": 0.5138909078940511,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0017114754099357996,
+      "loss": 1.2009,
+      "step": 5855
+    },
+    {
+      "epoch": 0.5139786774769536,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001711051731861959,
+      "loss": 1.205,
+      "step": 5856
+    },
+    {
+      "epoch": 0.5140664470598562,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0017106280477625546,
+      "loss": 1.178,
+      "step": 5857
+    },
+    {
+      "epoch": 0.5141542166427586,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0017102043576794023,
+      "loss": 1.2346,
+      "step": 5858
+    },
+    {
+      "epoch": 0.5142419862256611,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0017097806616543172,
+      "loss": 1.1765,
+      "step": 5859
+    },
+    {
+      "epoch": 0.5143297558085635,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0017093569597291176,
+      "loss": 1.2248,
+      "step": 5860
+    },
+    {
+      "epoch": 0.514417525391466,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0017089332519456214,
+      "loss": 1.1933,
+      "step": 5861
+    },
+    {
+      "epoch": 0.5145052949743686,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0017085095383456456,
+      "loss": 1.2237,
+      "step": 5862
+    },
+    {
+      "epoch": 0.514593064557271,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0017080858189710106,
+      "loss": 1.2071,
+      "step": 5863
+    },
+    {
+      "epoch": 0.5146808341401735,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0017076620938635352,
+      "loss": 1.2101,
+      "step": 5864
+    },
+    {
+      "epoch": 0.514768603723076,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0017072383630650388,
+      "loss": 1.1898,
+      "step": 5865
+    },
+    {
+      "epoch": 0.5148563733059784,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0017068146266173425,
+      "loss": 1.1323,
+      "step": 5866
+    },
+    {
+      "epoch": 0.514944142888881,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0017063908845622673,
+      "loss": 1.1967,
+      "step": 5867
+    },
+    {
+      "epoch": 0.5150319124717835,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0017059671369416354,
+      "loss": 1.2031,
+      "step": 5868
+    },
+    {
+      "epoch": 0.515119682054686,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.001705543383797268,
+      "loss": 1.1404,
+      "step": 5869
+    },
+    {
+      "epoch": 0.5152074516375884,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0017051196251709887,
+      "loss": 1.2557,
+      "step": 5870
+    },
+    {
+      "epoch": 0.5152952212204909,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0017046958611046203,
+      "loss": 1.1856,
+      "step": 5871
+    },
+    {
+      "epoch": 0.5153829908033933,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0017042720916399868,
+      "loss": 1.2141,
+      "step": 5872
+    },
+    {
+      "epoch": 0.5154707603862959,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0017038483168189131,
+      "loss": 1.2483,
+      "step": 5873
+    },
+    {
+      "epoch": 0.5155585299691984,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001703424536683223,
+      "loss": 1.1784,
+      "step": 5874
+    },
+    {
+      "epoch": 0.5156462995521008,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0017030007512747428,
+      "loss": 1.1865,
+      "step": 5875
+    },
+    {
+      "epoch": 0.5157340691350033,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0017025769606352982,
+      "loss": 1.2004,
+      "step": 5876
+    },
+    {
+      "epoch": 0.5158218387179058,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001702153164806715,
+      "loss": 1.2224,
+      "step": 5877
+    },
+    {
+      "epoch": 0.5159096083008083,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0017017293638308217,
+      "loss": 1.2389,
+      "step": 5878
+    },
+    {
+      "epoch": 0.5159973778837108,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0017013055577494445,
+      "loss": 1.1914,
+      "step": 5879
+    },
+    {
+      "epoch": 0.5160851474666133,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0017008817466044118,
+      "loss": 1.1712,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5161729170495157,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0017004579304375524,
+      "loss": 1.1601,
+      "step": 5881
+    },
+    {
+      "epoch": 0.5162606866324182,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.001700034109290694,
+      "loss": 1.1624,
+      "step": 5882
+    },
+    {
+      "epoch": 0.5163484562153208,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0016996102832056673,
+      "loss": 1.167,
+      "step": 5883
+    },
+    {
+      "epoch": 0.5164362257982232,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0016991864522243024,
+      "loss": 1.2413,
+      "step": 5884
+    },
+    {
+      "epoch": 0.5165239953811257,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0016987626163884292,
+      "loss": 1.1955,
+      "step": 5885
+    },
+    {
+      "epoch": 0.5166117649640282,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0016983387757398786,
+      "loss": 1.1993,
+      "step": 5886
+    },
+    {
+      "epoch": 0.5166995345469306,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0016979149303204815,
+      "loss": 1.2087,
+      "step": 5887
+    },
+    {
+      "epoch": 0.5167873041298332,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001697491080172071,
+      "loss": 1.1863,
+      "step": 5888
+    },
+    {
+      "epoch": 0.5168750737127357,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0016970672253364789,
+      "loss": 1.2547,
+      "step": 5889
+    },
+    {
+      "epoch": 0.5169628432956381,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001696643365855538,
+      "loss": 1.2017,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5170506128785406,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0016962195017710807,
+      "loss": 1.1522,
+      "step": 5891
+    },
+    {
+      "epoch": 0.5171383824614431,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0016957956331249422,
+      "loss": 1.22,
+      "step": 5892
+    },
+    {
+      "epoch": 0.5172261520443456,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0016953717599589551,
+      "loss": 1.1915,
+      "step": 5893
+    },
+    {
+      "epoch": 0.5173139216272481,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0016949478823149555,
+      "loss": 1.2239,
+      "step": 5894
+    },
+    {
+      "epoch": 0.5174016912101506,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.001694524000234778,
+      "loss": 1.2802,
+      "step": 5895
+    },
+    {
+      "epoch": 0.517489460793053,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.001694100113760257,
+      "loss": 1.1785,
+      "step": 5896
+    },
+    {
+      "epoch": 0.5175772303759555,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.00169367622293323,
+      "loss": 1.2512,
+      "step": 5897
+    },
+    {
+      "epoch": 0.5176649999588581,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0016932523277955324,
+      "loss": 1.2313,
+      "step": 5898
+    },
+    {
+      "epoch": 0.5177527695417605,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0016928284283890009,
+      "loss": 1.1816,
+      "step": 5899
+    },
+    {
+      "epoch": 0.517840539124663,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0016924045247554732,
+      "loss": 1.2514,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5179283087075655,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0016919806169367866,
+      "loss": 1.1811,
+      "step": 5901
+    },
+    {
+      "epoch": 0.5180160782904679,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016915567049747794,
+      "loss": 1.2001,
+      "step": 5902
+    },
+    {
+      "epoch": 0.5181038478733704,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0016911327889112902,
+      "loss": 1.1837,
+      "step": 5903
+    },
+    {
+      "epoch": 0.518191617456273,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0016907088687881566,
+      "loss": 1.2073,
+      "step": 5904
+    },
+    {
+      "epoch": 0.5182793870391754,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0016902849446472192,
+      "loss": 1.1584,
+      "step": 5905
+    },
+    {
+      "epoch": 0.5183671566220779,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0016898610165303174,
+      "loss": 1.1504,
+      "step": 5906
+    },
+    {
+      "epoch": 0.5184549262049803,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0016894370844792905,
+      "loss": 1.1864,
+      "step": 5907
+    },
+    {
+      "epoch": 0.5185426957878828,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0016890131485359796,
+      "loss": 1.2212,
+      "step": 5908
+    },
+    {
+      "epoch": 0.5186304653707854,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0016885892087422257,
+      "loss": 1.2034,
+      "step": 5909
+    },
+    {
+      "epoch": 0.5187182349536879,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0016881652651398692,
+      "loss": 1.1584,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5188060045365903,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0016877413177707522,
+      "loss": 1.1881,
+      "step": 5911
+    },
+    {
+      "epoch": 0.5188937741194928,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0016873173666767157,
+      "loss": 1.233,
+      "step": 5912
+    },
+    {
+      "epoch": 0.5189815437023952,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.001686893411899604,
+      "loss": 1.2495,
+      "step": 5913
+    },
+    {
+      "epoch": 0.5190693132852978,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0016864694534812577,
+      "loss": 1.188,
+      "step": 5914
+    },
+    {
+      "epoch": 0.5191570828682003,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.001686045491463521,
+      "loss": 1.2306,
+      "step": 5915
+    },
+    {
+      "epoch": 0.5192448524511027,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0016856215258882368,
+      "loss": 1.2367,
+      "step": 5916
+    },
+    {
+      "epoch": 0.5193326220340052,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0016851975567972488,
+      "loss": 1.172,
+      "step": 5917
+    },
+    {
+      "epoch": 0.5194203916169077,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0016847735842324013,
+      "loss": 1.2108,
+      "step": 5918
+    },
+    {
+      "epoch": 0.5195081611998102,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0016843496082355389,
+      "loss": 1.1407,
+      "step": 5919
+    },
+    {
+      "epoch": 0.5195959307827127,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0016839256288485058,
+      "loss": 1.2466,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5196837003656152,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0016835016461131476,
+      "loss": 1.1426,
+      "step": 5921
+    },
+    {
+      "epoch": 0.5197714699485176,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0016830776600713089,
+      "loss": 1.2075,
+      "step": 5922
+    },
+    {
+      "epoch": 0.5198592395314201,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0016826536707648366,
+      "loss": 1.2231,
+      "step": 5923
+    },
+    {
+      "epoch": 0.5199470091143227,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0016822296782355761,
+      "loss": 1.19,
+      "step": 5924
+    },
+    {
+      "epoch": 0.5200347786972251,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0016818056825253739,
+      "loss": 1.2029,
+      "step": 5925
+    },
+    {
+      "epoch": 0.5201225482801276,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0016813816836760767,
+      "loss": 1.1887,
+      "step": 5926
+    },
+    {
+      "epoch": 0.5202103178630301,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0016809576817295312,
+      "loss": 1.2285,
+      "step": 5927
+    },
+    {
+      "epoch": 0.5202980874459325,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0016805336767275853,
+      "loss": 1.1851,
+      "step": 5928
+    },
+    {
+      "epoch": 0.520385857028835,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001680109668712086,
+      "loss": 1.1849,
+      "step": 5929
+    },
+    {
+      "epoch": 0.5204736266117376,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.001679685657724882,
+      "loss": 1.1625,
+      "step": 5930
+    },
+    {
+      "epoch": 0.52056139619464,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0016792616438078211,
+      "loss": 1.1538,
+      "step": 5931
+    },
+    {
+      "epoch": 0.5206491657775425,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0016788376270027512,
+      "loss": 1.1891,
+      "step": 5932
+    },
+    {
+      "epoch": 0.520736935360445,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0016784136073515218,
+      "loss": 1.2512,
+      "step": 5933
+    },
+    {
+      "epoch": 0.5208247049433474,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0016779895848959819,
+      "loss": 1.2484,
+      "step": 5934
+    },
+    {
+      "epoch": 0.52091247452625,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0016775655596779808,
+      "loss": 1.179,
+      "step": 5935
+    },
+    {
+      "epoch": 0.5210002441091525,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0016771415317393679,
+      "loss": 1.2038,
+      "step": 5936
+    },
+    {
+      "epoch": 0.5210880136920549,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0016767175011219932,
+      "loss": 1.2503,
+      "step": 5937
+    },
+    {
+      "epoch": 0.5211757832749574,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001676293467867707,
+      "loss": 1.2412,
+      "step": 5938
+    },
+    {
+      "epoch": 0.5212635528578599,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0016758694320183594,
+      "loss": 1.2116,
+      "step": 5939
+    },
+    {
+      "epoch": 0.5213513224407624,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0016754453936158012,
+      "loss": 1.1987,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5214390920236649,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0016750213527018834,
+      "loss": 1.1392,
+      "step": 5941
+    },
+    {
+      "epoch": 0.5215268616065674,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0016745973093184572,
+      "loss": 1.2243,
+      "step": 5942
+    },
+    {
+      "epoch": 0.5216146311894698,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0016741732635073736,
+      "loss": 1.1448,
+      "step": 5943
+    },
+    {
+      "epoch": 0.5217024007723723,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0016737492153104845,
+      "loss": 1.1601,
+      "step": 5944
+    },
+    {
+      "epoch": 0.5217901703552749,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0016733251647696424,
+      "loss": 1.2327,
+      "step": 5945
+    },
+    {
+      "epoch": 0.5218779399381773,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0016729011119266985,
+      "loss": 1.2385,
+      "step": 5946
+    },
+    {
+      "epoch": 0.5219657095210798,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0016724770568235054,
+      "loss": 1.2016,
+      "step": 5947
+    },
+    {
+      "epoch": 0.5220534791039823,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.001672052999501916,
+      "loss": 1.1717,
+      "step": 5948
+    },
+    {
+      "epoch": 0.5221412486868847,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0016716289400037822,
+      "loss": 1.1706,
+      "step": 5949
+    },
+    {
+      "epoch": 0.5222290182697873,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.001671204878370958,
+      "loss": 1.1849,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5223167878526898,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0016707808146452963,
+      "loss": 1.2236,
+      "step": 5951
+    },
+    {
+      "epoch": 0.5224045574355922,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0016703567488686503,
+      "loss": 1.1704,
+      "step": 5952
+    },
+    {
+      "epoch": 0.5224923270184947,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.001669932681082874,
+      "loss": 1.2394,
+      "step": 5953
+    },
+    {
+      "epoch": 0.5225800966013971,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0016695086113298206,
+      "loss": 1.1929,
+      "step": 5954
+    },
+    {
+      "epoch": 0.5226678661842996,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0016690845396513444,
+      "loss": 1.2097,
+      "step": 5955
+    },
+    {
+      "epoch": 0.5227556357672022,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0016686604660892997,
+      "loss": 1.2398,
+      "step": 5956
+    },
+    {
+      "epoch": 0.5228434053501047,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.001668236390685541,
+      "loss": 1.1755,
+      "step": 5957
+    },
+    {
+      "epoch": 0.5229311749330071,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0016678123134819226,
+      "loss": 1.1669,
+      "step": 5958
+    },
+    {
+      "epoch": 0.5230189445159096,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0016673882345202991,
+      "loss": 1.2118,
+      "step": 5959
+    },
+    {
+      "epoch": 0.523106714098812,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0016669641538425263,
+      "loss": 1.2222,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5231944836817146,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001666540071490458,
+      "loss": 1.1956,
+      "step": 5961
+    },
+    {
+      "epoch": 0.5232822532646171,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0016661159875059508,
+      "loss": 1.2088,
+      "step": 5962
+    },
+    {
+      "epoch": 0.5233700228475195,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0016656919019308585,
+      "loss": 1.1622,
+      "step": 5963
+    },
+    {
+      "epoch": 0.523457792430422,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0016652678148070385,
+      "loss": 1.2541,
+      "step": 5964
+    },
+    {
+      "epoch": 0.5235455620133245,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0016648437261763446,
+      "loss": 1.216,
+      "step": 5965
+    },
+    {
+      "epoch": 0.523633331596227,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0016644196360806347,
+      "loss": 1.184,
+      "step": 5966
+    },
+    {
+      "epoch": 0.5237211011791295,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.001663995544561763,
+      "loss": 1.1758,
+      "step": 5967
+    },
+    {
+      "epoch": 0.523808870762032,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0016635714516615864,
+      "loss": 1.1795,
+      "step": 5968
+    },
+    {
+      "epoch": 0.5238966403449344,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0016631473574219614,
+      "loss": 1.2187,
+      "step": 5969
+    },
+    {
+      "epoch": 0.5239844099278369,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0016627232618847449,
+      "loss": 1.179,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5240721795107395,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.001662299165091792,
+      "loss": 1.3068,
+      "step": 5971
+    },
+    {
+      "epoch": 0.5241599490936419,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0016618750670849606,
+      "loss": 1.1545,
+      "step": 5972
+    },
+    {
+      "epoch": 0.5242477186765444,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0016614509679061063,
+      "loss": 1.2528,
+      "step": 5973
+    },
+    {
+      "epoch": 0.5243354882594469,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0016610268675970878,
+      "loss": 1.1369,
+      "step": 5974
+    },
+    {
+      "epoch": 0.5244232578423493,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0016606027661997603,
+      "loss": 1.2096,
+      "step": 5975
+    },
+    {
+      "epoch": 0.5245110274252519,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.001660178663755982,
+      "loss": 1.0937,
+      "step": 5976
+    },
+    {
+      "epoch": 0.5245987970081544,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0016597545603076103,
+      "loss": 1.1955,
+      "step": 5977
+    },
+    {
+      "epoch": 0.5246865665910568,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0016593304558965016,
+      "loss": 1.2321,
+      "step": 5978
+    },
+    {
+      "epoch": 0.5247743361739593,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0016589063505645138,
+      "loss": 1.1762,
+      "step": 5979
+    },
+    {
+      "epoch": 0.5248621057568618,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0016584822443535048,
+      "loss": 1.2473,
+      "step": 5980
+    },
+    {
+      "epoch": 0.5249498753397643,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0016580581373053314,
+      "loss": 1.1883,
+      "step": 5981
+    },
+    {
+      "epoch": 0.5250376449226668,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0016576340294618523,
+      "loss": 1.1568,
+      "step": 5982
+    },
+    {
+      "epoch": 0.5251254145055693,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0016572099208649242,
+      "loss": 1.1327,
+      "step": 5983
+    },
+    {
+      "epoch": 0.5252131840884717,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0016567858115564057,
+      "loss": 1.199,
+      "step": 5984
+    },
+    {
+      "epoch": 0.5253009536713742,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001656361701578154,
+      "loss": 1.1539,
+      "step": 5985
+    },
+    {
+      "epoch": 0.5253887232542767,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0016559375909720284,
+      "loss": 1.1824,
+      "step": 5986
+    },
+    {
+      "epoch": 0.5254764928371792,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0016555134797798853,
+      "loss": 1.1351,
+      "step": 5987
+    },
+    {
+      "epoch": 0.5255642624200817,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0016550893680435837,
+      "loss": 1.2035,
+      "step": 5988
+    },
+    {
+      "epoch": 0.5256520320029842,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0016546652558049816,
+      "loss": 1.2646,
+      "step": 5989
+    },
+    {
+      "epoch": 0.5257398015858866,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0016542411431059376,
+      "loss": 1.1536,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5258275711687891,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0016538170299883093,
+      "loss": 1.1688,
+      "step": 5991
+    },
+    {
+      "epoch": 0.5259153407516917,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0016533929164939554,
+      "loss": 1.1642,
+      "step": 5992
+    },
+    {
+      "epoch": 0.5260031103345941,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0016529688026647336,
+      "loss": 1.2201,
+      "step": 5993
+    },
+    {
+      "epoch": 0.5260908799174966,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0016525446885425031,
+      "loss": 1.2092,
+      "step": 5994
+    },
+    {
+      "epoch": 0.526178649500399,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0016521205741691217,
+      "loss": 1.1637,
+      "step": 5995
+    },
+    {
+      "epoch": 0.5262664190833015,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0016516964595864486,
+      "loss": 1.218,
+      "step": 5996
+    },
+    {
+      "epoch": 0.5263541886662041,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001651272344836341,
+      "loss": 1.2117,
+      "step": 5997
+    },
+    {
+      "epoch": 0.5264419582491066,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.001650848229960658,
+      "loss": 1.1804,
+      "step": 5998
+    },
+    {
+      "epoch": 0.526529727832009,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0016504241150012582,
+      "loss": 1.2124,
+      "step": 5999
+    },
+    {
+      "epoch": 0.5266174974149115,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0016500000000000002,
+      "loss": 1.2236,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5266174974149115,
+      "eval_loss": 1.1826441287994385,
+      "eval_runtime": 437.6578,
+      "eval_samples_per_second": 33.679,
+      "eval_steps_per_second": 8.42,
+      "step": 6000
+    },
+    {
+      "epoch": 0.526705266997814,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0016495758849987417,
+      "loss": 1.2329,
+      "step": 6001
+    },
+    {
+      "epoch": 0.5267930365807165,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0016491517700393424,
+      "loss": 1.22,
+      "step": 6002
+    },
+    {
+      "epoch": 0.526880806163619,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0016487276551636594,
+      "loss": 1.1706,
+      "step": 6003
+    },
+    {
+      "epoch": 0.5269685757465215,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0016483035404135523,
+      "loss": 1.219,
+      "step": 6004
+    },
+    {
+      "epoch": 0.5270563453294239,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0016478794258308787,
+      "loss": 1.1574,
+      "step": 6005
+    },
+    {
+      "epoch": 0.5271441149123264,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0016474553114574969,
+      "loss": 1.1855,
+      "step": 6006
+    },
+    {
+      "epoch": 0.527231884495229,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0016470311973352666,
+      "loss": 1.1979,
+      "step": 6007
+    },
+    {
+      "epoch": 0.5273196540781314,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001646607083506045,
+      "loss": 1.1581,
+      "step": 6008
+    },
+    {
+      "epoch": 0.5274074236610339,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0016461829700116911,
+      "loss": 1.2173,
+      "step": 6009
+    },
+    {
+      "epoch": 0.5274951932439363,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0016457588568940629,
+      "loss": 1.1897,
+      "step": 6010
+    },
+    {
+      "epoch": 0.5275829628268388,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0016453347441950188,
+      "loss": 1.2252,
+      "step": 6011
+    },
+    {
+      "epoch": 0.5276707324097413,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0016449106319564167,
+      "loss": 1.2289,
+      "step": 6012
+    },
+    {
+      "epoch": 0.5277585019926438,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0016444865202201154,
+      "loss": 1.2322,
+      "step": 6013
+    },
+    {
+      "epoch": 0.5278462715755463,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0016440624090279723,
+      "loss": 1.1854,
+      "step": 6014
+    },
+    {
+      "epoch": 0.5279340411584488,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.001643638298421846,
+      "loss": 1.2341,
+      "step": 6015
+    },
+    {
+      "epoch": 0.5280218107413512,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0016432141884435943,
+      "loss": 1.164,
+      "step": 6016
+    },
+    {
+      "epoch": 0.5281095803242537,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.001642790079135076,
+      "loss": 1.2079,
+      "step": 6017
+    },
+    {
+      "epoch": 0.5281973499071563,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.001642365970538148,
+      "loss": 1.2044,
+      "step": 6018
+    },
+    {
+      "epoch": 0.5282851194900587,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0016419418626946688,
+      "loss": 1.1996,
+      "step": 6019
+    },
+    {
+      "epoch": 0.5283728890729612,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0016415177556464956,
+      "loss": 1.2534,
+      "step": 6020
+    },
+    {
+      "epoch": 0.5284606586558637,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0016410936494354867,
+      "loss": 1.2127,
+      "step": 6021
+    },
+    {
+      "epoch": 0.5285484282387661,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0016406695441034988,
+      "loss": 1.2311,
+      "step": 6022
+    },
+    {
+      "epoch": 0.5286361978216687,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0016402454396923904,
+      "loss": 1.1849,
+      "step": 6023
+    },
+    {
+      "epoch": 0.5287239674045712,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0016398213362440182,
+      "loss": 1.1507,
+      "step": 6024
+    },
+    {
+      "epoch": 0.5288117369874736,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0016393972338002404,
+      "loss": 1.1292,
+      "step": 6025
+    },
+    {
+      "epoch": 0.5288995065703761,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0016389731324029126,
+      "loss": 1.1801,
+      "step": 6026
+    },
+    {
+      "epoch": 0.5289872761532786,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0016385490320938937,
+      "loss": 1.1779,
+      "step": 6027
+    },
+    {
+      "epoch": 0.5290750457361811,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0016381249329150396,
+      "loss": 1.2283,
+      "step": 6028
+    },
+    {
+      "epoch": 0.5291628153190836,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0016377008349082085,
+      "loss": 1.1734,
+      "step": 6029
+    },
+    {
+      "epoch": 0.5292505849019861,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0016372767381152558,
+      "loss": 1.1435,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5293383544848885,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0016368526425780386,
+      "loss": 1.1593,
+      "step": 6031
+    },
+    {
+      "epoch": 0.529426124067791,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0016364285483384138,
+      "loss": 1.2591,
+      "step": 6032
+    },
+    {
+      "epoch": 0.5295138936506936,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0016360044554382378,
+      "loss": 1.1593,
+      "step": 6033
+    },
+    {
+      "epoch": 0.529601663233596,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0016355803639193662,
+      "loss": 1.1442,
+      "step": 6034
+    },
+    {
+      "epoch": 0.5296894328164985,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0016351562738236558,
+      "loss": 1.2212,
+      "step": 6035
+    },
+    {
+      "epoch": 0.529777202399401,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0016347321851929624,
+      "loss": 1.1617,
+      "step": 6036
+    },
+    {
+      "epoch": 0.5298649719823034,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0016343080980691415,
+      "loss": 1.1788,
+      "step": 6037
+    },
+    {
+      "epoch": 0.529952741565206,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0016338840124940496,
+      "loss": 1.1635,
+      "step": 6038
+    },
+    {
+      "epoch": 0.5300405111481085,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0016334599285095421,
+      "loss": 1.2065,
+      "step": 6039
+    },
+    {
+      "epoch": 0.5301282807310109,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001633035846157474,
+      "loss": 1.2287,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5302160503139134,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0016326117654797009,
+      "loss": 1.1796,
+      "step": 6041
+    },
+    {
+      "epoch": 0.5303038198968159,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0016321876865180776,
+      "loss": 1.1575,
+      "step": 6042
+    },
+    {
+      "epoch": 0.5303915894797183,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0016317636093144596,
+      "loss": 1.141,
+      "step": 6043
+    },
+    {
+      "epoch": 0.5304793590626209,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016313395339107007,
+      "loss": 1.2192,
+      "step": 6044
+    },
+    {
+      "epoch": 0.5305671286455234,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0016309154603486563,
+      "loss": 1.2191,
+      "step": 6045
+    },
+    {
+      "epoch": 0.5306548982284258,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0016304913886701798,
+      "loss": 1.1893,
+      "step": 6046
+    },
+    {
+      "epoch": 0.5307426678113283,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0016300673189171264,
+      "loss": 1.2169,
+      "step": 6047
+    },
+    {
+      "epoch": 0.5308304373942307,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0016296432511313502,
+      "loss": 1.1733,
+      "step": 6048
+    },
+    {
+      "epoch": 0.5309182069771333,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0016292191853547041,
+      "loss": 1.2319,
+      "step": 6049
+    },
+    {
+      "epoch": 0.5310059765600358,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0016287951216290423,
+      "loss": 1.2229,
+      "step": 6050
+    },
+    {
+      "epoch": 0.5310937461429383,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.001628371059996218,
+      "loss": 1.2475,
+      "step": 6051
+    },
+    {
+      "epoch": 0.5311815157258407,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0016279470004980845,
+      "loss": 1.1824,
+      "step": 6052
+    },
+    {
+      "epoch": 0.5312692853087432,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0016275229431764952,
+      "loss": 1.1825,
+      "step": 6053
+    },
+    {
+      "epoch": 0.5313570548916458,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.001627098888073302,
+      "loss": 1.2794,
+      "step": 6054
+    },
+    {
+      "epoch": 0.5314448244745482,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0016266748352303582,
+      "loss": 1.2049,
+      "step": 6055
+    },
+    {
+      "epoch": 0.5315325940574507,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0016262507846895155,
+      "loss": 1.1904,
+      "step": 6056
+    },
+    {
+      "epoch": 0.5316203636403531,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0016258267364926264,
+      "loss": 1.1975,
+      "step": 6057
+    },
+    {
+      "epoch": 0.5317081332232556,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0016254026906815432,
+      "loss": 1.1282,
+      "step": 6058
+    },
+    {
+      "epoch": 0.5317959028061582,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0016249786472981166,
+      "loss": 1.2039,
+      "step": 6059
+    },
+    {
+      "epoch": 0.5318836723890606,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001624554606384199,
+      "loss": 1.1977,
+      "step": 6060
+    },
+    {
+      "epoch": 0.5319714419719631,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.001624130567981641,
+      "loss": 1.2141,
+      "step": 6061
+    },
+    {
+      "epoch": 0.5320592115548656,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0016237065321322935,
+      "loss": 1.1862,
+      "step": 6062
+    },
+    {
+      "epoch": 0.532146981137768,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0016232824988780072,
+      "loss": 1.2319,
+      "step": 6063
+    },
+    {
+      "epoch": 0.5322347507206706,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001622858468260633,
+      "loss": 1.1583,
+      "step": 6064
+    },
+    {
+      "epoch": 0.5323225203035731,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0016224344403220198,
+      "loss": 1.1346,
+      "step": 6065
+    },
+    {
+      "epoch": 0.5324102898864755,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0016220104151040181,
+      "loss": 1.1386,
+      "step": 6066
+    },
+    {
+      "epoch": 0.532498059469378,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0016215863926484781,
+      "loss": 1.211,
+      "step": 6067
+    },
+    {
+      "epoch": 0.5325858290522805,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0016211623729972487,
+      "loss": 1.1694,
+      "step": 6068
+    },
+    {
+      "epoch": 0.5326735986351829,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001620738356192179,
+      "loss": 1.1874,
+      "step": 6069
+    },
+    {
+      "epoch": 0.5327613682180855,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0016203143422751183,
+      "loss": 1.1789,
+      "step": 6070
+    },
+    {
+      "epoch": 0.532849137800988,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.001619890331287914,
+      "loss": 1.1887,
+      "step": 6071
+    },
+    {
+      "epoch": 0.5329369073838904,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0016194663232724153,
+      "loss": 1.1331,
+      "step": 6072
+    },
+    {
+      "epoch": 0.5330246769667929,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001619042318270469,
+      "loss": 1.1714,
+      "step": 6073
+    },
+    {
+      "epoch": 0.5331124465496954,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0016186183163239241,
+      "loss": 1.1899,
+      "step": 6074
+    },
+    {
+      "epoch": 0.5332002161325979,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0016181943174746263,
+      "loss": 1.1801,
+      "step": 6075
+    },
+    {
+      "epoch": 0.5332879857155004,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.001617770321764424,
+      "loss": 1.1737,
+      "step": 6076
+    },
+    {
+      "epoch": 0.5333757552984029,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0016173463292351634,
+      "loss": 1.1919,
+      "step": 6077
+    },
+    {
+      "epoch": 0.5334635248813053,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0016169223399286911,
+      "loss": 1.1951,
+      "step": 6078
+    },
+    {
+      "epoch": 0.5335512944642078,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0016164983538868528,
+      "loss": 1.1619,
+      "step": 6079
+    },
+    {
+      "epoch": 0.5336390640471104,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0016160743711514944,
+      "loss": 1.1679,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5337268336300128,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0016156503917644615,
+      "loss": 1.1544,
+      "step": 6081
+    },
+    {
+      "epoch": 0.5338146032129153,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.001615226415767599,
+      "loss": 1.1356,
+      "step": 6082
+    },
+    {
+      "epoch": 0.5339023727958178,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0016148024432027514,
+      "loss": 1.2222,
+      "step": 6083
+    },
+    {
+      "epoch": 0.5339901423787202,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0016143784741117638,
+      "loss": 1.1654,
+      "step": 6084
+    },
+    {
+      "epoch": 0.5340779119616228,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0016139545085364795,
+      "loss": 1.1409,
+      "step": 6085
+    },
+    {
+      "epoch": 0.5341656815445253,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0016135305465187423,
+      "loss": 1.2248,
+      "step": 6086
+    },
+    {
+      "epoch": 0.5342534511274277,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0016131065881003965,
+      "loss": 1.246,
+      "step": 6087
+    },
+    {
+      "epoch": 0.5343412207103302,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0016126826333232843,
+      "loss": 1.2109,
+      "step": 6088
+    },
+    {
+      "epoch": 0.5344289902932327,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0016122586822292482,
+      "loss": 1.1569,
+      "step": 6089
+    },
+    {
+      "epoch": 0.5345167598761352,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0016118347348601312,
+      "loss": 1.211,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5346045294590377,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0016114107912577748,
+      "loss": 1.1709,
+      "step": 6091
+    },
+    {
+      "epoch": 0.5346922990419402,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0016109868514640204,
+      "loss": 1.1282,
+      "step": 6092
+    },
+    {
+      "epoch": 0.5347800686248426,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0016105629155207099,
+      "loss": 1.1549,
+      "step": 6093
+    },
+    {
+      "epoch": 0.5348678382077451,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0016101389834696832,
+      "loss": 1.2166,
+      "step": 6094
+    },
+    {
+      "epoch": 0.5349556077906475,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.001609715055352781,
+      "loss": 1.2201,
+      "step": 6095
+    },
+    {
+      "epoch": 0.5350433773735501,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0016092911312118434,
+      "loss": 1.2744,
+      "step": 6096
+    },
+    {
+      "epoch": 0.5351311469564526,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0016088672110887102,
+      "loss": 1.1993,
+      "step": 6097
+    },
+    {
+      "epoch": 0.535218916539355,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0016084432950252206,
+      "loss": 1.17,
+      "step": 6098
+    },
+    {
+      "epoch": 0.5353066861222575,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0016080193830632133,
+      "loss": 1.2072,
+      "step": 6099
+    },
+    {
+      "epoch": 0.53539445570516,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.001607595475244527,
+      "loss": 1.2142,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5354822252880626,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0016071715716109993,
+      "loss": 1.1578,
+      "step": 6101
+    },
+    {
+      "epoch": 0.535569994870965,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001606747672204468,
+      "loss": 1.2198,
+      "step": 6102
+    },
+    {
+      "epoch": 0.5356577644538675,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00160632377706677,
+      "loss": 1.1644,
+      "step": 6103
+    },
+    {
+      "epoch": 0.5357455340367699,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001605899886239743,
+      "loss": 1.1905,
+      "step": 6104
+    },
+    {
+      "epoch": 0.5358333036196724,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0016054759997652225,
+      "loss": 1.1883,
+      "step": 6105
+    },
+    {
+      "epoch": 0.535921073202575,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0016050521176850445,
+      "loss": 1.1763,
+      "step": 6106
+    },
+    {
+      "epoch": 0.5360088427854774,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0016046282400410449,
+      "loss": 1.2033,
+      "step": 6107
+    },
+    {
+      "epoch": 0.5360966123683799,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.001604204366875058,
+      "loss": 1.2325,
+      "step": 6108
+    },
+    {
+      "epoch": 0.5361843819512824,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0016037804982289192,
+      "loss": 1.1775,
+      "step": 6109
+    },
+    {
+      "epoch": 0.5362721515341848,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0016033566341444624,
+      "loss": 1.2191,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5363599211170874,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0016029327746635215,
+      "loss": 1.1974,
+      "step": 6111
+    },
+    {
+      "epoch": 0.5364476906999899,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.001602508919827929,
+      "loss": 1.2442,
+      "step": 6112
+    },
+    {
+      "epoch": 0.5365354602828923,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.001602085069679519,
+      "loss": 1.2152,
+      "step": 6113
+    },
+    {
+      "epoch": 0.5366232298657948,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.001601661224260122,
+      "loss": 1.1253,
+      "step": 6114
+    },
+    {
+      "epoch": 0.5367109994486973,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0016012373836115715,
+      "loss": 1.2086,
+      "step": 6115
+    },
+    {
+      "epoch": 0.5367987690315998,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0016008135477756976,
+      "loss": 1.1453,
+      "step": 6116
+    },
+    {
+      "epoch": 0.5368865386145023,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0016003897167943327,
+      "loss": 1.1747,
+      "step": 6117
+    },
+    {
+      "epoch": 0.5369743081974048,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001599965890709306,
+      "loss": 1.138,
+      "step": 6118
+    },
+    {
+      "epoch": 0.5370620777803072,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0015995420695624485,
+      "loss": 1.1821,
+      "step": 6119
+    },
+    {
+      "epoch": 0.5371498473632097,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0015991182533955884,
+      "loss": 1.1671,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5372376169461123,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0015986944422505557,
+      "loss": 1.1383,
+      "step": 6121
+    },
+    {
+      "epoch": 0.5373253865290147,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0015982706361691787,
+      "loss": 1.1574,
+      "step": 6122
+    },
+    {
+      "epoch": 0.5374131561119172,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0015978468351932852,
+      "loss": 1.1801,
+      "step": 6123
+    },
+    {
+      "epoch": 0.5375009256948197,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0015974230393647024,
+      "loss": 1.1668,
+      "step": 6124
+    },
+    {
+      "epoch": 0.5375886952777221,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0015969992487252578,
+      "loss": 1.1966,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5376764648606246,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0015965754633167776,
+      "loss": 1.1818,
+      "step": 6126
+    },
+    {
+      "epoch": 0.5377642344435272,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0015961516831810873,
+      "loss": 1.2285,
+      "step": 6127
+    },
+    {
+      "epoch": 0.5378520040264296,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0015957279083600132,
+      "loss": 1.2416,
+      "step": 6128
+    },
+    {
+      "epoch": 0.5379397736093321,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0015953041388953799,
+      "loss": 1.1802,
+      "step": 6129
+    },
+    {
+      "epoch": 0.5380275431922346,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0015948803748290117,
+      "loss": 1.2045,
+      "step": 6130
+    },
+    {
+      "epoch": 0.538115312775137,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0015944566162027322,
+      "loss": 1.2394,
+      "step": 6131
+    },
+    {
+      "epoch": 0.5382030823580396,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001594032863058365,
+      "loss": 1.1509,
+      "step": 6132
+    },
+    {
+      "epoch": 0.5382908519409421,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0015936091154377331,
+      "loss": 1.1608,
+      "step": 6133
+    },
+    {
+      "epoch": 0.5383786215238445,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0015931853733826577,
+      "loss": 1.1442,
+      "step": 6134
+    },
+    {
+      "epoch": 0.538466391106747,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0015927616369349619,
+      "loss": 1.1767,
+      "step": 6135
+    },
+    {
+      "epoch": 0.5385541606896495,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0015923379061364657,
+      "loss": 1.129,
+      "step": 6136
+    },
+    {
+      "epoch": 0.538641930272552,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0015919141810289894,
+      "loss": 1.1244,
+      "step": 6137
+    },
+    {
+      "epoch": 0.5387296998554545,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0015914904616543544,
+      "loss": 1.1825,
+      "step": 6138
+    },
+    {
+      "epoch": 0.538817469438357,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.001591066748054379,
+      "loss": 1.1911,
+      "step": 6139
+    },
+    {
+      "epoch": 0.5389052390212594,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0015906430402708826,
+      "loss": 1.1746,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5389930086041619,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.001590219338345683,
+      "loss": 1.1888,
+      "step": 6141
+    },
+    {
+      "epoch": 0.5390807781870645,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0015897956423205982,
+      "loss": 1.1428,
+      "step": 6142
+    },
+    {
+      "epoch": 0.5391685477699669,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0015893719522374456,
+      "loss": 1.1885,
+      "step": 6143
+    },
+    {
+      "epoch": 0.5392563173528694,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001588948268138041,
+      "loss": 1.2071,
+      "step": 6144
+    },
+    {
+      "epoch": 0.5393440869357718,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0015885245900642015,
+      "loss": 1.101,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5394318565186743,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0015881009180577408,
+      "loss": 1.2732,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5395196261015769,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0015876772521604746,
+      "loss": 1.1889,
+      "step": 6147
+    },
+    {
+      "epoch": 0.5396073956844794,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0015872535924142176,
+      "loss": 1.1763,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5396951652673818,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0015868299388607824,
+      "loss": 1.1695,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5397829348502843,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0015864062915419827,
+      "loss": 1.1694,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5398707044331867,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00158598265049963,
+      "loss": 1.195,
+      "step": 6151
+    },
+    {
+      "epoch": 0.5399584740160892,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0015855590157755367,
+      "loss": 1.2528,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5400462435989918,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0015851353874115131,
+      "loss": 1.2017,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5401340131818942,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00158471176544937,
+      "loss": 1.1419,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5402217827647967,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.001584288149930918,
+      "loss": 1.1515,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5403095523476992,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0015838645408979655,
+      "loss": 1.2526,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5403973219306016,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0015834409383923202,
+      "loss": 1.163,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5404850915135042,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001583017342455792,
+      "loss": 1.1863,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5405728610964067,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0015825937531301864,
+      "loss": 1.2369,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5406606306793091,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0015821701704573114,
+      "loss": 1.2365,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5407484002622116,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0015817465944789715,
+      "loss": 1.1972,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5408361698451141,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001581323025236973,
+      "loss": 1.1861,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5409239394280166,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0015808994627731201,
+      "loss": 1.1823,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5410117090109191,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001580475907129217,
+      "loss": 1.1791,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5410994785938216,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0015800523583470662,
+      "loss": 1.2117,
+      "step": 6165
+    },
+    {
+      "epoch": 0.541187248176724,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0015796288164684714,
+      "loss": 1.1939,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5412750177596265,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.001579205281535233,
+      "loss": 1.204,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5413627873425291,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0015787817535891544,
+      "loss": 1.2298,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5414505569254315,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0015783582326720342,
+      "loss": 1.1736,
+      "step": 6169
+    },
+    {
+      "epoch": 0.541538326508334,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0015779347188256735,
+      "loss": 1.2245,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5416260960912365,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0015775112120918705,
+      "loss": 1.2119,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5417138656741389,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0015770877125124245,
+      "loss": 1.186,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5418016352570415,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0015766642201291321,
+      "loss": 1.1395,
+      "step": 6173
+    },
+    {
+      "epoch": 0.541889404839944,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0015762407349837917,
+      "loss": 1.1755,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5419771744228464,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0015758172571181985,
+      "loss": 1.2679,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5420649440057489,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001575393786574149,
+      "loss": 1.1841,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5421527135886514,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0015749703233934366,
+      "loss": 1.2218,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5422404831715539,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.001574546867617857,
+      "loss": 1.2051,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5423282527544564,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0015741234192892025,
+      "loss": 1.1963,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5424160223373589,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.001573699978449267,
+      "loss": 1.1433,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5425037919202613,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0015732765451398414,
+      "loss": 1.1547,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5425915615031638,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0015728531194027175,
+      "loss": 1.1695,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5426793310860663,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0015724297012796851,
+      "loss": 1.1888,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5427671006689688,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0015720062908125344,
+      "loss": 1.2037,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5428548702518713,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001571582888043054,
+      "loss": 1.2179,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5429426398347738,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0015711594930130329,
+      "loss": 1.1709,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5430304094176762,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.001570736105764257,
+      "loss": 1.165,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5431181790005787,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0015703127263385142,
+      "loss": 1.2589,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5432059485834813,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00156988935477759,
+      "loss": 1.1939,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5432937181663837,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0015694659911232695,
+      "loss": 1.1714,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5433814877492862,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0015690426354173373,
+      "loss": 1.1842,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5434692573321886,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0015686192877015764,
+      "loss": 1.2717,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5435570269150911,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0015681959480177698,
+      "loss": 1.1981,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5436447964979937,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0015677726164076995,
+      "loss": 1.2296,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5437325660808962,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0015673492929131466,
+      "loss": 1.177,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5438203356637986,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0015669259775758918,
+      "loss": 1.1738,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5439081052467011,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0015665026704377134,
+      "loss": 1.2296,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5439958748296035,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0015660793715403915,
+      "loss": 1.2486,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5440836444125061,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0015656560809257038,
+      "loss": 1.159,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5441714139954086,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.001565232798635427,
+      "loss": 1.1759,
+      "step": 6200
+    },
+    {
+      "epoch": 0.544259183578311,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0015648095247113379,
+      "loss": 1.1651,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5443469531612135,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.001564386259195211,
+      "loss": 1.1666,
+      "step": 6202
+    },
+    {
+      "epoch": 0.544434722744116,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0015639630021288223,
+      "loss": 1.1602,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5445224923270185,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0015635397535539444,
+      "loss": 1.1816,
+      "step": 6204
+    },
+    {
+      "epoch": 0.544610261909921,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0015631165135123507,
+      "loss": 1.206,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5446980314928235,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0015626932820458136,
+      "loss": 1.2081,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5447858010757259,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0015622700591961032,
+      "loss": 1.2119,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5448735706586284,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0015618468450049917,
+      "loss": 1.1715,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5449613402415309,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0015614236395142477,
+      "loss": 1.1457,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5450491098244334,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0015610004427656398,
+      "loss": 1.2001,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5451368794073359,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0015605772548009365,
+      "loss": 1.182,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5452246489902384,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.001560154075661904,
+      "loss": 1.1821,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5453124185731408,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0015597309053903088,
+      "loss": 1.1429,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5454001881560433,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0015593077440279161,
+      "loss": 1.1407,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5454879577389459,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0015588845916164901,
+      "loss": 1.1378,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5455757273218483,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0015584614481977943,
+      "loss": 1.2322,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5456634969047508,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.001558038313813591,
+      "loss": 1.1874,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5457512664876533,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0015576151885056428,
+      "loss": 1.1681,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5458390360705557,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00155719207231571,
+      "loss": 1.1887,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5459268056534583,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.001556768965285552,
+      "loss": 1.1835,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5460145752363608,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001556345867456929,
+      "loss": 1.2334,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5461023448192632,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0015559227788715976,
+      "loss": 1.1801,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5461901144021657,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.001555499699571316,
+      "loss": 1.2215,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5462778839850682,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00155507662959784,
+      "loss": 1.2031,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5463656535679707,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0015546535689929253,
+      "loss": 1.218,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5464534231508732,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.001554230517798326,
+      "loss": 1.1867,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5465411927337757,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0015538074760557952,
+      "loss": 1.2693,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5466289623166781,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0015533844438070862,
+      "loss": 1.1918,
+      "step": 6228
+    },
+    {
+      "epoch": 0.5467167318995806,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0015529614210939503,
+      "loss": 1.2335,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5468045014824832,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0015525384079581382,
+      "loss": 1.1971,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5468922710653856,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0015521154044413998,
+      "loss": 1.1628,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5469800406482881,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0015516924105854835,
+      "loss": 1.1257,
+      "step": 6232
+    },
+    {
+      "epoch": 0.5470678102311906,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0015512694264321373,
+      "loss": 1.1927,
+      "step": 6233
+    },
+    {
+      "epoch": 0.547155579814093,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0015508464520231078,
+      "loss": 1.1375,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5472433493969955,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.001550423487400142,
+      "loss": 1.1333,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5473311189798981,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001550000532604983,
+      "loss": 1.2687,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5474188885628005,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0015495775876793758,
+      "loss": 1.1878,
+      "step": 6237
+    },
+    {
+      "epoch": 0.547506658145703,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0015491546526650638,
+      "loss": 1.1995,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5475944277286054,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0015487317276037882,
+      "loss": 1.2238,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5476821973115079,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.001548308812537291,
+      "loss": 1.2511,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5477699668944105,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0015478859075073112,
+      "loss": 1.1773,
+      "step": 6241
+    },
+    {
+      "epoch": 0.547857736477313,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0015474630125555885,
+      "loss": 1.1703,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5479455060602154,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0015470401277238605,
+      "loss": 1.2217,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5480332756431179,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0015466172530538645,
+      "loss": 1.1443,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5481210452260203,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0015461943885873367,
+      "loss": 1.2284,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5482088148089229,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.001545771534366012,
+      "loss": 1.1882,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5482965843918254,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001545348690431624,
+      "loss": 1.1761,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5483843539747278,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0015449258568259068,
+      "loss": 1.1903,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5484721235576303,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0015445030335905915,
+      "loss": 1.1807,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5485598931405328,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0015440802207674098,
+      "loss": 1.1887,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5486476627234353,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0015436574183980904,
+      "loss": 1.1804,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5487354323063378,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0015432346265243635,
+      "loss": 1.2673,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5488232018892403,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0015428118451879566,
+      "loss": 1.1062,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5489109714721427,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0015423890744305966,
+      "loss": 1.1398,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5489987410550452,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0015419663142940084,
+      "loss": 1.1402,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5490865106379478,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0015415435648199179,
+      "loss": 1.2258,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5491742802208502,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0015411208260500487,
+      "loss": 1.2467,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5492620498037527,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0015406980980261224,
+      "loss": 1.1528,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5493498193866552,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0015402753807898621,
+      "loss": 1.1582,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5494375889695576,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0015398526743829874,
+      "loss": 1.1994,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5495253585524602,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0015394299788472176,
+      "loss": 1.2556,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5496131281353627,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.001539007294224272,
+      "loss": 1.1867,
+      "step": 6262
+    },
+    {
+      "epoch": 0.5497008977182651,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0015385846205558669,
+      "loss": 1.1549,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5497886673011676,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.001538161957883719,
+      "loss": 1.1977,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5498764368840701,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.001537739306249543,
+      "loss": 1.1624,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5499642064669725,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001537316665695054,
+      "loss": 1.1715,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5500519760498751,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0015368940362619635,
+      "loss": 1.2069,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5501397456327776,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0015364714179919839,
+      "loss": 1.2612,
+      "step": 6268
+    },
+    {
+      "epoch": 0.55022751521568,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0015360488109268267,
+      "loss": 1.1255,
+      "step": 6269
+    },
+    {
+      "epoch": 0.5503152847985825,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001535626215108201,
+      "loss": 1.1788,
+      "step": 6270
+    },
+    {
+      "epoch": 0.550403054381485,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001535203630577815,
+      "loss": 1.211,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5504908239643875,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0015347810573773768,
+      "loss": 1.2242,
+      "step": 6272
+    },
+    {
+      "epoch": 0.55057859354729,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0015343584955485922,
+      "loss": 1.1567,
+      "step": 6273
+    },
+    {
+      "epoch": 0.5506663631301925,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0015339359451331666,
+      "loss": 1.1769,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5507541327130949,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0015335134061728035,
+      "loss": 1.2169,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5508419022959974,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0015330908787092072,
+      "loss": 1.1668,
+      "step": 6276
+    },
+    {
+      "epoch": 0.5509296718789,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0015326683627840778,
+      "loss": 1.2519,
+      "step": 6277
+    },
+    {
+      "epoch": 0.5510174414618024,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0015322458584391167,
+      "loss": 1.1894,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5511052110447049,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0015318233657160235,
+      "loss": 1.1694,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5511929806276074,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001531400884656497,
+      "loss": 1.179,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5512807502105098,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.001530978415302233,
+      "loss": 1.2106,
+      "step": 6281
+    },
+    {
+      "epoch": 0.5513685197934124,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0015305559576949297,
+      "loss": 1.2047,
+      "step": 6282
+    },
+    {
+      "epoch": 0.5514562893763149,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00153013351187628,
+      "loss": 1.1516,
+      "step": 6283
+    },
+    {
+      "epoch": 0.5515440589592173,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0015297110778879786,
+      "loss": 1.2292,
+      "step": 6284
+    },
+    {
+      "epoch": 0.5516318285421198,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0015292886557717177,
+      "loss": 1.1481,
+      "step": 6285
+    },
+    {
+      "epoch": 0.5517195981250222,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0015288662455691892,
+      "loss": 1.1705,
+      "step": 6286
+    },
+    {
+      "epoch": 0.5518073677079248,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0015284438473220825,
+      "loss": 1.1854,
+      "step": 6287
+    },
+    {
+      "epoch": 0.5518951372908273,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0015280214610720868,
+      "loss": 1.1233,
+      "step": 6288
+    },
+    {
+      "epoch": 0.5519829068737298,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001527599086860891,
+      "loss": 1.1375,
+      "step": 6289
+    },
+    {
+      "epoch": 0.5520706764566322,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0015271767247301805,
+      "loss": 1.242,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5521584460395347,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0015267543747216417,
+      "loss": 1.1905,
+      "step": 6291
+    },
+    {
+      "epoch": 0.5522462156224371,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0015263320368769583,
+      "loss": 1.2025,
+      "step": 6292
+    },
+    {
+      "epoch": 0.5523339852053397,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0015259097112378136,
+      "loss": 1.1838,
+      "step": 6293
+    },
+    {
+      "epoch": 0.5524217547882422,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0015254873978458893,
+      "loss": 1.1738,
+      "step": 6294
+    },
+    {
+      "epoch": 0.5525095243711446,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001525065096742866,
+      "loss": 1.2225,
+      "step": 6295
+    },
+    {
+      "epoch": 0.5525972939540471,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0015246428079704237,
+      "loss": 1.2323,
+      "step": 6296
+    },
+    {
+      "epoch": 0.5526850635369496,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00152422053157024,
+      "loss": 1.114,
+      "step": 6297
+    },
+    {
+      "epoch": 0.5527728331198521,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0015237982675839917,
+      "loss": 1.1743,
+      "step": 6298
+    },
+    {
+      "epoch": 0.5528606027027546,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0015233760160533558,
+      "loss": 1.2142,
+      "step": 6299
+    },
+    {
+      "epoch": 0.5529483722856571,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0015229537770200057,
+      "loss": 1.1613,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5530361418685595,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0015225315505256153,
+      "loss": 1.1502,
+      "step": 6301
+    },
+    {
+      "epoch": 0.553123911451462,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0015221093366118562,
+      "loss": 1.1992,
+      "step": 6302
+    },
+    {
+      "epoch": 0.5532116810343646,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0015216871353204,
+      "loss": 1.1488,
+      "step": 6303
+    },
+    {
+      "epoch": 0.553299450617267,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0015212649466929153,
+      "loss": 1.2005,
+      "step": 6304
+    },
+    {
+      "epoch": 0.5533872202001695,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0015208427707710716,
+      "loss": 1.2089,
+      "step": 6305
+    },
+    {
+      "epoch": 0.553474989783072,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.001520420607596535,
+      "loss": 1.1906,
+      "step": 6306
+    },
+    {
+      "epoch": 0.5535627593659744,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0015199984572109716,
+      "loss": 1.2153,
+      "step": 6307
+    },
+    {
+      "epoch": 0.553650528948877,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001519576319656046,
+      "loss": 1.2561,
+      "step": 6308
+    },
+    {
+      "epoch": 0.5537382985317795,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.001519154194973422,
+      "loss": 1.1624,
+      "step": 6309
+    },
+    {
+      "epoch": 0.5538260681146819,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0015187320832047609,
+      "loss": 1.2236,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5539138376975844,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.001518309984391724,
+      "loss": 1.2115,
+      "step": 6311
+    },
+    {
+      "epoch": 0.5540016072804869,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0015178878985759705,
+      "loss": 1.1762,
+      "step": 6312
+    },
+    {
+      "epoch": 0.5540893768633894,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0015174658257991588,
+      "loss": 1.1615,
+      "step": 6313
+    },
+    {
+      "epoch": 0.5541771464462919,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0015170437661029457,
+      "loss": 1.1731,
+      "step": 6314
+    },
+    {
+      "epoch": 0.5542649160291944,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0015166217195289866,
+      "loss": 1.231,
+      "step": 6315
+    },
+    {
+      "epoch": 0.5543526856120968,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0015161996861189362,
+      "loss": 1.2212,
+      "step": 6316
+    },
+    {
+      "epoch": 0.5544404551949993,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0015157776659144473,
+      "loss": 1.2218,
+      "step": 6317
+    },
+    {
+      "epoch": 0.5545282247779019,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0015153556589571713,
+      "loss": 1.1686,
+      "step": 6318
+    },
+    {
+      "epoch": 0.5546159943608043,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0015149336652887592,
+      "loss": 1.2171,
+      "step": 6319
+    },
+    {
+      "epoch": 0.5547037639437068,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0015145116849508598,
+      "loss": 1.2342,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5547915335266093,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0015140897179851214,
+      "loss": 1.1894,
+      "step": 6321
+    },
+    {
+      "epoch": 0.5548793031095117,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0015136677644331897,
+      "loss": 1.1455,
+      "step": 6322
+    },
+    {
+      "epoch": 0.5549670726924142,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0015132458243367107,
+      "loss": 1.1692,
+      "step": 6323
+    },
+    {
+      "epoch": 0.5550548422753168,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0015128238977373273,
+      "loss": 1.1627,
+      "step": 6324
+    },
+    {
+      "epoch": 0.5551426118582192,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0015124019846766827,
+      "loss": 1.1449,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5552303814411217,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0015119800851964179,
+      "loss": 1.151,
+      "step": 6326
+    },
+    {
+      "epoch": 0.5553181510240242,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0015115581993381721,
+      "loss": 1.1691,
+      "step": 6327
+    },
+    {
+      "epoch": 0.5554059206069266,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.001511136327143584,
+      "loss": 1.1054,
+      "step": 6328
+    },
+    {
+      "epoch": 0.5554936901898292,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0015107144686542916,
+      "loss": 1.149,
+      "step": 6329
+    },
+    {
+      "epoch": 0.5555814597727317,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0015102926239119298,
+      "loss": 1.1962,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5556692293556341,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0015098707929581331,
+      "loss": 1.1762,
+      "step": 6331
+    },
+    {
+      "epoch": 0.5557569989385366,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001509448975834535,
+      "loss": 1.2402,
+      "step": 6332
+    },
+    {
+      "epoch": 0.555844768521439,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0015090271725827665,
+      "loss": 1.2026,
+      "step": 6333
+    },
+    {
+      "epoch": 0.5559325381043416,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0015086053832444582,
+      "loss": 1.1607,
+      "step": 6334
+    },
+    {
+      "epoch": 0.5560203076872441,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0015081836078612398,
+      "loss": 1.1899,
+      "step": 6335
+    },
+    {
+      "epoch": 0.5561080772701466,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0015077618464747373,
+      "loss": 1.1947,
+      "step": 6336
+    },
+    {
+      "epoch": 0.556195846853049,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0015073400991265785,
+      "loss": 1.1513,
+      "step": 6337
+    },
+    {
+      "epoch": 0.5562836164359515,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0015069183658583867,
+      "loss": 1.1301,
+      "step": 6338
+    },
+    {
+      "epoch": 0.556371386018854,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0015064966467117863,
+      "loss": 1.1245,
+      "step": 6339
+    },
+    {
+      "epoch": 0.5564591556017565,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0015060749417283993,
+      "loss": 1.2306,
+      "step": 6340
+    },
+    {
+      "epoch": 0.556546925184659,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0015056532509498458,
+      "loss": 1.2147,
+      "step": 6341
+    },
+    {
+      "epoch": 0.5566346947675614,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0015052315744177455,
+      "loss": 1.166,
+      "step": 6342
+    },
+    {
+      "epoch": 0.5567224643504639,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001504809912173716,
+      "loss": 1.184,
+      "step": 6343
+    },
+    {
+      "epoch": 0.5568102339333665,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0015043882642593737,
+      "loss": 1.208,
+      "step": 6344
+    },
+    {
+      "epoch": 0.556898003516269,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0015039666307163334,
+      "loss": 1.1931,
+      "step": 6345
+    },
+    {
+      "epoch": 0.5569857730991714,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0015035450115862086,
+      "loss": 1.211,
+      "step": 6346
+    },
+    {
+      "epoch": 0.5570735426820739,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001503123406910612,
+      "loss": 1.1789,
+      "step": 6347
+    },
+    {
+      "epoch": 0.5571613122649763,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0015027018167311536,
+      "loss": 1.2013,
+      "step": 6348
+    },
+    {
+      "epoch": 0.5572490818478788,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0015022802410894435,
+      "loss": 1.2266,
+      "step": 6349
+    },
+    {
+      "epoch": 0.5573368514307814,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001501858680027089,
+      "loss": 1.14,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5574246210136838,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0015014371335856965,
+      "loss": 1.2315,
+      "step": 6351
+    },
+    {
+      "epoch": 0.5575123905965863,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0015010156018068713,
+      "loss": 1.1814,
+      "step": 6352
+    },
+    {
+      "epoch": 0.5576001601794888,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0015005940847322167,
+      "loss": 1.2159,
+      "step": 6353
+    },
+    {
+      "epoch": 0.5576879297623912,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0015001725824033347,
+      "loss": 1.208,
+      "step": 6354
+    },
+    {
+      "epoch": 0.5577756993452938,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.001499751094861826,
+      "loss": 1.1614,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5578634689281963,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0014993296221492895,
+      "loss": 1.2138,
+      "step": 6356
+    },
+    {
+      "epoch": 0.5579512385110987,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0014989081643073234,
+      "loss": 1.1572,
+      "step": 6357
+    },
+    {
+      "epoch": 0.5580390080940012,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0014984867213775229,
+      "loss": 1.1705,
+      "step": 6358
+    },
+    {
+      "epoch": 0.5581267776769037,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0014980652934014841,
+      "loss": 1.2178,
+      "step": 6359
+    },
+    {
+      "epoch": 0.5582145472598062,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0014976438804208,
+      "loss": 1.1886,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5583023168427087,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0014972224824770614,
+      "loss": 1.1841,
+      "step": 6361
+    },
+    {
+      "epoch": 0.5583900864256112,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0014968010996118596,
+      "loss": 1.1725,
+      "step": 6362
+    },
+    {
+      "epoch": 0.5584778560085136,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0014963797318667833,
+      "loss": 1.1493,
+      "step": 6363
+    },
+    {
+      "epoch": 0.5585656255914161,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0014959583792834197,
+      "loss": 1.1721,
+      "step": 6364
+    },
+    {
+      "epoch": 0.5586533951743187,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0014955370419033537,
+      "loss": 1.2012,
+      "step": 6365
+    },
+    {
+      "epoch": 0.5587411647572211,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0014951157197681715,
+      "loss": 1.1987,
+      "step": 6366
+    },
+    {
+      "epoch": 0.5588289343401236,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0014946944129194545,
+      "loss": 1.1729,
+      "step": 6367
+    },
+    {
+      "epoch": 0.5589167039230261,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001494273121398784,
+      "loss": 1.1587,
+      "step": 6368
+    },
+    {
+      "epoch": 0.5590044735059285,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0014938518452477408,
+      "loss": 1.1279,
+      "step": 6369
+    },
+    {
+      "epoch": 0.5590922430888311,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0014934305845079024,
+      "loss": 1.2322,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5591800126717336,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0014930093392208454,
+      "loss": 1.1739,
+      "step": 6371
+    },
+    {
+      "epoch": 0.559267782254636,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.001492588109428146,
+      "loss": 1.2434,
+      "step": 6372
+    },
+    {
+      "epoch": 0.5593555518375385,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0014921668951713774,
+      "loss": 1.2339,
+      "step": 6373
+    },
+    {
+      "epoch": 0.559443321420441,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0014917456964921119,
+      "loss": 1.1908,
+      "step": 6374
+    },
+    {
+      "epoch": 0.5595310910033434,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0014913245134319194,
+      "loss": 1.1935,
+      "step": 6375
+    },
+    {
+      "epoch": 0.559618860586246,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0014909033460323699,
+      "loss": 1.1983,
+      "step": 6376
+    },
+    {
+      "epoch": 0.5597066301691485,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0014904821943350307,
+      "loss": 1.2365,
+      "step": 6377
+    },
+    {
+      "epoch": 0.5597943997520509,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0014900610583814676,
+      "loss": 1.1666,
+      "step": 6378
+    },
+    {
+      "epoch": 0.5598821693349534,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0014896399382132448,
+      "loss": 1.1454,
+      "step": 6379
+    },
+    {
+      "epoch": 0.5599699389178558,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.001489218833871926,
+      "loss": 1.2563,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5600577085007584,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001488797745399072,
+      "loss": 1.2089,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5601454780836609,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0014883766728362427,
+      "loss": 1.1981,
+      "step": 6382
+    },
+    {
+      "epoch": 0.5602332476665633,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0014879556162249964,
+      "loss": 1.2586,
+      "step": 6383
+    },
+    {
+      "epoch": 0.5603210172494658,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0014875345756068896,
+      "loss": 1.23,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5604087868323683,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0014871135510234775,
+      "loss": 1.181,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5604965564152709,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0014866925425163133,
+      "loss": 1.1506,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5605843259981733,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0014862715501269487,
+      "loss": 1.1312,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5606720955810758,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0014858505738969345,
+      "loss": 1.2007,
+      "step": 6388
+    },
+    {
+      "epoch": 0.5607598651639782,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001485429613867819,
+      "loss": 1.1438,
+      "step": 6389
+    },
+    {
+      "epoch": 0.5608476347468807,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0014850086700811493,
+      "loss": 1.2252,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5609354043297833,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0014845877425784713,
+      "loss": 1.1821,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5610231739126857,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0014841668314013285,
+      "loss": 1.1809,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5611109434955882,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0014837459365912636,
+      "loss": 1.1603,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5611987130784907,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0014833250581898164,
+      "loss": 1.1524,
+      "step": 6394
+    },
+    {
+      "epoch": 0.5612864826613931,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.001482904196238527,
+      "loss": 1.2611,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5613742522442957,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0014824833507789319,
+      "loss": 1.1781,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5614620218271982,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0014820625218525672,
+      "loss": 1.1606,
+      "step": 6397
+    },
+    {
+      "epoch": 0.5615497914101006,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0014816417095009674,
+      "loss": 1.2038,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5616375609930031,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0014812209137656644,
+      "loss": 1.1749,
+      "step": 6399
+    },
+    {
+      "epoch": 0.5617253305759056,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001480800134688189,
+      "loss": 1.227,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5618131001588081,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0014803793723100716,
+      "loss": 1.2232,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5619008697417106,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0014799586266728386,
+      "loss": 1.2032,
+      "step": 6402
+    },
+    {
+      "epoch": 0.5619886393246131,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0014795378978180167,
+      "loss": 1.1449,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5620764089075155,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0014791171857871297,
+      "loss": 1.1698,
+      "step": 6404
+    },
+    {
+      "epoch": 0.562164178490418,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0014786964906217,
+      "loss": 1.2111,
+      "step": 6405
+    },
+    {
+      "epoch": 0.5622519480733205,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0014782758123632492,
+      "loss": 1.1343,
+      "step": 6406
+    },
+    {
+      "epoch": 0.562339717656223,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0014778551510532963,
+      "loss": 1.2124,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5624274872391255,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0014774345067333585,
+      "loss": 1.1831,
+      "step": 6408
+    },
+    {
+      "epoch": 0.562515256822028,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0014770138794449519,
+      "loss": 1.2364,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5626030264049304,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0014765932692295909,
+      "loss": 1.2397,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5626907959878329,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0014761726761287888,
+      "loss": 1.1474,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5627785655707355,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0014757521001840554,
+      "loss": 1.1746,
+      "step": 6412
+    },
+    {
+      "epoch": 0.5628663351536379,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0014753315414369004,
+      "loss": 1.2141,
+      "step": 6413
+    },
+    {
+      "epoch": 0.5629541047365404,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.001474910999928831,
+      "loss": 1.1941,
+      "step": 6414
+    },
+    {
+      "epoch": 0.5630418743194429,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.001474490475701353,
+      "loss": 1.2217,
+      "step": 6415
+    },
+    {
+      "epoch": 0.5631296439023453,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0014740699687959707,
+      "loss": 1.1363,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5632174134852479,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0014736494792541865,
+      "loss": 1.131,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5633051830681504,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0014732290071175005,
+      "loss": 1.1526,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5633929526510528,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.001472808552427412,
+      "loss": 1.2008,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5634807222339553,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0014723881152254187,
+      "loss": 1.1515,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5635684918168578,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0014719676955530154,
+      "loss": 1.2366,
+      "step": 6421
+    },
+    {
+      "epoch": 0.5636562613997603,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0014715472934516965,
+      "loss": 1.1734,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5637440309826628,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0014711269089629532,
+      "loss": 1.1645,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5638318005655653,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0014707065421282763,
+      "loss": 1.2254,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5639195701484677,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0014702861929891547,
+      "loss": 1.2054,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5640073397313702,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0014698658615870741,
+      "loss": 1.2024,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5640951093142728,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001469445547963521,
+      "loss": 1.2008,
+      "step": 6427
+    },
+    {
+      "epoch": 0.5641828788971752,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0014690252521599772,
+      "loss": 1.1722,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5642706484800777,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0014686049742179247,
+      "loss": 1.1765,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5643584180629801,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.001468184714178844,
+      "loss": 1.1895,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5644461876458826,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0014677644720842125,
+      "loss": 1.1583,
+      "step": 6431
+    },
+    {
+      "epoch": 0.5645339572287851,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0014673442479755069,
+      "loss": 1.146,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5646217268116877,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0014669240418942009,
+      "loss": 1.1055,
+      "step": 6433
+    },
+    {
+      "epoch": 0.5647094963945901,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0014665038538817675,
+      "loss": 1.1772,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5647972659774926,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.001466083683979678,
+      "loss": 1.2191,
+      "step": 6435
+    },
+    {
+      "epoch": 0.564885035560395,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0014656635322294008,
+      "loss": 1.1748,
+      "step": 6436
+    },
+    {
+      "epoch": 0.5649728051432975,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0014652433986724039,
+      "loss": 1.1954,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5650605747262001,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0014648232833501524,
+      "loss": 1.1562,
+      "step": 6438
+    },
+    {
+      "epoch": 0.5651483443091025,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00146440318630411,
+      "loss": 1.1923,
+      "step": 6439
+    },
+    {
+      "epoch": 0.565236113892005,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001463983107575739,
+      "loss": 1.1443,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5653238834749075,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0014635630472064987,
+      "loss": 1.1385,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5654116530578099,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0014631430052378486,
+      "loss": 1.1702,
+      "step": 6442
+    },
+    {
+      "epoch": 0.5654994226407125,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0014627229817112443,
+      "loss": 1.1109,
+      "step": 6443
+    },
+    {
+      "epoch": 0.565587192223615,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0014623029766681404,
+      "loss": 1.2068,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5656749618065174,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0014618829901499903,
+      "loss": 1.1934,
+      "step": 6445
+    },
+    {
+      "epoch": 0.5657627313894199,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.001461463022198245,
+      "loss": 1.1864,
+      "step": 6446
+    },
+    {
+      "epoch": 0.5658505009723224,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001461043072854353,
+      "loss": 1.2161,
+      "step": 6447
+    },
+    {
+      "epoch": 0.5659382705552249,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001460623142159762,
+      "loss": 1.1406,
+      "step": 6448
+    },
+    {
+      "epoch": 0.5660260401381274,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0014602032301559175,
+      "loss": 1.1179,
+      "step": 6449
+    },
+    {
+      "epoch": 0.5661138097210299,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0014597833368842636,
+      "loss": 1.1474,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5662015793039323,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0014593634623862413,
+      "loss": 1.1589,
+      "step": 6451
+    },
+    {
+      "epoch": 0.5662893488868348,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0014589436067032912,
+      "loss": 1.1562,
+      "step": 6452
+    },
+    {
+      "epoch": 0.5663771184697374,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.001458523769876851,
+      "loss": 1.1687,
+      "step": 6453
+    },
+    {
+      "epoch": 0.5664648880526398,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0014581039519483573,
+      "loss": 1.1775,
+      "step": 6454
+    },
+    {
+      "epoch": 0.5665526576355423,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0014576841529592441,
+      "loss": 1.1994,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5666404272184448,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0014572643729509442,
+      "loss": 1.2083,
+      "step": 6456
+    },
+    {
+      "epoch": 0.5667281968013472,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001456844611964888,
+      "loss": 1.2278,
+      "step": 6457
+    },
+    {
+      "epoch": 0.5668159663842498,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0014564248700425046,
+      "loss": 1.1976,
+      "step": 6458
+    },
+    {
+      "epoch": 0.5669037359671523,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0014560051472252192,
+      "loss": 1.1578,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5669915055500547,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0014555854435544596,
+      "loss": 1.1847,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5670792751329572,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001455165759071647,
+      "loss": 1.1117,
+      "step": 6461
+    },
+    {
+      "epoch": 0.5671670447158597,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0014547460938182032,
+      "loss": 1.2152,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5672548142987621,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0014543264478355474,
+      "loss": 1.2081,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5673425838816647,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.001453906821165097,
+      "loss": 1.1085,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5674303534645672,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.001453487213848267,
+      "loss": 1.2243,
+      "step": 6465
+    },
+    {
+      "epoch": 0.5675181230474696,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0014530676259264723,
+      "loss": 1.1297,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5676058926303721,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0014526480574411228,
+      "loss": 1.1804,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5676936622132746,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0014522285084336297,
+      "loss": 1.1649,
+      "step": 6468
+    },
+    {
+      "epoch": 0.5677814317961771,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0014518089789454,
+      "loss": 1.1792,
+      "step": 6469
+    },
+    {
+      "epoch": 0.5678692013790796,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0014513894690178395,
+      "loss": 1.123,
+      "step": 6470
+    },
+    {
+      "epoch": 0.567956970961982,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0014509699786923532,
+      "loss": 1.2228,
+      "step": 6471
+    },
+    {
+      "epoch": 0.5680447405448845,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0014505505080103422,
+      "loss": 1.1719,
+      "step": 6472
+    },
+    {
+      "epoch": 0.568132510127787,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0014501310570132068,
+      "loss": 1.1645,
+      "step": 6473
+    },
+    {
+      "epoch": 0.5682202797106896,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0014497116257423453,
+      "loss": 1.1813,
+      "step": 6474
+    },
+    {
+      "epoch": 0.568308049293592,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001449292214239154,
+      "loss": 1.2241,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5683958188764945,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0014488728225450272,
+      "loss": 1.2724,
+      "step": 6476
+    },
+    {
+      "epoch": 0.568483588459397,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0014484534507013564,
+      "loss": 1.1721,
+      "step": 6477
+    },
+    {
+      "epoch": 0.5685713580422994,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001448034098749533,
+      "loss": 1.2187,
+      "step": 6478
+    },
+    {
+      "epoch": 0.568659127625202,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0014476147667309444,
+      "loss": 1.1692,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5687468972081045,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0014471954546869778,
+      "loss": 1.2007,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5688346667910069,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0014467761626590179,
+      "loss": 1.1827,
+      "step": 6481
+    },
+    {
+      "epoch": 0.5689224363739094,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0014463568906884457,
+      "loss": 1.1506,
+      "step": 6482
+    },
+    {
+      "epoch": 0.5690102059568118,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0014459376388166435,
+      "loss": 1.2087,
+      "step": 6483
+    },
+    {
+      "epoch": 0.5690979755397144,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0014455184070849884,
+      "loss": 1.1423,
+      "step": 6484
+    },
+    {
+      "epoch": 0.5691857451226169,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0014450991955348572,
+      "loss": 1.153,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5692735147055193,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0014446800042076252,
+      "loss": 1.1855,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5693612842884218,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001444260833144664,
+      "loss": 1.1741,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5694490538713243,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.001443841682387345,
+      "loss": 1.2187,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5695368234542267,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0014434225519770357,
+      "loss": 1.1736,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5696245930371293,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.001443003441955103,
+      "loss": 1.2535,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5697123626200318,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0014425843523629119,
+      "loss": 1.1821,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5698001322029342,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0014421652832418246,
+      "loss": 1.174,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5698879017858367,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0014417462346332014,
+      "loss": 1.1698,
+      "step": 6493
+    },
+    {
+      "epoch": 0.5699756713687392,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0014413272065784007,
+      "loss": 1.2362,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5700634409516417,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001440908199118779,
+      "loss": 1.1245,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5701512105345442,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001440489212295691,
+      "loss": 1.1902,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5702389801174467,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0014400702461504888,
+      "loss": 1.1254,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5703267497003491,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0014396513007245224,
+      "loss": 1.1656,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5704145192832516,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0014392323760591401,
+      "loss": 1.1992,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5705022888661542,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001438813472195688,
+      "loss": 1.1148,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5705022888661542,
+      "eval_loss": 1.1722602844238281,
+      "eval_runtime": 437.7292,
+      "eval_samples_per_second": 33.674,
+      "eval_steps_per_second": 8.418,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5705900584490566,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0014383945891755113,
+      "loss": 1.187,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5706778280319591,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0014379757270399512,
+      "loss": 1.1278,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5707655976148616,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001437556885830348,
+      "loss": 1.1504,
+      "step": 6503
+    },
+    {
+      "epoch": 0.570853367197764,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0014371380655880397,
+      "loss": 1.1369,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5709411367806666,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0014367192663543619,
+      "loss": 1.1784,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5710289063635691,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0014363004881706484,
+      "loss": 1.2018,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5711166759464715,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0014358817310782318,
+      "loss": 1.188,
+      "step": 6507
+    },
+    {
+      "epoch": 0.571204445529374,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00143546299511844,
+      "loss": 1.1936,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5712922151122765,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0014350442803326027,
+      "loss": 1.1837,
+      "step": 6509
+    },
+    {
+      "epoch": 0.571379984695179,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0014346255867620434,
+      "loss": 1.1955,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5714677542780815,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0014342069144480873,
+      "loss": 1.195,
+      "step": 6511
+    },
+    {
+      "epoch": 0.571555523860984,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0014337882634320545,
+      "loss": 1.1242,
+      "step": 6512
+    },
+    {
+      "epoch": 0.5716432934438864,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.001433369633755265,
+      "loss": 1.2227,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5717310630267889,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001432951025459035,
+      "loss": 1.1563,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5718188326096915,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0014325324385846805,
+      "loss": 1.1595,
+      "step": 6515
+    },
+    {
+      "epoch": 0.5719066021925939,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0014321138731735136,
+      "loss": 1.2171,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5719943717754964,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0014316953292668453,
+      "loss": 1.2011,
+      "step": 6517
+    },
+    {
+      "epoch": 0.5720821413583989,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0014312768069059843,
+      "loss": 1.1918,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5721699109413013,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.001430858306132237,
+      "loss": 1.1962,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5722576805242038,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.001430439826986907,
+      "loss": 1.1804,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5723454501071064,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0014300213695112981,
+      "loss": 1.198,
+      "step": 6521
+    },
+    {
+      "epoch": 0.5724332196900088,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0014296029337467096,
+      "loss": 1.158,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5725209892729113,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0014291845197344393,
+      "loss": 1.2386,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5726087588558137,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0014287661275157835,
+      "loss": 1.2123,
+      "step": 6524
+    },
+    {
+      "epoch": 0.5726965284387162,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0014283477571320352,
+      "loss": 1.2376,
+      "step": 6525
+    },
+    {
+      "epoch": 0.5727842980216188,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0014279294086244863,
+      "loss": 1.1893,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5728720676045213,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0014275110820344261,
+      "loss": 1.1707,
+      "step": 6527
+    },
+    {
+      "epoch": 0.5729598371874237,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0014270927774031419,
+      "loss": 1.2202,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5730476067703262,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0014266744947719185,
+      "loss": 1.1544,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5731353763532286,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0014262562341820384,
+      "loss": 1.1812,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5732231459361312,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.001425837995674783,
+      "loss": 1.1873,
+      "step": 6531
+    },
+    {
+      "epoch": 0.5733109155190337,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0014254197792914307,
+      "loss": 1.1328,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5733986851019361,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0014250015850732574,
+      "loss": 1.2488,
+      "step": 6533
+    },
+    {
+      "epoch": 0.5734864546848386,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0014245834130615373,
+      "loss": 1.1951,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5735742242677411,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0014241652632975422,
+      "loss": 1.0851,
+      "step": 6535
+    },
+    {
+      "epoch": 0.5736619938506436,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0014237471358225426,
+      "loss": 1.2339,
+      "step": 6536
+    },
+    {
+      "epoch": 0.5737497634335461,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0014233290306778052,
+      "loss": 1.1986,
+      "step": 6537
+    },
+    {
+      "epoch": 0.5738375330164486,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0014229109479045954,
+      "loss": 1.1767,
+      "step": 6538
+    },
+    {
+      "epoch": 0.573925302599351,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0014224928875441766,
+      "loss": 1.1588,
+      "step": 6539
+    },
+    {
+      "epoch": 0.5740130721822535,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0014220748496378095,
+      "loss": 1.1822,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5741008417651561,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0014216568342267529,
+      "loss": 1.1228,
+      "step": 6541
+    },
+    {
+      "epoch": 0.5741886113480585,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0014212388413522633,
+      "loss": 1.2022,
+      "step": 6542
+    },
+    {
+      "epoch": 0.574276380930961,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0014208208710555948,
+      "loss": 1.1496,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5743641505138635,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0014204029233779998,
+      "loss": 1.174,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5744519200967659,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0014199849983607276,
+      "loss": 1.235,
+      "step": 6545
+    },
+    {
+      "epoch": 0.5745396896796684,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0014195670960450261,
+      "loss": 1.1539,
+      "step": 6546
+    },
+    {
+      "epoch": 0.574627459262571,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0014191492164721403,
+      "loss": 1.1596,
+      "step": 6547
+    },
+    {
+      "epoch": 0.5747152288454734,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0014187313596833134,
+      "loss": 1.1661,
+      "step": 6548
+    },
+    {
+      "epoch": 0.5748029984283759,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0014183135257197863,
+      "loss": 1.2002,
+      "step": 6549
+    },
+    {
+      "epoch": 0.5748907680112784,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0014178957146227969,
+      "loss": 1.1735,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5749785375941808,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0014174779264335827,
+      "loss": 1.155,
+      "step": 6551
+    },
+    {
+      "epoch": 0.5750663071770834,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0014170601611933772,
+      "loss": 1.1719,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5751540767599859,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0014166424189434117,
+      "loss": 1.1812,
+      "step": 6553
+    },
+    {
+      "epoch": 0.5752418463428883,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0014162246997249167,
+      "loss": 1.1904,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5753296159257908,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0014158070035791184,
+      "loss": 1.1334,
+      "step": 6555
+    },
+    {
+      "epoch": 0.5754173855086933,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0014153893305472423,
+      "loss": 1.1606,
+      "step": 6556
+    },
+    {
+      "epoch": 0.5755051550915958,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0014149716806705107,
+      "loss": 1.1896,
+      "step": 6557
+    },
+    {
+      "epoch": 0.5755929246744983,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0014145540539901448,
+      "loss": 1.1903,
+      "step": 6558
+    },
+    {
+      "epoch": 0.5756806942574008,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0014141364505473617,
+      "loss": 1.1873,
+      "step": 6559
+    },
+    {
+      "epoch": 0.5757684638403032,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0014137188703833774,
+      "loss": 1.2167,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5758562334232057,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0014133013135394059,
+      "loss": 1.1179,
+      "step": 6561
+    },
+    {
+      "epoch": 0.5759440030061083,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0014128837800566586,
+      "loss": 1.2041,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5760317725890107,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0014124662699763436,
+      "loss": 1.1914,
+      "step": 6563
+    },
+    {
+      "epoch": 0.5761195421719132,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0014120487833396683,
+      "loss": 1.2151,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5762073117548157,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.001411631320187836,
+      "loss": 1.1646,
+      "step": 6565
+    },
+    {
+      "epoch": 0.5762950813377181,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0014112138805620496,
+      "loss": 1.2635,
+      "step": 6566
+    },
+    {
+      "epoch": 0.5763828509206207,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0014107964645035083,
+      "loss": 1.1508,
+      "step": 6567
+    },
+    {
+      "epoch": 0.5764706205035232,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0014103790720534095,
+      "loss": 1.1969,
+      "step": 6568
+    },
+    {
+      "epoch": 0.5765583900864256,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.001409961703252948,
+      "loss": 1.1657,
+      "step": 6569
+    },
+    {
+      "epoch": 0.5766461596693281,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0014095443581433162,
+      "loss": 1.2155,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5767339292522305,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0014091270367657055,
+      "loss": 1.187,
+      "step": 6571
+    },
+    {
+      "epoch": 0.576821698835133,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0014087097391613029,
+      "loss": 1.2271,
+      "step": 6572
+    },
+    {
+      "epoch": 0.5769094684180356,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0014082924653712945,
+      "loss": 1.1939,
+      "step": 6573
+    },
+    {
+      "epoch": 0.576997238000938,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.001407875215436863,
+      "loss": 1.1633,
+      "step": 6574
+    },
+    {
+      "epoch": 0.5770850075838405,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0014074579893991902,
+      "loss": 1.1841,
+      "step": 6575
+    },
+    {
+      "epoch": 0.577172777166743,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0014070407872994542,
+      "loss": 1.175,
+      "step": 6576
+    },
+    {
+      "epoch": 0.5772605467496454,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0014066236091788305,
+      "loss": 1.2139,
+      "step": 6577
+    },
+    {
+      "epoch": 0.577348316332548,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.001406206455078494,
+      "loss": 1.1748,
+      "step": 6578
+    },
+    {
+      "epoch": 0.5774360859154505,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0014057893250396158,
+      "loss": 1.1616,
+      "step": 6579
+    },
+    {
+      "epoch": 0.577523855498353,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0014053722191033644,
+      "loss": 1.1502,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5776116250812554,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.001404955137310908,
+      "loss": 1.261,
+      "step": 6581
+    },
+    {
+      "epoch": 0.5776993946641579,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0014045380797034091,
+      "loss": 1.1689,
+      "step": 6582
+    },
+    {
+      "epoch": 0.5777871642470604,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0014041210463220308,
+      "loss": 1.187,
+      "step": 6583
+    },
+    {
+      "epoch": 0.5778749338299629,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0014037040372079322,
+      "loss": 1.1499,
+      "step": 6584
+    },
+    {
+      "epoch": 0.5779627034128654,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0014032870524022711,
+      "loss": 1.1781,
+      "step": 6585
+    },
+    {
+      "epoch": 0.5780504729957678,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0014028700919462012,
+      "loss": 1.2142,
+      "step": 6586
+    },
+    {
+      "epoch": 0.5781382425786703,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0014024531558808758,
+      "loss": 1.16,
+      "step": 6587
+    },
+    {
+      "epoch": 0.5782260121615729,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0014020362442474438,
+      "loss": 1.1964,
+      "step": 6588
+    },
+    {
+      "epoch": 0.5783137817444753,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0014016193570870537,
+      "loss": 1.2038,
+      "step": 6589
+    },
+    {
+      "epoch": 0.5784015513273778,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0014012024944408499,
+      "loss": 1.1674,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5784893209102803,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001400785656349976,
+      "loss": 1.1208,
+      "step": 6591
+    },
+    {
+      "epoch": 0.5785770904931827,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0014003688428555715,
+      "loss": 1.1391,
+      "step": 6592
+    },
+    {
+      "epoch": 0.5786648600760853,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0013999520539987747,
+      "loss": 1.1053,
+      "step": 6593
+    },
+    {
+      "epoch": 0.5787526296589878,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0013995352898207205,
+      "loss": 1.1811,
+      "step": 6594
+    },
+    {
+      "epoch": 0.5788403992418902,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0013991185503625422,
+      "loss": 1.1793,
+      "step": 6595
+    },
+    {
+      "epoch": 0.5789281688247927,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0013987018356653705,
+      "loss": 1.1928,
+      "step": 6596
+    },
+    {
+      "epoch": 0.5790159384076952,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0013982851457703333,
+      "loss": 1.2151,
+      "step": 6597
+    },
+    {
+      "epoch": 0.5791037079905977,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0013978684807185561,
+      "loss": 1.22,
+      "step": 6598
+    },
+    {
+      "epoch": 0.5791914775735002,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0013974518405511622,
+      "loss": 1.2016,
+      "step": 6599
+    },
+    {
+      "epoch": 0.5792792471564027,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0013970352253092715,
+      "loss": 1.1353,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5793670167393051,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0013966186350340044,
+      "loss": 1.1568,
+      "step": 6601
+    },
+    {
+      "epoch": 0.5794547863222076,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0013962020697664747,
+      "loss": 1.1843,
+      "step": 6602
+    },
+    {
+      "epoch": 0.5795425559051101,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.001395785529547797,
+      "loss": 1.1162,
+      "step": 6603
+    },
+    {
+      "epoch": 0.5796303254880126,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0013953690144190814,
+      "loss": 1.1776,
+      "step": 6604
+    },
+    {
+      "epoch": 0.5797180950709151,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.001394952524421437,
+      "loss": 1.2004,
+      "step": 6605
+    },
+    {
+      "epoch": 0.5798058646538176,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0013945360595959686,
+      "loss": 1.1723,
+      "step": 6606
+    },
+    {
+      "epoch": 0.57989363423672,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0013941196199837808,
+      "loss": 1.191,
+      "step": 6607
+    },
+    {
+      "epoch": 0.5799814038196225,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0013937032056259733,
+      "loss": 1.2217,
+      "step": 6608
+    },
+    {
+      "epoch": 0.5800691734025251,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0013932868165636462,
+      "loss": 1.1823,
+      "step": 6609
+    },
+    {
+      "epoch": 0.5801569429854275,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0013928704528378937,
+      "loss": 1.2024,
+      "step": 6610
+    },
+    {
+      "epoch": 0.58024471256833,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0013924541144898104,
+      "loss": 1.2094,
+      "step": 6611
+    },
+    {
+      "epoch": 0.5803324821512325,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.001392037801560487,
+      "loss": 1.1543,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5804202517341349,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.001391621514091012,
+      "loss": 1.2433,
+      "step": 6613
+    },
+    {
+      "epoch": 0.5805080213170375,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0013912052521224713,
+      "loss": 1.2453,
+      "step": 6614
+    },
+    {
+      "epoch": 0.58059579089994,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0013907890156959481,
+      "loss": 1.2506,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5806835604828424,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0013903728048525231,
+      "loss": 1.1493,
+      "step": 6616
+    },
+    {
+      "epoch": 0.5807713300657449,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0013899566196332756,
+      "loss": 1.1852,
+      "step": 6617
+    },
+    {
+      "epoch": 0.5808590996486473,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0013895404600792805,
+      "loss": 1.1716,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5809468692315499,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0013891243262316114,
+      "loss": 1.1808,
+      "step": 6619
+    },
+    {
+      "epoch": 0.5810346388144524,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0013887082181313387,
+      "loss": 1.1694,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5811224083973549,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0013882921358195311,
+      "loss": 1.2075,
+      "step": 6621
+    },
+    {
+      "epoch": 0.5812101779802573,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0013878760793372548,
+      "loss": 1.142,
+      "step": 6622
+    },
+    {
+      "epoch": 0.5812979475631598,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0013874600487255718,
+      "loss": 1.2186,
+      "step": 6623
+    },
+    {
+      "epoch": 0.5813857171460624,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0013870440440255432,
+      "loss": 1.1511,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5814734867289648,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.001386628065278227,
+      "loss": 1.1749,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5815612563118673,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0013862121125246785,
+      "loss": 1.2003,
+      "step": 6626
+    },
+    {
+      "epoch": 0.5816490258947697,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0013857961858059513,
+      "loss": 1.2252,
+      "step": 6627
+    },
+    {
+      "epoch": 0.5817367954776722,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0013853802851630946,
+      "loss": 1.1847,
+      "step": 6628
+    },
+    {
+      "epoch": 0.5818245650605747,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.001384964410637157,
+      "loss": 1.1741,
+      "step": 6629
+    },
+    {
+      "epoch": 0.5819123346434772,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.001384548562269183,
+      "loss": 1.2139,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5820001042263797,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0013841327401002156,
+      "loss": 1.2527,
+      "step": 6631
+    },
+    {
+      "epoch": 0.5820878738092822,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0013837169441712953,
+      "loss": 1.2631,
+      "step": 6632
+    },
+    {
+      "epoch": 0.5821756433921846,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0013833011745234588,
+      "loss": 1.2431,
+      "step": 6633
+    },
+    {
+      "epoch": 0.5822634129750871,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.001382885431197741,
+      "loss": 1.1732,
+      "step": 6634
+    },
+    {
+      "epoch": 0.5823511825579897,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001382469714235174,
+      "loss": 1.1991,
+      "step": 6635
+    },
+    {
+      "epoch": 0.5824389521408921,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0013820540236767881,
+      "loss": 1.207,
+      "step": 6636
+    },
+    {
+      "epoch": 0.5825267217237946,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0013816383595636093,
+      "loss": 1.181,
+      "step": 6637
+    },
+    {
+      "epoch": 0.5826144913066971,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0013812227219366633,
+      "loss": 1.1459,
+      "step": 6638
+    },
+    {
+      "epoch": 0.5827022608895995,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0013808071108369704,
+      "loss": 1.2064,
+      "step": 6639
+    },
+    {
+      "epoch": 0.5827900304725021,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0013803915263055507,
+      "loss": 1.1676,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5828778000554046,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0013799759683834202,
+      "loss": 1.2055,
+      "step": 6641
+    },
+    {
+      "epoch": 0.582965569638307,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.001379560437111594,
+      "loss": 1.1583,
+      "step": 6642
+    },
+    {
+      "epoch": 0.5830533392212095,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0013791449325310822,
+      "loss": 1.1908,
+      "step": 6643
+    },
+    {
+      "epoch": 0.583141108804112,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0013787294546828942,
+      "loss": 1.1405,
+      "step": 6644
+    },
+    {
+      "epoch": 0.5832288783870145,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.001378314003608035,
+      "loss": 1.1627,
+      "step": 6645
+    },
+    {
+      "epoch": 0.583316647969917,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0013778985793475092,
+      "loss": 1.1758,
+      "step": 6646
+    },
+    {
+      "epoch": 0.5834044175528195,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0013774831819423168,
+      "loss": 1.1631,
+      "step": 6647
+    },
+    {
+      "epoch": 0.5834921871357219,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.001377067811433456,
+      "loss": 1.2112,
+      "step": 6648
+    },
+    {
+      "epoch": 0.5835799567186244,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0013766524678619225,
+      "loss": 1.1301,
+      "step": 6649
+    },
+    {
+      "epoch": 0.583667726301527,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0013762371512687087,
+      "loss": 1.2078,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5837554958844294,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0013758218616948042,
+      "loss": 1.1655,
+      "step": 6651
+    },
+    {
+      "epoch": 0.5838432654673319,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0013754065991811975,
+      "loss": 1.1804,
+      "step": 6652
+    },
+    {
+      "epoch": 0.5839310350502344,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0013749913637688732,
+      "loss": 1.2317,
+      "step": 6653
+    },
+    {
+      "epoch": 0.5840188046331368,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0013745761554988132,
+      "loss": 1.1907,
+      "step": 6654
+    },
+    {
+      "epoch": 0.5841065742160394,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0013741609744119968,
+      "loss": 1.1411,
+      "step": 6655
+    },
+    {
+      "epoch": 0.5841943437989419,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0013737458205494006,
+      "loss": 1.1553,
+      "step": 6656
+    },
+    {
+      "epoch": 0.5842821133818443,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0013733306939519988,
+      "loss": 1.172,
+      "step": 6657
+    },
+    {
+      "epoch": 0.5843698829647468,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.001372915594660763,
+      "loss": 1.2116,
+      "step": 6658
+    },
+    {
+      "epoch": 0.5844576525476493,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0013725005227166614,
+      "loss": 1.1522,
+      "step": 6659
+    },
+    {
+      "epoch": 0.5845454221305517,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0013720854781606607,
+      "loss": 1.1345,
+      "step": 6660
+    },
+    {
+      "epoch": 0.5846331917134543,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0013716704610337229,
+      "loss": 1.239,
+      "step": 6661
+    },
+    {
+      "epoch": 0.5847209612963568,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0013712554713768096,
+      "loss": 1.1829,
+      "step": 6662
+    },
+    {
+      "epoch": 0.5848087308792592,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0013708405092308783,
+      "loss": 1.1787,
+      "step": 6663
+    },
+    {
+      "epoch": 0.5848965004621617,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0013704255746368845,
+      "loss": 1.1498,
+      "step": 6664
+    },
+    {
+      "epoch": 0.5849842700450641,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0013700106676357798,
+      "loss": 1.2055,
+      "step": 6665
+    },
+    {
+      "epoch": 0.5850720396279667,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0013695957882685148,
+      "loss": 1.127,
+      "step": 6666
+    },
+    {
+      "epoch": 0.5851598092108692,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0013691809365760357,
+      "loss": 1.1633,
+      "step": 6667
+    },
+    {
+      "epoch": 0.5852475787937717,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0013687661125992873,
+      "loss": 1.1587,
+      "step": 6668
+    },
+    {
+      "epoch": 0.5853353483766741,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0013683513163792104,
+      "loss": 1.1445,
+      "step": 6669
+    },
+    {
+      "epoch": 0.5854231179595766,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0013679365479567446,
+      "loss": 1.2474,
+      "step": 6670
+    },
+    {
+      "epoch": 0.5855108875424792,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0013675218073728249,
+      "loss": 1.1838,
+      "step": 6671
+    },
+    {
+      "epoch": 0.5855986571253816,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0013671070946683851,
+      "loss": 1.2034,
+      "step": 6672
+    },
+    {
+      "epoch": 0.5856864267082841,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0013666924098843564,
+      "loss": 1.0969,
+      "step": 6673
+    },
+    {
+      "epoch": 0.5857741962911865,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0013662777530616653,
+      "loss": 1.2302,
+      "step": 6674
+    },
+    {
+      "epoch": 0.585861965874089,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0013658631242412377,
+      "loss": 1.1761,
+      "step": 6675
+    },
+    {
+      "epoch": 0.5859497354569916,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001365448523463995,
+      "loss": 1.176,
+      "step": 6676
+    },
+    {
+      "epoch": 0.586037505039894,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0013650339507708573,
+      "loss": 1.1412,
+      "step": 6677
+    },
+    {
+      "epoch": 0.5861252746227965,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0013646194062027414,
+      "loss": 1.1821,
+      "step": 6678
+    },
+    {
+      "epoch": 0.586213044205699,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0013642048898005609,
+      "loss": 1.1851,
+      "step": 6679
+    },
+    {
+      "epoch": 0.5863008137886014,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0013637904016052266,
+      "loss": 1.1822,
+      "step": 6680
+    },
+    {
+      "epoch": 0.586388583371504,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0013633759416576468,
+      "loss": 1.1519,
+      "step": 6681
+    },
+    {
+      "epoch": 0.5864763529544065,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.001362961509998728,
+      "loss": 1.2006,
+      "step": 6682
+    },
+    {
+      "epoch": 0.5865641225373089,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0013625471066693725,
+      "loss": 1.1577,
+      "step": 6683
+    },
+    {
+      "epoch": 0.5866518921202114,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0013621327317104798,
+      "loss": 1.154,
+      "step": 6684
+    },
+    {
+      "epoch": 0.5867396617031139,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0013617183851629482,
+      "loss": 1.202,
+      "step": 6685
+    },
+    {
+      "epoch": 0.5868274312860163,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0013613040670676707,
+      "loss": 1.2113,
+      "step": 6686
+    },
+    {
+      "epoch": 0.5869152008689189,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0013608897774655398,
+      "loss": 1.1577,
+      "step": 6687
+    },
+    {
+      "epoch": 0.5870029704518214,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0013604755163974433,
+      "loss": 1.1819,
+      "step": 6688
+    },
+    {
+      "epoch": 0.5870907400347238,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0013600612839042687,
+      "loss": 1.1824,
+      "step": 6689
+    },
+    {
+      "epoch": 0.5871785096176263,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0013596470800268975,
+      "loss": 1.224,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5872662792005288,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00135923290480621,
+      "loss": 1.1574,
+      "step": 6691
+    },
+    {
+      "epoch": 0.5873540487834313,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0013588187582830851,
+      "loss": 1.11,
+      "step": 6692
+    },
+    {
+      "epoch": 0.5874418183663338,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0013584046404983964,
+      "loss": 1.1514,
+      "step": 6693
+    },
+    {
+      "epoch": 0.5875295879492363,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0013579905514930163,
+      "loss": 1.2048,
+      "step": 6694
+    },
+    {
+      "epoch": 0.5876173575321387,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001357576491307813,
+      "loss": 1.2195,
+      "step": 6695
+    },
+    {
+      "epoch": 0.5877051271150412,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001357162459983653,
+      "loss": 1.2168,
+      "step": 6696
+    },
+    {
+      "epoch": 0.5877928966979438,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0013567484575613996,
+      "loss": 1.1229,
+      "step": 6697
+    },
+    {
+      "epoch": 0.5878806662808462,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0013563344840819133,
+      "loss": 1.25,
+      "step": 6698
+    },
+    {
+      "epoch": 0.5879684358637487,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0013559205395860508,
+      "loss": 1.1802,
+      "step": 6699
+    },
+    {
+      "epoch": 0.5880562054466512,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0013555066241146682,
+      "loss": 1.1616,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5881439750295536,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0013550927377086153,
+      "loss": 1.1636,
+      "step": 6701
+    },
+    {
+      "epoch": 0.5882317446124562,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0013546788804087433,
+      "loss": 1.1682,
+      "step": 6702
+    },
+    {
+      "epoch": 0.5883195141953587,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0013542650522558972,
+      "loss": 1.2086,
+      "step": 6703
+    },
+    {
+      "epoch": 0.5884072837782611,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0013538512532909198,
+      "loss": 1.1996,
+      "step": 6704
+    },
+    {
+      "epoch": 0.5884950533611636,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0013534374835546528,
+      "loss": 1.169,
+      "step": 6705
+    },
+    {
+      "epoch": 0.588582822944066,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001353023743087932,
+      "loss": 1.1751,
+      "step": 6706
+    },
+    {
+      "epoch": 0.5886705925269686,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.001352610031931593,
+      "loss": 1.1603,
+      "step": 6707
+    },
+    {
+      "epoch": 0.5887583621098711,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0013521963501264668,
+      "loss": 1.1373,
+      "step": 6708
+    },
+    {
+      "epoch": 0.5888461316927736,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0013517826977133833,
+      "loss": 1.2153,
+      "step": 6709
+    },
+    {
+      "epoch": 0.588933901275676,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0013513690747331668,
+      "loss": 1.1295,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5890216708585785,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0013509554812266406,
+      "loss": 1.1597,
+      "step": 6711
+    },
+    {
+      "epoch": 0.589109440441481,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0013505419172346257,
+      "loss": 1.2045,
+      "step": 6712
+    },
+    {
+      "epoch": 0.5891972100243835,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0013501283827979389,
+      "loss": 1.1683,
+      "step": 6713
+    },
+    {
+      "epoch": 0.589284979607286,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001349714877957394,
+      "loss": 1.1695,
+      "step": 6714
+    },
+    {
+      "epoch": 0.5893727491901884,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0013493014027538023,
+      "loss": 1.1957,
+      "step": 6715
+    },
+    {
+      "epoch": 0.5894605187730909,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0013488879572279727,
+      "loss": 1.2131,
+      "step": 6716
+    },
+    {
+      "epoch": 0.5895482883559934,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00134847454142071,
+      "loss": 1.1656,
+      "step": 6717
+    },
+    {
+      "epoch": 0.589636057938896,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0013480611553728168,
+      "loss": 1.1887,
+      "step": 6718
+    },
+    {
+      "epoch": 0.5897238275217984,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0013476477991250936,
+      "loss": 1.1974,
+      "step": 6719
+    },
+    {
+      "epoch": 0.5898115971047009,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0013472344727183357,
+      "loss": 1.18,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5898993666876033,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.001346821176193337,
+      "loss": 1.105,
+      "step": 6721
+    },
+    {
+      "epoch": 0.5899871362705058,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0013464079095908891,
+      "loss": 1.1723,
+      "step": 6722
+    },
+    {
+      "epoch": 0.5900749058534084,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001345994672951779,
+      "loss": 1.16,
+      "step": 6723
+    },
+    {
+      "epoch": 0.5901626754363108,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0013455814663167925,
+      "loss": 1.1991,
+      "step": 6724
+    },
+    {
+      "epoch": 0.5902504450192133,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0013451682897267102,
+      "loss": 1.1902,
+      "step": 6725
+    },
+    {
+      "epoch": 0.5903382146021158,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0013447551432223116,
+      "loss": 1.2524,
+      "step": 6726
+    },
+    {
+      "epoch": 0.5904259841850182,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0013443420268443725,
+      "loss": 1.2334,
+      "step": 6727
+    },
+    {
+      "epoch": 0.5905137537679208,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001343928940633666,
+      "loss": 1.1831,
+      "step": 6728
+    },
+    {
+      "epoch": 0.5906015233508233,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0013435158846309622,
+      "loss": 1.2259,
+      "step": 6729
+    },
+    {
+      "epoch": 0.5906892929337257,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0013431028588770278,
+      "loss": 1.1402,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5907770625166282,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0013426898634126263,
+      "loss": 1.1307,
+      "step": 6731
+    },
+    {
+      "epoch": 0.5908648320995307,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0013422768982785196,
+      "loss": 1.2026,
+      "step": 6732
+    },
+    {
+      "epoch": 0.5909526016824332,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0013418639635154658,
+      "loss": 1.128,
+      "step": 6733
+    },
+    {
+      "epoch": 0.5910403712653357,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0013414510591642195,
+      "loss": 1.1794,
+      "step": 6734
+    },
+    {
+      "epoch": 0.5911281408482382,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0013410381852655329,
+      "loss": 1.2261,
+      "step": 6735
+    },
+    {
+      "epoch": 0.5912159104311406,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0013406253418601546,
+      "loss": 1.1806,
+      "step": 6736
+    },
+    {
+      "epoch": 0.5913036800140431,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001340212528988831,
+      "loss": 1.1571,
+      "step": 6737
+    },
+    {
+      "epoch": 0.5913914495969457,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0013397997466923052,
+      "loss": 1.1411,
+      "step": 6738
+    },
+    {
+      "epoch": 0.5914792191798481,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0013393869950113167,
+      "loss": 1.1271,
+      "step": 6739
+    },
+    {
+      "epoch": 0.5915669887627506,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0013389742739866033,
+      "loss": 1.217,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5916547583456531,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0013385615836588973,
+      "loss": 1.144,
+      "step": 6741
+    },
+    {
+      "epoch": 0.5917425279285555,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0013381489240689316,
+      "loss": 1.271,
+      "step": 6742
+    },
+    {
+      "epoch": 0.591830297511458,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0013377362952574328,
+      "loss": 1.1815,
+      "step": 6743
+    },
+    {
+      "epoch": 0.5919180670943606,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0013373236972651266,
+      "loss": 1.1685,
+      "step": 6744
+    },
+    {
+      "epoch": 0.592005836677263,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0013369111301327339,
+      "loss": 1.1637,
+      "step": 6745
+    },
+    {
+      "epoch": 0.5920936062601655,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0013364985939009737,
+      "loss": 1.147,
+      "step": 6746
+    },
+    {
+      "epoch": 0.592181375843068,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.001336086088610562,
+      "loss": 1.1724,
+      "step": 6747
+    },
+    {
+      "epoch": 0.5922691454259704,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0013356736143022114,
+      "loss": 1.1324,
+      "step": 6748
+    },
+    {
+      "epoch": 0.592356915008873,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0013352611710166313,
+      "loss": 1.1162,
+      "step": 6749
+    },
+    {
+      "epoch": 0.5924446845917755,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.001334848758794528,
+      "loss": 1.1931,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5925324541746779,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0013344363776766057,
+      "loss": 1.2064,
+      "step": 6751
+    },
+    {
+      "epoch": 0.5926202237575804,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0013340240277035635,
+      "loss": 1.1467,
+      "step": 6752
+    },
+    {
+      "epoch": 0.5927079933404829,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0013336117089160998,
+      "loss": 1.2154,
+      "step": 6753
+    },
+    {
+      "epoch": 0.5927957629233854,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0013331994213549088,
+      "loss": 1.1327,
+      "step": 6754
+    },
+    {
+      "epoch": 0.5928835325062879,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.001332787165060681,
+      "loss": 1.1538,
+      "step": 6755
+    },
+    {
+      "epoch": 0.5929713020891904,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0013323749400741055,
+      "loss": 1.1738,
+      "step": 6756
+    },
+    {
+      "epoch": 0.5930590716720928,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0013319627464358664,
+      "loss": 1.1309,
+      "step": 6757
+    },
+    {
+      "epoch": 0.5931468412549953,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0013315505841866456,
+      "loss": 1.1185,
+      "step": 6758
+    },
+    {
+      "epoch": 0.5932346108378979,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0013311384533671222,
+      "loss": 1.1832,
+      "step": 6759
+    },
+    {
+      "epoch": 0.5933223804208003,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.001330726354017972,
+      "loss": 1.1471,
+      "step": 6760
+    },
+    {
+      "epoch": 0.5934101500037028,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.001330314286179867,
+      "loss": 1.1388,
+      "step": 6761
+    },
+    {
+      "epoch": 0.5934979195866052,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.001329902249893477,
+      "loss": 1.2121,
+      "step": 6762
+    },
+    {
+      "epoch": 0.5935856891695077,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0013294902451994688,
+      "loss": 1.1196,
+      "step": 6763
+    },
+    {
+      "epoch": 0.5936734587524103,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.001329078272138505,
+      "loss": 1.1655,
+      "step": 6764
+    },
+    {
+      "epoch": 0.5937612283353128,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0013286663307512459,
+      "loss": 1.15,
+      "step": 6765
+    },
+    {
+      "epoch": 0.5938489979182152,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0013282544210783485,
+      "loss": 1.1733,
+      "step": 6766
+    },
+    {
+      "epoch": 0.5939367675011177,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0013278425431604661,
+      "loss": 1.2254,
+      "step": 6767
+    },
+    {
+      "epoch": 0.5940245370840201,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001327430697038251,
+      "loss": 1.1162,
+      "step": 6768
+    },
+    {
+      "epoch": 0.5941123066669226,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0013270188827523492,
+      "loss": 1.1482,
+      "step": 6769
+    },
+    {
+      "epoch": 0.5942000762498252,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001326607100343406,
+      "loss": 1.1319,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5942878458327276,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0013261953498520619,
+      "loss": 1.1705,
+      "step": 6771
+    },
+    {
+      "epoch": 0.5943756154156301,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0013257836313189555,
+      "loss": 1.1317,
+      "step": 6772
+    },
+    {
+      "epoch": 0.5944633849985326,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0013253719447847223,
+      "loss": 1.2602,
+      "step": 6773
+    },
+    {
+      "epoch": 0.594551154581435,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0013249602902899935,
+      "loss": 1.1997,
+      "step": 6774
+    },
+    {
+      "epoch": 0.5946389241643376,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.001324548667875398,
+      "loss": 1.2098,
+      "step": 6775
+    },
+    {
+      "epoch": 0.5947266937472401,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0013241370775815612,
+      "loss": 1.1338,
+      "step": 6776
+    },
+    {
+      "epoch": 0.5948144633301425,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0013237255194491052,
+      "loss": 1.1814,
+      "step": 6777
+    },
+    {
+      "epoch": 0.594902232913045,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0013233139935186497,
+      "loss": 1.1611,
+      "step": 6778
+    },
+    {
+      "epoch": 0.5949900024959475,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0013229024998308106,
+      "loss": 1.2246,
+      "step": 6779
+    },
+    {
+      "epoch": 0.59507777207885,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0013224910384262002,
+      "loss": 1.1888,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5951655416617525,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0013220796093454284,
+      "loss": 1.1285,
+      "step": 6781
+    },
+    {
+      "epoch": 0.595253311244655,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0013216682126291016,
+      "loss": 1.1785,
+      "step": 6782
+    },
+    {
+      "epoch": 0.5953410808275574,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0013212568483178231,
+      "loss": 1.1864,
+      "step": 6783
+    },
+    {
+      "epoch": 0.5954288504104599,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.001320845516452193,
+      "loss": 1.1668,
+      "step": 6784
+    },
+    {
+      "epoch": 0.5955166199933625,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0013204342170728084,
+      "loss": 1.1141,
+      "step": 6785
+    },
+    {
+      "epoch": 0.5956043895762649,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0013200229502202622,
+      "loss": 1.2097,
+      "step": 6786
+    },
+    {
+      "epoch": 0.5956921591591674,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0013196117159351452,
+      "loss": 1.212,
+      "step": 6787
+    },
+    {
+      "epoch": 0.5957799287420699,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001319200514258044,
+      "loss": 1.2257,
+      "step": 6788
+    },
+    {
+      "epoch": 0.5958676983249723,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001318789345229544,
+      "loss": 1.1567,
+      "step": 6789
+    },
+    {
+      "epoch": 0.5959554679078749,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0013183782088902248,
+      "loss": 1.2497,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5960432374907774,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0013179671052806644,
+      "loss": 1.1258,
+      "step": 6791
+    },
+    {
+      "epoch": 0.5961310070736798,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.001317556034441436,
+      "loss": 1.1281,
+      "step": 6792
+    },
+    {
+      "epoch": 0.5962187766565823,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0013171449964131123,
+      "loss": 1.1671,
+      "step": 6793
+    },
+    {
+      "epoch": 0.5963065462394848,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0013167339912362602,
+      "loss": 1.2105,
+      "step": 6794
+    },
+    {
+      "epoch": 0.5963943158223873,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.001316323018951445,
+      "loss": 1.2116,
+      "step": 6795
+    },
+    {
+      "epoch": 0.5964820854052898,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001315912079599227,
+      "loss": 1.1563,
+      "step": 6796
+    },
+    {
+      "epoch": 0.5965698549881923,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.001315501173220165,
+      "loss": 1.2155,
+      "step": 6797
+    },
+    {
+      "epoch": 0.5966576245710947,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001315090299854814,
+      "loss": 1.1607,
+      "step": 6798
+    },
+    {
+      "epoch": 0.5967453941539972,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.001314679459543725,
+      "loss": 1.1468,
+      "step": 6799
+    },
+    {
+      "epoch": 0.5968331637368997,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0013142686523274463,
+      "loss": 1.1452,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5969209333198022,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0013138578782465233,
+      "loss": 1.2276,
+      "step": 6801
+    },
+    {
+      "epoch": 0.5970087029027047,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0013134471373414977,
+      "loss": 1.1902,
+      "step": 6802
+    },
+    {
+      "epoch": 0.5970964724856072,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.001313036429652908,
+      "loss": 1.1958,
+      "step": 6803
+    },
+    {
+      "epoch": 0.5971842420685096,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001312625755221289,
+      "loss": 1.1905,
+      "step": 6804
+    },
+    {
+      "epoch": 0.5972720116514121,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0013122151140871738,
+      "loss": 1.1934,
+      "step": 6805
+    },
+    {
+      "epoch": 0.5973597812343147,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0013118045062910897,
+      "loss": 1.1421,
+      "step": 6806
+    },
+    {
+      "epoch": 0.5974475508172171,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0013113939318735632,
+      "loss": 1.1579,
+      "step": 6807
+    },
+    {
+      "epoch": 0.5975353204001196,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0013109833908751154,
+      "loss": 1.2143,
+      "step": 6808
+    },
+    {
+      "epoch": 0.597623089983022,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0013105728833362658,
+      "loss": 1.1565,
+      "step": 6809
+    },
+    {
+      "epoch": 0.5977108595659245,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0013101624092975292,
+      "loss": 1.2205,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5977986291488271,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0013097519687994186,
+      "loss": 1.1543,
+      "step": 6811
+    },
+    {
+      "epoch": 0.5978863987317296,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0013093415618824417,
+      "loss": 1.1727,
+      "step": 6812
+    },
+    {
+      "epoch": 0.597974168314632,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0013089311885871046,
+      "loss": 1.1949,
+      "step": 6813
+    },
+    {
+      "epoch": 0.5980619378975345,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0013085208489539105,
+      "loss": 1.2246,
+      "step": 6814
+    },
+    {
+      "epoch": 0.5981497074804369,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001308110543023357,
+      "loss": 1.1133,
+      "step": 6815
+    },
+    {
+      "epoch": 0.5982374770633395,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0013077002708359407,
+      "loss": 1.2484,
+      "step": 6816
+    },
+    {
+      "epoch": 0.598325246646242,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0013072900324321525,
+      "loss": 1.1447,
+      "step": 6817
+    },
+    {
+      "epoch": 0.5984130162291444,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0013068798278524823,
+      "loss": 1.1981,
+      "step": 6818
+    },
+    {
+      "epoch": 0.5985007858120469,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0013064696571374159,
+      "loss": 1.1415,
+      "step": 6819
+    },
+    {
+      "epoch": 0.5985885553949494,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0013060595203274344,
+      "loss": 1.1742,
+      "step": 6820
+    },
+    {
+      "epoch": 0.598676324977852,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0013056494174630174,
+      "loss": 1.2231,
+      "step": 6821
+    },
+    {
+      "epoch": 0.5987640945607544,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0013052393485846405,
+      "loss": 1.1561,
+      "step": 6822
+    },
+    {
+      "epoch": 0.5988518641436569,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0013048293137327756,
+      "loss": 1.2442,
+      "step": 6823
+    },
+    {
+      "epoch": 0.5989396337265593,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0013044193129478922,
+      "loss": 1.2024,
+      "step": 6824
+    },
+    {
+      "epoch": 0.5990274033094618,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0013040093462704544,
+      "loss": 1.1612,
+      "step": 6825
+    },
+    {
+      "epoch": 0.5991151728923643,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001303599413740926,
+      "loss": 1.1794,
+      "step": 6826
+    },
+    {
+      "epoch": 0.5992029424752668,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0013031895153997645,
+      "loss": 1.1834,
+      "step": 6827
+    },
+    {
+      "epoch": 0.5992907120581693,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0013027796512874255,
+      "loss": 1.203,
+      "step": 6828
+    },
+    {
+      "epoch": 0.5993784816410718,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0013023698214443607,
+      "loss": 1.2028,
+      "step": 6829
+    },
+    {
+      "epoch": 0.5994662512239742,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0013019600259110192,
+      "loss": 1.1115,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5995540208068767,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0013015502647278465,
+      "loss": 1.1301,
+      "step": 6831
+    },
+    {
+      "epoch": 0.5996417903897793,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0013011405379352832,
+      "loss": 1.1864,
+      "step": 6832
+    },
+    {
+      "epoch": 0.5997295599726817,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.001300730845573769,
+      "loss": 1.1541,
+      "step": 6833
+    },
+    {
+      "epoch": 0.5998173295555842,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0013003211876837378,
+      "loss": 1.2194,
+      "step": 6834
+    },
+    {
+      "epoch": 0.5999050991384867,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0012999115643056221,
+      "loss": 1.1769,
+      "step": 6835
+    },
+    {
+      "epoch": 0.5999928687213891,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.00129950197547985,
+      "loss": 1.1447,
+      "step": 6836
+    },
+    {
+      "epoch": 0.6000806383042917,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0012990924212468457,
+      "loss": 1.1983,
+      "step": 6837
+    },
+    {
+      "epoch": 0.6001684078871942,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0012986829016470313,
+      "loss": 1.1921,
+      "step": 6838
+    },
+    {
+      "epoch": 0.6002561774700966,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0012982734167208243,
+      "loss": 1.1663,
+      "step": 6839
+    },
+    {
+      "epoch": 0.6003439470529991,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0012978639665086396,
+      "loss": 1.2064,
+      "step": 6840
+    },
+    {
+      "epoch": 0.6004317166359016,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0012974545510508878,
+      "loss": 1.1325,
+      "step": 6841
+    },
+    {
+      "epoch": 0.6005194862188041,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.001297045170387976,
+      "loss": 1.1607,
+      "step": 6842
+    },
+    {
+      "epoch": 0.6006072558017066,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0012966358245603104,
+      "loss": 1.1649,
+      "step": 6843
+    },
+    {
+      "epoch": 0.6006950253846091,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0012962265136082906,
+      "loss": 1.254,
+      "step": 6844
+    },
+    {
+      "epoch": 0.6007827949675115,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0012958172375723141,
+      "loss": 1.1603,
+      "step": 6845
+    },
+    {
+      "epoch": 0.600870564550414,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0012954079964927745,
+      "loss": 1.1788,
+      "step": 6846
+    },
+    {
+      "epoch": 0.6009583341333166,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0012949987904100628,
+      "loss": 1.2094,
+      "step": 6847
+    },
+    {
+      "epoch": 0.601046103716219,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0012945896193645658,
+      "loss": 1.2114,
+      "step": 6848
+    },
+    {
+      "epoch": 0.6011338732991215,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001294180483396667,
+      "loss": 1.1904,
+      "step": 6849
+    },
+    {
+      "epoch": 0.601221642882024,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.001293771382546747,
+      "loss": 1.1059,
+      "step": 6850
+    },
+    {
+      "epoch": 0.6013094124649264,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0012933623168551816,
+      "loss": 1.1635,
+      "step": 6851
+    },
+    {
+      "epoch": 0.6013971820478289,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001292953286362344,
+      "loss": 1.1235,
+      "step": 6852
+    },
+    {
+      "epoch": 0.6014849516307315,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.001292544291108605,
+      "loss": 1.2444,
+      "step": 6853
+    },
+    {
+      "epoch": 0.6015727212136339,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0012921353311343294,
+      "loss": 1.1733,
+      "step": 6854
+    },
+    {
+      "epoch": 0.6016604907965364,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.001291726406479881,
+      "loss": 1.1452,
+      "step": 6855
+    },
+    {
+      "epoch": 0.6017482603794388,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0012913175171856191,
+      "loss": 1.2193,
+      "step": 6856
+    },
+    {
+      "epoch": 0.6018360299623413,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0012909086632918985,
+      "loss": 1.2074,
+      "step": 6857
+    },
+    {
+      "epoch": 0.6019237995452439,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0012904998448390723,
+      "loss": 1.1404,
+      "step": 6858
+    },
+    {
+      "epoch": 0.6020115691281464,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0012900910618674887,
+      "loss": 1.2259,
+      "step": 6859
+    },
+    {
+      "epoch": 0.6020993387110488,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0012896823144174935,
+      "loss": 1.1565,
+      "step": 6860
+    },
+    {
+      "epoch": 0.6021871082939513,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.001289273602529428,
+      "loss": 1.1348,
+      "step": 6861
+    },
+    {
+      "epoch": 0.6022748778768537,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0012888649262436298,
+      "loss": 1.2166,
+      "step": 6862
+    },
+    {
+      "epoch": 0.6023626474597563,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0012884562856004356,
+      "loss": 1.1385,
+      "step": 6863
+    },
+    {
+      "epoch": 0.6024504170426588,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0012880476806401745,
+      "loss": 1.1808,
+      "step": 6864
+    },
+    {
+      "epoch": 0.6025381866255612,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0012876391114031759,
+      "loss": 1.1434,
+      "step": 6865
+    },
+    {
+      "epoch": 0.6026259562084637,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.001287230577929763,
+      "loss": 1.2709,
+      "step": 6866
+    },
+    {
+      "epoch": 0.6027137257913662,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0012868220802602567,
+      "loss": 1.202,
+      "step": 6867
+    },
+    {
+      "epoch": 0.6028014953742687,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.001286413618434974,
+      "loss": 1.147,
+      "step": 6868
+    },
+    {
+      "epoch": 0.6028892649571712,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0012860051924942282,
+      "loss": 1.1406,
+      "step": 6869
+    },
+    {
+      "epoch": 0.6029770345400737,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0012855968024783303,
+      "loss": 1.1262,
+      "step": 6870
+    },
+    {
+      "epoch": 0.6030648041229761,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0012851884484275857,
+      "loss": 1.1592,
+      "step": 6871
+    },
+    {
+      "epoch": 0.6031525737058786,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0012847801303822975,
+      "loss": 1.1824,
+      "step": 6872
+    },
+    {
+      "epoch": 0.6032403432887812,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0012843718483827658,
+      "loss": 1.1897,
+      "step": 6873
+    },
+    {
+      "epoch": 0.6033281128716836,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0012839636024692858,
+      "loss": 1.1865,
+      "step": 6874
+    },
+    {
+      "epoch": 0.6034158824545861,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00128355539268215,
+      "loss": 1.136,
+      "step": 6875
+    },
+    {
+      "epoch": 0.6035036520374886,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0012831472190616468,
+      "loss": 1.2101,
+      "step": 6876
+    },
+    {
+      "epoch": 0.603591421620391,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0012827390816480618,
+      "loss": 1.1872,
+      "step": 6877
+    },
+    {
+      "epoch": 0.6036791912032936,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0012823309804816758,
+      "loss": 1.2024,
+      "step": 6878
+    },
+    {
+      "epoch": 0.6037669607861961,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0012819229156027677,
+      "loss": 1.23,
+      "step": 6879
+    },
+    {
+      "epoch": 0.6038547303690985,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0012815148870516113,
+      "loss": 1.2162,
+      "step": 6880
+    },
+    {
+      "epoch": 0.603942499952001,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0012811068948684772,
+      "loss": 1.1488,
+      "step": 6881
+    },
+    {
+      "epoch": 0.6040302695349035,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0012806989390936334,
+      "loss": 1.1787,
+      "step": 6882
+    },
+    {
+      "epoch": 0.6041180391178059,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0012802910197673425,
+      "loss": 1.1544,
+      "step": 6883
+    },
+    {
+      "epoch": 0.6042058087007085,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0012798831369298653,
+      "loss": 1.1484,
+      "step": 6884
+    },
+    {
+      "epoch": 0.604293578283611,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0012794752906214583,
+      "loss": 1.1653,
+      "step": 6885
+    },
+    {
+      "epoch": 0.6043813478665134,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0012790674808823741,
+      "loss": 1.1532,
+      "step": 6886
+    },
+    {
+      "epoch": 0.6044691174494159,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0012786597077528617,
+      "loss": 1.195,
+      "step": 6887
+    },
+    {
+      "epoch": 0.6045568870323184,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0012782519712731665,
+      "loss": 1.1939,
+      "step": 6888
+    },
+    {
+      "epoch": 0.6046446566152209,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0012778442714835314,
+      "loss": 1.1468,
+      "step": 6889
+    },
+    {
+      "epoch": 0.6047324261981234,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0012774366084241937,
+      "loss": 1.1827,
+      "step": 6890
+    },
+    {
+      "epoch": 0.6048201957810259,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0012770289821353886,
+      "loss": 1.1516,
+      "step": 6891
+    },
+    {
+      "epoch": 0.6049079653639283,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0012766213926573473,
+      "loss": 1.1718,
+      "step": 6892
+    },
+    {
+      "epoch": 0.6049957349468308,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0012762138400302968,
+      "loss": 1.1408,
+      "step": 6893
+    },
+    {
+      "epoch": 0.6050835045297334,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0012758063242944615,
+      "loss": 1.2192,
+      "step": 6894
+    },
+    {
+      "epoch": 0.6051712741126358,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0012753988454900617,
+      "loss": 1.2055,
+      "step": 6895
+    },
+    {
+      "epoch": 0.6052590436955383,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0012749914036573131,
+      "loss": 1.1719,
+      "step": 6896
+    },
+    {
+      "epoch": 0.6053468132784408,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0012745839988364297,
+      "loss": 1.1574,
+      "step": 6897
+    },
+    {
+      "epoch": 0.6054345828613432,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0012741766310676197,
+      "loss": 1.1822,
+      "step": 6898
+    },
+    {
+      "epoch": 0.6055223524442458,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00127376930039109,
+      "loss": 1.1132,
+      "step": 6899
+    },
+    {
+      "epoch": 0.6056101220271483,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0012733620068470406,
+      "loss": 1.1966,
+      "step": 6900
+    },
+    {
+      "epoch": 0.6056978916100507,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0012729547504756716,
+      "loss": 1.2122,
+      "step": 6901
+    },
+    {
+      "epoch": 0.6057856611929532,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0012725475313171766,
+      "loss": 1.1796,
+      "step": 6902
+    },
+    {
+      "epoch": 0.6058734307758556,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0012721403494117462,
+      "loss": 1.2022,
+      "step": 6903
+    },
+    {
+      "epoch": 0.6059612003587582,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0012717332047995687,
+      "loss": 1.1928,
+      "step": 6904
+    },
+    {
+      "epoch": 0.6060489699416607,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0012713260975208273,
+      "loss": 1.1812,
+      "step": 6905
+    },
+    {
+      "epoch": 0.6061367395245632,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0012709190276157013,
+      "loss": 1.1573,
+      "step": 6906
+    },
+    {
+      "epoch": 0.6062245091074656,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0012705119951243677,
+      "loss": 1.1236,
+      "step": 6907
+    },
+    {
+      "epoch": 0.6063122786903681,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0012701050000869989,
+      "loss": 1.1884,
+      "step": 6908
+    },
+    {
+      "epoch": 0.6064000482732705,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0012696980425437629,
+      "loss": 1.1617,
+      "step": 6909
+    },
+    {
+      "epoch": 0.6064878178561731,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0012692911225348255,
+      "loss": 1.1964,
+      "step": 6910
+    },
+    {
+      "epoch": 0.6065755874390756,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.001268884240100348,
+      "loss": 1.1887,
+      "step": 6911
+    },
+    {
+      "epoch": 0.606663357021978,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0012684773952804877,
+      "loss": 1.2124,
+      "step": 6912
+    },
+    {
+      "epoch": 0.6067511266048805,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0012680705881153989,
+      "loss": 1.1838,
+      "step": 6913
+    },
+    {
+      "epoch": 0.606838896187783,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0012676638186452322,
+      "loss": 1.2104,
+      "step": 6914
+    },
+    {
+      "epoch": 0.6069266657706855,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0012672570869101333,
+      "loss": 1.2154,
+      "step": 6915
+    },
+    {
+      "epoch": 0.607014435353588,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.001266850392950246,
+      "loss": 1.1772,
+      "step": 6916
+    },
+    {
+      "epoch": 0.6071022049364905,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0012664437368057086,
+      "loss": 1.1934,
+      "step": 6917
+    },
+    {
+      "epoch": 0.6071899745193929,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0012660371185166568,
+      "loss": 1.1157,
+      "step": 6918
+    },
+    {
+      "epoch": 0.6072777441022954,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0012656305381232224,
+      "loss": 1.1927,
+      "step": 6919
+    },
+    {
+      "epoch": 0.607365513685198,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0012652239956655324,
+      "loss": 1.1763,
+      "step": 6920
+    },
+    {
+      "epoch": 0.6074532832681004,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.001264817491183712,
+      "loss": 1.1247,
+      "step": 6921
+    },
+    {
+      "epoch": 0.6075410528510029,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001264411024717881,
+      "loss": 1.2004,
+      "step": 6922
+    },
+    {
+      "epoch": 0.6076288224339054,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0012640045963081554,
+      "loss": 1.1899,
+      "step": 6923
+    },
+    {
+      "epoch": 0.6077165920168078,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00126359820599465,
+      "loss": 1.1362,
+      "step": 6924
+    },
+    {
+      "epoch": 0.6078043615997104,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0012631918538174723,
+      "loss": 1.2031,
+      "step": 6925
+    },
+    {
+      "epoch": 0.6078921311826129,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0012627855398167284,
+      "loss": 1.1759,
+      "step": 6926
+    },
+    {
+      "epoch": 0.6079799007655153,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.001262379264032519,
+      "loss": 1.189,
+      "step": 6927
+    },
+    {
+      "epoch": 0.6080676703484178,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.001261973026504943,
+      "loss": 1.169,
+      "step": 6928
+    },
+    {
+      "epoch": 0.6081554399313203,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001261566827274094,
+      "loss": 1.2582,
+      "step": 6929
+    },
+    {
+      "epoch": 0.6082432095142228,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0012611606663800621,
+      "loss": 1.1666,
+      "step": 6930
+    },
+    {
+      "epoch": 0.6083309790971253,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0012607545438629336,
+      "loss": 1.1363,
+      "step": 6931
+    },
+    {
+      "epoch": 0.6084187486800278,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0012603484597627913,
+      "loss": 1.2544,
+      "step": 6932
+    },
+    {
+      "epoch": 0.6085065182629302,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0012599424141197142,
+      "loss": 1.2262,
+      "step": 6933
+    },
+    {
+      "epoch": 0.6085942878458327,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001259536406973778,
+      "loss": 1.1832,
+      "step": 6934
+    },
+    {
+      "epoch": 0.6086820574287353,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.001259130438365053,
+      "loss": 1.1963,
+      "step": 6935
+    },
+    {
+      "epoch": 0.6087698270116377,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0012587245083336078,
+      "loss": 1.2135,
+      "step": 6936
+    },
+    {
+      "epoch": 0.6088575965945402,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.001258318616919505,
+      "loss": 1.1336,
+      "step": 6937
+    },
+    {
+      "epoch": 0.6089453661774427,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0012579127641628048,
+      "loss": 1.2023,
+      "step": 6938
+    },
+    {
+      "epoch": 0.6090331357603451,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0012575069501035633,
+      "loss": 1.1661,
+      "step": 6939
+    },
+    {
+      "epoch": 0.6091209053432476,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0012571011747818334,
+      "loss": 1.1614,
+      "step": 6940
+    },
+    {
+      "epoch": 0.6092086749261502,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0012566954382376623,
+      "loss": 1.2206,
+      "step": 6941
+    },
+    {
+      "epoch": 0.6092964445090526,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0012562897405110959,
+      "loss": 1.1611,
+      "step": 6942
+    },
+    {
+      "epoch": 0.6093842140919551,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0012558840816421734,
+      "loss": 1.2571,
+      "step": 6943
+    },
+    {
+      "epoch": 0.6094719836748576,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0012554784616709328,
+      "loss": 1.2077,
+      "step": 6944
+    },
+    {
+      "epoch": 0.60955975325776,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.001255072880637407,
+      "loss": 1.1752,
+      "step": 6945
+    },
+    {
+      "epoch": 0.6096475228406626,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0012546673385816255,
+      "loss": 1.1764,
+      "step": 6946
+    },
+    {
+      "epoch": 0.6097352924235651,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0012542618355436133,
+      "loss": 1.2131,
+      "step": 6947
+    },
+    {
+      "epoch": 0.6098230620064675,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0012538563715633925,
+      "loss": 1.1926,
+      "step": 6948
+    },
+    {
+      "epoch": 0.60991083158937,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0012534509466809798,
+      "loss": 1.2021,
+      "step": 6949
+    },
+    {
+      "epoch": 0.6099986011722724,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00125304556093639,
+      "loss": 1.1673,
+      "step": 6950
+    },
+    {
+      "epoch": 0.610086370755175,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0012526402143696322,
+      "loss": 1.1362,
+      "step": 6951
+    },
+    {
+      "epoch": 0.6101741403380775,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0012522349070207136,
+      "loss": 1.1588,
+      "step": 6952
+    },
+    {
+      "epoch": 0.61026190992098,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001251829638929635,
+      "loss": 1.1073,
+      "step": 6953
+    },
+    {
+      "epoch": 0.6103496795038824,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0012514244101363959,
+      "loss": 1.1571,
+      "step": 6954
+    },
+    {
+      "epoch": 0.6104374490867849,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0012510192206809905,
+      "loss": 1.1134,
+      "step": 6955
+    },
+    {
+      "epoch": 0.6105252186696875,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0012506140706034096,
+      "loss": 1.1792,
+      "step": 6956
+    },
+    {
+      "epoch": 0.6106129882525899,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0012502089599436392,
+      "loss": 1.1976,
+      "step": 6957
+    },
+    {
+      "epoch": 0.6107007578354924,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.001249803888741663,
+      "loss": 1.2081,
+      "step": 6958
+    },
+    {
+      "epoch": 0.6107885274183948,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.001249398857037459,
+      "loss": 1.1679,
+      "step": 6959
+    },
+    {
+      "epoch": 0.6108762970012973,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0012489938648710033,
+      "loss": 1.2241,
+      "step": 6960
+    },
+    {
+      "epoch": 0.6109640665841999,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0012485889122822661,
+      "loss": 1.1419,
+      "step": 6961
+    },
+    {
+      "epoch": 0.6110518361671023,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0012481839993112149,
+      "loss": 1.2416,
+      "step": 6962
+    },
+    {
+      "epoch": 0.6111396057500048,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0012477791259978127,
+      "loss": 1.1659,
+      "step": 6963
+    },
+    {
+      "epoch": 0.6112273753329073,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0012473742923820195,
+      "loss": 1.2044,
+      "step": 6964
+    },
+    {
+      "epoch": 0.6113151449158097,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0012469694985037909,
+      "loss": 1.1876,
+      "step": 6965
+    },
+    {
+      "epoch": 0.6114029144987122,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0012465647444030777,
+      "loss": 1.1839,
+      "step": 6966
+    },
+    {
+      "epoch": 0.6114906840816148,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0012461600301198284,
+      "loss": 1.1819,
+      "step": 6967
+    },
+    {
+      "epoch": 0.6115784536645172,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0012457553556939856,
+      "loss": 1.1562,
+      "step": 6968
+    },
+    {
+      "epoch": 0.6116662232474197,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0012453507211654905,
+      "loss": 1.2251,
+      "step": 6969
+    },
+    {
+      "epoch": 0.6117539928303222,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0012449461265742773,
+      "loss": 1.2239,
+      "step": 6970
+    },
+    {
+      "epoch": 0.6118417624132246,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0012445415719602787,
+      "loss": 1.245,
+      "step": 6971
+    },
+    {
+      "epoch": 0.6119295319961272,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0012441370573634235,
+      "loss": 1.104,
+      "step": 6972
+    },
+    {
+      "epoch": 0.6120173015790297,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0012437325828236338,
+      "loss": 1.1802,
+      "step": 6973
+    },
+    {
+      "epoch": 0.6121050711619321,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0012433281483808308,
+      "loss": 1.1566,
+      "step": 6974
+    },
+    {
+      "epoch": 0.6121928407448346,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0012429237540749311,
+      "loss": 1.1638,
+      "step": 6975
+    },
+    {
+      "epoch": 0.6122806103277371,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0012425193999458454,
+      "loss": 1.1968,
+      "step": 6976
+    },
+    {
+      "epoch": 0.6123683799106396,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0012421150860334833,
+      "loss": 1.1615,
+      "step": 6977
+    },
+    {
+      "epoch": 0.6124561494935421,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.001241710812377748,
+      "loss": 1.1369,
+      "step": 6978
+    },
+    {
+      "epoch": 0.6125439190764446,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00124130657901854,
+      "loss": 1.207,
+      "step": 6979
+    },
+    {
+      "epoch": 0.612631688659347,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0012409023859957556,
+      "loss": 1.2347,
+      "step": 6980
+    },
+    {
+      "epoch": 0.6127194582422495,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001240498233349287,
+      "loss": 1.1474,
+      "step": 6981
+    },
+    {
+      "epoch": 0.6128072278251521,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.001240094121119022,
+      "loss": 1.1375,
+      "step": 6982
+    },
+    {
+      "epoch": 0.6128949974080545,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0012396900493448452,
+      "loss": 1.1284,
+      "step": 6983
+    },
+    {
+      "epoch": 0.612982766990957,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0012392860180666374,
+      "loss": 1.1544,
+      "step": 6984
+    },
+    {
+      "epoch": 0.6130705365738595,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0012388820273242747,
+      "loss": 1.098,
+      "step": 6985
+    },
+    {
+      "epoch": 0.6131583061567619,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0012384780771576291,
+      "loss": 1.215,
+      "step": 6986
+    },
+    {
+      "epoch": 0.6132460757396645,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0012380741676065693,
+      "loss": 1.1714,
+      "step": 6987
+    },
+    {
+      "epoch": 0.613333845322567,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0012376702987109587,
+      "loss": 1.1642,
+      "step": 6988
+    },
+    {
+      "epoch": 0.6134216149054694,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0012372664705106587,
+      "loss": 1.1549,
+      "step": 6989
+    },
+    {
+      "epoch": 0.6135093844883719,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0012368626830455243,
+      "loss": 1.1274,
+      "step": 6990
+    },
+    {
+      "epoch": 0.6135971540712744,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0012364589363554087,
+      "loss": 1.1514,
+      "step": 6991
+    },
+    {
+      "epoch": 0.6136849236541768,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0012360552304801597,
+      "loss": 1.257,
+      "step": 6992
+    },
+    {
+      "epoch": 0.6137726932370794,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0012356515654596211,
+      "loss": 1.1884,
+      "step": 6993
+    },
+    {
+      "epoch": 0.6138604628199819,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0012352479413336342,
+      "loss": 1.1473,
+      "step": 6994
+    },
+    {
+      "epoch": 0.6139482324028843,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0012348443581420342,
+      "loss": 1.1769,
+      "step": 6995
+    },
+    {
+      "epoch": 0.6140360019857868,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.001234440815924653,
+      "loss": 1.2072,
+      "step": 6996
+    },
+    {
+      "epoch": 0.6141237715686892,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0012340373147213194,
+      "loss": 1.2169,
+      "step": 6997
+    },
+    {
+      "epoch": 0.6142115411515918,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001233633854571857,
+      "loss": 1.202,
+      "step": 6998
+    },
+    {
+      "epoch": 0.6142993107344943,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0012332304355160857,
+      "loss": 1.1562,
+      "step": 6999
+    },
+    {
+      "epoch": 0.6143870803173967,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0012328270575938213,
+      "loss": 1.2086,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6143870803173967,
+      "eval_loss": 1.1626135110855103,
+      "eval_runtime": 437.5685,
+      "eval_samples_per_second": 33.686,
+      "eval_steps_per_second": 8.422,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6144748499002992,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0012324237208448756,
+      "loss": 1.1551,
+      "step": 7001
+    },
+    {
+      "epoch": 0.6145626194832017,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0012320204253090563,
+      "loss": 1.0825,
+      "step": 7002
+    },
+    {
+      "epoch": 0.6146503890661043,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0012316171710261671,
+      "loss": 1.149,
+      "step": 7003
+    },
+    {
+      "epoch": 0.6147381586490067,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0012312139580360076,
+      "loss": 1.1136,
+      "step": 7004
+    },
+    {
+      "epoch": 0.6148259282319092,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0012308107863783732,
+      "loss": 1.216,
+      "step": 7005
+    },
+    {
+      "epoch": 0.6149136978148116,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0012304076560930562,
+      "loss": 1.1774,
+      "step": 7006
+    },
+    {
+      "epoch": 0.6150014673977141,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0012300045672198427,
+      "loss": 1.1624,
+      "step": 7007
+    },
+    {
+      "epoch": 0.6150892369806167,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0012296015197985168,
+      "loss": 1.1852,
+      "step": 7008
+    },
+    {
+      "epoch": 0.6151770065635191,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0012291985138688578,
+      "loss": 1.1813,
+      "step": 7009
+    },
+    {
+      "epoch": 0.6152647761464216,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0012287955494706402,
+      "loss": 1.1634,
+      "step": 7010
+    },
+    {
+      "epoch": 0.6153525457293241,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0012283926266436352,
+      "loss": 1.1131,
+      "step": 7011
+    },
+    {
+      "epoch": 0.6154403153122265,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0012279897454276097,
+      "loss": 1.2027,
+      "step": 7012
+    },
+    {
+      "epoch": 0.6155280848951291,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0012275869058623265,
+      "loss": 1.1554,
+      "step": 7013
+    },
+    {
+      "epoch": 0.6156158544780316,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0012271841079875441,
+      "loss": 1.1583,
+      "step": 7014
+    },
+    {
+      "epoch": 0.615703624060934,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0012267813518430175,
+      "loss": 1.1623,
+      "step": 7015
+    },
+    {
+      "epoch": 0.6157913936438365,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.001226378637468497,
+      "loss": 1.1714,
+      "step": 7016
+    },
+    {
+      "epoch": 0.615879163226739,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0012259759649037286,
+      "loss": 1.2085,
+      "step": 7017
+    },
+    {
+      "epoch": 0.6159669328096415,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.001225573334188455,
+      "loss": 1.2005,
+      "step": 7018
+    },
+    {
+      "epoch": 0.616054702392544,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0012251707453624135,
+      "loss": 1.1492,
+      "step": 7019
+    },
+    {
+      "epoch": 0.6161424719754465,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001224768198465339,
+      "loss": 1.1463,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6162302415583489,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0012243656935369603,
+      "loss": 1.1612,
+      "step": 7021
+    },
+    {
+      "epoch": 0.6163180111412514,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0012239632306170034,
+      "loss": 1.1562,
+      "step": 7022
+    },
+    {
+      "epoch": 0.6164057807241539,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0012235608097451904,
+      "loss": 1.217,
+      "step": 7023
+    },
+    {
+      "epoch": 0.6164935503070564,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0012231584309612374,
+      "loss": 1.1732,
+      "step": 7024
+    },
+    {
+      "epoch": 0.6165813198899589,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0012227560943048586,
+      "loss": 1.17,
+      "step": 7025
+    },
+    {
+      "epoch": 0.6166690894728614,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.001222353799815763,
+      "loss": 1.1783,
+      "step": 7026
+    },
+    {
+      "epoch": 0.6167568590557638,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0012219515475336556,
+      "loss": 1.1842,
+      "step": 7027
+    },
+    {
+      "epoch": 0.6168446286386663,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0012215493374982363,
+      "loss": 1.168,
+      "step": 7028
+    },
+    {
+      "epoch": 0.6169323982215689,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001221147169749202,
+      "loss": 1.1954,
+      "step": 7029
+    },
+    {
+      "epoch": 0.6170201678044713,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0012207450443262454,
+      "loss": 1.1376,
+      "step": 7030
+    },
+    {
+      "epoch": 0.6171079373873738,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0012203429612690545,
+      "loss": 1.1539,
+      "step": 7031
+    },
+    {
+      "epoch": 0.6171957069702763,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0012199409206173132,
+      "loss": 1.1465,
+      "step": 7032
+    },
+    {
+      "epoch": 0.6172834765531787,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.001219538922410701,
+      "loss": 1.1975,
+      "step": 7033
+    },
+    {
+      "epoch": 0.6173712461360813,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0012191369666888939,
+      "loss": 1.2356,
+      "step": 7034
+    },
+    {
+      "epoch": 0.6174590157189838,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0012187350534915637,
+      "loss": 1.1875,
+      "step": 7035
+    },
+    {
+      "epoch": 0.6175467853018862,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0012183331828583774,
+      "loss": 1.1797,
+      "step": 7036
+    },
+    {
+      "epoch": 0.6176345548847887,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0012179313548289976,
+      "loss": 1.1666,
+      "step": 7037
+    },
+    {
+      "epoch": 0.6177223244676912,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.001217529569443084,
+      "loss": 1.1897,
+      "step": 7038
+    },
+    {
+      "epoch": 0.6178100940505937,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0012171278267402903,
+      "loss": 1.1942,
+      "step": 7039
+    },
+    {
+      "epoch": 0.6178978636334962,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0012167261267602674,
+      "loss": 1.1504,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6179856332163987,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0012163244695426614,
+      "loss": 1.1722,
+      "step": 7041
+    },
+    {
+      "epoch": 0.6180734027993011,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0012159228551271144,
+      "loss": 1.2184,
+      "step": 7042
+    },
+    {
+      "epoch": 0.6181611723822036,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0012155212835532638,
+      "loss": 1.1419,
+      "step": 7043
+    },
+    {
+      "epoch": 0.6182489419651062,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0012151197548607433,
+      "loss": 1.1749,
+      "step": 7044
+    },
+    {
+      "epoch": 0.6183367115480086,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0012147182690891822,
+      "loss": 1.2015,
+      "step": 7045
+    },
+    {
+      "epoch": 0.6184244811309111,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.001214316826278206,
+      "loss": 1.209,
+      "step": 7046
+    },
+    {
+      "epoch": 0.6185122507138135,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0012139154264674346,
+      "loss": 1.1297,
+      "step": 7047
+    },
+    {
+      "epoch": 0.618600020296716,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0012135140696964857,
+      "loss": 1.2132,
+      "step": 7048
+    },
+    {
+      "epoch": 0.6186877898796185,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.001213112756004971,
+      "loss": 1.1788,
+      "step": 7049
+    },
+    {
+      "epoch": 0.618775559462521,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0012127114854324987,
+      "loss": 1.1229,
+      "step": 7050
+    },
+    {
+      "epoch": 0.6188633290454235,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0012123102580186718,
+      "loss": 1.171,
+      "step": 7051
+    },
+    {
+      "epoch": 0.618951098628326,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0012119090738030914,
+      "loss": 1.2009,
+      "step": 7052
+    },
+    {
+      "epoch": 0.6190388682112284,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0012115079328253514,
+      "loss": 1.1406,
+      "step": 7053
+    },
+    {
+      "epoch": 0.6191266377941309,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0012111068351250435,
+      "loss": 1.2315,
+      "step": 7054
+    },
+    {
+      "epoch": 0.6192144073770335,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0012107057807417545,
+      "loss": 1.1688,
+      "step": 7055
+    },
+    {
+      "epoch": 0.619302176959936,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0012103047697150666,
+      "loss": 1.1374,
+      "step": 7056
+    },
+    {
+      "epoch": 0.6193899465428384,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0012099038020845585,
+      "loss": 1.1608,
+      "step": 7057
+    },
+    {
+      "epoch": 0.6194777161257409,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0012095028778898035,
+      "loss": 1.1209,
+      "step": 7058
+    },
+    {
+      "epoch": 0.6195654857086433,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0012091019971703712,
+      "loss": 1.1819,
+      "step": 7059
+    },
+    {
+      "epoch": 0.6196532552915459,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0012087011599658279,
+      "loss": 1.2097,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6197410248744484,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0012083003663157336,
+      "loss": 1.1062,
+      "step": 7061
+    },
+    {
+      "epoch": 0.6198287944573508,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0012078996162596454,
+      "loss": 1.2385,
+      "step": 7062
+    },
+    {
+      "epoch": 0.6199165640402533,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001207498909837116,
+      "loss": 1.1682,
+      "step": 7063
+    },
+    {
+      "epoch": 0.6200043336231558,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0012070982470876927,
+      "loss": 1.1455,
+      "step": 7064
+    },
+    {
+      "epoch": 0.6200921032060583,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0012066976280509206,
+      "loss": 1.211,
+      "step": 7065
+    },
+    {
+      "epoch": 0.6201798727889608,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.001206297052766338,
+      "loss": 1.1755,
+      "step": 7066
+    },
+    {
+      "epoch": 0.6202676423718633,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0012058965212734813,
+      "loss": 1.1466,
+      "step": 7067
+    },
+    {
+      "epoch": 0.6203554119547657,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0012054960336118806,
+      "loss": 1.2203,
+      "step": 7068
+    },
+    {
+      "epoch": 0.6204431815376682,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0012050955898210622,
+      "loss": 1.1433,
+      "step": 7069
+    },
+    {
+      "epoch": 0.6205309511205708,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0012046951899405487,
+      "loss": 1.1716,
+      "step": 7070
+    },
+    {
+      "epoch": 0.6206187207034732,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0012042948340098582,
+      "loss": 1.2205,
+      "step": 7071
+    },
+    {
+      "epoch": 0.6207064902863757,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0012038945220685042,
+      "loss": 1.1083,
+      "step": 7072
+    },
+    {
+      "epoch": 0.6207942598692782,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0012034942541559951,
+      "loss": 1.1904,
+      "step": 7073
+    },
+    {
+      "epoch": 0.6208820294521806,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0012030940303118366,
+      "loss": 1.1444,
+      "step": 7074
+    },
+    {
+      "epoch": 0.6209697990350832,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0012026938505755292,
+      "loss": 1.1139,
+      "step": 7075
+    },
+    {
+      "epoch": 0.6210575686179857,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0012022937149865688,
+      "loss": 1.1869,
+      "step": 7076
+    },
+    {
+      "epoch": 0.6211453382008881,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0012018936235844474,
+      "loss": 1.1757,
+      "step": 7077
+    },
+    {
+      "epoch": 0.6212331077837906,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0012014935764086518,
+      "loss": 1.2213,
+      "step": 7078
+    },
+    {
+      "epoch": 0.6213208773666931,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0012010935734986662,
+      "loss": 1.1876,
+      "step": 7079
+    },
+    {
+      "epoch": 0.6214086469495955,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.001200693614893968,
+      "loss": 1.1851,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6214964165324981,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0012002937006340328,
+      "loss": 1.1814,
+      "step": 7081
+    },
+    {
+      "epoch": 0.6215841861154006,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0011998938307583296,
+      "loss": 1.1454,
+      "step": 7082
+    },
+    {
+      "epoch": 0.621671955698303,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0011994940053063248,
+      "loss": 1.1838,
+      "step": 7083
+    },
+    {
+      "epoch": 0.6217597252812055,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.001199094224317478,
+      "loss": 1.1797,
+      "step": 7084
+    },
+    {
+      "epoch": 0.621847494864108,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0011986944878312478,
+      "loss": 1.1387,
+      "step": 7085
+    },
+    {
+      "epoch": 0.6219352644470105,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.001198294795887086,
+      "loss": 1.179,
+      "step": 7086
+    },
+    {
+      "epoch": 0.622023034029913,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0011978951485244406,
+      "loss": 1.1901,
+      "step": 7087
+    },
+    {
+      "epoch": 0.6221108036128155,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.001197495545782755,
+      "loss": 1.1671,
+      "step": 7088
+    },
+    {
+      "epoch": 0.6221985731957179,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0011970959877014686,
+      "loss": 1.1593,
+      "step": 7089
+    },
+    {
+      "epoch": 0.6222863427786204,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0011966964743200161,
+      "loss": 1.1958,
+      "step": 7090
+    },
+    {
+      "epoch": 0.622374112361523,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0011962970056778282,
+      "loss": 1.1678,
+      "step": 7091
+    },
+    {
+      "epoch": 0.6224618819444254,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0011958975818143305,
+      "loss": 1.1732,
+      "step": 7092
+    },
+    {
+      "epoch": 0.6225496515273279,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001195498202768945,
+      "loss": 1.1991,
+      "step": 7093
+    },
+    {
+      "epoch": 0.6226374211102303,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0011950988685810878,
+      "loss": 1.1727,
+      "step": 7094
+    },
+    {
+      "epoch": 0.6227251906931328,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0011946995792901723,
+      "loss": 1.2144,
+      "step": 7095
+    },
+    {
+      "epoch": 0.6228129602760354,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0011943003349356074,
+      "loss": 1.1591,
+      "step": 7096
+    },
+    {
+      "epoch": 0.6229007298589379,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.001193901135556796,
+      "loss": 1.1556,
+      "step": 7097
+    },
+    {
+      "epoch": 0.6229884994418403,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0011935019811931376,
+      "loss": 1.119,
+      "step": 7098
+    },
+    {
+      "epoch": 0.6230762690247428,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.001193102871884028,
+      "loss": 1.1157,
+      "step": 7099
+    },
+    {
+      "epoch": 0.6231640386076452,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0011927038076688567,
+      "loss": 1.1643,
+      "step": 7100
+    },
+    {
+      "epoch": 0.6232518081905478,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0011923047885870105,
+      "loss": 1.1637,
+      "step": 7101
+    },
+    {
+      "epoch": 0.6233395777734503,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0011919058146778706,
+      "loss": 1.1633,
+      "step": 7102
+    },
+    {
+      "epoch": 0.6234273473563527,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0011915068859808138,
+      "loss": 1.139,
+      "step": 7103
+    },
+    {
+      "epoch": 0.6235151169392552,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0011911080025352133,
+      "loss": 1.1971,
+      "step": 7104
+    },
+    {
+      "epoch": 0.6236028865221577,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.001190709164380437,
+      "loss": 1.1467,
+      "step": 7105
+    },
+    {
+      "epoch": 0.6236906561050601,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0011903103715558492,
+      "loss": 1.1477,
+      "step": 7106
+    },
+    {
+      "epoch": 0.6237784256879627,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0011899116241008087,
+      "loss": 1.2083,
+      "step": 7107
+    },
+    {
+      "epoch": 0.6238661952708652,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0011895129220546709,
+      "loss": 1.142,
+      "step": 7108
+    },
+    {
+      "epoch": 0.6239539648537676,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001189114265456785,
+      "loss": 1.1692,
+      "step": 7109
+    },
+    {
+      "epoch": 0.6240417344366701,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0011887156543464977,
+      "loss": 1.1451,
+      "step": 7110
+    },
+    {
+      "epoch": 0.6241295040195726,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0011883170887631502,
+      "loss": 1.1768,
+      "step": 7111
+    },
+    {
+      "epoch": 0.6242172736024751,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001187918568746079,
+      "loss": 1.2109,
+      "step": 7112
+    },
+    {
+      "epoch": 0.6243050431853776,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.001187520094334617,
+      "loss": 1.2208,
+      "step": 7113
+    },
+    {
+      "epoch": 0.6243928127682801,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0011871216655680915,
+      "loss": 1.1855,
+      "step": 7114
+    },
+    {
+      "epoch": 0.6244805823511825,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0011867232824858257,
+      "loss": 1.1644,
+      "step": 7115
+    },
+    {
+      "epoch": 0.624568351934085,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0011863249451271398,
+      "loss": 1.1479,
+      "step": 7116
+    },
+    {
+      "epoch": 0.6246561215169876,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0011859266535313464,
+      "loss": 1.1568,
+      "step": 7117
+    },
+    {
+      "epoch": 0.62474389109989,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0011855284077377565,
+      "loss": 1.1773,
+      "step": 7118
+    },
+    {
+      "epoch": 0.6248316606827925,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0011851302077856747,
+      "loss": 1.2027,
+      "step": 7119
+    },
+    {
+      "epoch": 0.624919430265695,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0011847320537144023,
+      "loss": 1.184,
+      "step": 7120
+    },
+    {
+      "epoch": 0.6250071998485974,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0011843339455632348,
+      "loss": 1.2443,
+      "step": 7121
+    },
+    {
+      "epoch": 0.6250949694315,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0011839358833714648,
+      "loss": 1.1715,
+      "step": 7122
+    },
+    {
+      "epoch": 0.6251827390144025,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0011835378671783785,
+      "loss": 1.2384,
+      "step": 7123
+    },
+    {
+      "epoch": 0.6252705085973049,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0011831398970232588,
+      "loss": 1.1786,
+      "step": 7124
+    },
+    {
+      "epoch": 0.6253582781802074,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0011827419729453844,
+      "loss": 1.1308,
+      "step": 7125
+    },
+    {
+      "epoch": 0.6254460477631099,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0011823440949840286,
+      "loss": 1.1781,
+      "step": 7126
+    },
+    {
+      "epoch": 0.6255338173460124,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0011819462631784598,
+      "loss": 1.1174,
+      "step": 7127
+    },
+    {
+      "epoch": 0.6256215869289149,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0011815484775679435,
+      "loss": 1.1721,
+      "step": 7128
+    },
+    {
+      "epoch": 0.6257093565118174,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0011811507381917382,
+      "loss": 1.1676,
+      "step": 7129
+    },
+    {
+      "epoch": 0.6257971260947198,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0011807530450891,
+      "loss": 1.1665,
+      "step": 7130
+    },
+    {
+      "epoch": 0.6258848956776223,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0011803553982992797,
+      "loss": 1.1673,
+      "step": 7131
+    },
+    {
+      "epoch": 0.6259726652605248,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0011799577978615235,
+      "loss": 1.1472,
+      "step": 7132
+    },
+    {
+      "epoch": 0.6260604348434273,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0011795602438150726,
+      "loss": 1.1204,
+      "step": 7133
+    },
+    {
+      "epoch": 0.6261482044263298,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0011791627361991644,
+      "loss": 1.1843,
+      "step": 7134
+    },
+    {
+      "epoch": 0.6262359740092323,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0011787652750530303,
+      "loss": 1.1089,
+      "step": 7135
+    },
+    {
+      "epoch": 0.6263237435921347,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0011783678604158998,
+      "loss": 1.1411,
+      "step": 7136
+    },
+    {
+      "epoch": 0.6264115131750372,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001177970492326995,
+      "loss": 1.1829,
+      "step": 7137
+    },
+    {
+      "epoch": 0.6264992827579398,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0011775731708255348,
+      "loss": 1.2228,
+      "step": 7138
+    },
+    {
+      "epoch": 0.6265870523408422,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0011771758959507334,
+      "loss": 1.1749,
+      "step": 7139
+    },
+    {
+      "epoch": 0.6266748219237447,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0011767786677418007,
+      "loss": 1.1637,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6267625915066471,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0011763814862379405,
+      "loss": 1.1468,
+      "step": 7141
+    },
+    {
+      "epoch": 0.6268503610895496,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.001175984351478354,
+      "loss": 1.1494,
+      "step": 7142
+    },
+    {
+      "epoch": 0.6269381306724522,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.001175587263502236,
+      "loss": 1.1375,
+      "step": 7143
+    },
+    {
+      "epoch": 0.6270259002553547,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.001175190222348778,
+      "loss": 1.222,
+      "step": 7144
+    },
+    {
+      "epoch": 0.6271136698382571,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0011747932280571664,
+      "loss": 1.1345,
+      "step": 7145
+    },
+    {
+      "epoch": 0.6272014394211596,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0011743962806665827,
+      "loss": 1.1618,
+      "step": 7146
+    },
+    {
+      "epoch": 0.627289209004062,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0011739993802162047,
+      "loss": 1.21,
+      "step": 7147
+    },
+    {
+      "epoch": 0.6273769785869646,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.001173602526745204,
+      "loss": 1.1627,
+      "step": 7148
+    },
+    {
+      "epoch": 0.6274647481698671,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0011732057202927491,
+      "loss": 1.1676,
+      "step": 7149
+    },
+    {
+      "epoch": 0.6275525177527695,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.001172808960898003,
+      "loss": 1.1796,
+      "step": 7150
+    },
+    {
+      "epoch": 0.627640287335672,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0011724122486001243,
+      "loss": 1.1843,
+      "step": 7151
+    },
+    {
+      "epoch": 0.6277280569185745,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0011720155834382669,
+      "loss": 1.2094,
+      "step": 7152
+    },
+    {
+      "epoch": 0.627815826501477,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0011716189654515804,
+      "loss": 1.1981,
+      "step": 7153
+    },
+    {
+      "epoch": 0.6279035960843795,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.001171222394679209,
+      "loss": 1.1943,
+      "step": 7154
+    },
+    {
+      "epoch": 0.627991365667282,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0011708258711602925,
+      "loss": 1.1936,
+      "step": 7155
+    },
+    {
+      "epoch": 0.6280791352501844,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0011704293949339666,
+      "loss": 1.1402,
+      "step": 7156
+    },
+    {
+      "epoch": 0.6281669048330869,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.001170032966039362,
+      "loss": 1.1417,
+      "step": 7157
+    },
+    {
+      "epoch": 0.6282546744159895,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0011696365845156047,
+      "loss": 1.1775,
+      "step": 7158
+    },
+    {
+      "epoch": 0.6283424439988919,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0011692402504018156,
+      "loss": 1.1634,
+      "step": 7159
+    },
+    {
+      "epoch": 0.6284302135817944,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0011688439637371117,
+      "loss": 1.1678,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6285179831646969,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.001168447724560605,
+      "loss": 1.1607,
+      "step": 7161
+    },
+    {
+      "epoch": 0.6286057527475993,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0011680515329114017,
+      "loss": 1.1967,
+      "step": 7162
+    },
+    {
+      "epoch": 0.6286935223305018,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0011676553888286057,
+      "loss": 1.0864,
+      "step": 7163
+    },
+    {
+      "epoch": 0.6287812919134044,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.001167259292351314,
+      "loss": 1.1607,
+      "step": 7164
+    },
+    {
+      "epoch": 0.6288690614963068,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0011668632435186197,
+      "loss": 1.1733,
+      "step": 7165
+    },
+    {
+      "epoch": 0.6289568310792093,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0011664672423696123,
+      "loss": 1.1642,
+      "step": 7166
+    },
+    {
+      "epoch": 0.6290446006621118,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0011660712889433747,
+      "loss": 1.1231,
+      "step": 7167
+    },
+    {
+      "epoch": 0.6291323702450142,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0011656753832789855,
+      "loss": 1.1627,
+      "step": 7168
+    },
+    {
+      "epoch": 0.6292201398279168,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0011652795254155203,
+      "loss": 1.2064,
+      "step": 7169
+    },
+    {
+      "epoch": 0.6293079094108193,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001164883715392048,
+      "loss": 1.1959,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6293956789937217,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.001164487953247633,
+      "loss": 1.2448,
+      "step": 7171
+    },
+    {
+      "epoch": 0.6294834485766242,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0011640922390213364,
+      "loss": 1.2179,
+      "step": 7172
+    },
+    {
+      "epoch": 0.6295712181595267,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.001163696572752213,
+      "loss": 1.1581,
+      "step": 7173
+    },
+    {
+      "epoch": 0.6296589877424292,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0011633009544793134,
+      "loss": 1.1885,
+      "step": 7174
+    },
+    {
+      "epoch": 0.6297467573253317,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0011629053842416837,
+      "loss": 1.1907,
+      "step": 7175
+    },
+    {
+      "epoch": 0.6298345269082342,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0011625098620783652,
+      "loss": 1.1604,
+      "step": 7176
+    },
+    {
+      "epoch": 0.6299222964911366,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001162114388028395,
+      "loss": 1.2031,
+      "step": 7177
+    },
+    {
+      "epoch": 0.6300100660740391,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0011617189621308038,
+      "loss": 1.1562,
+      "step": 7178
+    },
+    {
+      "epoch": 0.6300978356569417,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0011613235844246195,
+      "loss": 1.1781,
+      "step": 7179
+    },
+    {
+      "epoch": 0.6301856052398441,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0011609282549488633,
+      "loss": 1.1495,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6302733748227466,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0011605329737425537,
+      "loss": 1.1638,
+      "step": 7181
+    },
+    {
+      "epoch": 0.630361144405649,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0011601377408447025,
+      "loss": 1.1707,
+      "step": 7182
+    },
+    {
+      "epoch": 0.6304489139885515,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.001159742556294318,
+      "loss": 1.1929,
+      "step": 7183
+    },
+    {
+      "epoch": 0.6305366835714541,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0011593474201304036,
+      "loss": 1.1718,
+      "step": 7184
+    },
+    {
+      "epoch": 0.6306244531543566,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0011589523323919568,
+      "loss": 1.1461,
+      "step": 7185
+    },
+    {
+      "epoch": 0.630712222737259,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0011585572931179721,
+      "loss": 1.1357,
+      "step": 7186
+    },
+    {
+      "epoch": 0.6307999923201615,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0011581623023474382,
+      "loss": 1.1554,
+      "step": 7187
+    },
+    {
+      "epoch": 0.630887761903064,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0011577673601193392,
+      "loss": 1.2128,
+      "step": 7188
+    },
+    {
+      "epoch": 0.6309755314859664,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0011573724664726537,
+      "loss": 1.2082,
+      "step": 7189
+    },
+    {
+      "epoch": 0.631063301068869,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0011569776214463566,
+      "loss": 1.1638,
+      "step": 7190
+    },
+    {
+      "epoch": 0.6311510706517715,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0011565828250794177,
+      "loss": 1.1235,
+      "step": 7191
+    },
+    {
+      "epoch": 0.6312388402346739,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0011561880774108013,
+      "loss": 1.1479,
+      "step": 7192
+    },
+    {
+      "epoch": 0.6313266098175764,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0011557933784794678,
+      "loss": 1.2079,
+      "step": 7193
+    },
+    {
+      "epoch": 0.6314143794004788,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0011553987283243726,
+      "loss": 1.1563,
+      "step": 7194
+    },
+    {
+      "epoch": 0.6315021489833814,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0011550041269844653,
+      "loss": 1.1656,
+      "step": 7195
+    },
+    {
+      "epoch": 0.6315899185662839,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0011546095744986926,
+      "loss": 1.2058,
+      "step": 7196
+    },
+    {
+      "epoch": 0.6316776881491863,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0011542150709059943,
+      "loss": 1.2202,
+      "step": 7197
+    },
+    {
+      "epoch": 0.6317654577320888,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0011538206162453074,
+      "loss": 1.167,
+      "step": 7198
+    },
+    {
+      "epoch": 0.6318532273149913,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0011534262105555618,
+      "loss": 1.171,
+      "step": 7199
+    },
+    {
+      "epoch": 0.6319409968978938,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0011530318538756847,
+      "loss": 1.1188,
+      "step": 7200
+    },
+    {
+      "epoch": 0.6320287664807963,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0011526375462445973,
+      "loss": 1.2036,
+      "step": 7201
+    },
+    {
+      "epoch": 0.6321165360636988,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.001152243287701216,
+      "loss": 1.1097,
+      "step": 7202
+    },
+    {
+      "epoch": 0.6322043056466012,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0011518490782844529,
+      "loss": 1.1917,
+      "step": 7203
+    },
+    {
+      "epoch": 0.6322920752295037,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0011514549180332144,
+      "loss": 1.1618,
+      "step": 7204
+    },
+    {
+      "epoch": 0.6323798448124063,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0011510608069864026,
+      "loss": 1.1459,
+      "step": 7205
+    },
+    {
+      "epoch": 0.6324676143953087,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0011506667451829154,
+      "loss": 1.1631,
+      "step": 7206
+    },
+    {
+      "epoch": 0.6325553839782112,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0011502727326616444,
+      "loss": 1.1211,
+      "step": 7207
+    },
+    {
+      "epoch": 0.6326431535611137,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0011498787694614778,
+      "loss": 1.1253,
+      "step": 7208
+    },
+    {
+      "epoch": 0.6327309231440161,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0011494848556212972,
+      "loss": 1.1424,
+      "step": 7209
+    },
+    {
+      "epoch": 0.6328186927269187,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0011490909911799814,
+      "loss": 1.1541,
+      "step": 7210
+    },
+    {
+      "epoch": 0.6329064623098212,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0011486971761764024,
+      "loss": 1.155,
+      "step": 7211
+    },
+    {
+      "epoch": 0.6329942318927236,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001148303410649429,
+      "loss": 1.0851,
+      "step": 7212
+    },
+    {
+      "epoch": 0.6330820014756261,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0011479096946379237,
+      "loss": 1.1443,
+      "step": 7213
+    },
+    {
+      "epoch": 0.6331697710585286,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0011475160281807445,
+      "loss": 1.1484,
+      "step": 7214
+    },
+    {
+      "epoch": 0.6332575406414311,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.001147122411316745,
+      "loss": 1.1804,
+      "step": 7215
+    },
+    {
+      "epoch": 0.6333453102243336,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001146728844084774,
+      "loss": 1.2011,
+      "step": 7216
+    },
+    {
+      "epoch": 0.6334330798072361,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0011463353265236747,
+      "loss": 1.1652,
+      "step": 7217
+    },
+    {
+      "epoch": 0.6335208493901385,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.001145941858672286,
+      "loss": 1.2062,
+      "step": 7218
+    },
+    {
+      "epoch": 0.633608618973041,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0011455484405694413,
+      "loss": 1.1947,
+      "step": 7219
+    },
+    {
+      "epoch": 0.6336963885559435,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0011451550722539696,
+      "loss": 1.1659,
+      "step": 7220
+    },
+    {
+      "epoch": 0.633784158138846,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0011447617537646943,
+      "loss": 1.1516,
+      "step": 7221
+    },
+    {
+      "epoch": 0.6338719277217485,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0011443684851404356,
+      "loss": 1.202,
+      "step": 7222
+    },
+    {
+      "epoch": 0.633959697304651,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0011439752664200058,
+      "loss": 1.1446,
+      "step": 7223
+    },
+    {
+      "epoch": 0.6340474668875534,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0011435820976422157,
+      "loss": 1.1788,
+      "step": 7224
+    },
+    {
+      "epoch": 0.6341352364704559,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0011431889788458677,
+      "loss": 1.1853,
+      "step": 7225
+    },
+    {
+      "epoch": 0.6342230060533585,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0011427959100697629,
+      "loss": 1.1176,
+      "step": 7226
+    },
+    {
+      "epoch": 0.6343107756362609,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0011424028913526946,
+      "loss": 1.1809,
+      "step": 7227
+    },
+    {
+      "epoch": 0.6343985452191634,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.001142009922733453,
+      "loss": 1.1996,
+      "step": 7228
+    },
+    {
+      "epoch": 0.6344863148020659,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0011416170042508216,
+      "loss": 1.1812,
+      "step": 7229
+    },
+    {
+      "epoch": 0.6345740843849683,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0011412241359435808,
+      "loss": 1.1516,
+      "step": 7230
+    },
+    {
+      "epoch": 0.6346618539678709,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0011408313178505039,
+      "loss": 1.1472,
+      "step": 7231
+    },
+    {
+      "epoch": 0.6347496235507734,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0011404385500103618,
+      "loss": 1.1636,
+      "step": 7232
+    },
+    {
+      "epoch": 0.6348373931336758,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0011400458324619183,
+      "loss": 1.159,
+      "step": 7233
+    },
+    {
+      "epoch": 0.6349251627165783,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0011396531652439338,
+      "loss": 1.1447,
+      "step": 7234
+    },
+    {
+      "epoch": 0.6350129322994807,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0011392605483951618,
+      "loss": 1.1615,
+      "step": 7235
+    },
+    {
+      "epoch": 0.6351007018823833,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0011388679819543531,
+      "loss": 1.1763,
+      "step": 7236
+    },
+    {
+      "epoch": 0.6351884714652858,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001138475465960252,
+      "loss": 1.1982,
+      "step": 7237
+    },
+    {
+      "epoch": 0.6352762410481883,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0011380830004515986,
+      "loss": 1.1802,
+      "step": 7238
+    },
+    {
+      "epoch": 0.6353640106310907,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0011376905854671277,
+      "loss": 1.1532,
+      "step": 7239
+    },
+    {
+      "epoch": 0.6354517802139932,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0011372982210455685,
+      "loss": 1.1243,
+      "step": 7240
+    },
+    {
+      "epoch": 0.6355395497968958,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0011369059072256467,
+      "loss": 1.1266,
+      "step": 7241
+    },
+    {
+      "epoch": 0.6356273193797982,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0011365136440460815,
+      "loss": 1.2035,
+      "step": 7242
+    },
+    {
+      "epoch": 0.6357150889627007,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0011361214315455876,
+      "loss": 1.1416,
+      "step": 7243
+    },
+    {
+      "epoch": 0.6358028585456031,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0011357292697628753,
+      "loss": 1.1719,
+      "step": 7244
+    },
+    {
+      "epoch": 0.6358906281285056,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.001135337158736649,
+      "loss": 1.2165,
+      "step": 7245
+    },
+    {
+      "epoch": 0.6359783977114081,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0011349450985056087,
+      "loss": 1.1613,
+      "step": 7246
+    },
+    {
+      "epoch": 0.6360661672943106,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0011345530891084493,
+      "loss": 1.1289,
+      "step": 7247
+    },
+    {
+      "epoch": 0.6361539368772131,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0011341611305838606,
+      "loss": 1.1489,
+      "step": 7248
+    },
+    {
+      "epoch": 0.6362417064601156,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0011337692229705274,
+      "loss": 1.1339,
+      "step": 7249
+    },
+    {
+      "epoch": 0.636329476043018,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0011333773663071288,
+      "loss": 1.1312,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6364172456259205,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0011329855606323402,
+      "loss": 1.1126,
+      "step": 7251
+    },
+    {
+      "epoch": 0.6365050152088231,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0011325938059848315,
+      "loss": 1.1681,
+      "step": 7252
+    },
+    {
+      "epoch": 0.6365927847917255,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0011322021024032663,
+      "loss": 1.1388,
+      "step": 7253
+    },
+    {
+      "epoch": 0.636680554374628,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0011318104499263048,
+      "loss": 1.1661,
+      "step": 7254
+    },
+    {
+      "epoch": 0.6367683239575305,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0011314188485926017,
+      "loss": 1.1006,
+      "step": 7255
+    },
+    {
+      "epoch": 0.6368560935404329,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0011310272984408059,
+      "loss": 1.1824,
+      "step": 7256
+    },
+    {
+      "epoch": 0.6369438631233355,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0011306357995095628,
+      "loss": 1.1186,
+      "step": 7257
+    },
+    {
+      "epoch": 0.637031632706238,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.001130244351837511,
+      "loss": 1.1357,
+      "step": 7258
+    },
+    {
+      "epoch": 0.6371194022891404,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0011298529554632855,
+      "loss": 1.1606,
+      "step": 7259
+    },
+    {
+      "epoch": 0.6372071718720429,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0011294616104255148,
+      "loss": 1.1812,
+      "step": 7260
+    },
+    {
+      "epoch": 0.6372949414549454,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0011290703167628237,
+      "loss": 1.1791,
+      "step": 7261
+    },
+    {
+      "epoch": 0.6373827110378479,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0011286790745138307,
+      "loss": 1.1093,
+      "step": 7262
+    },
+    {
+      "epoch": 0.6374704806207504,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.001128287883717151,
+      "loss": 1.1883,
+      "step": 7263
+    },
+    {
+      "epoch": 0.6375582502036529,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0011278967444113923,
+      "loss": 1.1696,
+      "step": 7264
+    },
+    {
+      "epoch": 0.6376460197865553,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001127505656635159,
+      "loss": 1.1262,
+      "step": 7265
+    },
+    {
+      "epoch": 0.6377337893694578,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0011271146204270503,
+      "loss": 1.1378,
+      "step": 7266
+    },
+    {
+      "epoch": 0.6378215589523604,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0011267236358256595,
+      "loss": 1.1303,
+      "step": 7267
+    },
+    {
+      "epoch": 0.6379093285352628,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0011263327028695753,
+      "loss": 1.2141,
+      "step": 7268
+    },
+    {
+      "epoch": 0.6379970981181653,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0011259418215973812,
+      "loss": 1.2336,
+      "step": 7269
+    },
+    {
+      "epoch": 0.6380848677010678,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001125550992047656,
+      "loss": 1.1659,
+      "step": 7270
+    },
+    {
+      "epoch": 0.6381726372839702,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0011251602142589726,
+      "loss": 1.1649,
+      "step": 7271
+    },
+    {
+      "epoch": 0.6382604068668727,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0011247694882698992,
+      "loss": 1.1968,
+      "step": 7272
+    },
+    {
+      "epoch": 0.6383481764497753,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0011243788141189995,
+      "loss": 1.1923,
+      "step": 7273
+    },
+    {
+      "epoch": 0.6384359460326777,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0011239881918448308,
+      "loss": 1.1587,
+      "step": 7274
+    },
+    {
+      "epoch": 0.6385237156155802,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.001123597621485946,
+      "loss": 1.2302,
+      "step": 7275
+    },
+    {
+      "epoch": 0.6386114851984827,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.001123207103080893,
+      "loss": 1.1734,
+      "step": 7276
+    },
+    {
+      "epoch": 0.6386992547813851,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.001122816636668215,
+      "loss": 1.1333,
+      "step": 7277
+    },
+    {
+      "epoch": 0.6387870243642877,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.001122426222286449,
+      "loss": 1.1407,
+      "step": 7278
+    },
+    {
+      "epoch": 0.6388747939471902,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0011220358599741272,
+      "loss": 1.1355,
+      "step": 7279
+    },
+    {
+      "epoch": 0.6389625635300926,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0011216455497697767,
+      "loss": 1.136,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6390503331129951,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0011212552917119205,
+      "loss": 1.1687,
+      "step": 7281
+    },
+    {
+      "epoch": 0.6391381026958975,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0011208650858390745,
+      "loss": 1.2206,
+      "step": 7282
+    },
+    {
+      "epoch": 0.6392258722788001,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001120474932189751,
+      "loss": 1.1859,
+      "step": 7283
+    },
+    {
+      "epoch": 0.6393136418617026,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001120084830802456,
+      "loss": 1.1483,
+      "step": 7284
+    },
+    {
+      "epoch": 0.639401411444605,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001119694781715692,
+      "loss": 1.1752,
+      "step": 7285
+    },
+    {
+      "epoch": 0.6394891810275075,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0011193047849679541,
+      "loss": 1.2025,
+      "step": 7286
+    },
+    {
+      "epoch": 0.63957695061041,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0011189148405977344,
+      "loss": 1.1522,
+      "step": 7287
+    },
+    {
+      "epoch": 0.6396647201933126,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0011185249486435182,
+      "loss": 1.1891,
+      "step": 7288
+    },
+    {
+      "epoch": 0.639752489776215,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0011181351091437875,
+      "loss": 1.1109,
+      "step": 7289
+    },
+    {
+      "epoch": 0.6398402593591175,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0011177453221370161,
+      "loss": 1.1646,
+      "step": 7290
+    },
+    {
+      "epoch": 0.6399280289420199,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0011173555876616763,
+      "loss": 1.1242,
+      "step": 7291
+    },
+    {
+      "epoch": 0.6400157985249224,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0011169659057562317,
+      "loss": 1.1304,
+      "step": 7292
+    },
+    {
+      "epoch": 0.640103568107825,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0011165762764591435,
+      "loss": 1.1663,
+      "step": 7293
+    },
+    {
+      "epoch": 0.6401913376907274,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.001116186699808866,
+      "loss": 1.1809,
+      "step": 7294
+    },
+    {
+      "epoch": 0.6402791072736299,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.001115797175843849,
+      "loss": 1.1356,
+      "step": 7295
+    },
+    {
+      "epoch": 0.6403668768565324,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0011154077046025368,
+      "loss": 1.1455,
+      "step": 7296
+    },
+    {
+      "epoch": 0.6404546464394348,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0011150182861233689,
+      "loss": 1.1849,
+      "step": 7297
+    },
+    {
+      "epoch": 0.6405424160223374,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0011146289204447797,
+      "loss": 1.2536,
+      "step": 7298
+    },
+    {
+      "epoch": 0.6406301856052399,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0011142396076051974,
+      "loss": 1.1973,
+      "step": 7299
+    },
+    {
+      "epoch": 0.6407179551881423,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0011138503476430463,
+      "loss": 1.1742,
+      "step": 7300
+    },
+    {
+      "epoch": 0.6408057247710448,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0011134611405967444,
+      "loss": 1.1528,
+      "step": 7301
+    },
+    {
+      "epoch": 0.6408934943539473,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0011130719865047045,
+      "loss": 1.1716,
+      "step": 7302
+    },
+    {
+      "epoch": 0.6409812639368497,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0011126828854053357,
+      "loss": 1.1243,
+      "step": 7303
+    },
+    {
+      "epoch": 0.6410690335197523,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0011122938373370394,
+      "loss": 1.1636,
+      "step": 7304
+    },
+    {
+      "epoch": 0.6411568031026548,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0011119048423382143,
+      "loss": 1.1518,
+      "step": 7305
+    },
+    {
+      "epoch": 0.6412445726855572,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0011115159004472515,
+      "loss": 1.1176,
+      "step": 7306
+    },
+    {
+      "epoch": 0.6413323422684597,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0011111270117025388,
+      "loss": 1.1836,
+      "step": 7307
+    },
+    {
+      "epoch": 0.6414201118513622,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0011107381761424583,
+      "loss": 1.1192,
+      "step": 7308
+    },
+    {
+      "epoch": 0.6415078814342647,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0011103493938053855,
+      "loss": 1.1399,
+      "step": 7309
+    },
+    {
+      "epoch": 0.6415956510171672,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0011099606647296926,
+      "loss": 1.1455,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6416834206000697,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0011095719889537452,
+      "loss": 1.1812,
+      "step": 7311
+    },
+    {
+      "epoch": 0.6417711901829721,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0011091833665159043,
+      "loss": 1.2031,
+      "step": 7312
+    },
+    {
+      "epoch": 0.6418589597658746,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0011087947974545251,
+      "loss": 1.2279,
+      "step": 7313
+    },
+    {
+      "epoch": 0.6419467293487772,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.001108406281807958,
+      "loss": 1.1942,
+      "step": 7314
+    },
+    {
+      "epoch": 0.6420344989316796,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0011080178196145479,
+      "loss": 1.1859,
+      "step": 7315
+    },
+    {
+      "epoch": 0.6421222685145821,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.001107629410912634,
+      "loss": 1.1638,
+      "step": 7316
+    },
+    {
+      "epoch": 0.6422100380974846,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0011072410557405512,
+      "loss": 1.1869,
+      "step": 7317
+    },
+    {
+      "epoch": 0.642297807680387,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0011068527541366292,
+      "loss": 1.1768,
+      "step": 7318
+    },
+    {
+      "epoch": 0.6423855772632896,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0011064645061391908,
+      "loss": 1.1563,
+      "step": 7319
+    },
+    {
+      "epoch": 0.6424733468461921,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0011060763117865557,
+      "loss": 1.1811,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6425611164290945,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.001105688171117036,
+      "loss": 1.2627,
+      "step": 7321
+    },
+    {
+      "epoch": 0.642648886011997,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00110530008416894,
+      "loss": 1.1634,
+      "step": 7322
+    },
+    {
+      "epoch": 0.6427366555948995,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0011049120509805705,
+      "loss": 1.1544,
+      "step": 7323
+    },
+    {
+      "epoch": 0.642824425177802,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0011045240715902248,
+      "loss": 1.1144,
+      "step": 7324
+    },
+    {
+      "epoch": 0.6429121947607045,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0011041361460361946,
+      "loss": 1.1491,
+      "step": 7325
+    },
+    {
+      "epoch": 0.642999964343607,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0011037482743567668,
+      "loss": 1.1259,
+      "step": 7326
+    },
+    {
+      "epoch": 0.6430877339265094,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0011033604565902232,
+      "loss": 1.1499,
+      "step": 7327
+    },
+    {
+      "epoch": 0.6431755035094119,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0011029726927748397,
+      "loss": 1.148,
+      "step": 7328
+    },
+    {
+      "epoch": 0.6432632730923143,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0011025849829488866,
+      "loss": 1.1747,
+      "step": 7329
+    },
+    {
+      "epoch": 0.6433510426752169,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0011021973271506304,
+      "loss": 1.2126,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6434388122581194,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0011018097254183296,
+      "loss": 1.2099,
+      "step": 7331
+    },
+    {
+      "epoch": 0.6435265818410218,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0011014221777902406,
+      "loss": 1.1144,
+      "step": 7332
+    },
+    {
+      "epoch": 0.6436143514239243,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0011010346843046113,
+      "loss": 1.1617,
+      "step": 7333
+    },
+    {
+      "epoch": 0.6437021210068268,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0011006472449996872,
+      "loss": 1.1637,
+      "step": 7334
+    },
+    {
+      "epoch": 0.6437898905897294,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001100259859913706,
+      "loss": 1.1474,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6438776601726318,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0010998725290849007,
+      "loss": 1.1841,
+      "step": 7336
+    },
+    {
+      "epoch": 0.6439654297555343,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001099485252551501,
+      "loss": 1.1419,
+      "step": 7337
+    },
+    {
+      "epoch": 0.6440531993384367,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0010990980303517285,
+      "loss": 1.2143,
+      "step": 7338
+    },
+    {
+      "epoch": 0.6441409689213392,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0010987108625238004,
+      "loss": 1.2047,
+      "step": 7339
+    },
+    {
+      "epoch": 0.6442287385042418,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0010983237491059292,
+      "loss": 1.1395,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6443165080871442,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0010979366901363201,
+      "loss": 1.1266,
+      "step": 7341
+    },
+    {
+      "epoch": 0.6444042776700467,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0010975496856531766,
+      "loss": 1.1079,
+      "step": 7342
+    },
+    {
+      "epoch": 0.6444920472529492,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0010971627356946923,
+      "loss": 1.1667,
+      "step": 7343
+    },
+    {
+      "epoch": 0.6445798168358516,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001096775840299059,
+      "loss": 1.2228,
+      "step": 7344
+    },
+    {
+      "epoch": 0.6446675864187542,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0010963889995044608,
+      "loss": 1.152,
+      "step": 7345
+    },
+    {
+      "epoch": 0.6447553560016567,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0010960022133490776,
+      "loss": 1.1455,
+      "step": 7346
+    },
+    {
+      "epoch": 0.6448431255845591,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0010956154818710848,
+      "loss": 1.1353,
+      "step": 7347
+    },
+    {
+      "epoch": 0.6449308951674616,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0010952288051086496,
+      "loss": 1.171,
+      "step": 7348
+    },
+    {
+      "epoch": 0.6450186647503641,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0010948421830999364,
+      "loss": 1.1642,
+      "step": 7349
+    },
+    {
+      "epoch": 0.6451064343332666,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0010944556158831032,
+      "loss": 1.1893,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6451942039161691,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0010940691034963027,
+      "loss": 1.159,
+      "step": 7351
+    },
+    {
+      "epoch": 0.6452819734990716,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0010936826459776816,
+      "loss": 1.1697,
+      "step": 7352
+    },
+    {
+      "epoch": 0.645369743081974,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0010932962433653824,
+      "loss": 1.164,
+      "step": 7353
+    },
+    {
+      "epoch": 0.6454575126648765,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0010929098956975413,
+      "loss": 1.1506,
+      "step": 7354
+    },
+    {
+      "epoch": 0.6455452822477791,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0010925236030122885,
+      "loss": 1.1291,
+      "step": 7355
+    },
+    {
+      "epoch": 0.6456330518306815,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0010921373653477505,
+      "loss": 1.1586,
+      "step": 7356
+    },
+    {
+      "epoch": 0.645720821413584,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0010917511827420476,
+      "loss": 1.2451,
+      "step": 7357
+    },
+    {
+      "epoch": 0.6458085909964865,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001091365055233294,
+      "loss": 1.1703,
+      "step": 7358
+    },
+    {
+      "epoch": 0.6458963605793889,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.001090978982859599,
+      "loss": 1.1519,
+      "step": 7359
+    },
+    {
+      "epoch": 0.6459841301622914,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0010905929656590664,
+      "loss": 1.1431,
+      "step": 7360
+    },
+    {
+      "epoch": 0.646071899745194,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0010902070036697949,
+      "loss": 1.1881,
+      "step": 7361
+    },
+    {
+      "epoch": 0.6461596693280964,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.001089821096929877,
+      "loss": 1.152,
+      "step": 7362
+    },
+    {
+      "epoch": 0.6462474389109989,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0010894352454774003,
+      "loss": 1.1487,
+      "step": 7363
+    },
+    {
+      "epoch": 0.6463352084939014,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0010890494493504467,
+      "loss": 1.2132,
+      "step": 7364
+    },
+    {
+      "epoch": 0.6464229780768038,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0010886637085870928,
+      "loss": 1.1879,
+      "step": 7365
+    },
+    {
+      "epoch": 0.6465107476597064,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0010882780232254093,
+      "loss": 1.1915,
+      "step": 7366
+    },
+    {
+      "epoch": 0.6465985172426089,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0010878923933034632,
+      "loss": 1.2057,
+      "step": 7367
+    },
+    {
+      "epoch": 0.6466862868255113,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0010875068188593133,
+      "loss": 1.2067,
+      "step": 7368
+    },
+    {
+      "epoch": 0.6467740564084138,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0010871212999310148,
+      "loss": 1.1513,
+      "step": 7369
+    },
+    {
+      "epoch": 0.6468618259913163,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.001086735836556617,
+      "loss": 1.2339,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6469495955742188,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0010863504287741636,
+      "loss": 1.1471,
+      "step": 7371
+    },
+    {
+      "epoch": 0.6470373651571213,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0010859650766216918,
+      "loss": 1.1353,
+      "step": 7372
+    },
+    {
+      "epoch": 0.6471251347400238,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0010855797801372356,
+      "loss": 1.1863,
+      "step": 7373
+    },
+    {
+      "epoch": 0.6472129043229262,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0010851945393588224,
+      "loss": 1.2163,
+      "step": 7374
+    },
+    {
+      "epoch": 0.6473006739058287,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0010848093543244725,
+      "loss": 1.1878,
+      "step": 7375
+    },
+    {
+      "epoch": 0.6473884434887313,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0010844242250722031,
+      "loss": 1.1544,
+      "step": 7376
+    },
+    {
+      "epoch": 0.6474762130716337,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0010840391516400243,
+      "loss": 1.1305,
+      "step": 7377
+    },
+    {
+      "epoch": 0.6475639826545362,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0010836541340659431,
+      "loss": 1.2138,
+      "step": 7378
+    },
+    {
+      "epoch": 0.6476517522374386,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0010832691723879572,
+      "loss": 1.2414,
+      "step": 7379
+    },
+    {
+      "epoch": 0.6477395218203411,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0010828842666440615,
+      "loss": 1.1234,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6478272914032437,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.001082499416872245,
+      "loss": 1.1633,
+      "step": 7381
+    },
+    {
+      "epoch": 0.6479150609861462,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0010821146231104908,
+      "loss": 1.1926,
+      "step": 7382
+    },
+    {
+      "epoch": 0.6480028305690486,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0010817298853967757,
+      "loss": 1.1721,
+      "step": 7383
+    },
+    {
+      "epoch": 0.6480906001519511,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0010813452037690728,
+      "loss": 1.1369,
+      "step": 7384
+    },
+    {
+      "epoch": 0.6481783697348535,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0010809605782653479,
+      "loss": 1.1265,
+      "step": 7385
+    },
+    {
+      "epoch": 0.648266139317756,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001080576008923563,
+      "loss": 1.1657,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6483539089006586,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0010801914957816717,
+      "loss": 1.192,
+      "step": 7387
+    },
+    {
+      "epoch": 0.648441678483561,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0010798070388776258,
+      "loss": 1.1861,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6485294480664635,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0010794226382493698,
+      "loss": 1.1926,
+      "step": 7389
+    },
+    {
+      "epoch": 0.648617217649366,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001079038293934841,
+      "loss": 1.232,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6487049872322684,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0010786540059719732,
+      "loss": 1.1828,
+      "step": 7391
+    },
+    {
+      "epoch": 0.648792756815171,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0010782697743986947,
+      "loss": 1.1692,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6488805263980735,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.001077885599252927,
+      "loss": 1.1621,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6489682959809759,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0010775014805725872,
+      "loss": 1.2018,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6490560655638784,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0010771174183955858,
+      "loss": 1.1678,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6491438351467809,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0010767334127598283,
+      "loss": 1.1907,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6492316047296834,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0010763494637032148,
+      "loss": 1.1496,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6493193743125859,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0010759655712636392,
+      "loss": 1.138,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6494071438954884,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0010755817354789906,
+      "loss": 1.1731,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6494949134783908,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.001075197956387152,
+      "loss": 1.1525,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6495826830612933,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0010748142340260012,
+      "loss": 1.2266,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6496704526441959,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0010744305684334085,
+      "loss": 1.1505,
+      "step": 7402
+    },
+    {
+      "epoch": 0.6497582222270983,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0010740469596472418,
+      "loss": 1.1368,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6498459918100008,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0010736634077053613,
+      "loss": 1.102,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6499337613929033,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.001073279912645622,
+      "loss": 1.1294,
+      "step": 7405
+    },
+    {
+      "epoch": 0.6500215309758057,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0010728964745058734,
+      "loss": 1.0945,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6501093005587083,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0010725130933239586,
+      "loss": 1.1717,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6501970701416108,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0010721297691377168,
+      "loss": 1.2247,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6502848397245132,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0010717465019849813,
+      "loss": 1.2197,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6503726093074157,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0010713632919035771,
+      "loss": 1.1236,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6504603788903182,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0010709801389313263,
+      "loss": 1.1849,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6505481484732206,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.001070597043106045,
+      "loss": 1.1478,
+      "step": 7412
+    },
+    {
+      "epoch": 0.6506359180561232,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0010702140044655433,
+      "loss": 1.1462,
+      "step": 7413
+    },
+    {
+      "epoch": 0.6507236876390257,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0010698310230476248,
+      "loss": 1.1684,
+      "step": 7414
+    },
+    {
+      "epoch": 0.6508114572219281,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.001069448098890089,
+      "loss": 1.1847,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6508992268048306,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0010690652320307283,
+      "loss": 1.2024,
+      "step": 7416
+    },
+    {
+      "epoch": 0.650986996387733,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0010686824225073308,
+      "loss": 1.1414,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6510747659706356,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0010682996703576778,
+      "loss": 1.1509,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6511625355535381,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.001067916975619546,
+      "loss": 1.189,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6512503051364406,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0010675343383307052,
+      "loss": 1.1827,
+      "step": 7420
+    },
+    {
+      "epoch": 0.651338074719343,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0010671517585289212,
+      "loss": 1.1685,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6514258443022455,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0010667692362519518,
+      "loss": 1.1659,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6515136138851481,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0010663867715375514,
+      "loss": 1.159,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6516013834680505,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0010660043644234668,
+      "loss": 1.1856,
+      "step": 7424
+    },
+    {
+      "epoch": 0.651689153050953,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.001065622014947442,
+      "loss": 1.2044,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6517769226338554,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001065239723147211,
+      "loss": 1.1827,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6518646922167579,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0010648574890605048,
+      "loss": 1.1387,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6519524617996605,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0010644753127250501,
+      "loss": 1.1702,
+      "step": 7428
+    },
+    {
+      "epoch": 0.652040231382563,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0010640931941785657,
+      "loss": 1.1393,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6521280009654654,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.001063711133458764,
+      "loss": 1.1745,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6522157705483679,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0010633291306033537,
+      "loss": 1.1578,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6523035401312703,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0010629471856500369,
+      "loss": 1.1797,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6523913097141729,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0010625652986365102,
+      "loss": 1.192,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6524790792970754,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0010621834696004642,
+      "loss": 1.154,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6525668488799778,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0010618016985795837,
+      "loss": 1.1647,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6526546184628803,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.001061419985611548,
+      "loss": 1.1697,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6527423880457828,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.001061038330734031,
+      "loss": 1.1335,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6528301576286853,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0010606567339847004,
+      "loss": 1.142,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6529179272115878,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0010602751954012178,
+      "loss": 1.2087,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6530056967944903,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0010598937150212411,
+      "loss": 1.178,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6530934663773927,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.001059512292882419,
+      "loss": 1.1701,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6531812359602952,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.001059130929022397,
+      "loss": 1.1621,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6532690055431977,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0010587496234788145,
+      "loss": 1.1816,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6533567751261002,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0010583683762893047,
+      "loss": 1.1756,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6534445447090027,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0010579871874914957,
+      "loss": 1.1811,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6535323142919052,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0010576060571230084,
+      "loss": 1.1811,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6536200838748076,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0010572249852214585,
+      "loss": 1.1828,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6537078534577101,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0010568439718244584,
+      "loss": 1.1525,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6537956230406127,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0010564630169696107,
+      "loss": 1.1948,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6538833926235151,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0010560821206945146,
+      "loss": 1.1936,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6539711622064176,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0010557012830367631,
+      "loss": 1.1798,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6540589317893201,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0010553205040339441,
+      "loss": 1.1943,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6541467013722225,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0010549397837236378,
+      "loss": 1.1061,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6542344709551251,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0010545591221434204,
+      "loss": 1.1858,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6543222405380276,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0010541785193308613,
+      "loss": 1.1468,
+      "step": 7455
+    },
+    {
+      "epoch": 0.65441001012093,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0010537979753235255,
+      "loss": 1.2189,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6544977797038325,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0010534174901589697,
+      "loss": 1.1345,
+      "step": 7457
+    },
+    {
+      "epoch": 0.654585549286735,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0010530370638747472,
+      "loss": 1.1063,
+      "step": 7458
+    },
+    {
+      "epoch": 0.6546733188696375,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.001052656696508405,
+      "loss": 1.1867,
+      "step": 7459
+    },
+    {
+      "epoch": 0.65476108845254,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0010522763880974832,
+      "loss": 1.1661,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6548488580354425,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0010518961386795163,
+      "loss": 1.1643,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6549366276183449,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0010515159482920344,
+      "loss": 1.1617,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6550243972012474,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00105113581697256,
+      "loss": 1.1695,
+      "step": 7463
+    },
+    {
+      "epoch": 0.65511216678415,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0010507557447586113,
+      "loss": 1.2014,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6551999363670524,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0010503757316876992,
+      "loss": 1.2037,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6552877059499549,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0010499957777973298,
+      "loss": 1.1516,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6553754755328574,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0010496158831250027,
+      "loss": 1.1988,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6554632451157598,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0010492360477082129,
+      "loss": 1.1087,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6555510146986623,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0010488562715844476,
+      "loss": 1.1695,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6556387842815649,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00104847655479119,
+      "loss": 1.1787,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6557265538644673,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0010480968973659163,
+      "loss": 1.2096,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6558143234473698,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0010477172993460975,
+      "loss": 1.1342,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6559020930302722,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0010473377607691978,
+      "loss": 1.1352,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6559898626131747,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0010469582816726765,
+      "loss": 1.1584,
+      "step": 7474
+    },
+    {
+      "epoch": 0.6560776321960773,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0010465788620939865,
+      "loss": 1.1493,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6561654017789798,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.001046199502070576,
+      "loss": 1.1838,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6562531713618822,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0010458202016398843,
+      "loss": 1.1783,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6563409409447847,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0010454409608393485,
+      "loss": 1.0811,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6564287105276871,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0010450617797063985,
+      "loss": 1.2077,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6565164801105897,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0010446826582784568,
+      "loss": 1.1302,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6566042496934922,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0010443035965929414,
+      "loss": 1.1415,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6566920192763946,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0010439245946872642,
+      "loss": 1.2063,
+      "step": 7482
+    },
+    {
+      "epoch": 0.6567797888592971,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0010435456525988316,
+      "loss": 1.1844,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6568675584421996,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.001043166770365044,
+      "loss": 1.1675,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6569553280251021,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0010427879480232947,
+      "loss": 1.1046,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6570430976080046,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.001042409185610972,
+      "loss": 1.2051,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6571308671909071,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0010420304831654586,
+      "loss": 1.1564,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6572186367738095,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.001041651840724131,
+      "loss": 1.211,
+      "step": 7488
+    },
+    {
+      "epoch": 0.657306406356712,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0010412732583243596,
+      "loss": 1.1553,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6573941759396146,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0010408947360035091,
+      "loss": 1.2304,
+      "step": 7490
+    },
+    {
+      "epoch": 0.657481945522517,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0010405162737989384,
+      "loss": 1.1062,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6575697151054195,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0010401378717479995,
+      "loss": 1.1875,
+      "step": 7492
+    },
+    {
+      "epoch": 0.657657484688322,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0010397595298880394,
+      "loss": 1.1181,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6577452542712244,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0010393812482563992,
+      "loss": 1.197,
+      "step": 7494
+    },
+    {
+      "epoch": 0.657833023854127,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0010390030268904139,
+      "loss": 1.1492,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6579207934370295,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0010386248658274123,
+      "loss": 1.1897,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6580085630199319,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0010382467651047166,
+      "loss": 1.1576,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6580963326028344,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.001037868724759645,
+      "loss": 1.157,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6581841021857369,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.001037490744829509,
+      "loss": 1.2018,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6582718717686393,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.001037112825351612,
+      "loss": 1.1524,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6582718717686393,
+      "eval_loss": 1.1542067527770996,
+      "eval_runtime": 437.527,
+      "eval_samples_per_second": 33.689,
+      "eval_steps_per_second": 8.422,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6583596413515419,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0010367349663632543,
+      "loss": 1.1367,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6584474109344444,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0010363571679017287,
+      "loss": 1.1746,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6585351805173468,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.001035979430004323,
+      "loss": 1.1149,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6586229501002493,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.001035601752708318,
+      "loss": 1.1887,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6587107196831518,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0010352241360509884,
+      "loss": 1.0863,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6587984892660543,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0010348465800696042,
+      "loss": 1.1251,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6588862588489568,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0010344690848014286,
+      "loss": 1.1696,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6589740284318593,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0010340916502837186,
+      "loss": 1.1502,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6590617980147617,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001033714276553726,
+      "loss": 1.2118,
+      "step": 7509
+    },
+    {
+      "epoch": 0.6591495675976642,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0010333369636486955,
+      "loss": 1.1022,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6592373371805668,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0010329597116058673,
+      "loss": 1.179,
+      "step": 7511
+    },
+    {
+      "epoch": 0.6593251067634692,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0010325825204624739,
+      "loss": 1.1378,
+      "step": 7512
+    },
+    {
+      "epoch": 0.6594128763463717,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.001032205390255742,
+      "loss": 1.1667,
+      "step": 7513
+    },
+    {
+      "epoch": 0.6595006459292742,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.001031828321022894,
+      "loss": 1.1835,
+      "step": 7514
+    },
+    {
+      "epoch": 0.6595884155121766,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.001031451312801145,
+      "loss": 1.1409,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6596761850950792,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.001031074365627704,
+      "loss": 1.1511,
+      "step": 7516
+    },
+    {
+      "epoch": 0.6597639546779817,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0010306974795397735,
+      "loss": 1.2515,
+      "step": 7517
+    },
+    {
+      "epoch": 0.6598517242608841,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0010303206545745512,
+      "loss": 1.1652,
+      "step": 7518
+    },
+    {
+      "epoch": 0.6599394938437866,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0010299438907692294,
+      "loss": 1.1844,
+      "step": 7519
+    },
+    {
+      "epoch": 0.660027263426689,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0010295671881609915,
+      "loss": 1.2179,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6601150330095916,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0010291905467870173,
+      "loss": 1.1929,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6602028025924941,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0010288139666844796,
+      "loss": 1.1631,
+      "step": 7522
+    },
+    {
+      "epoch": 0.6602905721753966,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.001028437447890546,
+      "loss": 1.1859,
+      "step": 7523
+    },
+    {
+      "epoch": 0.660378341758299,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0010280609904423762,
+      "loss": 1.1383,
+      "step": 7524
+    },
+    {
+      "epoch": 0.6604661113412015,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.001027684594377126,
+      "loss": 1.1695,
+      "step": 7525
+    },
+    {
+      "epoch": 0.6605538809241039,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0010273082597319438,
+      "loss": 1.2506,
+      "step": 7526
+    },
+    {
+      "epoch": 0.6606416505070065,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0010269319865439733,
+      "loss": 1.1903,
+      "step": 7527
+    },
+    {
+      "epoch": 0.660729420089909,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0010265557748503492,
+      "loss": 1.2044,
+      "step": 7528
+    },
+    {
+      "epoch": 0.6608171896728114,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0010261796246882034,
+      "loss": 1.1101,
+      "step": 7529
+    },
+    {
+      "epoch": 0.6609049592557139,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0010258035360946613,
+      "loss": 1.2722,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6609927288386164,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0010254275091068398,
+      "loss": 1.1531,
+      "step": 7531
+    },
+    {
+      "epoch": 0.661080498421519,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0010250515437618516,
+      "loss": 1.1621,
+      "step": 7532
+    },
+    {
+      "epoch": 0.6611682680044214,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0010246756400968035,
+      "loss": 1.1826,
+      "step": 7533
+    },
+    {
+      "epoch": 0.6612560375873239,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0010242997981487953,
+      "loss": 1.162,
+      "step": 7534
+    },
+    {
+      "epoch": 0.6613438071702263,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.001023924017954922,
+      "loss": 1.0968,
+      "step": 7535
+    },
+    {
+      "epoch": 0.6614315767531288,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0010235482995522701,
+      "loss": 1.1565,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6615193463360314,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0010231726429779221,
+      "loss": 1.1429,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6616071159189338,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0010227970482689544,
+      "loss": 1.1905,
+      "step": 7538
+    },
+    {
+      "epoch": 0.6616948855018363,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0010224215154624363,
+      "loss": 1.1633,
+      "step": 7539
+    },
+    {
+      "epoch": 0.6617826550847388,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.001022046044595431,
+      "loss": 1.1435,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6618704246676412,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0010216706357049968,
+      "loss": 1.1591,
+      "step": 7541
+    },
+    {
+      "epoch": 0.6619581942505438,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.001021295288828185,
+      "loss": 1.1754,
+      "step": 7542
+    },
+    {
+      "epoch": 0.6620459638334463,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0010209200040020403,
+      "loss": 1.1583,
+      "step": 7543
+    },
+    {
+      "epoch": 0.6621337334163487,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.001020544781263602,
+      "loss": 1.1559,
+      "step": 7544
+    },
+    {
+      "epoch": 0.6622215029992512,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001020169620649903,
+      "loss": 1.2208,
+      "step": 7545
+    },
+    {
+      "epoch": 0.6623092725821537,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0010197945221979703,
+      "loss": 1.1291,
+      "step": 7546
+    },
+    {
+      "epoch": 0.6623970421650562,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0010194194859448253,
+      "loss": 1.1563,
+      "step": 7547
+    },
+    {
+      "epoch": 0.6624848117479587,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0010190445119274809,
+      "loss": 1.1922,
+      "step": 7548
+    },
+    {
+      "epoch": 0.6625725813308612,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.001018669600182947,
+      "loss": 1.1005,
+      "step": 7549
+    },
+    {
+      "epoch": 0.6626603509137636,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0010182947507482263,
+      "loss": 1.1426,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6627481204966661,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0010179199636603134,
+      "loss": 1.1628,
+      "step": 7551
+    },
+    {
+      "epoch": 0.6628358900795686,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.001017545238956199,
+      "loss": 1.1081,
+      "step": 7552
+    },
+    {
+      "epoch": 0.6629236596624711,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0010171705766728672,
+      "loss": 1.172,
+      "step": 7553
+    },
+    {
+      "epoch": 0.6630114292453736,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.001016795976847296,
+      "loss": 1.2612,
+      "step": 7554
+    },
+    {
+      "epoch": 0.6630991988282761,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0010164214395164558,
+      "loss": 1.1134,
+      "step": 7555
+    },
+    {
+      "epoch": 0.6631869684111785,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0010160469647173128,
+      "loss": 1.1254,
+      "step": 7556
+    },
+    {
+      "epoch": 0.663274737994081,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0010156725524868258,
+      "loss": 1.193,
+      "step": 7557
+    },
+    {
+      "epoch": 0.6633625075769836,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0010152982028619481,
+      "loss": 1.1669,
+      "step": 7558
+    },
+    {
+      "epoch": 0.663450277159886,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.001014923915879626,
+      "loss": 1.2107,
+      "step": 7559
+    },
+    {
+      "epoch": 0.6635380467427885,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.001014549691576801,
+      "loss": 1.1502,
+      "step": 7560
+    },
+    {
+      "epoch": 0.663625816325691,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001014175529990407,
+      "loss": 1.1681,
+      "step": 7561
+    },
+    {
+      "epoch": 0.6637135859085934,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0010138014311573725,
+      "loss": 1.217,
+      "step": 7562
+    },
+    {
+      "epoch": 0.663801355491496,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0010134273951146193,
+      "loss": 1.2235,
+      "step": 7563
+    },
+    {
+      "epoch": 0.6638891250743985,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0010130534218990633,
+      "loss": 1.2234,
+      "step": 7564
+    },
+    {
+      "epoch": 0.6639768946573009,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0010126795115476144,
+      "loss": 1.1258,
+      "step": 7565
+    },
+    {
+      "epoch": 0.6640646642402034,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0010123056640971764,
+      "loss": 1.1986,
+      "step": 7566
+    },
+    {
+      "epoch": 0.6641524338231058,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0010119318795846453,
+      "loss": 1.1504,
+      "step": 7567
+    },
+    {
+      "epoch": 0.6642402034060084,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.001011558158046913,
+      "loss": 1.2351,
+      "step": 7568
+    },
+    {
+      "epoch": 0.6643279729889109,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0010111844995208646,
+      "loss": 1.169,
+      "step": 7569
+    },
+    {
+      "epoch": 0.6644157425718134,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0010108109040433786,
+      "loss": 1.1816,
+      "step": 7570
+    },
+    {
+      "epoch": 0.6645035121547158,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0010104373716513272,
+      "loss": 1.1892,
+      "step": 7571
+    },
+    {
+      "epoch": 0.6645912817376183,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0010100639023815764,
+      "loss": 1.1593,
+      "step": 7572
+    },
+    {
+      "epoch": 0.6646790513205209,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0010096904962709863,
+      "loss": 1.1053,
+      "step": 7573
+    },
+    {
+      "epoch": 0.6647668209034233,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0010093171533564114,
+      "loss": 1.1776,
+      "step": 7574
+    },
+    {
+      "epoch": 0.6648545904863258,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0010089438736746977,
+      "loss": 1.188,
+      "step": 7575
+    },
+    {
+      "epoch": 0.6649423600692282,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.001008570657262687,
+      "loss": 1.1419,
+      "step": 7576
+    },
+    {
+      "epoch": 0.6650301296521307,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0010081975041572146,
+      "loss": 1.1453,
+      "step": 7577
+    },
+    {
+      "epoch": 0.6651178992350333,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0010078244143951092,
+      "loss": 1.1999,
+      "step": 7578
+    },
+    {
+      "epoch": 0.6652056688179357,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0010074513880131927,
+      "loss": 1.1856,
+      "step": 7579
+    },
+    {
+      "epoch": 0.6652934384008382,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0010070784250482822,
+      "loss": 1.1643,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6653812079837407,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0010067055255371874,
+      "loss": 1.1458,
+      "step": 7581
+    },
+    {
+      "epoch": 0.6654689775666431,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0010063326895167117,
+      "loss": 1.1482,
+      "step": 7582
+    },
+    {
+      "epoch": 0.6655567471495456,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0010059599170236522,
+      "loss": 1.1127,
+      "step": 7583
+    },
+    {
+      "epoch": 0.6656445167324482,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.001005587208094801,
+      "loss": 1.1735,
+      "step": 7584
+    },
+    {
+      "epoch": 0.6657322863153506,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0010052145627669424,
+      "loss": 1.1611,
+      "step": 7585
+    },
+    {
+      "epoch": 0.6658200558982531,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0010048419810768556,
+      "loss": 1.1656,
+      "step": 7586
+    },
+    {
+      "epoch": 0.6659078254811556,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0010044694630613122,
+      "loss": 1.1401,
+      "step": 7587
+    },
+    {
+      "epoch": 0.665995595064058,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0010040970087570778,
+      "loss": 1.1751,
+      "step": 7588
+    },
+    {
+      "epoch": 0.6660833646469606,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.001003724618200914,
+      "loss": 1.1848,
+      "step": 7589
+    },
+    {
+      "epoch": 0.6661711342298631,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0010033522914295729,
+      "loss": 1.1616,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6662589038127655,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0010029800284798017,
+      "loss": 1.1242,
+      "step": 7591
+    },
+    {
+      "epoch": 0.666346673395668,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0010026078293883414,
+      "loss": 1.1533,
+      "step": 7592
+    },
+    {
+      "epoch": 0.6664344429785705,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0010022356941919277,
+      "loss": 1.1897,
+      "step": 7593
+    },
+    {
+      "epoch": 0.666522212561473,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0010018636229272872,
+      "loss": 1.1415,
+      "step": 7594
+    },
+    {
+      "epoch": 0.6666099821443755,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0010014916156311428,
+      "loss": 1.1156,
+      "step": 7595
+    },
+    {
+      "epoch": 0.666697751727278,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0010011196723402095,
+      "loss": 1.1491,
+      "step": 7596
+    },
+    {
+      "epoch": 0.6667855213101804,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0010007477930911973,
+      "loss": 1.2224,
+      "step": 7597
+    },
+    {
+      "epoch": 0.6668732908930829,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.001000375977920809,
+      "loss": 1.1837,
+      "step": 7598
+    },
+    {
+      "epoch": 0.6669610604759855,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0010000042268657411,
+      "loss": 1.1603,
+      "step": 7599
+    },
+    {
+      "epoch": 0.6670488300588879,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0009996325399626842,
+      "loss": 1.1426,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6671365996417904,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0009992609172483227,
+      "loss": 1.1939,
+      "step": 7601
+    },
+    {
+      "epoch": 0.6672243692246929,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0009988893587593334,
+      "loss": 1.1789,
+      "step": 7602
+    },
+    {
+      "epoch": 0.6673121388075953,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.000998517864532388,
+      "loss": 1.1577,
+      "step": 7603
+    },
+    {
+      "epoch": 0.6673999083904979,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0009981464346041514,
+      "loss": 1.1166,
+      "step": 7604
+    },
+    {
+      "epoch": 0.6674876779734004,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0009977750690112832,
+      "loss": 1.1571,
+      "step": 7605
+    },
+    {
+      "epoch": 0.6675754475563028,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0009974037677904345,
+      "loss": 1.19,
+      "step": 7606
+    },
+    {
+      "epoch": 0.6676632171392053,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0009970325309782517,
+      "loss": 1.1368,
+      "step": 7607
+    },
+    {
+      "epoch": 0.6677509867221078,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0009966613586113737,
+      "loss": 1.1521,
+      "step": 7608
+    },
+    {
+      "epoch": 0.6678387563050102,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0009962902507264359,
+      "loss": 1.1355,
+      "step": 7609
+    },
+    {
+      "epoch": 0.6679265258879128,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0009959192073600628,
+      "loss": 1.1424,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6680142954708153,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.000995548228548876,
+      "loss": 1.1445,
+      "step": 7611
+    },
+    {
+      "epoch": 0.6681020650537177,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0009951773143294896,
+      "loss": 1.1468,
+      "step": 7612
+    },
+    {
+      "epoch": 0.6681898346366202,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.000994806464738512,
+      "loss": 1.1588,
+      "step": 7613
+    },
+    {
+      "epoch": 0.6682776042195226,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.000994435679812543,
+      "loss": 1.157,
+      "step": 7614
+    },
+    {
+      "epoch": 0.6683653738024252,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0009940649595881782,
+      "loss": 1.1971,
+      "step": 7615
+    },
+    {
+      "epoch": 0.6684531433853277,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.000993694304102007,
+      "loss": 1.1247,
+      "step": 7616
+    },
+    {
+      "epoch": 0.6685409129682301,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0009933237133906115,
+      "loss": 1.1375,
+      "step": 7617
+    },
+    {
+      "epoch": 0.6686286825511326,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.000992953187490566,
+      "loss": 1.155,
+      "step": 7618
+    },
+    {
+      "epoch": 0.6687164521340351,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0009925827264384418,
+      "loss": 1.1406,
+      "step": 7619
+    },
+    {
+      "epoch": 0.6688042217169377,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.000992212330270801,
+      "loss": 1.2041,
+      "step": 7620
+    },
+    {
+      "epoch": 0.6688919912998401,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0009918419990242013,
+      "loss": 1.1697,
+      "step": 7621
+    },
+    {
+      "epoch": 0.6689797608827426,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0009914717327351914,
+      "loss": 1.1396,
+      "step": 7622
+    },
+    {
+      "epoch": 0.669067530465645,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0009911015314403159,
+      "loss": 1.1588,
+      "step": 7623
+    },
+    {
+      "epoch": 0.6691553000485475,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0009907313951761118,
+      "loss": 1.1473,
+      "step": 7624
+    },
+    {
+      "epoch": 0.6692430696314501,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0009903613239791113,
+      "loss": 1.1636,
+      "step": 7625
+    },
+    {
+      "epoch": 0.6693308392143525,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0009899913178858372,
+      "loss": 1.1066,
+      "step": 7626
+    },
+    {
+      "epoch": 0.669418608797255,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009896213769328087,
+      "loss": 1.1912,
+      "step": 7627
+    },
+    {
+      "epoch": 0.6695063783801575,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009892515011565373,
+      "loss": 1.2135,
+      "step": 7628
+    },
+    {
+      "epoch": 0.6695941479630599,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0009888816905935284,
+      "loss": 1.2189,
+      "step": 7629
+    },
+    {
+      "epoch": 0.6696819175459625,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0009885119452802806,
+      "loss": 1.0823,
+      "step": 7630
+    },
+    {
+      "epoch": 0.669769687128865,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0009881422652532866,
+      "loss": 1.1698,
+      "step": 7631
+    },
+    {
+      "epoch": 0.6698574567117674,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0009877726505490324,
+      "loss": 1.1805,
+      "step": 7632
+    },
+    {
+      "epoch": 0.6699452262946699,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0009874031012039968,
+      "loss": 1.1808,
+      "step": 7633
+    },
+    {
+      "epoch": 0.6700329958775724,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0009870336172546536,
+      "loss": 1.1631,
+      "step": 7634
+    },
+    {
+      "epoch": 0.6701207654604749,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0009866641987374691,
+      "loss": 1.1339,
+      "step": 7635
+    },
+    {
+      "epoch": 0.6702085350433774,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0009862948456889034,
+      "loss": 1.1956,
+      "step": 7636
+    },
+    {
+      "epoch": 0.6702963046262799,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0009859255581454107,
+      "loss": 1.1833,
+      "step": 7637
+    },
+    {
+      "epoch": 0.6703840742091823,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0009855563361434374,
+      "loss": 1.1233,
+      "step": 7638
+    },
+    {
+      "epoch": 0.6704718437920848,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.000985187179719424,
+      "loss": 1.1722,
+      "step": 7639
+    },
+    {
+      "epoch": 0.6705596133749873,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0009848180889098065,
+      "loss": 1.162,
+      "step": 7640
+    },
+    {
+      "epoch": 0.6706473829578898,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0009844490637510113,
+      "loss": 1.173,
+      "step": 7641
+    },
+    {
+      "epoch": 0.6707351525407923,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.00098408010427946,
+      "loss": 1.1737,
+      "step": 7642
+    },
+    {
+      "epoch": 0.6708229221236948,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0009837112105315674,
+      "loss": 1.1646,
+      "step": 7643
+    },
+    {
+      "epoch": 0.6709106917065972,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0009833423825437425,
+      "loss": 1.0963,
+      "step": 7644
+    },
+    {
+      "epoch": 0.6709984612894997,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.000982973620352386,
+      "loss": 1.1699,
+      "step": 7645
+    },
+    {
+      "epoch": 0.6710862308724023,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.000982604923993894,
+      "loss": 1.1752,
+      "step": 7646
+    },
+    {
+      "epoch": 0.6711740004553047,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0009822362935046553,
+      "loss": 1.1581,
+      "step": 7647
+    },
+    {
+      "epoch": 0.6712617700382072,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0009818677289210524,
+      "loss": 1.1628,
+      "step": 7648
+    },
+    {
+      "epoch": 0.6713495396211097,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0009814992302794607,
+      "loss": 1.1635,
+      "step": 7649
+    },
+    {
+      "epoch": 0.6714373092040121,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0009811307976162497,
+      "loss": 1.1884,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6715250787869147,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0009807624309677828,
+      "loss": 1.1399,
+      "step": 7651
+    },
+    {
+      "epoch": 0.6716128483698172,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0009803941303704162,
+      "loss": 1.2139,
+      "step": 7652
+    },
+    {
+      "epoch": 0.6717006179527196,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0009800258958604989,
+      "loss": 1.1877,
+      "step": 7653
+    },
+    {
+      "epoch": 0.6717883875356221,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0009796577274743747,
+      "loss": 1.1546,
+      "step": 7654
+    },
+    {
+      "epoch": 0.6718761571185246,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0009792896252483806,
+      "loss": 1.2116,
+      "step": 7655
+    },
+    {
+      "epoch": 0.6719639267014271,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000978921589218847,
+      "loss": 1.1654,
+      "step": 7656
+    },
+    {
+      "epoch": 0.6720516962843296,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0009785536194220966,
+      "loss": 1.2392,
+      "step": 7657
+    },
+    {
+      "epoch": 0.672139465867232,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0009781857158944474,
+      "loss": 1.1494,
+      "step": 7658
+    },
+    {
+      "epoch": 0.6722272354501345,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.000977817878672209,
+      "loss": 1.1945,
+      "step": 7659
+    },
+    {
+      "epoch": 0.672315005033037,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0009774501077916876,
+      "loss": 1.1894,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6724027746159396,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0009770824032891787,
+      "loss": 1.1601,
+      "step": 7661
+    },
+    {
+      "epoch": 0.672490544198842,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0009767147652009743,
+      "loss": 1.1335,
+      "step": 7662
+    },
+    {
+      "epoch": 0.6725783137817445,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0009763471935633583,
+      "loss": 1.1646,
+      "step": 7663
+    },
+    {
+      "epoch": 0.672666083364647,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0009759796884126091,
+      "loss": 1.1935,
+      "step": 7664
+    },
+    {
+      "epoch": 0.6727538529475494,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0009756122497849975,
+      "loss": 1.1513,
+      "step": 7665
+    },
+    {
+      "epoch": 0.6728416225304519,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0009752448777167882,
+      "loss": 1.2074,
+      "step": 7666
+    },
+    {
+      "epoch": 0.6729293921133545,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0009748775722442397,
+      "loss": 1.1713,
+      "step": 7667
+    },
+    {
+      "epoch": 0.6730171616962569,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000974510333403604,
+      "loss": 1.1759,
+      "step": 7668
+    },
+    {
+      "epoch": 0.6731049312791594,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0009741431612311248,
+      "loss": 1.1839,
+      "step": 7669
+    },
+    {
+      "epoch": 0.6731927008620618,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0009737760557630415,
+      "loss": 1.1683,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6732804704449643,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0009734090170355861,
+      "loss": 1.1198,
+      "step": 7671
+    },
+    {
+      "epoch": 0.6733682400278669,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000973042045084984,
+      "loss": 1.1495,
+      "step": 7672
+    },
+    {
+      "epoch": 0.6734560096107693,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0009726751399474528,
+      "loss": 1.1711,
+      "step": 7673
+    },
+    {
+      "epoch": 0.6735437791936718,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0009723083016592053,
+      "loss": 1.2031,
+      "step": 7674
+    },
+    {
+      "epoch": 0.6736315487765743,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0009719415302564471,
+      "loss": 1.1721,
+      "step": 7675
+    },
+    {
+      "epoch": 0.6737193183594767,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0009715748257753773,
+      "loss": 1.2159,
+      "step": 7676
+    },
+    {
+      "epoch": 0.6738070879423793,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0009712081882521875,
+      "loss": 1.2307,
+      "step": 7677
+    },
+    {
+      "epoch": 0.6738948575252818,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0009708416177230634,
+      "loss": 1.1604,
+      "step": 7678
+    },
+    {
+      "epoch": 0.6739826271081842,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0009704751142241843,
+      "loss": 1.1815,
+      "step": 7679
+    },
+    {
+      "epoch": 0.6740703966910867,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0009701086777917229,
+      "loss": 1.0932,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6741581662739892,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0009697423084618447,
+      "loss": 1.168,
+      "step": 7681
+    },
+    {
+      "epoch": 0.6742459358568917,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0009693760062707093,
+      "loss": 1.1439,
+      "step": 7682
+    },
+    {
+      "epoch": 0.6743337054397942,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0009690097712544691,
+      "loss": 1.1661,
+      "step": 7683
+    },
+    {
+      "epoch": 0.6744214750226967,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0009686436034492695,
+      "loss": 1.1484,
+      "step": 7684
+    },
+    {
+      "epoch": 0.6745092446055991,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0009682775028912504,
+      "loss": 1.202,
+      "step": 7685
+    },
+    {
+      "epoch": 0.6745970141885016,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0009679114696165442,
+      "loss": 1.1418,
+      "step": 7686
+    },
+    {
+      "epoch": 0.6746847837714042,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0009675455036612772,
+      "loss": 1.143,
+      "step": 7687
+    },
+    {
+      "epoch": 0.6747725533543066,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.000967179605061569,
+      "loss": 1.1094,
+      "step": 7688
+    },
+    {
+      "epoch": 0.6748603229372091,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0009668137738535309,
+      "loss": 1.1096,
+      "step": 7689
+    },
+    {
+      "epoch": 0.6749480925201116,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0009664480100732707,
+      "loss": 1.1566,
+      "step": 7690
+    },
+    {
+      "epoch": 0.675035862103014,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.000966082313756888,
+      "loss": 1.1705,
+      "step": 7691
+    },
+    {
+      "epoch": 0.6751236316859165,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.000965716684940474,
+      "loss": 1.1592,
+      "step": 7692
+    },
+    {
+      "epoch": 0.6752114012688191,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0009653511236601156,
+      "loss": 1.1951,
+      "step": 7693
+    },
+    {
+      "epoch": 0.6752991708517215,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0009649856299518925,
+      "loss": 1.1524,
+      "step": 7694
+    },
+    {
+      "epoch": 0.675386940434624,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0009646202038518777,
+      "loss": 1.2181,
+      "step": 7695
+    },
+    {
+      "epoch": 0.6754747100175265,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0009642548453961363,
+      "loss": 1.1886,
+      "step": 7696
+    },
+    {
+      "epoch": 0.6755624796004289,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0009638895546207283,
+      "loss": 1.1406,
+      "step": 7697
+    },
+    {
+      "epoch": 0.6756502491833315,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0009635243315617063,
+      "loss": 1.1723,
+      "step": 7698
+    },
+    {
+      "epoch": 0.675738018766234,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0009631591762551168,
+      "loss": 1.2226,
+      "step": 7699
+    },
+    {
+      "epoch": 0.6758257883491364,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0009627940887369988,
+      "loss": 1.1631,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6759135579320389,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.000962429069043385,
+      "loss": 1.193,
+      "step": 7701
+    },
+    {
+      "epoch": 0.6760013275149414,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0009620641172103018,
+      "loss": 1.162,
+      "step": 7702
+    },
+    {
+      "epoch": 0.6760890970978439,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0009616992332737685,
+      "loss": 1.2011,
+      "step": 7703
+    },
+    {
+      "epoch": 0.6761768666807464,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009613344172697968,
+      "loss": 1.1268,
+      "step": 7704
+    },
+    {
+      "epoch": 0.6762646362636489,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0009609696692343934,
+      "loss": 1.1433,
+      "step": 7705
+    },
+    {
+      "epoch": 0.6763524058465513,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0009606049892035571,
+      "loss": 1.1454,
+      "step": 7706
+    },
+    {
+      "epoch": 0.6764401754294538,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0009602403772132812,
+      "loss": 1.1823,
+      "step": 7707
+    },
+    {
+      "epoch": 0.6765279450123564,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0009598758332995505,
+      "loss": 1.142,
+      "step": 7708
+    },
+    {
+      "epoch": 0.6766157145952588,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0009595113574983433,
+      "loss": 1.1685,
+      "step": 7709
+    },
+    {
+      "epoch": 0.6767034841781613,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0009591469498456342,
+      "loss": 1.2022,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6767912537610637,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0009587826103773877,
+      "loss": 1.1981,
+      "step": 7711
+    },
+    {
+      "epoch": 0.6768790233439662,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.000958418339129562,
+      "loss": 1.1564,
+      "step": 7712
+    },
+    {
+      "epoch": 0.6769667929268688,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0009580541361381101,
+      "loss": 1.1694,
+      "step": 7713
+    },
+    {
+      "epoch": 0.6770545625097713,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0009576900014389766,
+      "loss": 1.1816,
+      "step": 7714
+    },
+    {
+      "epoch": 0.6771423320926737,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0009573259350681015,
+      "loss": 1.1547,
+      "step": 7715
+    },
+    {
+      "epoch": 0.6772301016755762,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0009569619370614152,
+      "loss": 1.2021,
+      "step": 7716
+    },
+    {
+      "epoch": 0.6773178712584786,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0009565980074548434,
+      "loss": 1.1702,
+      "step": 7717
+    },
+    {
+      "epoch": 0.6774056408413812,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0009562341462843047,
+      "loss": 1.165,
+      "step": 7718
+    },
+    {
+      "epoch": 0.6774934104242837,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0009558703535857109,
+      "loss": 1.1516,
+      "step": 7719
+    },
+    {
+      "epoch": 0.6775811800071861,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0009555066293949663,
+      "loss": 1.1533,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6776689495900886,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0009551429737479696,
+      "loss": 1.1167,
+      "step": 7721
+    },
+    {
+      "epoch": 0.6777567191729911,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0009547793866806117,
+      "loss": 1.1483,
+      "step": 7722
+    },
+    {
+      "epoch": 0.6778444887558935,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0009544158682287785,
+      "loss": 1.099,
+      "step": 7723
+    },
+    {
+      "epoch": 0.6779322583387961,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0009540524184283457,
+      "loss": 1.1579,
+      "step": 7724
+    },
+    {
+      "epoch": 0.6780200279216986,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.000953689037315186,
+      "loss": 1.1459,
+      "step": 7725
+    },
+    {
+      "epoch": 0.678107797504601,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.000953325724925163,
+      "loss": 1.1634,
+      "step": 7726
+    },
+    {
+      "epoch": 0.6781955670875035,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0009529624812941349,
+      "loss": 1.1859,
+      "step": 7727
+    },
+    {
+      "epoch": 0.678283336670406,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0009525993064579512,
+      "loss": 1.1511,
+      "step": 7728
+    },
+    {
+      "epoch": 0.6783711062533085,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.000952236200452456,
+      "loss": 1.1826,
+      "step": 7729
+    },
+    {
+      "epoch": 0.678458875836211,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0009518731633134881,
+      "loss": 1.21,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6785466454191135,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0009515101950768759,
+      "loss": 1.1254,
+      "step": 7731
+    },
+    {
+      "epoch": 0.6786344150020159,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0009511472957784438,
+      "loss": 1.1918,
+      "step": 7732
+    },
+    {
+      "epoch": 0.6787221845849184,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0009507844654540083,
+      "loss": 1.191,
+      "step": 7733
+    },
+    {
+      "epoch": 0.678809954167821,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00095042170413938,
+      "loss": 1.1622,
+      "step": 7734
+    },
+    {
+      "epoch": 0.6788977237507234,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0009500590118703608,
+      "loss": 1.118,
+      "step": 7735
+    },
+    {
+      "epoch": 0.6789854933336259,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0009496963886827476,
+      "loss": 1.22,
+      "step": 7736
+    },
+    {
+      "epoch": 0.6790732629165284,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0009493338346123299,
+      "loss": 1.1956,
+      "step": 7737
+    },
+    {
+      "epoch": 0.6791610324994308,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0009489713496948904,
+      "loss": 1.1584,
+      "step": 7738
+    },
+    {
+      "epoch": 0.6792488020823334,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0009486089339662048,
+      "loss": 1.1471,
+      "step": 7739
+    },
+    {
+      "epoch": 0.6793365716652359,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0009482465874620421,
+      "loss": 1.1977,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6794243412481383,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0009478843102181646,
+      "loss": 1.2106,
+      "step": 7741
+    },
+    {
+      "epoch": 0.6795121108310408,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0009475221022703281,
+      "loss": 1.1361,
+      "step": 7742
+    },
+    {
+      "epoch": 0.6795998804139433,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0009471599636542799,
+      "loss": 1.1064,
+      "step": 7743
+    },
+    {
+      "epoch": 0.6796876499968458,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0009467978944057623,
+      "loss": 1.1828,
+      "step": 7744
+    },
+    {
+      "epoch": 0.6797754195797483,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0009464358945605101,
+      "loss": 1.0835,
+      "step": 7745
+    },
+    {
+      "epoch": 0.6798631891626508,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0009460739641542517,
+      "loss": 1.1395,
+      "step": 7746
+    },
+    {
+      "epoch": 0.6799509587455532,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0009457121032227075,
+      "loss": 1.1495,
+      "step": 7747
+    },
+    {
+      "epoch": 0.6800387283284557,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0009453503118015917,
+      "loss": 1.1995,
+      "step": 7748
+    },
+    {
+      "epoch": 0.6801264979113582,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0009449885899266117,
+      "loss": 1.1618,
+      "step": 7749
+    },
+    {
+      "epoch": 0.6802142674942607,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0009446269376334691,
+      "loss": 1.1612,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6803020370771632,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0009442653549578565,
+      "loss": 1.2035,
+      "step": 7751
+    },
+    {
+      "epoch": 0.6803898066600657,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0009439038419354605,
+      "loss": 1.1417,
+      "step": 7752
+    },
+    {
+      "epoch": 0.6804775762429681,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0009435423986019617,
+      "loss": 1.1134,
+      "step": 7753
+    },
+    {
+      "epoch": 0.6805653458258706,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0009431810249930333,
+      "loss": 1.1726,
+      "step": 7754
+    },
+    {
+      "epoch": 0.6806531154087732,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0009428197211443405,
+      "loss": 1.1233,
+      "step": 7755
+    },
+    {
+      "epoch": 0.6807408849916756,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0009424584870915431,
+      "loss": 1.0944,
+      "step": 7756
+    },
+    {
+      "epoch": 0.6808286545745781,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0009420973228702932,
+      "loss": 1.1806,
+      "step": 7757
+    },
+    {
+      "epoch": 0.6809164241574805,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0009417362285162372,
+      "loss": 1.1325,
+      "step": 7758
+    },
+    {
+      "epoch": 0.681004193740383,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0009413752040650124,
+      "loss": 1.1478,
+      "step": 7759
+    },
+    {
+      "epoch": 0.6810919633232856,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0009410142495522508,
+      "loss": 1.2035,
+      "step": 7760
+    },
+    {
+      "epoch": 0.681179732906188,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0009406533650135775,
+      "loss": 1.1652,
+      "step": 7761
+    },
+    {
+      "epoch": 0.6812675024890905,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0009402925504846111,
+      "loss": 1.1365,
+      "step": 7762
+    },
+    {
+      "epoch": 0.681355272071993,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0009399318060009614,
+      "loss": 1.1572,
+      "step": 7763
+    },
+    {
+      "epoch": 0.6814430416548954,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009395711315982327,
+      "loss": 1.2206,
+      "step": 7764
+    },
+    {
+      "epoch": 0.681530811237798,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0009392105273120223,
+      "loss": 1.1487,
+      "step": 7765
+    },
+    {
+      "epoch": 0.6816185808207005,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0009388499931779209,
+      "loss": 1.1278,
+      "step": 7766
+    },
+    {
+      "epoch": 0.681706350403603,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.000938489529231511,
+      "loss": 1.1254,
+      "step": 7767
+    },
+    {
+      "epoch": 0.6817941199865054,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0009381291355083689,
+      "loss": 1.1666,
+      "step": 7768
+    },
+    {
+      "epoch": 0.6818818895694079,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0009377688120440647,
+      "loss": 1.1531,
+      "step": 7769
+    },
+    {
+      "epoch": 0.6819696591523104,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009374085588741604,
+      "loss": 1.1257,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6820574287352129,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0009370483760342119,
+      "loss": 1.1738,
+      "step": 7771
+    },
+    {
+      "epoch": 0.6821451983181154,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0009366882635597676,
+      "loss": 1.1972,
+      "step": 7772
+    },
+    {
+      "epoch": 0.6822329679010178,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0009363282214863693,
+      "loss": 1.2191,
+      "step": 7773
+    },
+    {
+      "epoch": 0.6823207374839203,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0009359682498495521,
+      "loss": 1.149,
+      "step": 7774
+    },
+    {
+      "epoch": 0.6824085070668229,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.000935608348684843,
+      "loss": 1.1587,
+      "step": 7775
+    },
+    {
+      "epoch": 0.6824962766497253,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000935248518027763,
+      "loss": 1.1649,
+      "step": 7776
+    },
+    {
+      "epoch": 0.6825840462326278,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0009348887579138264,
+      "loss": 1.1573,
+      "step": 7777
+    },
+    {
+      "epoch": 0.6826718158155303,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0009345290683785399,
+      "loss": 1.1891,
+      "step": 7778
+    },
+    {
+      "epoch": 0.6827595853984327,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0009341694494574031,
+      "loss": 1.1742,
+      "step": 7779
+    },
+    {
+      "epoch": 0.6828473549813352,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0009338099011859087,
+      "loss": 1.1488,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6829351245642378,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0009334504235995444,
+      "loss": 1.1522,
+      "step": 7781
+    },
+    {
+      "epoch": 0.6830228941471402,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0009330910167337875,
+      "loss": 1.1659,
+      "step": 7782
+    },
+    {
+      "epoch": 0.6831106637300427,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0009327316806241106,
+      "loss": 1.1656,
+      "step": 7783
+    },
+    {
+      "epoch": 0.6831984333129452,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009323724153059786,
+      "loss": 1.1937,
+      "step": 7784
+    },
+    {
+      "epoch": 0.6832862028958476,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0009320132208148503,
+      "loss": 1.0912,
+      "step": 7785
+    },
+    {
+      "epoch": 0.6833739724787502,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0009316540971861756,
+      "loss": 1.2041,
+      "step": 7786
+    },
+    {
+      "epoch": 0.6834617420616527,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0009312950444553992,
+      "loss": 1.2052,
+      "step": 7787
+    },
+    {
+      "epoch": 0.6835495116445551,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0009309360626579583,
+      "loss": 1.1474,
+      "step": 7788
+    },
+    {
+      "epoch": 0.6836372812274576,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0009305771518292824,
+      "loss": 1.1178,
+      "step": 7789
+    },
+    {
+      "epoch": 0.6837250508103601,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0009302183120047954,
+      "loss": 1.1986,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6838128203932626,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0009298595432199128,
+      "loss": 1.1676,
+      "step": 7791
+    },
+    {
+      "epoch": 0.6839005899761651,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0009295008455100439,
+      "loss": 1.1287,
+      "step": 7792
+    },
+    {
+      "epoch": 0.6839883595590676,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0009291422189105911,
+      "loss": 1.1446,
+      "step": 7793
+    },
+    {
+      "epoch": 0.68407612914197,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0009287836634569485,
+      "loss": 1.1741,
+      "step": 7794
+    },
+    {
+      "epoch": 0.6841638987248725,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0009284251791845048,
+      "loss": 1.1546,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6842516683077751,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0009280667661286405,
+      "loss": 1.1526,
+      "step": 7796
+    },
+    {
+      "epoch": 0.6843394378906775,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0009277084243247306,
+      "loss": 1.1517,
+      "step": 7797
+    },
+    {
+      "epoch": 0.68442720747358,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0009273501538081407,
+      "loss": 1.0987,
+      "step": 7798
+    },
+    {
+      "epoch": 0.6845149770564825,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.000926991954614231,
+      "loss": 1.209,
+      "step": 7799
+    },
+    {
+      "epoch": 0.6846027466393849,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0009266338267783543,
+      "loss": 1.1941,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6846905162222875,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0009262757703358576,
+      "loss": 1.1794,
+      "step": 7801
+    },
+    {
+      "epoch": 0.68477828580519,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0009259177853220784,
+      "loss": 1.1494,
+      "step": 7802
+    },
+    {
+      "epoch": 0.6848660553880924,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0009255598717723486,
+      "loss": 1.1274,
+      "step": 7803
+    },
+    {
+      "epoch": 0.6849538249709949,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0009252020297219928,
+      "loss": 1.1584,
+      "step": 7804
+    },
+    {
+      "epoch": 0.6850415945538973,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0009248442592063297,
+      "loss": 1.2084,
+      "step": 7805
+    },
+    {
+      "epoch": 0.6851293641367998,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000924486560260668,
+      "loss": 1.151,
+      "step": 7806
+    },
+    {
+      "epoch": 0.6852171337197024,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0009241289329203124,
+      "loss": 1.139,
+      "step": 7807
+    },
+    {
+      "epoch": 0.6853049033026049,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0009237713772205589,
+      "loss": 1.1686,
+      "step": 7808
+    },
+    {
+      "epoch": 0.6853926728855073,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0009234138931966975,
+      "loss": 1.1458,
+      "step": 7809
+    },
+    {
+      "epoch": 0.6854804424684098,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0009230564808840087,
+      "loss": 1.1502,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6855682120513122,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0009226991403177694,
+      "loss": 1.1437,
+      "step": 7811
+    },
+    {
+      "epoch": 0.6856559816342148,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0009223418715332473,
+      "loss": 1.1714,
+      "step": 7812
+    },
+    {
+      "epoch": 0.6857437512171173,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0009219846745657038,
+      "loss": 1.1479,
+      "step": 7813
+    },
+    {
+      "epoch": 0.6858315208000197,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009216275494503916,
+      "loss": 1.1003,
+      "step": 7814
+    },
+    {
+      "epoch": 0.6859192903829222,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0009212704962225583,
+      "loss": 1.2119,
+      "step": 7815
+    },
+    {
+      "epoch": 0.6860070599658247,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0009209135149174437,
+      "loss": 1.1742,
+      "step": 7816
+    },
+    {
+      "epoch": 0.6860948295487272,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.000920556605570281,
+      "loss": 1.132,
+      "step": 7817
+    },
+    {
+      "epoch": 0.6861825991316297,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0009201997682162942,
+      "loss": 1.1431,
+      "step": 7818
+    },
+    {
+      "epoch": 0.6862703687145322,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0009198430028907028,
+      "loss": 1.1604,
+      "step": 7819
+    },
+    {
+      "epoch": 0.6863581382974346,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0009194863096287177,
+      "loss": 1.1313,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6864459078803371,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0009191296884655437,
+      "loss": 1.1142,
+      "step": 7821
+    },
+    {
+      "epoch": 0.6865336774632397,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0009187731394363774,
+      "loss": 1.1423,
+      "step": 7822
+    },
+    {
+      "epoch": 0.6866214470461421,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0009184166625764088,
+      "loss": 1.1423,
+      "step": 7823
+    },
+    {
+      "epoch": 0.6867092166290446,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0009180602579208209,
+      "loss": 1.1432,
+      "step": 7824
+    },
+    {
+      "epoch": 0.6867969862119471,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0009177039255047901,
+      "loss": 1.1728,
+      "step": 7825
+    },
+    {
+      "epoch": 0.6868847557948495,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0009173476653634835,
+      "loss": 1.1396,
+      "step": 7826
+    },
+    {
+      "epoch": 0.6869725253777521,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0009169914775320636,
+      "loss": 1.1726,
+      "step": 7827
+    },
+    {
+      "epoch": 0.6870602949606546,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0009166353620456843,
+      "loss": 1.2212,
+      "step": 7828
+    },
+    {
+      "epoch": 0.687148064543557,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009162793189394936,
+      "loss": 1.159,
+      "step": 7829
+    },
+    {
+      "epoch": 0.6872358341264595,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0009159233482486299,
+      "loss": 1.1691,
+      "step": 7830
+    },
+    {
+      "epoch": 0.687323603709362,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0009155674500082278,
+      "loss": 1.1411,
+      "step": 7831
+    },
+    {
+      "epoch": 0.6874113732922645,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0009152116242534129,
+      "loss": 1.2069,
+      "step": 7832
+    },
+    {
+      "epoch": 0.687499142875167,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0009148558710193026,
+      "loss": 1.1356,
+      "step": 7833
+    },
+    {
+      "epoch": 0.6875869124580695,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.000914500190341009,
+      "loss": 1.2197,
+      "step": 7834
+    },
+    {
+      "epoch": 0.6876746820409719,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0009141445822536363,
+      "loss": 1.1316,
+      "step": 7835
+    },
+    {
+      "epoch": 0.6877624516238744,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0009137890467922824,
+      "loss": 1.1261,
+      "step": 7836
+    },
+    {
+      "epoch": 0.6878502212067769,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.000913433583992036,
+      "loss": 1.1897,
+      "step": 7837
+    },
+    {
+      "epoch": 0.6879379907896794,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0009130781938879804,
+      "loss": 1.1852,
+      "step": 7838
+    },
+    {
+      "epoch": 0.6880257603725819,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0009127228765151913,
+      "loss": 1.192,
+      "step": 7839
+    },
+    {
+      "epoch": 0.6881135299554844,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0009123676319087368,
+      "loss": 1.127,
+      "step": 7840
+    },
+    {
+      "epoch": 0.6882012995383868,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0009120124601036785,
+      "loss": 1.12,
+      "step": 7841
+    },
+    {
+      "epoch": 0.6882890691212893,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0009116573611350704,
+      "loss": 1.22,
+      "step": 7842
+    },
+    {
+      "epoch": 0.6883768387041919,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0009113023350379591,
+      "loss": 1.1146,
+      "step": 7843
+    },
+    {
+      "epoch": 0.6884646082870943,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.000910947381847385,
+      "loss": 1.194,
+      "step": 7844
+    },
+    {
+      "epoch": 0.6885523778699968,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009105925015983797,
+      "loss": 1.1791,
+      "step": 7845
+    },
+    {
+      "epoch": 0.6886401474528993,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0009102376943259686,
+      "loss": 1.1493,
+      "step": 7846
+    },
+    {
+      "epoch": 0.6887279170358017,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0009098829600651703,
+      "loss": 1.1358,
+      "step": 7847
+    },
+    {
+      "epoch": 0.6888156866187043,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0009095282988509954,
+      "loss": 1.0898,
+      "step": 7848
+    },
+    {
+      "epoch": 0.6889034562016068,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0009091737107184473,
+      "loss": 1.1643,
+      "step": 7849
+    },
+    {
+      "epoch": 0.6889912257845092,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009088191957025221,
+      "loss": 1.141,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6890789953674117,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0009084647538382098,
+      "loss": 1.2213,
+      "step": 7851
+    },
+    {
+      "epoch": 0.6891667649503141,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000908110385160493,
+      "loss": 1.1476,
+      "step": 7852
+    },
+    {
+      "epoch": 0.6892545345332167,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0009077560897043451,
+      "loss": 1.1282,
+      "step": 7853
+    },
+    {
+      "epoch": 0.6893423041161192,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0009074018675047343,
+      "loss": 1.1878,
+      "step": 7854
+    },
+    {
+      "epoch": 0.6894300736990217,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0009070477185966208,
+      "loss": 1.1365,
+      "step": 7855
+    },
+    {
+      "epoch": 0.6895178432819241,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0009066936430149584,
+      "loss": 1.1499,
+      "step": 7856
+    },
+    {
+      "epoch": 0.6896056128648266,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0009063396407946915,
+      "loss": 1.1404,
+      "step": 7857
+    },
+    {
+      "epoch": 0.6896933824477292,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0009059857119707598,
+      "loss": 1.1857,
+      "step": 7858
+    },
+    {
+      "epoch": 0.6897811520306316,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0009056318565780943,
+      "loss": 1.1286,
+      "step": 7859
+    },
+    {
+      "epoch": 0.6898689216135341,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0009052780746516193,
+      "loss": 1.1534,
+      "step": 7860
+    },
+    {
+      "epoch": 0.6899566911964365,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0009049243662262517,
+      "loss": 1.1646,
+      "step": 7861
+    },
+    {
+      "epoch": 0.690044460779339,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0009045707313369008,
+      "loss": 1.1905,
+      "step": 7862
+    },
+    {
+      "epoch": 0.6901322303622415,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0009042171700184694,
+      "loss": 1.0952,
+      "step": 7863
+    },
+    {
+      "epoch": 0.690219999945144,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0009038636823058527,
+      "loss": 1.2066,
+      "step": 7864
+    },
+    {
+      "epoch": 0.6903077695280465,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0009035102682339379,
+      "loss": 1.246,
+      "step": 7865
+    },
+    {
+      "epoch": 0.690395539110949,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0009031569278376059,
+      "loss": 1.1462,
+      "step": 7866
+    },
+    {
+      "epoch": 0.6904833086938514,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0009028036611517299,
+      "loss": 1.1576,
+      "step": 7867
+    },
+    {
+      "epoch": 0.6905710782767539,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0009024504682111767,
+      "loss": 1.1296,
+      "step": 7868
+    },
+    {
+      "epoch": 0.6906588478596565,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0009020973490508038,
+      "loss": 1.18,
+      "step": 7869
+    },
+    {
+      "epoch": 0.6907466174425589,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0009017443037054629,
+      "loss": 1.1658,
+      "step": 7870
+    },
+    {
+      "epoch": 0.6908343870254614,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0009013913322099993,
+      "loss": 1.1464,
+      "step": 7871
+    },
+    {
+      "epoch": 0.6909221566083639,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0009010384345992486,
+      "loss": 1.1643,
+      "step": 7872
+    },
+    {
+      "epoch": 0.6910099261912663,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0009006856109080411,
+      "loss": 1.145,
+      "step": 7873
+    },
+    {
+      "epoch": 0.6910976957741689,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0009003328611711989,
+      "loss": 1.1348,
+      "step": 7874
+    },
+    {
+      "epoch": 0.6911854653570714,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008999801854235374,
+      "loss": 1.164,
+      "step": 7875
+    },
+    {
+      "epoch": 0.6912732349399738,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008996275836998635,
+      "loss": 1.1317,
+      "step": 7876
+    },
+    {
+      "epoch": 0.6913610045228763,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.000899275056034978,
+      "loss": 1.1937,
+      "step": 7877
+    },
+    {
+      "epoch": 0.6914487741057788,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0008989226024636739,
+      "loss": 1.1793,
+      "step": 7878
+    },
+    {
+      "epoch": 0.6915365436886813,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008985702230207373,
+      "loss": 1.1312,
+      "step": 7879
+    },
+    {
+      "epoch": 0.6916243132715838,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0008982179177409468,
+      "loss": 1.1587,
+      "step": 7880
+    },
+    {
+      "epoch": 0.6917120828544863,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008978656866590721,
+      "loss": 1.1438,
+      "step": 7881
+    },
+    {
+      "epoch": 0.6917998524373887,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008975135298098785,
+      "loss": 1.1543,
+      "step": 7882
+    },
+    {
+      "epoch": 0.6918876220202912,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0008971614472281228,
+      "loss": 1.1608,
+      "step": 7883
+    },
+    {
+      "epoch": 0.6919753916031938,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008968094389485529,
+      "loss": 1.184,
+      "step": 7884
+    },
+    {
+      "epoch": 0.6920631611860962,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008964575050059109,
+      "loss": 1.1414,
+      "step": 7885
+    },
+    {
+      "epoch": 0.6921509307689987,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008961056454349318,
+      "loss": 1.1749,
+      "step": 7886
+    },
+    {
+      "epoch": 0.6922387003519012,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000895753860270343,
+      "loss": 1.2071,
+      "step": 7887
+    },
+    {
+      "epoch": 0.6923264699348036,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008954021495468632,
+      "loss": 1.1777,
+      "step": 7888
+    },
+    {
+      "epoch": 0.6924142395177061,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008950505132992053,
+      "loss": 1.1578,
+      "step": 7889
+    },
+    {
+      "epoch": 0.6925020091006087,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008946989515620748,
+      "loss": 1.1667,
+      "step": 7890
+    },
+    {
+      "epoch": 0.6925897786835111,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000894347464370169,
+      "loss": 1.1379,
+      "step": 7891
+    },
+    {
+      "epoch": 0.6926775482664136,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008939960517581787,
+      "loss": 1.1023,
+      "step": 7892
+    },
+    {
+      "epoch": 0.692765317849316,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008936447137607867,
+      "loss": 1.1582,
+      "step": 7893
+    },
+    {
+      "epoch": 0.6928530874322185,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008932934504126686,
+      "loss": 1.1507,
+      "step": 7894
+    },
+    {
+      "epoch": 0.6929408570151211,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008929422617484934,
+      "loss": 1.1835,
+      "step": 7895
+    },
+    {
+      "epoch": 0.6930286265980236,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008925911478029209,
+      "loss": 1.1464,
+      "step": 7896
+    },
+    {
+      "epoch": 0.693116396180926,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008922401086106051,
+      "loss": 1.1411,
+      "step": 7897
+    },
+    {
+      "epoch": 0.6932041657638285,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008918891442061923,
+      "loss": 1.1695,
+      "step": 7898
+    },
+    {
+      "epoch": 0.693291935346731,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.000891538254624322,
+      "loss": 1.121,
+      "step": 7899
+    },
+    {
+      "epoch": 0.6933797049296335,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.000891187439899624,
+      "loss": 1.1367,
+      "step": 7900
+    },
+    {
+      "epoch": 0.693467474512536,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008908367000667231,
+      "loss": 1.1596,
+      "step": 7901
+    },
+    {
+      "epoch": 0.6935552440954385,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008904860351602363,
+      "loss": 1.1344,
+      "step": 7902
+    },
+    {
+      "epoch": 0.6936430136783409,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008901354452147732,
+      "loss": 1.1349,
+      "step": 7903
+    },
+    {
+      "epoch": 0.6937307832612434,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008897849302649344,
+      "loss": 1.1195,
+      "step": 7904
+    },
+    {
+      "epoch": 0.693818552844146,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008894344903453151,
+      "loss": 1.1102,
+      "step": 7905
+    },
+    {
+      "epoch": 0.6939063224270484,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.000889084125490502,
+      "loss": 1.1432,
+      "step": 7906
+    },
+    {
+      "epoch": 0.6939940920099509,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008887338357350757,
+      "loss": 1.1167,
+      "step": 7907
+    },
+    {
+      "epoch": 0.6940818615928533,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008883836211136069,
+      "loss": 1.1585,
+      "step": 7908
+    },
+    {
+      "epoch": 0.6941696311757558,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008880334816606614,
+      "loss": 1.2015,
+      "step": 7909
+    },
+    {
+      "epoch": 0.6942574007586584,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0008876834174107963,
+      "loss": 1.2039,
+      "step": 7910
+    },
+    {
+      "epoch": 0.6943451703415608,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008873334283985613,
+      "loss": 1.1902,
+      "step": 7911
+    },
+    {
+      "epoch": 0.6944329399244633,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008869835146584993,
+      "loss": 1.2057,
+      "step": 7912
+    },
+    {
+      "epoch": 0.6945207095073658,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008866336762251453,
+      "loss": 1.141,
+      "step": 7913
+    },
+    {
+      "epoch": 0.6946084790902682,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008862839131330268,
+      "loss": 1.1679,
+      "step": 7914
+    },
+    {
+      "epoch": 0.6946962486731708,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008859342254166649,
+      "loss": 1.137,
+      "step": 7915
+    },
+    {
+      "epoch": 0.6947840182560733,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008855846131105711,
+      "loss": 1.1661,
+      "step": 7916
+    },
+    {
+      "epoch": 0.6948717878389757,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0008852350762492515,
+      "loss": 1.1702,
+      "step": 7917
+    },
+    {
+      "epoch": 0.6949595574218782,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008848856148672039,
+      "loss": 1.1293,
+      "step": 7918
+    },
+    {
+      "epoch": 0.6950473270047807,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.000884536228998919,
+      "loss": 1.1934,
+      "step": 7919
+    },
+    {
+      "epoch": 0.6951350965876831,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008841869186788791,
+      "loss": 1.1548,
+      "step": 7920
+    },
+    {
+      "epoch": 0.6952228661705857,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008838376839415597,
+      "loss": 1.1796,
+      "step": 7921
+    },
+    {
+      "epoch": 0.6953106357534882,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008834885248214302,
+      "loss": 1.1513,
+      "step": 7922
+    },
+    {
+      "epoch": 0.6953984053363906,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00088313944135295,
+      "loss": 1.1515,
+      "step": 7923
+    },
+    {
+      "epoch": 0.6954861749192931,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008827904335705726,
+      "loss": 1.1361,
+      "step": 7924
+    },
+    {
+      "epoch": 0.6955739445021956,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008824415015087437,
+      "loss": 1.1863,
+      "step": 7925
+    },
+    {
+      "epoch": 0.6956617140850981,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0008820926452019019,
+      "loss": 1.1436,
+      "step": 7926
+    },
+    {
+      "epoch": 0.6957494836680006,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008817438646844772,
+      "loss": 1.1111,
+      "step": 7927
+    },
+    {
+      "epoch": 0.6958372532509031,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008813951599908934,
+      "loss": 1.2014,
+      "step": 7928
+    },
+    {
+      "epoch": 0.6959250228338055,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008810465311555658,
+      "loss": 1.2127,
+      "step": 7929
+    },
+    {
+      "epoch": 0.696012792416708,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008806979782129029,
+      "loss": 1.1955,
+      "step": 7930
+    },
+    {
+      "epoch": 0.6961005619996106,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008803495011973059,
+      "loss": 1.2014,
+      "step": 7931
+    },
+    {
+      "epoch": 0.696188331582513,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008800011001431676,
+      "loss": 1.1769,
+      "step": 7932
+    },
+    {
+      "epoch": 0.6962761011654155,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000879652775084874,
+      "loss": 1.0934,
+      "step": 7933
+    },
+    {
+      "epoch": 0.696363870748318,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0008793045260568038,
+      "loss": 1.1541,
+      "step": 7934
+    },
+    {
+      "epoch": 0.6964516403312204,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.000878956353093327,
+      "loss": 1.2017,
+      "step": 7935
+    },
+    {
+      "epoch": 0.696539409914123,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0008786082562288073,
+      "loss": 1.1835,
+      "step": 7936
+    },
+    {
+      "epoch": 0.6966271794970255,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0008782602354976003,
+      "loss": 1.1298,
+      "step": 7937
+    },
+    {
+      "epoch": 0.6967149490799279,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008779122909340551,
+      "loss": 1.1772,
+      "step": 7938
+    },
+    {
+      "epoch": 0.6968027186628304,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008775644225725112,
+      "loss": 1.1627,
+      "step": 7939
+    },
+    {
+      "epoch": 0.6968904882457329,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008772166304473027,
+      "loss": 1.2279,
+      "step": 7940
+    },
+    {
+      "epoch": 0.6969782578286354,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0008768689145927541,
+      "loss": 1.2426,
+      "step": 7941
+    },
+    {
+      "epoch": 0.6970660274115379,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008765212750431859,
+      "loss": 1.1309,
+      "step": 7942
+    },
+    {
+      "epoch": 0.6971537969944404,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0008761737118329066,
+      "loss": 1.0997,
+      "step": 7943
+    },
+    {
+      "epoch": 0.6972415665773428,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0008758262249962201,
+      "loss": 1.1593,
+      "step": 7944
+    },
+    {
+      "epoch": 0.6973293361602453,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008754788145674219,
+      "loss": 1.1673,
+      "step": 7945
+    },
+    {
+      "epoch": 0.6974171057431477,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008751314805808008,
+      "loss": 1.2096,
+      "step": 7946
+    },
+    {
+      "epoch": 0.6975048753260503,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008747842230706358,
+      "loss": 1.1555,
+      "step": 7947
+    },
+    {
+      "epoch": 0.6975926449089528,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008744370420712006,
+      "loss": 1.1892,
+      "step": 7948
+    },
+    {
+      "epoch": 0.6976804144918552,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008740899376167607,
+      "loss": 1.1588,
+      "step": 7949
+    },
+    {
+      "epoch": 0.6977681840747577,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008737429097415743,
+      "loss": 1.1726,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6978559536576602,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008733959584798905,
+      "loss": 1.168,
+      "step": 7951
+    },
+    {
+      "epoch": 0.6979437232405628,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.000873049083865953,
+      "loss": 1.2057,
+      "step": 7952
+    },
+    {
+      "epoch": 0.6980314928234652,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008727022859339966,
+      "loss": 1.0888,
+      "step": 7953
+    },
+    {
+      "epoch": 0.6981192624063677,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0008723555647182494,
+      "loss": 1.2037,
+      "step": 7954
+    },
+    {
+      "epoch": 0.6982070319892701,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0008720089202529307,
+      "loss": 1.1685,
+      "step": 7955
+    },
+    {
+      "epoch": 0.6982948015721726,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0008716623525722532,
+      "loss": 1.1794,
+      "step": 7956
+    },
+    {
+      "epoch": 0.6983825711550752,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0008713158617104218,
+      "loss": 1.1843,
+      "step": 7957
+    },
+    {
+      "epoch": 0.6984703407379776,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0008709694477016341,
+      "loss": 1.2038,
+      "step": 7958
+    },
+    {
+      "epoch": 0.6985581103208801,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.000870623110580079,
+      "loss": 1.142,
+      "step": 7959
+    },
+    {
+      "epoch": 0.6986458799037826,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000870276850379939,
+      "loss": 1.1956,
+      "step": 7960
+    },
+    {
+      "epoch": 0.698733649486685,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008699306671353886,
+      "loss": 1.2102,
+      "step": 7961
+    },
+    {
+      "epoch": 0.6988214190695876,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0008695845608805951,
+      "loss": 1.1742,
+      "step": 7962
+    },
+    {
+      "epoch": 0.6989091886524901,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0008692385316497171,
+      "loss": 1.1309,
+      "step": 7963
+    },
+    {
+      "epoch": 0.6989969582353925,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008688925794769069,
+      "loss": 1.1515,
+      "step": 7964
+    },
+    {
+      "epoch": 0.699084727818295,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008685467043963085,
+      "loss": 1.1973,
+      "step": 7965
+    },
+    {
+      "epoch": 0.6991724974011975,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008682009064420587,
+      "loss": 1.1156,
+      "step": 7966
+    },
+    {
+      "epoch": 0.6992602669841,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008678551856482854,
+      "loss": 1.119,
+      "step": 7967
+    },
+    {
+      "epoch": 0.6993480365670025,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0008675095420491107,
+      "loss": 1.198,
+      "step": 7968
+    },
+    {
+      "epoch": 0.699435806149905,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.000867163975678648,
+      "loss": 1.1736,
+      "step": 7969
+    },
+    {
+      "epoch": 0.6995235757328074,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0008668184865710039,
+      "loss": 1.1241,
+      "step": 7970
+    },
+    {
+      "epoch": 0.6996113453157099,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0008664730747602752,
+      "loss": 1.121,
+      "step": 7971
+    },
+    {
+      "epoch": 0.6996991148986125,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0008661277402805542,
+      "loss": 1.1474,
+      "step": 7972
+    },
+    {
+      "epoch": 0.6997868844815149,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0008657824831659242,
+      "loss": 1.1391,
+      "step": 7973
+    },
+    {
+      "epoch": 0.6998746540644174,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008654373034504596,
+      "loss": 1.0934,
+      "step": 7974
+    },
+    {
+      "epoch": 0.6999624236473199,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008650922011682288,
+      "loss": 1.1186,
+      "step": 7975
+    },
+    {
+      "epoch": 0.7000501932302223,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0008647471763532921,
+      "loss": 1.1194,
+      "step": 7976
+    },
+    {
+      "epoch": 0.7001379628131248,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0008644022290397023,
+      "loss": 1.1592,
+      "step": 7977
+    },
+    {
+      "epoch": 0.7002257323960274,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008640573592615037,
+      "loss": 1.1302,
+      "step": 7978
+    },
+    {
+      "epoch": 0.7003135019789298,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.000863712567052734,
+      "loss": 1.141,
+      "step": 7979
+    },
+    {
+      "epoch": 0.7004012715618323,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008633678524474225,
+      "loss": 1.1636,
+      "step": 7980
+    },
+    {
+      "epoch": 0.7004890411447348,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0008630232154795916,
+      "loss": 1.1401,
+      "step": 7981
+    },
+    {
+      "epoch": 0.7005768107276372,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008626786561832554,
+      "loss": 1.1493,
+      "step": 7982
+    },
+    {
+      "epoch": 0.7006645803105398,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008623341745924205,
+      "loss": 1.2085,
+      "step": 7983
+    },
+    {
+      "epoch": 0.7007523498934423,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008619897707410859,
+      "loss": 1.2312,
+      "step": 7984
+    },
+    {
+      "epoch": 0.7008401194763447,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0008616454446632436,
+      "loss": 1.2128,
+      "step": 7985
+    },
+    {
+      "epoch": 0.7009278890592472,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008613011963928759,
+      "loss": 1.1734,
+      "step": 7986
+    },
+    {
+      "epoch": 0.7010156586421497,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008609570259639594,
+      "loss": 1.121,
+      "step": 7987
+    },
+    {
+      "epoch": 0.7011034282250522,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008606129334104624,
+      "loss": 1.1304,
+      "step": 7988
+    },
+    {
+      "epoch": 0.7011911978079547,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0008602689187663459,
+      "loss": 1.1924,
+      "step": 7989
+    },
+    {
+      "epoch": 0.7012789673908572,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0008599249820655618,
+      "loss": 1.1726,
+      "step": 7990
+    },
+    {
+      "epoch": 0.7013667369737596,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008595811233420553,
+      "loss": 1.155,
+      "step": 7991
+    },
+    {
+      "epoch": 0.7014545065566621,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0008592373426297648,
+      "loss": 1.1438,
+      "step": 7992
+    },
+    {
+      "epoch": 0.7015422761395647,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008588936399626204,
+      "loss": 1.1263,
+      "step": 7993
+    },
+    {
+      "epoch": 0.7016300457224671,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.000858550015374543,
+      "loss": 1.1801,
+      "step": 7994
+    },
+    {
+      "epoch": 0.7017178153053696,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008582064688994472,
+      "loss": 1.1011,
+      "step": 7995
+    },
+    {
+      "epoch": 0.701805584888272,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008578630005712402,
+      "loss": 1.1515,
+      "step": 7996
+    },
+    {
+      "epoch": 0.7018933544711745,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.000857519610423821,
+      "loss": 1.1174,
+      "step": 7997
+    },
+    {
+      "epoch": 0.7019811240540771,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008571762984910803,
+      "loss": 1.1864,
+      "step": 7998
+    },
+    {
+      "epoch": 0.7020688936369796,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008568330648069019,
+      "loss": 1.1275,
+      "step": 7999
+    },
+    {
+      "epoch": 0.702156663219882,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008564899094051616,
+      "loss": 1.1177,
+      "step": 8000
+    },
+    {
+      "epoch": 0.702156663219882,
+      "eval_loss": 1.1471246480941772,
+      "eval_runtime": 437.8001,
+      "eval_samples_per_second": 33.668,
+      "eval_steps_per_second": 8.417,
+      "step": 8000
+    },
+    {
+      "epoch": 0.7022444328027845,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008561468323197276,
+      "loss": 1.1678,
+      "step": 8001
+    },
+    {
+      "epoch": 0.7023322023856869,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008558038335844595,
+      "loss": 1.153,
+      "step": 8002
+    },
+    {
+      "epoch": 0.7024199719685894,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008554609132332109,
+      "loss": 1.1802,
+      "step": 8003
+    },
+    {
+      "epoch": 0.702507741551492,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0008551180712998264,
+      "loss": 1.1259,
+      "step": 8004
+    },
+    {
+      "epoch": 0.7025955111343944,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0008547753078181436,
+      "loss": 1.1597,
+      "step": 8005
+    },
+    {
+      "epoch": 0.7026832807172969,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008544326228219906,
+      "loss": 1.1624,
+      "step": 8006
+    },
+    {
+      "epoch": 0.7027710503001994,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008540900163451903,
+      "loss": 1.1409,
+      "step": 8007
+    },
+    {
+      "epoch": 0.7028588198831018,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008537474884215557,
+      "loss": 1.1886,
+      "step": 8008
+    },
+    {
+      "epoch": 0.7029465894660044,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008534050390848943,
+      "loss": 1.1641,
+      "step": 8009
+    },
+    {
+      "epoch": 0.7030343590489069,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008530626683690029,
+      "loss": 1.1501,
+      "step": 8010
+    },
+    {
+      "epoch": 0.7031221286318093,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008527203763076725,
+      "loss": 1.1932,
+      "step": 8011
+    },
+    {
+      "epoch": 0.7032098982147118,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008523781629346866,
+      "loss": 1.2357,
+      "step": 8012
+    },
+    {
+      "epoch": 0.7032976677976143,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008520360282838198,
+      "loss": 1.1678,
+      "step": 8013
+    },
+    {
+      "epoch": 0.7033854373805168,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.00085169397238884,
+      "loss": 1.1612,
+      "step": 8014
+    },
+    {
+      "epoch": 0.7034732069634193,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.000851351995283506,
+      "loss": 1.2211,
+      "step": 8015
+    },
+    {
+      "epoch": 0.7035609765463218,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0008510100970015701,
+      "loss": 1.2001,
+      "step": 8016
+    },
+    {
+      "epoch": 0.7036487461292242,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008506682775767766,
+      "loss": 1.1788,
+      "step": 8017
+    },
+    {
+      "epoch": 0.7037365157121267,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008503265370428609,
+      "loss": 1.1875,
+      "step": 8018
+    },
+    {
+      "epoch": 0.7038242852950293,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000849984875433552,
+      "loss": 1.0967,
+      "step": 8019
+    },
+    {
+      "epoch": 0.7039120548779317,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0008496432927825701,
+      "loss": 1.0976,
+      "step": 8020
+    },
+    {
+      "epoch": 0.7039998244608342,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.000849301789123629,
+      "loss": 1.1261,
+      "step": 8021
+    },
+    {
+      "epoch": 0.7040875940437367,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.000848960364490432,
+      "loss": 1.1449,
+      "step": 8022
+    },
+    {
+      "epoch": 0.7041753636266391,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008486190189166782,
+      "loss": 1.1959,
+      "step": 8023
+    },
+    {
+      "epoch": 0.7042631332095417,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008482777524360566,
+      "loss": 1.1847,
+      "step": 8024
+    },
+    {
+      "epoch": 0.7043509027924442,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008479365650822481,
+      "loss": 1.1565,
+      "step": 8025
+    },
+    {
+      "epoch": 0.7044386723753466,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008475954568889271,
+      "loss": 1.1588,
+      "step": 8026
+    },
+    {
+      "epoch": 0.7045264419582491,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008472544278897595,
+      "loss": 1.1422,
+      "step": 8027
+    },
+    {
+      "epoch": 0.7046142115411516,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0008469134781184042,
+      "loss": 1.1482,
+      "step": 8028
+    },
+    {
+      "epoch": 0.704701981124054,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008465726076085102,
+      "loss": 1.1184,
+      "step": 8029
+    },
+    {
+      "epoch": 0.7047897507069566,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008462318163937211,
+      "loss": 1.1375,
+      "step": 8030
+    },
+    {
+      "epoch": 0.7048775202898591,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008458911045076714,
+      "loss": 1.1433,
+      "step": 8031
+    },
+    {
+      "epoch": 0.7049652898727615,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0008455504719839877,
+      "loss": 1.1473,
+      "step": 8032
+    },
+    {
+      "epoch": 0.705053059455664,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0008452099188562895,
+      "loss": 1.1667,
+      "step": 8033
+    },
+    {
+      "epoch": 0.7051408290385665,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0008448694451581879,
+      "loss": 1.1975,
+      "step": 8034
+    },
+    {
+      "epoch": 0.705228598621469,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0008445290509232862,
+      "loss": 1.1357,
+      "step": 8035
+    },
+    {
+      "epoch": 0.7053163682043715,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008441887361851809,
+      "loss": 1.1275,
+      "step": 8036
+    },
+    {
+      "epoch": 0.705404137787274,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0008438485009774582,
+      "loss": 1.1822,
+      "step": 8037
+    },
+    {
+      "epoch": 0.7054919073701764,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0008435083453336988,
+      "loss": 1.1445,
+      "step": 8038
+    },
+    {
+      "epoch": 0.7055796769530789,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008431682692874746,
+      "loss": 1.1483,
+      "step": 8039
+    },
+    {
+      "epoch": 0.7056674465359815,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00084282827287235,
+      "loss": 1.0875,
+      "step": 8040
+    },
+    {
+      "epoch": 0.7057552161188839,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0008424883561218808,
+      "loss": 1.2037,
+      "step": 8041
+    },
+    {
+      "epoch": 0.7058429857017864,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008421485190696152,
+      "loss": 1.18,
+      "step": 8042
+    },
+    {
+      "epoch": 0.7059307552846888,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0008418087617490945,
+      "loss": 1.1512,
+      "step": 8043
+    },
+    {
+      "epoch": 0.7060185248675913,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0008414690841938518,
+      "loss": 1.1785,
+      "step": 8044
+    },
+    {
+      "epoch": 0.7061062944504939,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0008411294864374107,
+      "loss": 1.2018,
+      "step": 8045
+    },
+    {
+      "epoch": 0.7061940640333964,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008407899685132891,
+      "loss": 1.1908,
+      "step": 8046
+    },
+    {
+      "epoch": 0.7062818336162988,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0008404505304549953,
+      "loss": 1.1615,
+      "step": 8047
+    },
+    {
+      "epoch": 0.7063696031992013,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008401111722960315,
+      "loss": 1.2101,
+      "step": 8048
+    },
+    {
+      "epoch": 0.7064573727821037,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0008397718940698902,
+      "loss": 1.1683,
+      "step": 8049
+    },
+    {
+      "epoch": 0.7065451423650063,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008394326958100569,
+      "loss": 1.1499,
+      "step": 8050
+    },
+    {
+      "epoch": 0.7066329119479088,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008390935775500091,
+      "loss": 1.1874,
+      "step": 8051
+    },
+    {
+      "epoch": 0.7067206815308112,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0008387545393232167,
+      "loss": 1.1311,
+      "step": 8052
+    },
+    {
+      "epoch": 0.7068084511137137,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008384155811631413,
+      "loss": 1.1082,
+      "step": 8053
+    },
+    {
+      "epoch": 0.7068962206966162,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008380767031032369,
+      "loss": 1.1117,
+      "step": 8054
+    },
+    {
+      "epoch": 0.7069839902795187,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0008377379051769492,
+      "loss": 1.1786,
+      "step": 8055
+    },
+    {
+      "epoch": 0.7070717598624212,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008373991874177169,
+      "loss": 1.0997,
+      "step": 8056
+    },
+    {
+      "epoch": 0.7071595294453237,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008370605498589688,
+      "loss": 1.1927,
+      "step": 8057
+    },
+    {
+      "epoch": 0.7072472990282261,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008367219925341279,
+      "loss": 1.1912,
+      "step": 8058
+    },
+    {
+      "epoch": 0.7073350686111286,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008363835154766086,
+      "loss": 1.1395,
+      "step": 8059
+    },
+    {
+      "epoch": 0.7074228381940311,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008360451187198172,
+      "loss": 1.1543,
+      "step": 8060
+    },
+    {
+      "epoch": 0.7075106077769336,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008357068022971514,
+      "loss": 1.1077,
+      "step": 8061
+    },
+    {
+      "epoch": 0.7075983773598361,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008353685662420017,
+      "loss": 1.1065,
+      "step": 8062
+    },
+    {
+      "epoch": 0.7076861469427386,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0008350304105877523,
+      "loss": 1.1348,
+      "step": 8063
+    },
+    {
+      "epoch": 0.707773916525641,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008346923353677765,
+      "loss": 1.1112,
+      "step": 8064
+    },
+    {
+      "epoch": 0.7078616861085435,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008343543406154406,
+      "loss": 1.1745,
+      "step": 8065
+    },
+    {
+      "epoch": 0.7079494556914461,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008340164263641042,
+      "loss": 1.1502,
+      "step": 8066
+    },
+    {
+      "epoch": 0.7080372252743485,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0008336785926471177,
+      "loss": 1.0929,
+      "step": 8067
+    },
+    {
+      "epoch": 0.708124994857251,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.000833340839497825,
+      "loss": 1.1477,
+      "step": 8068
+    },
+    {
+      "epoch": 0.7082127644401535,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008330031669495591,
+      "loss": 1.1251,
+      "step": 8069
+    },
+    {
+      "epoch": 0.7083005340230559,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008326655750356482,
+      "loss": 1.122,
+      "step": 8070
+    },
+    {
+      "epoch": 0.7083883036059585,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008323280637894109,
+      "loss": 1.1204,
+      "step": 8071
+    },
+    {
+      "epoch": 0.708476073188861,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0008319906332441587,
+      "loss": 1.1384,
+      "step": 8072
+    },
+    {
+      "epoch": 0.7085638427717634,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008316532834331938,
+      "loss": 1.1414,
+      "step": 8073
+    },
+    {
+      "epoch": 0.7086516123546659,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.000831316014389812,
+      "loss": 1.1786,
+      "step": 8074
+    },
+    {
+      "epoch": 0.7087393819375684,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0008309788261473005,
+      "loss": 1.1688,
+      "step": 8075
+    },
+    {
+      "epoch": 0.7088271515204709,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0008306417187389383,
+      "loss": 1.1575,
+      "step": 8076
+    },
+    {
+      "epoch": 0.7089149211033734,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008303046921979958,
+      "loss": 1.1474,
+      "step": 8077
+    },
+    {
+      "epoch": 0.7090026906862759,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008299677465577369,
+      "loss": 1.1274,
+      "step": 8078
+    },
+    {
+      "epoch": 0.7090904602691783,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0008296308818514173,
+      "loss": 1.1735,
+      "step": 8079
+    },
+    {
+      "epoch": 0.7091782298520808,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008292940981122831,
+      "loss": 1.1973,
+      "step": 8080
+    },
+    {
+      "epoch": 0.7092659994349834,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008289573953735742,
+      "loss": 1.1888,
+      "step": 8081
+    },
+    {
+      "epoch": 0.7093537690178858,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.000828620773668521,
+      "loss": 1.1099,
+      "step": 8082
+    },
+    {
+      "epoch": 0.7094415386007883,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0008282842330303484,
+      "loss": 1.1515,
+      "step": 8083
+    },
+    {
+      "epoch": 0.7095293081836908,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0008279477734922702,
+      "loss": 1.1285,
+      "step": 8084
+    },
+    {
+      "epoch": 0.7096170777665932,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008276113950874941,
+      "loss": 1.1546,
+      "step": 8085
+    },
+    {
+      "epoch": 0.7097048473494957,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008272750978492191,
+      "loss": 1.1336,
+      "step": 8086
+    },
+    {
+      "epoch": 0.7097926169323983,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0008269388818106372,
+      "loss": 1.1816,
+      "step": 8087
+    },
+    {
+      "epoch": 0.7098803865153007,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008266027470049304,
+      "loss": 1.1219,
+      "step": 8088
+    },
+    {
+      "epoch": 0.7099681560982032,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008262666934652745,
+      "loss": 1.1989,
+      "step": 8089
+    },
+    {
+      "epoch": 0.7100559256811056,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0008259307212248367,
+      "loss": 1.1362,
+      "step": 8090
+    },
+    {
+      "epoch": 0.7101436952640081,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008255948303167765,
+      "loss": 1.1496,
+      "step": 8091
+    },
+    {
+      "epoch": 0.7102314648469107,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008252590207742435,
+      "loss": 1.2594,
+      "step": 8092
+    },
+    {
+      "epoch": 0.7103192344298132,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0008249232926303822,
+      "loss": 1.0901,
+      "step": 8093
+    },
+    {
+      "epoch": 0.7104070040127156,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0008245876459183275,
+      "loss": 1.2021,
+      "step": 8094
+    },
+    {
+      "epoch": 0.7104947735956181,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008242520806712065,
+      "loss": 1.2186,
+      "step": 8095
+    },
+    {
+      "epoch": 0.7105825431785205,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0008239165969221373,
+      "loss": 1.1477,
+      "step": 8096
+    },
+    {
+      "epoch": 0.7106703127614231,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0008235811947042315,
+      "loss": 1.1506,
+      "step": 8097
+    },
+    {
+      "epoch": 0.7107580823443256,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008232458740505917,
+      "loss": 1.1307,
+      "step": 8098
+    },
+    {
+      "epoch": 0.710845851927228,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0008229106349943131,
+      "loss": 1.1794,
+      "step": 8099
+    },
+    {
+      "epoch": 0.7109336215101305,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0008225754775684821,
+      "loss": 1.1802,
+      "step": 8100
+    },
+    {
+      "epoch": 0.711021391093033,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008222404018061773,
+      "loss": 1.1269,
+      "step": 8101
+    },
+    {
+      "epoch": 0.7111091606759355,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008219054077404695,
+      "loss": 1.1665,
+      "step": 8102
+    },
+    {
+      "epoch": 0.711196930258838,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0008215704954044215,
+      "loss": 1.1491,
+      "step": 8103
+    },
+    {
+      "epoch": 0.7112846998417405,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008212356648310878,
+      "loss": 1.1515,
+      "step": 8104
+    },
+    {
+      "epoch": 0.7113724694246429,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0008209009160535145,
+      "loss": 1.1225,
+      "step": 8105
+    },
+    {
+      "epoch": 0.7114602390075454,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0008205662491047403,
+      "loss": 1.1575,
+      "step": 8106
+    },
+    {
+      "epoch": 0.711548008590448,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0008202316640177957,
+      "loss": 1.1575,
+      "step": 8107
+    },
+    {
+      "epoch": 0.7116357781733504,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008198971608257024,
+      "loss": 1.1593,
+      "step": 8108
+    },
+    {
+      "epoch": 0.7117235477562529,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008195627395614746,
+      "loss": 1.1837,
+      "step": 8109
+    },
+    {
+      "epoch": 0.7118113173391554,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0008192284002581185,
+      "loss": 1.1095,
+      "step": 8110
+    },
+    {
+      "epoch": 0.7118990869220578,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008188941429486329,
+      "loss": 1.1565,
+      "step": 8111
+    },
+    {
+      "epoch": 0.7119868565049604,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008185599676660053,
+      "loss": 1.1888,
+      "step": 8112
+    },
+    {
+      "epoch": 0.7120746260878629,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008182258744432202,
+      "loss": 1.1901,
+      "step": 8113
+    },
+    {
+      "epoch": 0.7121623956707653,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0008178918633132502,
+      "loss": 1.15,
+      "step": 8114
+    },
+    {
+      "epoch": 0.7122501652536678,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008175579343090604,
+      "loss": 1.1865,
+      "step": 8115
+    },
+    {
+      "epoch": 0.7123379348365703,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008172240874636088,
+      "loss": 1.172,
+      "step": 8116
+    },
+    {
+      "epoch": 0.7124257044194727,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008168903228098444,
+      "loss": 1.1987,
+      "step": 8117
+    },
+    {
+      "epoch": 0.7125134740023753,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0008165566403807089,
+      "loss": 1.1877,
+      "step": 8118
+    },
+    {
+      "epoch": 0.7126012435852778,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.000816223040209136,
+      "loss": 1.1775,
+      "step": 8119
+    },
+    {
+      "epoch": 0.7126890131681802,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0008158895223280488,
+      "loss": 1.1176,
+      "step": 8120
+    },
+    {
+      "epoch": 0.7127767827510827,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008155560867703657,
+      "loss": 1.2036,
+      "step": 8121
+    },
+    {
+      "epoch": 0.7128645523339852,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008152227335689951,
+      "loss": 1.0962,
+      "step": 8122
+    },
+    {
+      "epoch": 0.7129523219168877,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0008148894627568375,
+      "loss": 1.1895,
+      "step": 8123
+    },
+    {
+      "epoch": 0.7130400914997902,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0008145562743667857,
+      "loss": 1.1046,
+      "step": 8124
+    },
+    {
+      "epoch": 0.7131278610826927,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0008142231684317241,
+      "loss": 1.1973,
+      "step": 8125
+    },
+    {
+      "epoch": 0.7132156306655951,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.000813890144984529,
+      "loss": 1.1623,
+      "step": 8126
+    },
+    {
+      "epoch": 0.7133034002484976,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008135572040580678,
+      "loss": 1.2352,
+      "step": 8127
+    },
+    {
+      "epoch": 0.7133911698314002,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008132243456852008,
+      "loss": 1.1858,
+      "step": 8128
+    },
+    {
+      "epoch": 0.7134789394143026,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008128915698987803,
+      "loss": 1.1735,
+      "step": 8129
+    },
+    {
+      "epoch": 0.7135667089972051,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0008125588767316499,
+      "loss": 1.1336,
+      "step": 8130
+    },
+    {
+      "epoch": 0.7136544785801076,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008122262662166442,
+      "loss": 1.1738,
+      "step": 8131
+    },
+    {
+      "epoch": 0.71374224816301,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008118937383865913,
+      "loss": 1.1372,
+      "step": 8132
+    },
+    {
+      "epoch": 0.7138300177459126,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008115612932743096,
+      "loss": 1.144,
+      "step": 8133
+    },
+    {
+      "epoch": 0.7139177873288151,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008112289309126115,
+      "loss": 1.1825,
+      "step": 8134
+    },
+    {
+      "epoch": 0.7140055569117175,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008108966513342989,
+      "loss": 1.1263,
+      "step": 8135
+    },
+    {
+      "epoch": 0.71409332649462,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008105644545721665,
+      "loss": 1.1545,
+      "step": 8136
+    },
+    {
+      "epoch": 0.7141810960775224,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008102323406590008,
+      "loss": 1.2117,
+      "step": 8137
+    },
+    {
+      "epoch": 0.714268865660425,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.000809900309627581,
+      "loss": 1.1631,
+      "step": 8138
+    },
+    {
+      "epoch": 0.7143566352433275,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.000809568361510676,
+      "loss": 1.1727,
+      "step": 8139
+    },
+    {
+      "epoch": 0.71444440482623,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0008092364963410484,
+      "loss": 1.1887,
+      "step": 8140
+    },
+    {
+      "epoch": 0.7145321744091324,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008089047141514517,
+      "loss": 1.1596,
+      "step": 8141
+    },
+    {
+      "epoch": 0.7146199439920349,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0008085730149746324,
+      "loss": 1.1526,
+      "step": 8142
+    },
+    {
+      "epoch": 0.7147077135749373,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008082413988433257,
+      "loss": 1.1069,
+      "step": 8143
+    },
+    {
+      "epoch": 0.7147954831578399,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.000807909865790263,
+      "loss": 1.1478,
+      "step": 8144
+    },
+    {
+      "epoch": 0.7148832527407424,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0008075784158481648,
+      "loss": 1.1218,
+      "step": 8145
+    },
+    {
+      "epoch": 0.7149710223236448,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0008072470490497438,
+      "loss": 1.1542,
+      "step": 8146
+    },
+    {
+      "epoch": 0.7150587919065473,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0008069157654277044,
+      "loss": 1.1688,
+      "step": 8147
+    },
+    {
+      "epoch": 0.7151465614894498,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008065845650147425,
+      "loss": 1.1451,
+      "step": 8148
+    },
+    {
+      "epoch": 0.7152343310723523,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008062534478435471,
+      "loss": 1.1853,
+      "step": 8149
+    },
+    {
+      "epoch": 0.7153221006552548,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008059224139467984,
+      "loss": 1.1953,
+      "step": 8150
+    },
+    {
+      "epoch": 0.7154098702381573,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008055914633571669,
+      "loss": 1.1066,
+      "step": 8151
+    },
+    {
+      "epoch": 0.7154976398210597,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008052605961073172,
+      "loss": 1.1849,
+      "step": 8152
+    },
+    {
+      "epoch": 0.7155854094039622,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000804929812229904,
+      "loss": 1.1605,
+      "step": 8153
+    },
+    {
+      "epoch": 0.7156731789868648,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0008045991117575744,
+      "loss": 1.1168,
+      "step": 8154
+    },
+    {
+      "epoch": 0.7157609485697672,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0008042684947229675,
+      "loss": 1.0826,
+      "step": 8155
+    },
+    {
+      "epoch": 0.7158487181526697,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0008039379611587137,
+      "loss": 1.217,
+      "step": 8156
+    },
+    {
+      "epoch": 0.7159364877355722,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0008036075110974354,
+      "loss": 1.1792,
+      "step": 8157
+    },
+    {
+      "epoch": 0.7160242573184746,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008032771445717475,
+      "loss": 1.146,
+      "step": 8158
+    },
+    {
+      "epoch": 0.7161120269013772,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008029468616142544,
+      "loss": 1.1615,
+      "step": 8159
+    },
+    {
+      "epoch": 0.7161997964842797,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008026166622575546,
+      "loss": 1.1208,
+      "step": 8160
+    },
+    {
+      "epoch": 0.7162875660671821,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0008022865465342372,
+      "loss": 1.1549,
+      "step": 8161
+    },
+    {
+      "epoch": 0.7163753356500846,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008019565144768839,
+      "loss": 1.1125,
+      "step": 8162
+    },
+    {
+      "epoch": 0.7164631052329871,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008016265661180661,
+      "loss": 1.1935,
+      "step": 8163
+    },
+    {
+      "epoch": 0.7165508748158896,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0008012967014903499,
+      "loss": 1.1977,
+      "step": 8164
+    },
+    {
+      "epoch": 0.7166386443987921,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008009669206262915,
+      "loss": 1.1273,
+      "step": 8165
+    },
+    {
+      "epoch": 0.7167264139816946,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008006372235584383,
+      "loss": 1.1459,
+      "step": 8166
+    },
+    {
+      "epoch": 0.716814183564597,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008003076103193302,
+      "loss": 1.1948,
+      "step": 8167
+    },
+    {
+      "epoch": 0.7169019531474995,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.000799978080941499,
+      "loss": 1.1416,
+      "step": 8168
+    },
+    {
+      "epoch": 0.716989722730402,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0007996486354574679,
+      "loss": 1.1834,
+      "step": 8169
+    },
+    {
+      "epoch": 0.7170774923133045,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0007993192738997521,
+      "loss": 1.154,
+      "step": 8170
+    },
+    {
+      "epoch": 0.717165261896207,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0007989899963008575,
+      "loss": 1.2251,
+      "step": 8171
+    },
+    {
+      "epoch": 0.7172530314791095,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0007986608026932832,
+      "loss": 1.1749,
+      "step": 8172
+    },
+    {
+      "epoch": 0.7173408010620119,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.000798331693109519,
+      "loss": 1.1721,
+      "step": 8173
+    },
+    {
+      "epoch": 0.7174285706449144,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007980026675820468,
+      "loss": 1.1117,
+      "step": 8174
+    },
+    {
+      "epoch": 0.717516340227817,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0007976737261433401,
+      "loss": 1.202,
+      "step": 8175
+    },
+    {
+      "epoch": 0.7176041098107194,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007973448688258642,
+      "loss": 1.2084,
+      "step": 8176
+    },
+    {
+      "epoch": 0.7176918793936219,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0007970160956620764,
+      "loss": 1.1368,
+      "step": 8177
+    },
+    {
+      "epoch": 0.7177796489765244,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0007966874066844246,
+      "loss": 1.1516,
+      "step": 8178
+    },
+    {
+      "epoch": 0.7178674185594268,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000796358801925349,
+      "loss": 1.1201,
+      "step": 8179
+    },
+    {
+      "epoch": 0.7179551881423294,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0007960302814172819,
+      "loss": 1.1424,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7180429577252319,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.000795701845192648,
+      "loss": 1.1603,
+      "step": 8181
+    },
+    {
+      "epoch": 0.7181307273081343,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.000795373493283861,
+      "loss": 1.1256,
+      "step": 8182
+    },
+    {
+      "epoch": 0.7182184968910368,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.000795045225723328,
+      "loss": 1.2039,
+      "step": 8183
+    },
+    {
+      "epoch": 0.7183062664739392,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0007947170425434489,
+      "loss": 1.1409,
+      "step": 8184
+    },
+    {
+      "epoch": 0.7183940360568418,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.000794388943776614,
+      "loss": 1.1557,
+      "step": 8185
+    },
+    {
+      "epoch": 0.7184818056397443,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007940609294552045,
+      "loss": 1.1527,
+      "step": 8186
+    },
+    {
+      "epoch": 0.7185695752226468,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0007937329996115941,
+      "loss": 1.1221,
+      "step": 8187
+    },
+    {
+      "epoch": 0.7186573448055492,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0007934051542781492,
+      "loss": 1.2086,
+      "step": 8188
+    },
+    {
+      "epoch": 0.7187451143884517,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0007930773934872262,
+      "loss": 1.1355,
+      "step": 8189
+    },
+    {
+      "epoch": 0.7188328839713543,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0007927497172711735,
+      "loss": 1.1576,
+      "step": 8190
+    },
+    {
+      "epoch": 0.7189206535542567,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0007924221256623317,
+      "loss": 1.1539,
+      "step": 8191
+    },
+    {
+      "epoch": 0.7190084231371592,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0007920946186930327,
+      "loss": 1.1644,
+      "step": 8192
+    },
+    {
+      "epoch": 0.7190961927200616,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0007917671963956004,
+      "loss": 1.1196,
+      "step": 8193
+    },
+    {
+      "epoch": 0.7191839623029641,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00079143985880235,
+      "loss": 1.1343,
+      "step": 8194
+    },
+    {
+      "epoch": 0.7192717318858667,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007911126059455882,
+      "loss": 1.1543,
+      "step": 8195
+    },
+    {
+      "epoch": 0.7193595014687691,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0007907854378576139,
+      "loss": 1.1318,
+      "step": 8196
+    },
+    {
+      "epoch": 0.7194472710516716,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0007904583545707176,
+      "loss": 1.1311,
+      "step": 8197
+    },
+    {
+      "epoch": 0.7195350406345741,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007901313561171801,
+      "loss": 1.1459,
+      "step": 8198
+    },
+    {
+      "epoch": 0.7196228102174765,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0007898044425292756,
+      "loss": 1.1687,
+      "step": 8199
+    },
+    {
+      "epoch": 0.719710579800379,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0007894776138392688,
+      "loss": 1.1381,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7197983493832816,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0007891508700794172,
+      "loss": 1.1794,
+      "step": 8201
+    },
+    {
+      "epoch": 0.719886118966184,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000788824211281968,
+      "loss": 1.2086,
+      "step": 8202
+    },
+    {
+      "epoch": 0.7199738885490865,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0007884976374791613,
+      "loss": 1.1001,
+      "step": 8203
+    },
+    {
+      "epoch": 0.720061658131989,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0007881711487032298,
+      "loss": 1.2339,
+      "step": 8204
+    },
+    {
+      "epoch": 0.7201494277148914,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0007878447449863954,
+      "loss": 1.1392,
+      "step": 8205
+    },
+    {
+      "epoch": 0.720237197297794,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0007875184263608733,
+      "loss": 1.2256,
+      "step": 8206
+    },
+    {
+      "epoch": 0.7203249668806965,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0007871921928588699,
+      "loss": 1.0912,
+      "step": 8207
+    },
+    {
+      "epoch": 0.7204127364635989,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.000786866044512583,
+      "loss": 1.1243,
+      "step": 8208
+    },
+    {
+      "epoch": 0.7205005060465014,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007865399813542027,
+      "loss": 1.1164,
+      "step": 8209
+    },
+    {
+      "epoch": 0.7205882756294039,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0007862140034159094,
+      "loss": 1.1818,
+      "step": 8210
+    },
+    {
+      "epoch": 0.7206760452123064,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.000785888110729876,
+      "loss": 1.1882,
+      "step": 8211
+    },
+    {
+      "epoch": 0.7207638147952089,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.000785562303328267,
+      "loss": 1.2095,
+      "step": 8212
+    },
+    {
+      "epoch": 0.7208515843781114,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0007852365812432384,
+      "loss": 1.1593,
+      "step": 8213
+    },
+    {
+      "epoch": 0.7209393539610138,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0007849109445069373,
+      "loss": 1.1206,
+      "step": 8214
+    },
+    {
+      "epoch": 0.7210271235439163,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0007845853931515028,
+      "loss": 1.1798,
+      "step": 8215
+    },
+    {
+      "epoch": 0.7211148931268189,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007842599272090665,
+      "loss": 1.1498,
+      "step": 8216
+    },
+    {
+      "epoch": 0.7212026627097213,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0007839345467117492,
+      "loss": 1.1779,
+      "step": 8217
+    },
+    {
+      "epoch": 0.7212904322926238,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0007836092516916656,
+      "loss": 1.1119,
+      "step": 8218
+    },
+    {
+      "epoch": 0.7213782018755263,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0007832840421809205,
+      "loss": 1.158,
+      "step": 8219
+    },
+    {
+      "epoch": 0.7214659714584287,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0007829589182116111,
+      "loss": 1.1138,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7215537410413313,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.000782633879815826,
+      "loss": 1.133,
+      "step": 8221
+    },
+    {
+      "epoch": 0.7216415106242338,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0007823089270256448,
+      "loss": 1.1864,
+      "step": 8222
+    },
+    {
+      "epoch": 0.7217292802071362,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0007819840598731387,
+      "loss": 1.2169,
+      "step": 8223
+    },
+    {
+      "epoch": 0.7218170497900387,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0007816592783903723,
+      "loss": 1.1469,
+      "step": 8224
+    },
+    {
+      "epoch": 0.7219048193729412,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0007813345826093985,
+      "loss": 1.1467,
+      "step": 8225
+    },
+    {
+      "epoch": 0.7219925889558436,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0007810099725622646,
+      "loss": 1.1864,
+      "step": 8226
+    },
+    {
+      "epoch": 0.7220803585387462,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0007806854482810077,
+      "loss": 1.1445,
+      "step": 8227
+    },
+    {
+      "epoch": 0.7221681281216487,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0007803610097976579,
+      "loss": 1.1971,
+      "step": 8228
+    },
+    {
+      "epoch": 0.7222558977045511,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007800366571442351,
+      "loss": 1.1671,
+      "step": 8229
+    },
+    {
+      "epoch": 0.7223436672874536,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0007797123903527519,
+      "loss": 1.1107,
+      "step": 8230
+    },
+    {
+      "epoch": 0.722431436870356,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0007793882094552123,
+      "loss": 1.2199,
+      "step": 8231
+    },
+    {
+      "epoch": 0.7225192064532586,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0007790641144836118,
+      "loss": 1.1662,
+      "step": 8232
+    },
+    {
+      "epoch": 0.7226069760361611,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007787401054699366,
+      "loss": 1.1398,
+      "step": 8233
+    },
+    {
+      "epoch": 0.7226947456190635,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0007784161824461656,
+      "loss": 1.1298,
+      "step": 8234
+    },
+    {
+      "epoch": 0.722782515201966,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.000778092345444269,
+      "loss": 1.1341,
+      "step": 8235
+    },
+    {
+      "epoch": 0.7228702847848685,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0007777685944962081,
+      "loss": 1.1177,
+      "step": 8236
+    },
+    {
+      "epoch": 0.722958054367771,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0007774449296339354,
+      "loss": 1.2303,
+      "step": 8237
+    },
+    {
+      "epoch": 0.7230458239506735,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0007771213508893956,
+      "loss": 1.1695,
+      "step": 8238
+    },
+    {
+      "epoch": 0.723133593533576,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007767978582945247,
+      "loss": 1.2148,
+      "step": 8239
+    },
+    {
+      "epoch": 0.7232213631164784,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0007764744518812505,
+      "loss": 1.1626,
+      "step": 8240
+    },
+    {
+      "epoch": 0.7233091326993809,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0007761511316814913,
+      "loss": 1.0917,
+      "step": 8241
+    },
+    {
+      "epoch": 0.7233969022822835,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0007758278977271574,
+      "loss": 1.1469,
+      "step": 8242
+    },
+    {
+      "epoch": 0.723484671865186,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0007755047500501514,
+      "loss": 1.1313,
+      "step": 8243
+    },
+    {
+      "epoch": 0.7235724414480884,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.000775181688682366,
+      "loss": 1.1902,
+      "step": 8244
+    },
+    {
+      "epoch": 0.7236602110309909,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0007748587136556868,
+      "loss": 1.122,
+      "step": 8245
+    },
+    {
+      "epoch": 0.7237479806138933,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0007745358250019896,
+      "loss": 1.2049,
+      "step": 8246
+    },
+    {
+      "epoch": 0.7238357501967959,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007742130227531425,
+      "loss": 1.1116,
+      "step": 8247
+    },
+    {
+      "epoch": 0.7239235197796984,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0007738903069410052,
+      "loss": 1.1346,
+      "step": 8248
+    },
+    {
+      "epoch": 0.7240112893626008,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0007735676775974277,
+      "loss": 1.2077,
+      "step": 8249
+    },
+    {
+      "epoch": 0.7240990589455033,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0007732451347542522,
+      "loss": 1.1489,
+      "step": 8250
+    },
+    {
+      "epoch": 0.7241868285284058,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0007729226784433129,
+      "loss": 1.1526,
+      "step": 8251
+    },
+    {
+      "epoch": 0.7242745981113083,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007726003086964353,
+      "loss": 1.152,
+      "step": 8252
+    },
+    {
+      "epoch": 0.7243623676942108,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0007722780255454353,
+      "loss": 1.1602,
+      "step": 8253
+    },
+    {
+      "epoch": 0.7244501372771133,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0007719558290221205,
+      "loss": 1.1117,
+      "step": 8254
+    },
+    {
+      "epoch": 0.7245379068600157,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000771633719158292,
+      "loss": 1.1001,
+      "step": 8255
+    },
+    {
+      "epoch": 0.7246256764429182,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0007713116959857395,
+      "loss": 1.1352,
+      "step": 8256
+    },
+    {
+      "epoch": 0.7247134460258207,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007709897595362459,
+      "loss": 1.2054,
+      "step": 8257
+    },
+    {
+      "epoch": 0.7248012156087232,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0007706679098415848,
+      "loss": 1.1485,
+      "step": 8258
+    },
+    {
+      "epoch": 0.7248889851916257,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0007703461469335219,
+      "loss": 1.1161,
+      "step": 8259
+    },
+    {
+      "epoch": 0.7249767547745282,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0007700244708438139,
+      "loss": 1.1407,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7250645243574306,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0007697028816042086,
+      "loss": 1.2015,
+      "step": 8261
+    },
+    {
+      "epoch": 0.7251522939403331,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0007693813792464454,
+      "loss": 1.1444,
+      "step": 8262
+    },
+    {
+      "epoch": 0.7252400635232357,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0007690599638022559,
+      "loss": 1.1668,
+      "step": 8263
+    },
+    {
+      "epoch": 0.7253278331061381,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.000768738635303362,
+      "loss": 1.0985,
+      "step": 8264
+    },
+    {
+      "epoch": 0.7254156026890406,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.000768417393781478,
+      "loss": 1.1922,
+      "step": 8265
+    },
+    {
+      "epoch": 0.7255033722719431,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000768096239268309,
+      "loss": 1.1707,
+      "step": 8266
+    },
+    {
+      "epoch": 0.7255911418548455,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.000767775171795552,
+      "loss": 1.1993,
+      "step": 8267
+    },
+    {
+      "epoch": 0.7256789114377481,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0007674541913948943,
+      "loss": 1.2496,
+      "step": 8268
+    },
+    {
+      "epoch": 0.7257666810206506,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.000767133298098016,
+      "loss": 1.1862,
+      "step": 8269
+    },
+    {
+      "epoch": 0.725854450603553,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0007668124919365877,
+      "loss": 1.1618,
+      "step": 8270
+    },
+    {
+      "epoch": 0.7259422201864555,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0007664917729422718,
+      "loss": 1.1247,
+      "step": 8271
+    },
+    {
+      "epoch": 0.726029989769358,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0007661711411467226,
+      "loss": 1.154,
+      "step": 8272
+    },
+    {
+      "epoch": 0.7261177593522605,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0007658505965815843,
+      "loss": 1.1557,
+      "step": 8273
+    },
+    {
+      "epoch": 0.726205528935163,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007655301392784929,
+      "loss": 1.1444,
+      "step": 8274
+    },
+    {
+      "epoch": 0.7262932985180655,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0007652097692690783,
+      "loss": 1.1556,
+      "step": 8275
+    },
+    {
+      "epoch": 0.7263810681009679,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0007648894865849581,
+      "loss": 1.2098,
+      "step": 8276
+    },
+    {
+      "epoch": 0.7264688376838704,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0007645692912577433,
+      "loss": 1.1254,
+      "step": 8277
+    },
+    {
+      "epoch": 0.726556607266773,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0007642491833190359,
+      "loss": 1.1025,
+      "step": 8278
+    },
+    {
+      "epoch": 0.7266443768496754,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00076392916280043,
+      "loss": 1.1741,
+      "step": 8279
+    },
+    {
+      "epoch": 0.7267321464325779,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.000763609229733509,
+      "loss": 1.1587,
+      "step": 8280
+    },
+    {
+      "epoch": 0.7268199160154803,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00076328938414985,
+      "loss": 1.1053,
+      "step": 8281
+    },
+    {
+      "epoch": 0.7269076855983828,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0007629696260810201,
+      "loss": 1.213,
+      "step": 8282
+    },
+    {
+      "epoch": 0.7269954551812853,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007626499555585788,
+      "loss": 1.1238,
+      "step": 8283
+    },
+    {
+      "epoch": 0.7270832247641879,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.000762330372614075,
+      "loss": 1.1886,
+      "step": 8284
+    },
+    {
+      "epoch": 0.7271709943470903,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0007620108772790515,
+      "loss": 1.1812,
+      "step": 8285
+    },
+    {
+      "epoch": 0.7272587639299928,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0007616914695850407,
+      "loss": 1.1265,
+      "step": 8286
+    },
+    {
+      "epoch": 0.7273465335128952,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0007613721495635677,
+      "loss": 1.1719,
+      "step": 8287
+    },
+    {
+      "epoch": 0.7274343030957977,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0007610529172461466,
+      "loss": 1.1754,
+      "step": 8288
+    },
+    {
+      "epoch": 0.7275220726787003,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0007607337726642853,
+      "loss": 1.1208,
+      "step": 8289
+    },
+    {
+      "epoch": 0.7276098422616027,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0007604147158494822,
+      "loss": 1.1825,
+      "step": 8290
+    },
+    {
+      "epoch": 0.7276976118445052,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0007600957468332271,
+      "loss": 1.1893,
+      "step": 8291
+    },
+    {
+      "epoch": 0.7277853814274077,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.000759776865647,
+      "loss": 1.1803,
+      "step": 8292
+    },
+    {
+      "epoch": 0.7278731510103101,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007594580723222739,
+      "loss": 1.1355,
+      "step": 8293
+    },
+    {
+      "epoch": 0.7279609205932127,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0007591393668905125,
+      "loss": 1.175,
+      "step": 8294
+    },
+    {
+      "epoch": 0.7280486901761152,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0007588207493831705,
+      "loss": 1.1554,
+      "step": 8295
+    },
+    {
+      "epoch": 0.7281364597590176,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0007585022198316944,
+      "loss": 1.1618,
+      "step": 8296
+    },
+    {
+      "epoch": 0.7282242293419201,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0007581837782675214,
+      "loss": 1.1288,
+      "step": 8297
+    },
+    {
+      "epoch": 0.7283119989248226,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007578654247220812,
+      "loss": 1.1434,
+      "step": 8298
+    },
+    {
+      "epoch": 0.7283997685077251,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0007575471592267939,
+      "loss": 1.1366,
+      "step": 8299
+    },
+    {
+      "epoch": 0.7284875380906276,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0007572289818130703,
+      "loss": 1.0952,
+      "step": 8300
+    },
+    {
+      "epoch": 0.7285753076735301,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0007569108925123134,
+      "loss": 1.1477,
+      "step": 8301
+    },
+    {
+      "epoch": 0.7286630772564325,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0007565928913559179,
+      "loss": 1.1287,
+      "step": 8302
+    },
+    {
+      "epoch": 0.728750846839335,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007562749783752693,
+      "loss": 1.1622,
+      "step": 8303
+    },
+    {
+      "epoch": 0.7288386164222376,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0007559571536017433,
+      "loss": 1.1765,
+      "step": 8304
+    },
+    {
+      "epoch": 0.72892638600514,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000755639417066709,
+      "loss": 1.1332,
+      "step": 8305
+    },
+    {
+      "epoch": 0.7290141555880425,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0007553217688015261,
+      "loss": 1.1759,
+      "step": 8306
+    },
+    {
+      "epoch": 0.729101925170945,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007550042088375439,
+      "loss": 1.1812,
+      "step": 8307
+    },
+    {
+      "epoch": 0.7291896947538474,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.000754686737206105,
+      "loss": 1.1298,
+      "step": 8308
+    },
+    {
+      "epoch": 0.7292774643367499,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0007543693539385427,
+      "loss": 1.1134,
+      "step": 8309
+    },
+    {
+      "epoch": 0.7293652339196525,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0007540520590661809,
+      "loss": 1.1361,
+      "step": 8310
+    },
+    {
+      "epoch": 0.7294530035025549,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0007537348526203367,
+      "loss": 1.1192,
+      "step": 8311
+    },
+    {
+      "epoch": 0.7295407730854574,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007534177346323156,
+      "loss": 1.1915,
+      "step": 8312
+    },
+    {
+      "epoch": 0.7296285426683599,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0007531007051334163,
+      "loss": 1.1566,
+      "step": 8313
+    },
+    {
+      "epoch": 0.7297163122512623,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0007527837641549288,
+      "loss": 1.1459,
+      "step": 8314
+    },
+    {
+      "epoch": 0.7298040818341649,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0007524669117281337,
+      "loss": 1.1931,
+      "step": 8315
+    },
+    {
+      "epoch": 0.7298918514170674,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0007521501478843029,
+      "loss": 1.191,
+      "step": 8316
+    },
+    {
+      "epoch": 0.7299796209999698,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0007518334726546998,
+      "loss": 1.1467,
+      "step": 8317
+    },
+    {
+      "epoch": 0.7300673905828723,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00075151688607058,
+      "loss": 1.1848,
+      "step": 8318
+    },
+    {
+      "epoch": 0.7301551601657748,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0007512003881631875,
+      "loss": 1.172,
+      "step": 8319
+    },
+    {
+      "epoch": 0.7302429297486773,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0007508839789637605,
+      "loss": 1.1661,
+      "step": 8320
+    },
+    {
+      "epoch": 0.7303306993315798,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.000750567658503527,
+      "loss": 1.1335,
+      "step": 8321
+    },
+    {
+      "epoch": 0.7304184689144823,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0007502514268137075,
+      "loss": 1.1349,
+      "step": 8322
+    },
+    {
+      "epoch": 0.7305062384973847,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0007499352839255113,
+      "loss": 1.1712,
+      "step": 8323
+    },
+    {
+      "epoch": 0.7305940080802872,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0007496192298701409,
+      "loss": 1.1382,
+      "step": 8324
+    },
+    {
+      "epoch": 0.7306817776631898,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0007493032646787903,
+      "loss": 1.1627,
+      "step": 8325
+    },
+    {
+      "epoch": 0.7307695472460922,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.000748987388382644,
+      "loss": 1.1848,
+      "step": 8326
+    },
+    {
+      "epoch": 0.7308573168289947,
+      "grad_norm": 0.125,
+      "learning_rate": 0.000748671601012877,
+      "loss": 1.1663,
+      "step": 8327
+    },
+    {
+      "epoch": 0.7309450864118971,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007483559026006565,
+      "loss": 1.2134,
+      "step": 8328
+    },
+    {
+      "epoch": 0.7310328559947996,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0007480402931771409,
+      "loss": 1.1005,
+      "step": 8329
+    },
+    {
+      "epoch": 0.7311206255777022,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007477247727734797,
+      "loss": 1.2171,
+      "step": 8330
+    },
+    {
+      "epoch": 0.7312083951606047,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.000747409341420813,
+      "loss": 1.2324,
+      "step": 8331
+    },
+    {
+      "epoch": 0.7312961647435071,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0007470939991502729,
+      "loss": 1.189,
+      "step": 8332
+    },
+    {
+      "epoch": 0.7313839343264096,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0007467787459929826,
+      "loss": 1.1615,
+      "step": 8333
+    },
+    {
+      "epoch": 0.731471703909312,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0007464635819800561,
+      "loss": 1.2332,
+      "step": 8334
+    },
+    {
+      "epoch": 0.7315594734922146,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.000746148507142599,
+      "loss": 1.1637,
+      "step": 8335
+    },
+    {
+      "epoch": 0.7316472430751171,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0007458335215117078,
+      "loss": 1.1804,
+      "step": 8336
+    },
+    {
+      "epoch": 0.7317350126580195,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0007455186251184705,
+      "loss": 1.1391,
+      "step": 8337
+    },
+    {
+      "epoch": 0.731822782240922,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0007452038179939666,
+      "loss": 1.2426,
+      "step": 8338
+    },
+    {
+      "epoch": 0.7319105518238245,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0007448891001692651,
+      "loss": 1.1535,
+      "step": 8339
+    },
+    {
+      "epoch": 0.7319983214067269,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.000744574471675428,
+      "loss": 1.1503,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7320860909896295,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0007442599325435082,
+      "loss": 1.1845,
+      "step": 8341
+    },
+    {
+      "epoch": 0.732173860572532,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007439454828045496,
+      "loss": 1.0727,
+      "step": 8342
+    },
+    {
+      "epoch": 0.7322616301554344,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0007436311224895862,
+      "loss": 1.1311,
+      "step": 8343
+    },
+    {
+      "epoch": 0.7323493997383369,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0007433168516296445,
+      "loss": 1.1518,
+      "step": 8344
+    },
+    {
+      "epoch": 0.7324371693212394,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0007430026702557426,
+      "loss": 1.1488,
+      "step": 8345
+    },
+    {
+      "epoch": 0.7325249389041419,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.000742688578398888,
+      "loss": 1.1479,
+      "step": 8346
+    },
+    {
+      "epoch": 0.7326127084870444,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0007423745760900807,
+      "loss": 1.1437,
+      "step": 8347
+    },
+    {
+      "epoch": 0.7327004780699469,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007420606633603113,
+      "loss": 1.1732,
+      "step": 8348
+    },
+    {
+      "epoch": 0.7327882476528493,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000741746840240562,
+      "loss": 1.1703,
+      "step": 8349
+    },
+    {
+      "epoch": 0.7328760172357518,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0007414331067618064,
+      "loss": 1.1776,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7329637868186544,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007411194629550075,
+      "loss": 1.1625,
+      "step": 8351
+    },
+    {
+      "epoch": 0.7330515564015568,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007408059088511213,
+      "loss": 1.1393,
+      "step": 8352
+    },
+    {
+      "epoch": 0.7331393259844593,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0007404924444810944,
+      "loss": 1.1328,
+      "step": 8353
+    },
+    {
+      "epoch": 0.7332270955673618,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0007401790698758646,
+      "loss": 1.1462,
+      "step": 8354
+    },
+    {
+      "epoch": 0.7333148651502642,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0007398657850663607,
+      "loss": 1.1509,
+      "step": 8355
+    },
+    {
+      "epoch": 0.7334026347331668,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0007395525900835026,
+      "loss": 1.1044,
+      "step": 8356
+    },
+    {
+      "epoch": 0.7334904043160693,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007392394849582016,
+      "loss": 1.1303,
+      "step": 8357
+    },
+    {
+      "epoch": 0.7335781738989717,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007389264697213595,
+      "loss": 1.1039,
+      "step": 8358
+    },
+    {
+      "epoch": 0.7336659434818742,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0007386135444038698,
+      "loss": 1.1484,
+      "step": 8359
+    },
+    {
+      "epoch": 0.7337537130647767,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.000738300709036617,
+      "loss": 1.1153,
+      "step": 8360
+    },
+    {
+      "epoch": 0.7338414826476792,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.000737987963650477,
+      "loss": 1.187,
+      "step": 8361
+    },
+    {
+      "epoch": 0.7339292522305817,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0007376753082763167,
+      "loss": 1.1616,
+      "step": 8362
+    },
+    {
+      "epoch": 0.7340170218134842,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0007373627429449932,
+      "loss": 1.1584,
+      "step": 8363
+    },
+    {
+      "epoch": 0.7341047913963866,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0007370502676873553,
+      "loss": 1.1824,
+      "step": 8364
+    },
+    {
+      "epoch": 0.7341925609792891,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0007367378825342446,
+      "loss": 1.2153,
+      "step": 8365
+    },
+    {
+      "epoch": 0.7342803305621916,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0007364255875164908,
+      "loss": 1.1826,
+      "step": 8366
+    },
+    {
+      "epoch": 0.7343681001450941,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0007361133826649166,
+      "loss": 1.1497,
+      "step": 8367
+    },
+    {
+      "epoch": 0.7344558697279966,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0007358012680103356,
+      "loss": 1.1012,
+      "step": 8368
+    },
+    {
+      "epoch": 0.734543639310899,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0007354892435835526,
+      "loss": 1.1287,
+      "step": 8369
+    },
+    {
+      "epoch": 0.7346314088938015,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0007351773094153622,
+      "loss": 1.1146,
+      "step": 8370
+    },
+    {
+      "epoch": 0.734719178476704,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0007348654655365516,
+      "loss": 1.163,
+      "step": 8371
+    },
+    {
+      "epoch": 0.7348069480596066,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0007345537119778986,
+      "loss": 1.1396,
+      "step": 8372
+    },
+    {
+      "epoch": 0.734894717642509,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0007342420487701723,
+      "loss": 1.1828,
+      "step": 8373
+    },
+    {
+      "epoch": 0.7349824872254115,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0007339304759441316,
+      "loss": 1.1609,
+      "step": 8374
+    },
+    {
+      "epoch": 0.735070256808314,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0007336189935305286,
+      "loss": 1.1658,
+      "step": 8375
+    },
+    {
+      "epoch": 0.7351580263912164,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0007333076015601049,
+      "loss": 1.206,
+      "step": 8376
+    },
+    {
+      "epoch": 0.735245795974119,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0007329963000635943,
+      "loss": 1.1918,
+      "step": 8377
+    },
+    {
+      "epoch": 0.7353335655570215,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00073268508907172,
+      "loss": 1.2066,
+      "step": 8378
+    },
+    {
+      "epoch": 0.7354213351399239,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007323739686151978,
+      "loss": 1.1498,
+      "step": 8379
+    },
+    {
+      "epoch": 0.7355091047228264,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0007320629387247341,
+      "loss": 1.1217,
+      "step": 8380
+    },
+    {
+      "epoch": 0.7355968743057288,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0007317519994310266,
+      "loss": 1.0902,
+      "step": 8381
+    },
+    {
+      "epoch": 0.7356846438886314,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.000731441150764763,
+      "loss": 1.1749,
+      "step": 8382
+    },
+    {
+      "epoch": 0.7357724134715339,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007311303927566233,
+      "loss": 1.1661,
+      "step": 8383
+    },
+    {
+      "epoch": 0.7358601830544363,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0007308197254372779,
+      "loss": 1.1385,
+      "step": 8384
+    },
+    {
+      "epoch": 0.7359479526373388,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0007305091488373889,
+      "loss": 1.1733,
+      "step": 8385
+    },
+    {
+      "epoch": 0.7360357222202413,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0007301986629876085,
+      "loss": 1.1791,
+      "step": 8386
+    },
+    {
+      "epoch": 0.7361234918031438,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0007298882679185806,
+      "loss": 1.1444,
+      "step": 8387
+    },
+    {
+      "epoch": 0.7362112613860463,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0007295779636609399,
+      "loss": 1.2301,
+      "step": 8388
+    },
+    {
+      "epoch": 0.7362990309689488,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0007292677502453127,
+      "loss": 1.1757,
+      "step": 8389
+    },
+    {
+      "epoch": 0.7363868005518512,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.000728957627702315,
+      "loss": 1.1578,
+      "step": 8390
+    },
+    {
+      "epoch": 0.7364745701347537,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0007286475960625551,
+      "loss": 1.1385,
+      "step": 8391
+    },
+    {
+      "epoch": 0.7365623397176563,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0007283376553566318,
+      "loss": 1.191,
+      "step": 8392
+    },
+    {
+      "epoch": 0.7366501093005587,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0007280278056151355,
+      "loss": 1.1525,
+      "step": 8393
+    },
+    {
+      "epoch": 0.7367378788834612,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0007277180468686464,
+      "loss": 1.1553,
+      "step": 8394
+    },
+    {
+      "epoch": 0.7368256484663637,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0007274083791477362,
+      "loss": 1.1601,
+      "step": 8395
+    },
+    {
+      "epoch": 0.7369134180492661,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0007270988024829695,
+      "loss": 1.1663,
+      "step": 8396
+    },
+    {
+      "epoch": 0.7370011876321686,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0007267893169048987,
+      "loss": 1.1766,
+      "step": 8397
+    },
+    {
+      "epoch": 0.7370889572150712,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0007264799224440694,
+      "loss": 1.1366,
+      "step": 8398
+    },
+    {
+      "epoch": 0.7371767267979736,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0007261706191310177,
+      "loss": 1.1434,
+      "step": 8399
+    },
+    {
+      "epoch": 0.7372644963808761,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0007258614069962702,
+      "loss": 1.1159,
+      "step": 8400
+    },
+    {
+      "epoch": 0.7373522659637786,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0007255522860703458,
+      "loss": 1.102,
+      "step": 8401
+    },
+    {
+      "epoch": 0.737440035546681,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0007252432563837525,
+      "loss": 1.1612,
+      "step": 8402
+    },
+    {
+      "epoch": 0.7375278051295836,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0007249343179669906,
+      "loss": 1.1391,
+      "step": 8403
+    },
+    {
+      "epoch": 0.7376155747124861,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007246254708505514,
+      "loss": 1.1784,
+      "step": 8404
+    },
+    {
+      "epoch": 0.7377033442953885,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0007243167150649166,
+      "loss": 1.148,
+      "step": 8405
+    },
+    {
+      "epoch": 0.737791113878291,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0007240080506405593,
+      "loss": 1.1294,
+      "step": 8406
+    },
+    {
+      "epoch": 0.7378788834611935,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0007236994776079433,
+      "loss": 1.1732,
+      "step": 8407
+    },
+    {
+      "epoch": 0.737966653044096,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0007233909959975243,
+      "loss": 1.107,
+      "step": 8408
+    },
+    {
+      "epoch": 0.7380544226269985,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0007230826058397471,
+      "loss": 1.1739,
+      "step": 8409
+    },
+    {
+      "epoch": 0.738142192209901,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0007227743071650491,
+      "loss": 1.1803,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7382299617928034,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.000722466100003858,
+      "loss": 1.1353,
+      "step": 8411
+    },
+    {
+      "epoch": 0.7383177313757059,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0007221579843865928,
+      "loss": 1.1819,
+      "step": 8412
+    },
+    {
+      "epoch": 0.7384055009586085,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0007218499603436639,
+      "loss": 1.1574,
+      "step": 8413
+    },
+    {
+      "epoch": 0.7384932705415109,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0007215420279054706,
+      "loss": 1.1329,
+      "step": 8414
+    },
+    {
+      "epoch": 0.7385810401244134,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0007212341871024054,
+      "loss": 1.1885,
+      "step": 8415
+    },
+    {
+      "epoch": 0.7386688097073159,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0007209264379648515,
+      "loss": 1.1419,
+      "step": 8416
+    },
+    {
+      "epoch": 0.7387565792902183,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0007206187805231819,
+      "loss": 1.1329,
+      "step": 8417
+    },
+    {
+      "epoch": 0.7388443488731209,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007203112148077609,
+      "loss": 1.1462,
+      "step": 8418
+    },
+    {
+      "epoch": 0.7389321184560234,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0007200037408489448,
+      "loss": 1.1127,
+      "step": 8419
+    },
+    {
+      "epoch": 0.7390198880389258,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0007196963586770799,
+      "loss": 1.1505,
+      "step": 8420
+    },
+    {
+      "epoch": 0.7391076576218283,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007193890683225031,
+      "loss": 1.1477,
+      "step": 8421
+    },
+    {
+      "epoch": 0.7391954272047307,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0007190818698155427,
+      "loss": 1.1684,
+      "step": 8422
+    },
+    {
+      "epoch": 0.7392831967876332,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0007187747631865186,
+      "loss": 1.1236,
+      "step": 8423
+    },
+    {
+      "epoch": 0.7393709663705358,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.000718467748465741,
+      "loss": 1.1462,
+      "step": 8424
+    },
+    {
+      "epoch": 0.7394587359534383,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0007181608256835099,
+      "loss": 1.2067,
+      "step": 8425
+    },
+    {
+      "epoch": 0.7395465055363407,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0007178539948701185,
+      "loss": 1.1058,
+      "step": 8426
+    },
+    {
+      "epoch": 0.7396342751192432,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0007175472560558498,
+      "loss": 1.1649,
+      "step": 8427
+    },
+    {
+      "epoch": 0.7397220447021456,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0007172406092709776,
+      "loss": 1.1227,
+      "step": 8428
+    },
+    {
+      "epoch": 0.7398098142850482,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.000716934054545766,
+      "loss": 1.113,
+      "step": 8429
+    },
+    {
+      "epoch": 0.7398975838679507,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0007166275919104714,
+      "loss": 1.1899,
+      "step": 8430
+    },
+    {
+      "epoch": 0.7399853534508531,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0007163212213953403,
+      "loss": 1.1824,
+      "step": 8431
+    },
+    {
+      "epoch": 0.7400731230337556,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0007160149430306107,
+      "loss": 1.1019,
+      "step": 8432
+    },
+    {
+      "epoch": 0.7401608926166581,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0007157087568465103,
+      "loss": 1.166,
+      "step": 8433
+    },
+    {
+      "epoch": 0.7402486621995606,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0007154026628732585,
+      "loss": 1.1094,
+      "step": 8434
+    },
+    {
+      "epoch": 0.7403364317824631,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007150966611410663,
+      "loss": 1.1777,
+      "step": 8435
+    },
+    {
+      "epoch": 0.7404242013653656,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000714790751680134,
+      "loss": 1.1778,
+      "step": 8436
+    },
+    {
+      "epoch": 0.740511970948268,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0007144849345206543,
+      "loss": 1.1165,
+      "step": 8437
+    },
+    {
+      "epoch": 0.7405997405311705,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0007141792096928099,
+      "loss": 1.1769,
+      "step": 8438
+    },
+    {
+      "epoch": 0.7406875101140731,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0007138735772267746,
+      "loss": 1.1514,
+      "step": 8439
+    },
+    {
+      "epoch": 0.7407752796969755,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0007135680371527136,
+      "loss": 1.1674,
+      "step": 8440
+    },
+    {
+      "epoch": 0.740863049279878,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007132625895007817,
+      "loss": 1.1506,
+      "step": 8441
+    },
+    {
+      "epoch": 0.7409508188627805,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0007129572343011256,
+      "loss": 1.1312,
+      "step": 8442
+    },
+    {
+      "epoch": 0.7410385884456829,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0007126519715838829,
+      "loss": 1.099,
+      "step": 8443
+    },
+    {
+      "epoch": 0.7411263580285855,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.000712346801379182,
+      "loss": 1.1059,
+      "step": 8444
+    },
+    {
+      "epoch": 0.741214127611488,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0007120417237171409,
+      "loss": 1.1142,
+      "step": 8445
+    },
+    {
+      "epoch": 0.7413018971943904,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0007117367386278708,
+      "loss": 1.1303,
+      "step": 8446
+    },
+    {
+      "epoch": 0.7413896667772929,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0007114318461414725,
+      "loss": 1.1078,
+      "step": 8447
+    },
+    {
+      "epoch": 0.7414774363601954,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0007111270462880369,
+      "loss": 1.1315,
+      "step": 8448
+    },
+    {
+      "epoch": 0.7415652059430978,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007108223390976468,
+      "loss": 1.1788,
+      "step": 8449
+    },
+    {
+      "epoch": 0.7416529755260004,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0007105177246003757,
+      "loss": 1.1856,
+      "step": 8450
+    },
+    {
+      "epoch": 0.7417407451089029,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0007102132028262878,
+      "loss": 1.2113,
+      "step": 8451
+    },
+    {
+      "epoch": 0.7418285146918053,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0007099087738054388,
+      "loss": 1.1909,
+      "step": 8452
+    },
+    {
+      "epoch": 0.7419162842747078,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007096044375678735,
+      "loss": 1.1496,
+      "step": 8453
+    },
+    {
+      "epoch": 0.7420040538576103,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0007093001941436293,
+      "loss": 1.207,
+      "step": 8454
+    },
+    {
+      "epoch": 0.7420918234405128,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0007089960435627339,
+      "loss": 1.1298,
+      "step": 8455
+    },
+    {
+      "epoch": 0.7421795930234153,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0007086919858552055,
+      "loss": 1.1545,
+      "step": 8456
+    },
+    {
+      "epoch": 0.7422673626063178,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0007083880210510536,
+      "loss": 1.12,
+      "step": 8457
+    },
+    {
+      "epoch": 0.7423551321892202,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0007080841491802781,
+      "loss": 1.1787,
+      "step": 8458
+    },
+    {
+      "epoch": 0.7424429017721227,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0007077803702728707,
+      "loss": 1.1311,
+      "step": 8459
+    },
+    {
+      "epoch": 0.7425306713550253,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0007074766843588121,
+      "loss": 1.1326,
+      "step": 8460
+    },
+    {
+      "epoch": 0.7426184409379277,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0007071730914680754,
+      "loss": 1.1026,
+      "step": 8461
+    },
+    {
+      "epoch": 0.7427062105208302,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0007068695916306239,
+      "loss": 1.1872,
+      "step": 8462
+    },
+    {
+      "epoch": 0.7427939801037327,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0007065661848764121,
+      "loss": 1.1355,
+      "step": 8463
+    },
+    {
+      "epoch": 0.7428817496866351,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0007062628712353852,
+      "loss": 1.1146,
+      "step": 8464
+    },
+    {
+      "epoch": 0.7429695192695377,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007059596507374778,
+      "loss": 1.1837,
+      "step": 8465
+    },
+    {
+      "epoch": 0.7430572888524402,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000705656523412618,
+      "loss": 1.1358,
+      "step": 8466
+    },
+    {
+      "epoch": 0.7431450584353426,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0007053534892907232,
+      "loss": 1.1312,
+      "step": 8467
+    },
+    {
+      "epoch": 0.7432328280182451,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007050505484017008,
+      "loss": 1.1287,
+      "step": 8468
+    },
+    {
+      "epoch": 0.7433205976011475,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0007047477007754502,
+      "loss": 1.1703,
+      "step": 8469
+    },
+    {
+      "epoch": 0.7434083671840501,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0007044449464418615,
+      "loss": 1.1449,
+      "step": 8470
+    },
+    {
+      "epoch": 0.7434961367669526,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0007041422854308154,
+      "loss": 1.1556,
+      "step": 8471
+    },
+    {
+      "epoch": 0.743583906349855,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.000703839717772183,
+      "loss": 1.1889,
+      "step": 8472
+    },
+    {
+      "epoch": 0.7436716759327575,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0007035372434958266,
+      "loss": 1.1807,
+      "step": 8473
+    },
+    {
+      "epoch": 0.74375944551566,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0007032348626315994,
+      "loss": 1.1338,
+      "step": 8474
+    },
+    {
+      "epoch": 0.7438472150985626,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0007029325752093449,
+      "loss": 1.1493,
+      "step": 8475
+    },
+    {
+      "epoch": 0.743934984681465,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0007026303812588979,
+      "loss": 1.127,
+      "step": 8476
+    },
+    {
+      "epoch": 0.7440227542643675,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0007023282808100837,
+      "loss": 1.1338,
+      "step": 8477
+    },
+    {
+      "epoch": 0.74411052384727,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0007020262738927187,
+      "loss": 1.1619,
+      "step": 8478
+    },
+    {
+      "epoch": 0.7441982934301724,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0007017243605366097,
+      "loss": 1.1385,
+      "step": 8479
+    },
+    {
+      "epoch": 0.7442860630130749,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0007014225407715538,
+      "loss": 1.1603,
+      "step": 8480
+    },
+    {
+      "epoch": 0.7443738325959774,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0007011208146273398,
+      "loss": 1.181,
+      "step": 8481
+    },
+    {
+      "epoch": 0.7444616021788799,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0007008191821337468,
+      "loss": 1.1028,
+      "step": 8482
+    },
+    {
+      "epoch": 0.7445493717617824,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0007005176433205454,
+      "loss": 1.1458,
+      "step": 8483
+    },
+    {
+      "epoch": 0.7446371413446848,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0007002161982174951,
+      "loss": 1.1529,
+      "step": 8484
+    },
+    {
+      "epoch": 0.7447249109275873,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0006999148468543475,
+      "loss": 1.2336,
+      "step": 8485
+    },
+    {
+      "epoch": 0.7448126805104899,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0006996135892608461,
+      "loss": 1.1366,
+      "step": 8486
+    },
+    {
+      "epoch": 0.7449004500933923,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006993124254667227,
+      "loss": 1.0917,
+      "step": 8487
+    },
+    {
+      "epoch": 0.7449882196762948,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.000699011355501701,
+      "loss": 1.1885,
+      "step": 8488
+    },
+    {
+      "epoch": 0.7450759892591973,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0006987103793954959,
+      "loss": 1.236,
+      "step": 8489
+    },
+    {
+      "epoch": 0.7451637588420997,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.000698409497177812,
+      "loss": 1.1423,
+      "step": 8490
+    },
+    {
+      "epoch": 0.7452515284250023,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000698108708878346,
+      "loss": 1.1744,
+      "step": 8491
+    },
+    {
+      "epoch": 0.7453392980079048,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006978080145267837,
+      "loss": 1.1729,
+      "step": 8492
+    },
+    {
+      "epoch": 0.7454270675908072,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0006975074141528025,
+      "loss": 1.1777,
+      "step": 8493
+    },
+    {
+      "epoch": 0.7455148371737097,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006972069077860709,
+      "loss": 1.1564,
+      "step": 8494
+    },
+    {
+      "epoch": 0.7456026067566122,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0006969064954562474,
+      "loss": 1.1593,
+      "step": 8495
+    },
+    {
+      "epoch": 0.7456903763395147,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006966061771929816,
+      "loss": 1.195,
+      "step": 8496
+    },
+    {
+      "epoch": 0.7457781459224172,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0006963059530259139,
+      "loss": 1.1528,
+      "step": 8497
+    },
+    {
+      "epoch": 0.7458659155053197,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006960058229846752,
+      "loss": 1.151,
+      "step": 8498
+    },
+    {
+      "epoch": 0.7459536850882221,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0006957057870988866,
+      "loss": 1.132,
+      "step": 8499
+    },
+    {
+      "epoch": 0.7460414546711246,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0006954058453981609,
+      "loss": 1.1894,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7460414546711246,
+      "eval_loss": 1.141708254814148,
+      "eval_runtime": 437.7426,
+      "eval_samples_per_second": 33.673,
+      "eval_steps_per_second": 8.418,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7461292242540272,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006951059979121013,
+      "loss": 1.1957,
+      "step": 8501
+    },
+    {
+      "epoch": 0.7462169938369296,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006948062446703011,
+      "loss": 1.1382,
+      "step": 8502
+    },
+    {
+      "epoch": 0.7463047634198321,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0006945065857023457,
+      "loss": 1.1155,
+      "step": 8503
+    },
+    {
+      "epoch": 0.7463925330027346,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0006942070210378092,
+      "loss": 1.1686,
+      "step": 8504
+    },
+    {
+      "epoch": 0.746480302585637,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0006939075507062576,
+      "loss": 1.1396,
+      "step": 8505
+    },
+    {
+      "epoch": 0.7465680721685395,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0006936081747372477,
+      "loss": 1.0948,
+      "step": 8506
+    },
+    {
+      "epoch": 0.7466558417514421,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.000693308893160327,
+      "loss": 1.1383,
+      "step": 8507
+    },
+    {
+      "epoch": 0.7467436113343445,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0006930097060050329,
+      "loss": 1.1544,
+      "step": 8508
+    },
+    {
+      "epoch": 0.746831380917247,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0006927106133008943,
+      "loss": 1.1371,
+      "step": 8509
+    },
+    {
+      "epoch": 0.7469191505001495,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0006924116150774306,
+      "loss": 1.1288,
+      "step": 8510
+    },
+    {
+      "epoch": 0.7470069200830519,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006921127113641511,
+      "loss": 1.1467,
+      "step": 8511
+    },
+    {
+      "epoch": 0.7470946896659545,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.000691813902190557,
+      "loss": 1.1482,
+      "step": 8512
+    },
+    {
+      "epoch": 0.747182459248857,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0006915151875861392,
+      "loss": 1.0911,
+      "step": 8513
+    },
+    {
+      "epoch": 0.7472702288317594,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00069121656758038,
+      "loss": 1.1823,
+      "step": 8514
+    },
+    {
+      "epoch": 0.7473579984146619,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0006909180422027524,
+      "loss": 1.1376,
+      "step": 8515
+    },
+    {
+      "epoch": 0.7474457679975643,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.000690619611482718,
+      "loss": 1.1867,
+      "step": 8516
+    },
+    {
+      "epoch": 0.7475335375804669,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0006903212754497325,
+      "loss": 1.1473,
+      "step": 8517
+    },
+    {
+      "epoch": 0.7476213071633694,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0006900230341332401,
+      "loss": 1.1112,
+      "step": 8518
+    },
+    {
+      "epoch": 0.7477090767462719,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006897248875626755,
+      "loss": 1.1499,
+      "step": 8519
+    },
+    {
+      "epoch": 0.7477968463291743,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0006894268357674648,
+      "loss": 1.164,
+      "step": 8520
+    },
+    {
+      "epoch": 0.7478846159120768,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0006891288787770247,
+      "loss": 1.1484,
+      "step": 8521
+    },
+    {
+      "epoch": 0.7479723854949794,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0006888310166207628,
+      "loss": 1.1421,
+      "step": 8522
+    },
+    {
+      "epoch": 0.7480601550778818,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0006885332493280757,
+      "loss": 1.1863,
+      "step": 8523
+    },
+    {
+      "epoch": 0.7481479246607843,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0006882355769283529,
+      "loss": 1.164,
+      "step": 8524
+    },
+    {
+      "epoch": 0.7482356942436867,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0006879379994509728,
+      "loss": 1.1772,
+      "step": 8525
+    },
+    {
+      "epoch": 0.7483234638265892,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0006876405169253056,
+      "loss": 1.1712,
+      "step": 8526
+    },
+    {
+      "epoch": 0.7484112334094918,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0006873431293807116,
+      "loss": 1.1831,
+      "step": 8527
+    },
+    {
+      "epoch": 0.7484990029923942,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0006870458368465417,
+      "loss": 1.1263,
+      "step": 8528
+    },
+    {
+      "epoch": 0.7485867725752967,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0006867486393521375,
+      "loss": 1.1902,
+      "step": 8529
+    },
+    {
+      "epoch": 0.7486745421581992,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0006864515369268314,
+      "loss": 1.1848,
+      "step": 8530
+    },
+    {
+      "epoch": 0.7487623117411016,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006861545295999457,
+      "loss": 1.0883,
+      "step": 8531
+    },
+    {
+      "epoch": 0.7488500813240042,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006858576174007943,
+      "loss": 1.1079,
+      "step": 8532
+    },
+    {
+      "epoch": 0.7489378509069067,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0006855608003586812,
+      "loss": 1.1182,
+      "step": 8533
+    },
+    {
+      "epoch": 0.7490256204898091,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0006852640785029013,
+      "loss": 1.1827,
+      "step": 8534
+    },
+    {
+      "epoch": 0.7491133900727116,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0006849674518627392,
+      "loss": 1.1932,
+      "step": 8535
+    },
+    {
+      "epoch": 0.7492011596556141,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0006846709204674707,
+      "loss": 1.1675,
+      "step": 8536
+    },
+    {
+      "epoch": 0.7492889292385165,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0006843744843463638,
+      "loss": 1.1222,
+      "step": 8537
+    },
+    {
+      "epoch": 0.7493766988214191,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0006840781435286738,
+      "loss": 1.2118,
+      "step": 8538
+    },
+    {
+      "epoch": 0.7494644684043216,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.000683781898043649,
+      "loss": 1.1358,
+      "step": 8539
+    },
+    {
+      "epoch": 0.749552237987224,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006834857479205279,
+      "loss": 1.1248,
+      "step": 8540
+    },
+    {
+      "epoch": 0.7496400075701265,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006831896931885393,
+      "loss": 1.153,
+      "step": 8541
+    },
+    {
+      "epoch": 0.749727777153029,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0006828937338769027,
+      "loss": 1.1219,
+      "step": 8542
+    },
+    {
+      "epoch": 0.7498155467359315,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0006825978700148276,
+      "loss": 1.1341,
+      "step": 8543
+    },
+    {
+      "epoch": 0.749903316318834,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0006823021016315149,
+      "loss": 1.1411,
+      "step": 8544
+    },
+    {
+      "epoch": 0.7499910859017365,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0006820064287561558,
+      "loss": 1.1503,
+      "step": 8545
+    },
+    {
+      "epoch": 0.7500788554846389,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0006817108514179321,
+      "loss": 1.2033,
+      "step": 8546
+    },
+    {
+      "epoch": 0.7501666250675414,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0006814153696460159,
+      "loss": 1.1501,
+      "step": 8547
+    },
+    {
+      "epoch": 0.750254394650444,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006811199834695703,
+      "loss": 1.1597,
+      "step": 8548
+    },
+    {
+      "epoch": 0.7503421642333464,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000680824692917749,
+      "loss": 1.1415,
+      "step": 8549
+    },
+    {
+      "epoch": 0.7504299338162489,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0006805294980196953,
+      "loss": 1.1668,
+      "step": 8550
+    },
+    {
+      "epoch": 0.7505177033991514,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0006802343988045441,
+      "loss": 1.1531,
+      "step": 8551
+    },
+    {
+      "epoch": 0.7506054729820538,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006799393953014207,
+      "loss": 1.1298,
+      "step": 8552
+    },
+    {
+      "epoch": 0.7506932425649564,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0006796444875394405,
+      "loss": 1.1164,
+      "step": 8553
+    },
+    {
+      "epoch": 0.7507810121478589,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0006793496755477103,
+      "loss": 1.109,
+      "step": 8554
+    },
+    {
+      "epoch": 0.7508687817307613,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0006790549593553262,
+      "loss": 1.1383,
+      "step": 8555
+    },
+    {
+      "epoch": 0.7509565513136638,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0006787603389913753,
+      "loss": 1.1799,
+      "step": 8556
+    },
+    {
+      "epoch": 0.7510443208965663,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0006784658144849367,
+      "loss": 1.111,
+      "step": 8557
+    },
+    {
+      "epoch": 0.7511320904794688,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006781713858650777,
+      "loss": 1.1155,
+      "step": 8558
+    },
+    {
+      "epoch": 0.7512198600623713,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006778770531608576,
+      "loss": 1.1247,
+      "step": 8559
+    },
+    {
+      "epoch": 0.7513076296452738,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006775828164013257,
+      "loss": 1.1699,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7513953992281762,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0006772886756155227,
+      "loss": 1.1305,
+      "step": 8561
+    },
+    {
+      "epoch": 0.7514831688110787,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.000676994630832478,
+      "loss": 1.1612,
+      "step": 8562
+    },
+    {
+      "epoch": 0.7515709383939811,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0006767006820812135,
+      "loss": 1.1629,
+      "step": 8563
+    },
+    {
+      "epoch": 0.7516587079768837,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006764068293907404,
+      "loss": 1.0811,
+      "step": 8564
+    },
+    {
+      "epoch": 0.7517464775597862,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006761130727900611,
+      "loss": 1.1527,
+      "step": 8565
+    },
+    {
+      "epoch": 0.7518342471426886,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0006758194123081677,
+      "loss": 1.1406,
+      "step": 8566
+    },
+    {
+      "epoch": 0.7519220167255911,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0006755258479740439,
+      "loss": 1.1875,
+      "step": 8567
+    },
+    {
+      "epoch": 0.7520097863084936,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006752323798166631,
+      "loss": 1.0801,
+      "step": 8568
+    },
+    {
+      "epoch": 0.7520975558913962,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0006749390078649899,
+      "loss": 1.1579,
+      "step": 8569
+    },
+    {
+      "epoch": 0.7521853254742986,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006746457321479779,
+      "loss": 1.2005,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7522730950572011,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0006743525526945731,
+      "loss": 1.1657,
+      "step": 8571
+    },
+    {
+      "epoch": 0.7523608646401035,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0006740594695337106,
+      "loss": 1.1037,
+      "step": 8572
+    },
+    {
+      "epoch": 0.752448634223006,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0006737664826943175,
+      "loss": 1.159,
+      "step": 8573
+    },
+    {
+      "epoch": 0.7525364038059086,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0006734735922053093,
+      "loss": 1.1044,
+      "step": 8574
+    },
+    {
+      "epoch": 0.752624173388811,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0006731807980955936,
+      "loss": 1.1352,
+      "step": 8575
+    },
+    {
+      "epoch": 0.7527119429717135,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006728881003940677,
+      "loss": 1.1613,
+      "step": 8576
+    },
+    {
+      "epoch": 0.752799712554616,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.000672595499129621,
+      "loss": 1.1515,
+      "step": 8577
+    },
+    {
+      "epoch": 0.7528874821375184,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006723029943311305,
+      "loss": 1.1349,
+      "step": 8578
+    },
+    {
+      "epoch": 0.752975251720421,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.000672010586027466,
+      "loss": 1.1826,
+      "step": 8579
+    },
+    {
+      "epoch": 0.7530630213033235,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006717182742474871,
+      "loss": 1.17,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7531507908862259,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0006714260590200437,
+      "loss": 1.0976,
+      "step": 8581
+    },
+    {
+      "epoch": 0.7532385604691284,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0006711339403739761,
+      "loss": 1.1327,
+      "step": 8582
+    },
+    {
+      "epoch": 0.7533263300520309,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0006708419183381156,
+      "loss": 1.1697,
+      "step": 8583
+    },
+    {
+      "epoch": 0.7534140996349334,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.000670549992941283,
+      "loss": 1.1126,
+      "step": 8584
+    },
+    {
+      "epoch": 0.7535018692178359,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0006702581642122914,
+      "loss": 1.1667,
+      "step": 8585
+    },
+    {
+      "epoch": 0.7535896388007384,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0006699664321799413,
+      "loss": 1.1885,
+      "step": 8586
+    },
+    {
+      "epoch": 0.7536774083836408,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.000669674796873027,
+      "loss": 1.1701,
+      "step": 8587
+    },
+    {
+      "epoch": 0.7537651779665433,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0006693832583203317,
+      "loss": 1.132,
+      "step": 8588
+    },
+    {
+      "epoch": 0.7538529475494458,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0006690918165506282,
+      "loss": 1.1461,
+      "step": 8589
+    },
+    {
+      "epoch": 0.7539407171323483,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0006688004715926813,
+      "loss": 1.1358,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7540284867152508,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0006685092234752454,
+      "loss": 1.176,
+      "step": 8591
+    },
+    {
+      "epoch": 0.7541162562981533,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0006682180722270656,
+      "loss": 1.1259,
+      "step": 8592
+    },
+    {
+      "epoch": 0.7542040258810557,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006679270178768776,
+      "loss": 1.1567,
+      "step": 8593
+    },
+    {
+      "epoch": 0.7542917954639582,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0006676360604534069,
+      "loss": 1.1975,
+      "step": 8594
+    },
+    {
+      "epoch": 0.7543795650468608,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0006673451999853699,
+      "loss": 1.1658,
+      "step": 8595
+    },
+    {
+      "epoch": 0.7544673346297632,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0006670544365014735,
+      "loss": 1.2089,
+      "step": 8596
+    },
+    {
+      "epoch": 0.7545551042126657,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.000666763770030415,
+      "loss": 1.1224,
+      "step": 8597
+    },
+    {
+      "epoch": 0.7546428737955682,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0006664732006008818,
+      "loss": 1.1679,
+      "step": 8598
+    },
+    {
+      "epoch": 0.7547306433784706,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.000666182728241552,
+      "loss": 1.1787,
+      "step": 8599
+    },
+    {
+      "epoch": 0.7548184129613732,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006658923529810947,
+      "loss": 1.1164,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7549061825442757,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0006656020748481679,
+      "loss": 1.1526,
+      "step": 8601
+    },
+    {
+      "epoch": 0.7549939521271781,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0006653118938714211,
+      "loss": 1.137,
+      "step": 8602
+    },
+    {
+      "epoch": 0.7550817217100806,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0006650218100794942,
+      "loss": 1.1438,
+      "step": 8603
+    },
+    {
+      "epoch": 0.755169491292983,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006647318235010173,
+      "loss": 1.1472,
+      "step": 8604
+    },
+    {
+      "epoch": 0.7552572608758856,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0006644419341646113,
+      "loss": 1.1967,
+      "step": 8605
+    },
+    {
+      "epoch": 0.7553450304587881,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0006641521420988859,
+      "loss": 1.1401,
+      "step": 8606
+    },
+    {
+      "epoch": 0.7554328000416906,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0006638624473324437,
+      "loss": 1.1187,
+      "step": 8607
+    },
+    {
+      "epoch": 0.755520569624593,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006635728498938766,
+      "loss": 1.1579,
+      "step": 8608
+    },
+    {
+      "epoch": 0.7556083392074955,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006632833498117655,
+      "loss": 1.1882,
+      "step": 8609
+    },
+    {
+      "epoch": 0.7556961087903981,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0006629939471146835,
+      "loss": 1.2035,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7557838783733005,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006627046418311937,
+      "loss": 1.1847,
+      "step": 8611
+    },
+    {
+      "epoch": 0.755871647956203,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0006624154339898494,
+      "loss": 1.1543,
+      "step": 8612
+    },
+    {
+      "epoch": 0.7559594175391054,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006621263236191938,
+      "loss": 1.162,
+      "step": 8613
+    },
+    {
+      "epoch": 0.7560471871220079,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0006618373107477613,
+      "loss": 1.1451,
+      "step": 8614
+    },
+    {
+      "epoch": 0.7561349567049105,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006615483954040763,
+      "loss": 1.1697,
+      "step": 8615
+    },
+    {
+      "epoch": 0.756222726287813,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0006612595776166534,
+      "loss": 1.1166,
+      "step": 8616
+    },
+    {
+      "epoch": 0.7563104958707154,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.000660970857413998,
+      "loss": 1.1201,
+      "step": 8617
+    },
+    {
+      "epoch": 0.7563982654536179,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0006606822348246057,
+      "loss": 1.1736,
+      "step": 8618
+    },
+    {
+      "epoch": 0.7564860350365203,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0006603937098769622,
+      "loss": 1.2015,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7565738046194228,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0006601052825995443,
+      "loss": 1.1369,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7566615742023254,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0006598169530208177,
+      "loss": 1.1814,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7567493437852278,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0006595287211692399,
+      "loss": 1.1684,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7568371133681303,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006592405870732581,
+      "loss": 1.1205,
+      "step": 8623
+    },
+    {
+      "epoch": 0.7569248829510328,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0006589525507613109,
+      "loss": 1.2387,
+      "step": 8624
+    },
+    {
+      "epoch": 0.7570126525339352,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0006586646122618248,
+      "loss": 1.1295,
+      "step": 8625
+    },
+    {
+      "epoch": 0.7571004221168378,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.000658376771603219,
+      "loss": 1.148,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7571881916997403,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.000658089028813902,
+      "loss": 1.1662,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7572759612826427,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006578013839222738,
+      "loss": 1.0862,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7573637308655452,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006575138369567227,
+      "loss": 1.223,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7574515004484477,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0006572263879456285,
+      "loss": 1.1459,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7575392700313502,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0006569390369173619,
+      "loss": 1.0904,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7576270396142527,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0006566517839002835,
+      "loss": 1.1778,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7577148091971552,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0006563646289227429,
+      "loss": 1.1191,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7578025787800576,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006560775720130819,
+      "loss": 1.1139,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7578903483629601,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0006557906131996318,
+      "loss": 1.1556,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7579781179458627,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0006555037525107148,
+      "loss": 1.1601,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7580658875287651,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0006552169899746413,
+      "loss": 1.1592,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7581536571116676,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0006549303256197154,
+      "loss": 1.1812,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7582414266945701,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0006546437594742293,
+      "loss": 1.2025,
+      "step": 8639
+    },
+    {
+      "epoch": 0.7583291962774725,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0006543572915664654,
+      "loss": 1.1903,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7584169658603751,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006540709219246973,
+      "loss": 1.148,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7585047354432776,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0006537846505771884,
+      "loss": 1.1705,
+      "step": 8642
+    },
+    {
+      "epoch": 0.75859250502618,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0006534984775521929,
+      "loss": 1.2103,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7586802746090825,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0006532124028779549,
+      "loss": 1.0952,
+      "step": 8644
+    },
+    {
+      "epoch": 0.758768044191985,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0006529264265827081,
+      "loss": 1.1011,
+      "step": 8645
+    },
+    {
+      "epoch": 0.7588558137748874,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.000652640548694678,
+      "loss": 1.1525,
+      "step": 8646
+    },
+    {
+      "epoch": 0.75894358335779,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0006523547692420795,
+      "loss": 1.1269,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7590313529406925,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0006520690882531176,
+      "loss": 1.1823,
+      "step": 8648
+    },
+    {
+      "epoch": 0.7591191225235949,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006517835057559882,
+      "loss": 1.2419,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7592068921064974,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0006514980217788768,
+      "loss": 1.1298,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7592946616893999,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006512126363499604,
+      "loss": 1.1652,
+      "step": 8651
+    },
+    {
+      "epoch": 0.7593824312723024,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006509273494974044,
+      "loss": 1.1391,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7594702008552049,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0006506421612493657,
+      "loss": 1.1167,
+      "step": 8653
+    },
+    {
+      "epoch": 0.7595579704381074,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006503570716339916,
+      "loss": 1.1395,
+      "step": 8654
+    },
+    {
+      "epoch": 0.7596457400210098,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0006500720806794189,
+      "loss": 1.161,
+      "step": 8655
+    },
+    {
+      "epoch": 0.7597335096039123,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000649787188413776,
+      "loss": 1.127,
+      "step": 8656
+    },
+    {
+      "epoch": 0.7598212791868149,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0006495023948651791,
+      "loss": 1.0858,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7599090487697173,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0006492177000617374,
+      "loss": 1.1626,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7599968183526198,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0006489331040315492,
+      "loss": 1.1348,
+      "step": 8659
+    },
+    {
+      "epoch": 0.7600845879355222,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0006486486068027023,
+      "loss": 1.1339,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7601723575184247,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006483642084032757,
+      "loss": 1.1278,
+      "step": 8661
+    },
+    {
+      "epoch": 0.7602601271013273,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006480799088613384,
+      "loss": 1.1566,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7603478966842298,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0006477957082049501,
+      "loss": 1.1427,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7604356662671322,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006475116064621594,
+      "loss": 1.1486,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7605234358500347,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0006472276036610066,
+      "loss": 1.1566,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7606112054329371,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0006469436998295217,
+      "loss": 1.2041,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7606989750158397,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0006466598949957245,
+      "loss": 1.1775,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7607867445987422,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0006463761891876258,
+      "loss": 1.1453,
+      "step": 8668
+    },
+    {
+      "epoch": 0.7608745141816446,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006460925824332261,
+      "loss": 1.1149,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7609622837645471,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0006458090747605163,
+      "loss": 1.1587,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7610500533474496,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006455256661974776,
+      "loss": 1.1349,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7611378229303521,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0006452423567720807,
+      "loss": 1.2067,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7612255925132546,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0006449591465122878,
+      "loss": 1.1151,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7613133620961571,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0006446760354460502,
+      "loss": 1.1831,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7614011316790595,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0006443930236013106,
+      "loss": 1.1651,
+      "step": 8675
+    },
+    {
+      "epoch": 0.761488901261962,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0006441101110060002,
+      "loss": 1.169,
+      "step": 8676
+    },
+    {
+      "epoch": 0.7615766708448645,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0006438272976880412,
+      "loss": 1.0868,
+      "step": 8677
+    },
+    {
+      "epoch": 0.761664440427767,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0006435445836753472,
+      "loss": 1.1515,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7617522100106695,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0006432619689958208,
+      "loss": 1.2236,
+      "step": 8679
+    },
+    {
+      "epoch": 0.761839979593572,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006429794536773543,
+      "loss": 1.1629,
+      "step": 8680
+    },
+    {
+      "epoch": 0.7619277491764744,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0006426970377478309,
+      "loss": 1.1795,
+      "step": 8681
+    },
+    {
+      "epoch": 0.7620155187593769,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0006424147212351245,
+      "loss": 1.1077,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7621032883422795,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0006421325041670985,
+      "loss": 1.1042,
+      "step": 8683
+    },
+    {
+      "epoch": 0.7621910579251819,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.000641850386571606,
+      "loss": 1.1544,
+      "step": 8684
+    },
+    {
+      "epoch": 0.7622788275080844,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0006415683684764912,
+      "loss": 1.1374,
+      "step": 8685
+    },
+    {
+      "epoch": 0.7623665970909869,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006412864499095885,
+      "loss": 1.138,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7624543666738893,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0006410046308987218,
+      "loss": 1.1721,
+      "step": 8687
+    },
+    {
+      "epoch": 0.7625421362567919,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0006407229114717057,
+      "loss": 1.1118,
+      "step": 8688
+    },
+    {
+      "epoch": 0.7626299058396944,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0006404412916563445,
+      "loss": 1.1628,
+      "step": 8689
+    },
+    {
+      "epoch": 0.7627176754225968,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000640159771480434,
+      "loss": 1.1699,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7628054450054993,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0006398783509717576,
+      "loss": 1.146,
+      "step": 8691
+    },
+    {
+      "epoch": 0.7628932145884018,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0006395970301580912,
+      "loss": 1.1182,
+      "step": 8692
+    },
+    {
+      "epoch": 0.7629809841713043,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0006393158090672,
+      "loss": 1.1061,
+      "step": 8693
+    },
+    {
+      "epoch": 0.7630687537542068,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006390346877268394,
+      "loss": 1.1479,
+      "step": 8694
+    },
+    {
+      "epoch": 0.7631565233371093,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.000638753666164755,
+      "loss": 1.1956,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7632442929200117,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006384727444086824,
+      "loss": 1.0852,
+      "step": 8696
+    },
+    {
+      "epoch": 0.7633320625029142,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000638191922486347,
+      "loss": 1.1165,
+      "step": 8697
+    },
+    {
+      "epoch": 0.7634198320858168,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0006379112004254663,
+      "loss": 1.1223,
+      "step": 8698
+    },
+    {
+      "epoch": 0.7635076016687192,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0006376305782537448,
+      "loss": 1.1657,
+      "step": 8699
+    },
+    {
+      "epoch": 0.7635953712516217,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0006373500559988796,
+      "loss": 1.1898,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7636831408345242,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0006370696336885571,
+      "loss": 1.1642,
+      "step": 8701
+    },
+    {
+      "epoch": 0.7637709104174266,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0006367893113504542,
+      "loss": 1.1641,
+      "step": 8702
+    },
+    {
+      "epoch": 0.7638586800003291,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0006365090890122367,
+      "loss": 1.148,
+      "step": 8703
+    },
+    {
+      "epoch": 0.7639464495832317,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.000636228966701562,
+      "loss": 1.2143,
+      "step": 8704
+    },
+    {
+      "epoch": 0.7640342191661341,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0006359489444460768,
+      "loss": 1.1437,
+      "step": 8705
+    },
+    {
+      "epoch": 0.7641219887490366,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0006356690222734185,
+      "loss": 1.1094,
+      "step": 8706
+    },
+    {
+      "epoch": 0.764209758331939,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006353892002112143,
+      "loss": 1.132,
+      "step": 8707
+    },
+    {
+      "epoch": 0.7642975279148415,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0006351094782870814,
+      "loss": 1.1507,
+      "step": 8708
+    },
+    {
+      "epoch": 0.7643852974977441,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0006348298565286273,
+      "loss": 1.1495,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7644730670806466,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0006345503349634499,
+      "loss": 1.1036,
+      "step": 8710
+    },
+    {
+      "epoch": 0.764560836663549,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.000634270913619136,
+      "loss": 1.0846,
+      "step": 8711
+    },
+    {
+      "epoch": 0.7646486062464515,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0006339915925232639,
+      "loss": 1.116,
+      "step": 8712
+    },
+    {
+      "epoch": 0.7647363758293539,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0006337123717034016,
+      "loss": 1.1801,
+      "step": 8713
+    },
+    {
+      "epoch": 0.7648241454122565,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0006334332511871073,
+      "loss": 1.0995,
+      "step": 8714
+    },
+    {
+      "epoch": 0.764911914995159,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0006331542310019283,
+      "loss": 1.1235,
+      "step": 8715
+    },
+    {
+      "epoch": 0.7649996845780614,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0006328753111754033,
+      "loss": 1.1325,
+      "step": 8716
+    },
+    {
+      "epoch": 0.7650874541609639,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006325964917350601,
+      "loss": 1.1824,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7651752237438664,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0006323177727084184,
+      "loss": 1.1909,
+      "step": 8718
+    },
+    {
+      "epoch": 0.765262993326769,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0006320391541229852,
+      "loss": 1.1964,
+      "step": 8719
+    },
+    {
+      "epoch": 0.7653507629096714,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0006317606360062598,
+      "loss": 1.138,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7654385324925739,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0006314822183857307,
+      "loss": 1.1818,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7655263020754763,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0006312039012888768,
+      "loss": 1.1296,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7656140716583788,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0006309256847431663,
+      "loss": 1.1908,
+      "step": 8723
+    },
+    {
+      "epoch": 0.7657018412412814,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006306475687760584,
+      "loss": 1.166,
+      "step": 8724
+    },
+    {
+      "epoch": 0.7657896108241838,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0006303695534150023,
+      "loss": 1.1742,
+      "step": 8725
+    },
+    {
+      "epoch": 0.7658773804070863,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0006300916386874369,
+      "loss": 1.1255,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7659651499899888,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0006298138246207906,
+      "loss": 1.1758,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7660529195728912,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0006295361112424834,
+      "loss": 1.1753,
+      "step": 8728
+    },
+    {
+      "epoch": 0.7661406891557937,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0006292584985799248,
+      "loss": 1.1,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7662284587386963,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0006289809866605132,
+      "loss": 1.0972,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7663162283215987,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.000628703575511638,
+      "loss": 1.1394,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7664039979045012,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0006284262651606788,
+      "loss": 1.1526,
+      "step": 8732
+    },
+    {
+      "epoch": 0.7664917674874037,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0006281490556350052,
+      "loss": 1.1279,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7665795370703061,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006278719469619769,
+      "loss": 1.1102,
+      "step": 8734
+    },
+    {
+      "epoch": 0.7666673066532087,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0006275949391689429,
+      "loss": 1.1428,
+      "step": 8735
+    },
+    {
+      "epoch": 0.7667550762361112,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0006273180322832427,
+      "loss": 1.1647,
+      "step": 8736
+    },
+    {
+      "epoch": 0.7668428458190136,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0006270412263322063,
+      "loss": 1.1026,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7669306154019161,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.000626764521343153,
+      "loss": 1.154,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7670183849848186,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0006264879173433929,
+      "loss": 1.0877,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7671061545677211,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0006262114143602255,
+      "loss": 1.2151,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7671939241506236,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0006259350124209409,
+      "loss": 1.1715,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7672816937335261,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0006256587115528183,
+      "loss": 1.1832,
+      "step": 8742
+    },
+    {
+      "epoch": 0.7673694633164285,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0006253825117831276,
+      "loss": 1.1227,
+      "step": 8743
+    },
+    {
+      "epoch": 0.767457232899331,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006251064131391289,
+      "loss": 1.1478,
+      "step": 8744
+    },
+    {
+      "epoch": 0.7675450024822336,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0006248304156480718,
+      "loss": 1.1023,
+      "step": 8745
+    },
+    {
+      "epoch": 0.767632772065136,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0006245545193371967,
+      "loss": 1.1231,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7677205416480385,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0006242787242337323,
+      "loss": 1.2054,
+      "step": 8747
+    },
+    {
+      "epoch": 0.767808311230941,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006240030303648995,
+      "loss": 1.1875,
+      "step": 8748
+    },
+    {
+      "epoch": 0.7678960808138434,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006237274377579085,
+      "loss": 1.1708,
+      "step": 8749
+    },
+    {
+      "epoch": 0.767983850396746,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0006234519464399583,
+      "loss": 1.1638,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7680716199796485,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0006231765564382392,
+      "loss": 1.097,
+      "step": 8751
+    },
+    {
+      "epoch": 0.7681593895625509,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006229012677799307,
+      "loss": 1.1916,
+      "step": 8752
+    },
+    {
+      "epoch": 0.7682471591454534,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006226260804922038,
+      "loss": 1.0675,
+      "step": 8753
+    },
+    {
+      "epoch": 0.7683349287283558,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.000622350994602217,
+      "loss": 1.1247,
+      "step": 8754
+    },
+    {
+      "epoch": 0.7684226983112584,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006220760101371209,
+      "loss": 1.1291,
+      "step": 8755
+    },
+    {
+      "epoch": 0.7685104678941609,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0006218011271240554,
+      "loss": 1.1078,
+      "step": 8756
+    },
+    {
+      "epoch": 0.7685982374770634,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0006215263455901501,
+      "loss": 1.1702,
+      "step": 8757
+    },
+    {
+      "epoch": 0.7686860070599658,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006212516655625253,
+      "loss": 1.1328,
+      "step": 8758
+    },
+    {
+      "epoch": 0.7687737766428683,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0006209770870682904,
+      "loss": 1.1639,
+      "step": 8759
+    },
+    {
+      "epoch": 0.7688615462257707,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0006207026101345453,
+      "loss": 1.1599,
+      "step": 8760
+    },
+    {
+      "epoch": 0.7689493158086733,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0006204282347883803,
+      "loss": 1.1493,
+      "step": 8761
+    },
+    {
+      "epoch": 0.7690370853915758,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0006201539610568743,
+      "loss": 1.1697,
+      "step": 8762
+    },
+    {
+      "epoch": 0.7691248549744782,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0006198797889670975,
+      "loss": 1.1701,
+      "step": 8763
+    },
+    {
+      "epoch": 0.7692126245573807,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0006196057185461095,
+      "loss": 1.1368,
+      "step": 8764
+    },
+    {
+      "epoch": 0.7693003941402832,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0006193317498209604,
+      "loss": 1.1793,
+      "step": 8765
+    },
+    {
+      "epoch": 0.7693881637231857,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000619057882818689,
+      "loss": 1.1404,
+      "step": 8766
+    },
+    {
+      "epoch": 0.7694759333060882,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0006187841175663255,
+      "loss": 1.1415,
+      "step": 8767
+    },
+    {
+      "epoch": 0.7695637028889907,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0006185104540908888,
+      "loss": 1.155,
+      "step": 8768
+    },
+    {
+      "epoch": 0.7696514724718931,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0006182368924193897,
+      "loss": 1.1097,
+      "step": 8769
+    },
+    {
+      "epoch": 0.7697392420547956,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0006179634325788265,
+      "loss": 1.1739,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7698270116376982,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0006176900745961888,
+      "loss": 1.1454,
+      "step": 8771
+    },
+    {
+      "epoch": 0.7699147812206006,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0006174168184984562,
+      "loss": 1.1491,
+      "step": 8772
+    },
+    {
+      "epoch": 0.7700025508035031,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0006171436643125982,
+      "loss": 1.1386,
+      "step": 8773
+    },
+    {
+      "epoch": 0.7700903203864056,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0006168706120655734,
+      "loss": 1.1506,
+      "step": 8774
+    },
+    {
+      "epoch": 0.770178089969308,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0006165976617843312,
+      "loss": 1.1295,
+      "step": 8775
+    },
+    {
+      "epoch": 0.7702658595522106,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0006163248134958108,
+      "loss": 1.1491,
+      "step": 8776
+    },
+    {
+      "epoch": 0.7703536291351131,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0006160520672269419,
+      "loss": 1.1225,
+      "step": 8777
+    },
+    {
+      "epoch": 0.7704413987180155,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0006157794230046417,
+      "loss": 1.1135,
+      "step": 8778
+    },
+    {
+      "epoch": 0.770529168300918,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0006155068808558209,
+      "loss": 1.2143,
+      "step": 8779
+    },
+    {
+      "epoch": 0.7706169378838205,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0006152344408073779,
+      "loss": 1.1266,
+      "step": 8780
+    },
+    {
+      "epoch": 0.770704707466723,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.000614962102886201,
+      "loss": 1.1008,
+      "step": 8781
+    },
+    {
+      "epoch": 0.7707924770496255,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.000614689867119169,
+      "loss": 1.1411,
+      "step": 8782
+    },
+    {
+      "epoch": 0.770880246632528,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0006144177335331507,
+      "loss": 1.1368,
+      "step": 8783
+    },
+    {
+      "epoch": 0.7709680162154304,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0006141457021550044,
+      "loss": 1.1618,
+      "step": 8784
+    },
+    {
+      "epoch": 0.7710557857983329,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.000613873773011579,
+      "loss": 1.1666,
+      "step": 8785
+    },
+    {
+      "epoch": 0.7711435553812354,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0006136019461297123,
+      "loss": 1.1844,
+      "step": 8786
+    },
+    {
+      "epoch": 0.7712313249641379,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0006133302215362324,
+      "loss": 1.2051,
+      "step": 8787
+    },
+    {
+      "epoch": 0.7713190945470404,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0006130585992579578,
+      "loss": 1.1258,
+      "step": 8788
+    },
+    {
+      "epoch": 0.7714068641299429,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0006127870793216966,
+      "loss": 1.1952,
+      "step": 8789
+    },
+    {
+      "epoch": 0.7714946337128453,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0006125156617542467,
+      "loss": 1.2428,
+      "step": 8790
+    },
+    {
+      "epoch": 0.7715824032957478,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006122443465823957,
+      "loss": 1.0708,
+      "step": 8791
+    },
+    {
+      "epoch": 0.7716701728786504,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0006119731338329219,
+      "loss": 1.1709,
+      "step": 8792
+    },
+    {
+      "epoch": 0.7717579424615528,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0006117020235325921,
+      "loss": 1.121,
+      "step": 8793
+    },
+    {
+      "epoch": 0.7718457120444553,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0006114310157081644,
+      "loss": 1.1171,
+      "step": 8794
+    },
+    {
+      "epoch": 0.7719334816273578,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0006111601103863859,
+      "loss": 1.0908,
+      "step": 8795
+    },
+    {
+      "epoch": 0.7720212512102602,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.000610889307593994,
+      "loss": 1.1811,
+      "step": 8796
+    },
+    {
+      "epoch": 0.7721090207931628,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0006106186073577164,
+      "loss": 1.2217,
+      "step": 8797
+    },
+    {
+      "epoch": 0.7721967903760653,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0006103480097042687,
+      "loss": 1.1678,
+      "step": 8798
+    },
+    {
+      "epoch": 0.7722845599589677,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0006100775146603592,
+      "loss": 1.1732,
+      "step": 8799
+    },
+    {
+      "epoch": 0.7723723295418702,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0006098071222526847,
+      "loss": 1.1286,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7724600991247726,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000609536832507931,
+      "loss": 1.1574,
+      "step": 8801
+    },
+    {
+      "epoch": 0.7725478687076752,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.000609266645452775,
+      "loss": 1.177,
+      "step": 8802
+    },
+    {
+      "epoch": 0.7726356382905777,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0006089965611138831,
+      "loss": 1.1774,
+      "step": 8803
+    },
+    {
+      "epoch": 0.7727234078734802,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0006087265795179119,
+      "loss": 1.1507,
+      "step": 8804
+    },
+    {
+      "epoch": 0.7728111774563826,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006084567006915069,
+      "loss": 1.1198,
+      "step": 8805
+    },
+    {
+      "epoch": 0.7728989470392851,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0006081869246613044,
+      "loss": 1.1679,
+      "step": 8806
+    },
+    {
+      "epoch": 0.7729867166221877,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0006079172514539301,
+      "loss": 1.1337,
+      "step": 8807
+    },
+    {
+      "epoch": 0.7730744862050901,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0006076476810959997,
+      "loss": 1.1922,
+      "step": 8808
+    },
+    {
+      "epoch": 0.7731622557879926,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0006073782136141189,
+      "loss": 1.1247,
+      "step": 8809
+    },
+    {
+      "epoch": 0.773250025370895,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0006071088490348831,
+      "loss": 1.0949,
+      "step": 8810
+    },
+    {
+      "epoch": 0.7733377949537975,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0006068395873848771,
+      "loss": 1.157,
+      "step": 8811
+    },
+    {
+      "epoch": 0.7734255645367001,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0006065704286906769,
+      "loss": 1.1218,
+      "step": 8812
+    },
+    {
+      "epoch": 0.7735133341196025,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0006063013729788461,
+      "loss": 1.138,
+      "step": 8813
+    },
+    {
+      "epoch": 0.773601103702505,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0006060324202759403,
+      "loss": 1.1659,
+      "step": 8814
+    },
+    {
+      "epoch": 0.7736888732854075,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0006057635706085037,
+      "loss": 1.1615,
+      "step": 8815
+    },
+    {
+      "epoch": 0.7737766428683099,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0006054948240030713,
+      "loss": 1.174,
+      "step": 8816
+    },
+    {
+      "epoch": 0.7738644124512124,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0006052261804861665,
+      "loss": 1.136,
+      "step": 8817
+    },
+    {
+      "epoch": 0.773952182034115,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0006049576400843034,
+      "loss": 1.1736,
+      "step": 8818
+    },
+    {
+      "epoch": 0.7740399516170174,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006046892028239864,
+      "loss": 1.1277,
+      "step": 8819
+    },
+    {
+      "epoch": 0.7741277211999199,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0006044208687317096,
+      "loss": 1.1682,
+      "step": 8820
+    },
+    {
+      "epoch": 0.7742154907828224,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0006041526378339553,
+      "loss": 1.184,
+      "step": 8821
+    },
+    {
+      "epoch": 0.7743032603657248,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0006038845101571976,
+      "loss": 1.1191,
+      "step": 8822
+    },
+    {
+      "epoch": 0.7743910299486274,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0006036164857278993,
+      "loss": 1.1231,
+      "step": 8823
+    },
+    {
+      "epoch": 0.7744787995315299,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0006033485645725141,
+      "loss": 1.1072,
+      "step": 8824
+    },
+    {
+      "epoch": 0.7745665691144323,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0006030807467174836,
+      "loss": 1.1581,
+      "step": 8825
+    },
+    {
+      "epoch": 0.7746543386973348,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0006028130321892412,
+      "loss": 1.1083,
+      "step": 8826
+    },
+    {
+      "epoch": 0.7747421082802373,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0006025454210142089,
+      "loss": 1.1567,
+      "step": 8827
+    },
+    {
+      "epoch": 0.7748298778631398,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.000602277913218799,
+      "loss": 1.1968,
+      "step": 8828
+    },
+    {
+      "epoch": 0.7749176474460423,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0006020105088294133,
+      "loss": 1.1695,
+      "step": 8829
+    },
+    {
+      "epoch": 0.7750054170289448,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0006017432078724438,
+      "loss": 1.1544,
+      "step": 8830
+    },
+    {
+      "epoch": 0.7750931866118472,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0006014760103742724,
+      "loss": 1.1603,
+      "step": 8831
+    },
+    {
+      "epoch": 0.7751809561947497,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0006012089163612694,
+      "loss": 1.2235,
+      "step": 8832
+    },
+    {
+      "epoch": 0.7752687257776523,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0006009419258597967,
+      "loss": 1.1794,
+      "step": 8833
+    },
+    {
+      "epoch": 0.7753564953605547,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0006006750388962049,
+      "loss": 1.1652,
+      "step": 8834
+    },
+    {
+      "epoch": 0.7754442649434572,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0006004082554968348,
+      "loss": 1.1797,
+      "step": 8835
+    },
+    {
+      "epoch": 0.7755320345263597,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0006001415756880172,
+      "loss": 1.1387,
+      "step": 8836
+    },
+    {
+      "epoch": 0.7756198041092621,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005998749994960717,
+      "loss": 1.1246,
+      "step": 8837
+    },
+    {
+      "epoch": 0.7757075736921647,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005996085269473081,
+      "loss": 1.1578,
+      "step": 8838
+    },
+    {
+      "epoch": 0.7757953432750672,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005993421580680276,
+      "loss": 1.1078,
+      "step": 8839
+    },
+    {
+      "epoch": 0.7758831128579696,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005990758928845185,
+      "loss": 1.0945,
+      "step": 8840
+    },
+    {
+      "epoch": 0.7759708824408721,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005988097314230604,
+      "loss": 1.1454,
+      "step": 8841
+    },
+    {
+      "epoch": 0.7760586520237746,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005985436737099224,
+      "loss": 1.1117,
+      "step": 8842
+    },
+    {
+      "epoch": 0.776146421606677,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.000598277719771364,
+      "loss": 1.1197,
+      "step": 8843
+    },
+    {
+      "epoch": 0.7762341911895796,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005980118696336325,
+      "loss": 1.1312,
+      "step": 8844
+    },
+    {
+      "epoch": 0.7763219607724821,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005977461233229671,
+      "loss": 1.1226,
+      "step": 8845
+    },
+    {
+      "epoch": 0.7764097303553845,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005974804808655959,
+      "loss": 1.1957,
+      "step": 8846
+    },
+    {
+      "epoch": 0.776497499938287,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005972149422877364,
+      "loss": 1.1272,
+      "step": 8847
+    },
+    {
+      "epoch": 0.7765852695211894,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005969495076155966,
+      "loss": 1.1628,
+      "step": 8848
+    },
+    {
+      "epoch": 0.776673039104092,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005966841768753737,
+      "loss": 1.1514,
+      "step": 8849
+    },
+    {
+      "epoch": 0.7767608086869945,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005964189500932549,
+      "loss": 1.1393,
+      "step": 8850
+    },
+    {
+      "epoch": 0.776848578269897,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005961538272954169,
+      "loss": 1.188,
+      "step": 8851
+    },
+    {
+      "epoch": 0.7769363478527994,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005958888085080262,
+      "loss": 1.2155,
+      "step": 8852
+    },
+    {
+      "epoch": 0.7770241174357019,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005956238937572391,
+      "loss": 1.0922,
+      "step": 8853
+    },
+    {
+      "epoch": 0.7771118870186045,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005953590830692017,
+      "loss": 1.0895,
+      "step": 8854
+    },
+    {
+      "epoch": 0.7771996566015069,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005950943764700503,
+      "loss": 1.1665,
+      "step": 8855
+    },
+    {
+      "epoch": 0.7772874261844094,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005948297739859094,
+      "loss": 1.149,
+      "step": 8856
+    },
+    {
+      "epoch": 0.7773751957673118,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005945652756428947,
+      "loss": 1.1409,
+      "step": 8857
+    },
+    {
+      "epoch": 0.7774629653502143,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005943008814671109,
+      "loss": 1.1349,
+      "step": 8858
+    },
+    {
+      "epoch": 0.7775507349331169,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005940365914846536,
+      "loss": 1.1617,
+      "step": 8859
+    },
+    {
+      "epoch": 0.7776385045160193,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005937724057216062,
+      "loss": 1.1498,
+      "step": 8860
+    },
+    {
+      "epoch": 0.7777262740989218,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005935083242040429,
+      "loss": 1.1729,
+      "step": 8861
+    },
+    {
+      "epoch": 0.7778140436818243,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005932443469580276,
+      "loss": 1.0901,
+      "step": 8862
+    },
+    {
+      "epoch": 0.7779018132647267,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005929804740096144,
+      "loss": 1.0708,
+      "step": 8863
+    },
+    {
+      "epoch": 0.7779895828476293,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005927167053848457,
+      "loss": 1.0949,
+      "step": 8864
+    },
+    {
+      "epoch": 0.7780773524305318,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005924530411097547,
+      "loss": 1.1382,
+      "step": 8865
+    },
+    {
+      "epoch": 0.7781651220134342,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005921894812103639,
+      "loss": 1.1688,
+      "step": 8866
+    },
+    {
+      "epoch": 0.7782528915963367,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005919260257126862,
+      "loss": 1.1107,
+      "step": 8867
+    },
+    {
+      "epoch": 0.7783406611792392,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005916626746427225,
+      "loss": 1.1678,
+      "step": 8868
+    },
+    {
+      "epoch": 0.7784284307621416,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005913994280264654,
+      "loss": 1.1564,
+      "step": 8869
+    },
+    {
+      "epoch": 0.7785162003450442,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005911362858898962,
+      "loss": 1.2149,
+      "step": 8870
+    },
+    {
+      "epoch": 0.7786039699279467,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005908732482589863,
+      "loss": 1.1861,
+      "step": 8871
+    },
+    {
+      "epoch": 0.7786917395108491,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005906103151596955,
+      "loss": 1.2011,
+      "step": 8872
+    },
+    {
+      "epoch": 0.7787795090937516,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005903474866179749,
+      "loss": 1.1424,
+      "step": 8873
+    },
+    {
+      "epoch": 0.7788672786766541,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005900847626597646,
+      "loss": 1.1502,
+      "step": 8874
+    },
+    {
+      "epoch": 0.7789550482595566,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005898221433109946,
+      "loss": 1.159,
+      "step": 8875
+    },
+    {
+      "epoch": 0.7790428178424591,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.000589559628597584,
+      "loss": 1.127,
+      "step": 8876
+    },
+    {
+      "epoch": 0.7791305874253616,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005892972185454419,
+      "loss": 1.1401,
+      "step": 8877
+    },
+    {
+      "epoch": 0.779218357008264,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005890349131804672,
+      "loss": 1.1947,
+      "step": 8878
+    },
+    {
+      "epoch": 0.7793061265911665,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005887727125285489,
+      "loss": 1.1445,
+      "step": 8879
+    },
+    {
+      "epoch": 0.7793938961740691,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005885106166155645,
+      "loss": 1.0948,
+      "step": 8880
+    },
+    {
+      "epoch": 0.7794816657569715,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005882486254673823,
+      "loss": 1.1607,
+      "step": 8881
+    },
+    {
+      "epoch": 0.779569435339874,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005879867391098599,
+      "loss": 1.1134,
+      "step": 8882
+    },
+    {
+      "epoch": 0.7796572049227765,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005877249575688439,
+      "loss": 1.1484,
+      "step": 8883
+    },
+    {
+      "epoch": 0.7797449745056789,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005874632808701714,
+      "loss": 1.1193,
+      "step": 8884
+    },
+    {
+      "epoch": 0.7798327440885815,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005872017090396685,
+      "loss": 1.1354,
+      "step": 8885
+    },
+    {
+      "epoch": 0.779920513671484,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005869402421031521,
+      "loss": 1.1567,
+      "step": 8886
+    },
+    {
+      "epoch": 0.7800082832543864,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005866788800864277,
+      "loss": 1.1459,
+      "step": 8887
+    },
+    {
+      "epoch": 0.7800960528372889,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005864176230152901,
+      "loss": 1.1526,
+      "step": 8888
+    },
+    {
+      "epoch": 0.7801838224201914,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005861564709155244,
+      "loss": 1.1497,
+      "step": 8889
+    },
+    {
+      "epoch": 0.7802715920030939,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005858954238129061,
+      "loss": 1.1187,
+      "step": 8890
+    },
+    {
+      "epoch": 0.7803593615859964,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005856344817331989,
+      "loss": 1.1662,
+      "step": 8891
+    },
+    {
+      "epoch": 0.7804471311688989,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005853736447021567,
+      "loss": 1.15,
+      "step": 8892
+    },
+    {
+      "epoch": 0.7805349007518013,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005851129127455234,
+      "loss": 1.1881,
+      "step": 8893
+    },
+    {
+      "epoch": 0.7806226703347038,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005848522858890324,
+      "loss": 1.1766,
+      "step": 8894
+    },
+    {
+      "epoch": 0.7807104399176064,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005845917641584058,
+      "loss": 1.1765,
+      "step": 8895
+    },
+    {
+      "epoch": 0.7807982095005088,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005843313475793566,
+      "loss": 1.1793,
+      "step": 8896
+    },
+    {
+      "epoch": 0.7808859790834113,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005840710361775867,
+      "loss": 1.1694,
+      "step": 8897
+    },
+    {
+      "epoch": 0.7809737486663137,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005838108299787878,
+      "loss": 1.1063,
+      "step": 8898
+    },
+    {
+      "epoch": 0.7810615182492162,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005835507290086412,
+      "loss": 1.1678,
+      "step": 8899
+    },
+    {
+      "epoch": 0.7811492878321187,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005832907332928179,
+      "loss": 1.1383,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7812370574150213,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005830308428569787,
+      "loss": 1.1127,
+      "step": 8901
+    },
+    {
+      "epoch": 0.7813248269979237,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005827710577267737,
+      "loss": 1.2,
+      "step": 8902
+    },
+    {
+      "epoch": 0.7814125965808262,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005825113779278422,
+      "loss": 1.1176,
+      "step": 8903
+    },
+    {
+      "epoch": 0.7815003661637286,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005822518034858139,
+      "loss": 1.2235,
+      "step": 8904
+    },
+    {
+      "epoch": 0.7815881357466311,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005819923344263076,
+      "loss": 1.1418,
+      "step": 8905
+    },
+    {
+      "epoch": 0.7816759053295337,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005817329707749325,
+      "loss": 1.111,
+      "step": 8906
+    },
+    {
+      "epoch": 0.7817636749124361,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005814737125572858,
+      "loss": 1.1578,
+      "step": 8907
+    },
+    {
+      "epoch": 0.7818514444953386,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005812145597989558,
+      "loss": 1.1602,
+      "step": 8908
+    },
+    {
+      "epoch": 0.7819392140782411,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005809555125255194,
+      "loss": 1.1149,
+      "step": 8909
+    },
+    {
+      "epoch": 0.7820269836611435,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005806965707625446,
+      "loss": 1.1882,
+      "step": 8910
+    },
+    {
+      "epoch": 0.7821147532440461,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000580437734535587,
+      "loss": 1.1413,
+      "step": 8911
+    },
+    {
+      "epoch": 0.7822025228269486,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005801790038701928,
+      "loss": 1.1901,
+      "step": 8912
+    },
+    {
+      "epoch": 0.782290292409851,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.000579920378791898,
+      "loss": 1.1908,
+      "step": 8913
+    },
+    {
+      "epoch": 0.7823780619927535,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005796618593262279,
+      "loss": 1.1151,
+      "step": 8914
+    },
+    {
+      "epoch": 0.782465831575656,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005794034454986969,
+      "loss": 1.1254,
+      "step": 8915
+    },
+    {
+      "epoch": 0.7825536011585585,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005791451373348096,
+      "loss": 1.1627,
+      "step": 8916
+    },
+    {
+      "epoch": 0.782641370741461,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005788869348600603,
+      "loss": 1.1722,
+      "step": 8917
+    },
+    {
+      "epoch": 0.7827291403243635,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005786288380999326,
+      "loss": 1.17,
+      "step": 8918
+    },
+    {
+      "epoch": 0.7828169099072659,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005783708470798987,
+      "loss": 1.1488,
+      "step": 8919
+    },
+    {
+      "epoch": 0.7829046794901684,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005781129618254222,
+      "loss": 1.1738,
+      "step": 8920
+    },
+    {
+      "epoch": 0.782992449073071,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005778551823619552,
+      "loss": 1.1267,
+      "step": 8921
+    },
+    {
+      "epoch": 0.7830802186559734,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005775975087149399,
+      "loss": 1.1721,
+      "step": 8922
+    },
+    {
+      "epoch": 0.7831679882388759,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005773399409098066,
+      "loss": 1.1295,
+      "step": 8923
+    },
+    {
+      "epoch": 0.7832557578217784,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.000577082478971977,
+      "loss": 1.2333,
+      "step": 8924
+    },
+    {
+      "epoch": 0.7833435274046808,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005768251229268615,
+      "loss": 1.161,
+      "step": 8925
+    },
+    {
+      "epoch": 0.7834312969875833,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005765678727998602,
+      "loss": 1.1363,
+      "step": 8926
+    },
+    {
+      "epoch": 0.7835190665704859,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005763107286163622,
+      "loss": 1.1456,
+      "step": 8927
+    },
+    {
+      "epoch": 0.7836068361533883,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005760536904017469,
+      "loss": 1.0977,
+      "step": 8928
+    },
+    {
+      "epoch": 0.7836946057362908,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.000575796758181383,
+      "loss": 1.1844,
+      "step": 8929
+    },
+    {
+      "epoch": 0.7837823753191933,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0005755399319806288,
+      "loss": 1.1188,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7838701449020957,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005752832118248317,
+      "loss": 1.1393,
+      "step": 8931
+    },
+    {
+      "epoch": 0.7839579144849983,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.000575026597739329,
+      "loss": 1.1289,
+      "step": 8932
+    },
+    {
+      "epoch": 0.7840456840679008,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005747700897494484,
+      "loss": 1.0853,
+      "step": 8933
+    },
+    {
+      "epoch": 0.7841334536508032,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005745136878805047,
+      "loss": 1.126,
+      "step": 8934
+    },
+    {
+      "epoch": 0.7842212232337057,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005742573921578047,
+      "loss": 1.1316,
+      "step": 8935
+    },
+    {
+      "epoch": 0.7843089928166082,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0005740012026066437,
+      "loss": 1.0678,
+      "step": 8936
+    },
+    {
+      "epoch": 0.7843967623995107,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005737451192523065,
+      "loss": 1.182,
+      "step": 8937
+    },
+    {
+      "epoch": 0.7844845319824132,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005734891421200677,
+      "loss": 1.1269,
+      "step": 8938
+    },
+    {
+      "epoch": 0.7845723015653157,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005732332712351904,
+      "loss": 1.1515,
+      "step": 8939
+    },
+    {
+      "epoch": 0.7846600711482181,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005729775066229291,
+      "loss": 1.1858,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7847478407311206,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005727218483085266,
+      "loss": 1.1307,
+      "step": 8941
+    },
+    {
+      "epoch": 0.7848356103140232,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005724662963172147,
+      "loss": 1.1158,
+      "step": 8942
+    },
+    {
+      "epoch": 0.7849233798969256,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005722108506742158,
+      "loss": 1.1597,
+      "step": 8943
+    },
+    {
+      "epoch": 0.7850111494798281,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005719555114047414,
+      "loss": 1.1196,
+      "step": 8944
+    },
+    {
+      "epoch": 0.7850989190627305,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005717002785339928,
+      "loss": 1.1897,
+      "step": 8945
+    },
+    {
+      "epoch": 0.785186688645633,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005714451520871596,
+      "loss": 1.1579,
+      "step": 8946
+    },
+    {
+      "epoch": 0.7852744582285356,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005711901320894224,
+      "loss": 1.1368,
+      "step": 8947
+    },
+    {
+      "epoch": 0.785362227811438,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005709352185659504,
+      "loss": 1.0869,
+      "step": 8948
+    },
+    {
+      "epoch": 0.7854499973943405,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005706804115419027,
+      "loss": 1.1871,
+      "step": 8949
+    },
+    {
+      "epoch": 0.785537766977243,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005704257110424278,
+      "loss": 1.1747,
+      "step": 8950
+    },
+    {
+      "epoch": 0.7856255365601454,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005701711170926635,
+      "loss": 1.1055,
+      "step": 8951
+    },
+    {
+      "epoch": 0.785713306143048,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005699166297177374,
+      "loss": 1.1589,
+      "step": 8952
+    },
+    {
+      "epoch": 0.7858010757259505,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005696622489427665,
+      "loss": 1.1669,
+      "step": 8953
+    },
+    {
+      "epoch": 0.785888845308853,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005694079747928565,
+      "loss": 1.1098,
+      "step": 8954
+    },
+    {
+      "epoch": 0.7859766148917554,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005691538072931039,
+      "loss": 1.2223,
+      "step": 8955
+    },
+    {
+      "epoch": 0.7860643844746579,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005688997464685937,
+      "loss": 1.151,
+      "step": 8956
+    },
+    {
+      "epoch": 0.7861521540575603,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005686457923444014,
+      "loss": 1.086,
+      "step": 8957
+    },
+    {
+      "epoch": 0.7862399236404629,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00056839194494559,
+      "loss": 1.118,
+      "step": 8958
+    },
+    {
+      "epoch": 0.7863276932233654,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005681382042972138,
+      "loss": 1.1427,
+      "step": 8959
+    },
+    {
+      "epoch": 0.7864154628062678,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005678845704243165,
+      "loss": 1.1143,
+      "step": 8960
+    },
+    {
+      "epoch": 0.7865032323891703,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005676310433519306,
+      "loss": 1.179,
+      "step": 8961
+    },
+    {
+      "epoch": 0.7865910019720728,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005673776231050778,
+      "loss": 1.2031,
+      "step": 8962
+    },
+    {
+      "epoch": 0.7866787715549753,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005671243097087698,
+      "loss": 1.1817,
+      "step": 8963
+    },
+    {
+      "epoch": 0.7867665411378778,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005668711031880077,
+      "loss": 1.162,
+      "step": 8964
+    },
+    {
+      "epoch": 0.7868543107207803,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005666180035677822,
+      "loss": 1.1508,
+      "step": 8965
+    },
+    {
+      "epoch": 0.7869420803036827,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005663650108730728,
+      "loss": 1.1568,
+      "step": 8966
+    },
+    {
+      "epoch": 0.7870298498865852,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005661121251288492,
+      "loss": 1.1954,
+      "step": 8967
+    },
+    {
+      "epoch": 0.7871176194694878,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00056585934636007,
+      "loss": 1.1566,
+      "step": 8968
+    },
+    {
+      "epoch": 0.7872053890523902,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005656066745916837,
+      "loss": 1.1391,
+      "step": 8969
+    },
+    {
+      "epoch": 0.7872931586352927,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005653541098486278,
+      "loss": 1.1338,
+      "step": 8970
+    },
+    {
+      "epoch": 0.7873809282181952,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005651016521558294,
+      "loss": 1.1456,
+      "step": 8971
+    },
+    {
+      "epoch": 0.7874686978010976,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005648493015382054,
+      "loss": 1.093,
+      "step": 8972
+    },
+    {
+      "epoch": 0.7875564673840002,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.000564597058020662,
+      "loss": 1.0946,
+      "step": 8973
+    },
+    {
+      "epoch": 0.7876442369669027,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005643449216280938,
+      "loss": 1.114,
+      "step": 8974
+    },
+    {
+      "epoch": 0.7877320065498051,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005640928923853859,
+      "loss": 1.1166,
+      "step": 8975
+    },
+    {
+      "epoch": 0.7878197761327076,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.000563840970317413,
+      "loss": 1.1709,
+      "step": 8976
+    },
+    {
+      "epoch": 0.7879075457156101,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005635891554490391,
+      "loss": 1.1168,
+      "step": 8977
+    },
+    {
+      "epoch": 0.7879953152985126,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005633374478051165,
+      "loss": 1.1582,
+      "step": 8978
+    },
+    {
+      "epoch": 0.7880830848814151,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005630858474104874,
+      "loss": 1.1707,
+      "step": 8979
+    },
+    {
+      "epoch": 0.7881708544643176,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005628343542899855,
+      "loss": 1.1654,
+      "step": 8980
+    },
+    {
+      "epoch": 0.78825862404722,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005625829684684306,
+      "loss": 1.1088,
+      "step": 8981
+    },
+    {
+      "epoch": 0.7883463936301225,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005623316899706341,
+      "loss": 1.1514,
+      "step": 8982
+    },
+    {
+      "epoch": 0.788434163213025,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005620805188213961,
+      "loss": 1.1086,
+      "step": 8983
+    },
+    {
+      "epoch": 0.7885219327959275,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005618294550455066,
+      "loss": 1.1266,
+      "step": 8984
+    },
+    {
+      "epoch": 0.78860970237883,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005615784986677438,
+      "loss": 1.1967,
+      "step": 8985
+    },
+    {
+      "epoch": 0.7886974719617325,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005613276497128766,
+      "loss": 1.1246,
+      "step": 8986
+    },
+    {
+      "epoch": 0.7887852415446349,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005610769082056627,
+      "loss": 1.1034,
+      "step": 8987
+    },
+    {
+      "epoch": 0.7888730111275374,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005608262741708492,
+      "loss": 1.1201,
+      "step": 8988
+    },
+    {
+      "epoch": 0.78896078071044,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005605757476331729,
+      "loss": 1.143,
+      "step": 8989
+    },
+    {
+      "epoch": 0.7890485502933424,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005603253286173595,
+      "loss": 1.1463,
+      "step": 8990
+    },
+    {
+      "epoch": 0.7891363198762449,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005600750171481248,
+      "loss": 1.0934,
+      "step": 8991
+    },
+    {
+      "epoch": 0.7892240894591473,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005598248132501734,
+      "loss": 1.1333,
+      "step": 8992
+    },
+    {
+      "epoch": 0.7893118590420498,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005595747169481991,
+      "loss": 1.1286,
+      "step": 8993
+    },
+    {
+      "epoch": 0.7893996286249524,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005593247282668856,
+      "loss": 1.2306,
+      "step": 8994
+    },
+    {
+      "epoch": 0.7894873982078549,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005590748472309059,
+      "loss": 1.1688,
+      "step": 8995
+    },
+    {
+      "epoch": 0.7895751677907573,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005588250738649226,
+      "loss": 1.1168,
+      "step": 8996
+    },
+    {
+      "epoch": 0.7896629373736598,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005585754081935864,
+      "loss": 1.1813,
+      "step": 8997
+    },
+    {
+      "epoch": 0.7897507069565622,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005583258502415391,
+      "loss": 1.1295,
+      "step": 8998
+    },
+    {
+      "epoch": 0.7898384765394648,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005580764000334103,
+      "loss": 1.1088,
+      "step": 8999
+    },
+    {
+      "epoch": 0.7899262461223673,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005578270575938212,
+      "loss": 1.1384,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7899262461223673,
+      "eval_loss": 1.137856125831604,
+      "eval_runtime": 437.8028,
+      "eval_samples_per_second": 33.668,
+      "eval_steps_per_second": 8.417,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7900140157052697,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005575778229473794,
+      "loss": 1.227,
+      "step": 9001
+    },
+    {
+      "epoch": 0.7901017852881722,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.000557328696118684,
+      "loss": 1.1171,
+      "step": 9002
+    },
+    {
+      "epoch": 0.7901895548710747,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.000557079677132323,
+      "loss": 1.148,
+      "step": 9003
+    },
+    {
+      "epoch": 0.7902773244539772,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005568307660128736,
+      "loss": 1.1277,
+      "step": 9004
+    },
+    {
+      "epoch": 0.7903650940368797,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005565819627849019,
+      "loss": 1.1717,
+      "step": 9005
+    },
+    {
+      "epoch": 0.7904528636197822,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005563332674729641,
+      "loss": 1.1228,
+      "step": 9006
+    },
+    {
+      "epoch": 0.7905406332026846,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005560846801016052,
+      "loss": 1.2144,
+      "step": 9007
+    },
+    {
+      "epoch": 0.7906284027855871,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005558362006953605,
+      "loss": 1.1107,
+      "step": 9008
+    },
+    {
+      "epoch": 0.7907161723684896,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005555878292787528,
+      "loss": 1.1852,
+      "step": 9009
+    },
+    {
+      "epoch": 0.7908039419513921,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005553395658762958,
+      "loss": 1.1522,
+      "step": 9010
+    },
+    {
+      "epoch": 0.7908917115342946,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005550914105124924,
+      "loss": 1.1175,
+      "step": 9011
+    },
+    {
+      "epoch": 0.7909794811171971,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005548433632118351,
+      "loss": 1.1853,
+      "step": 9012
+    },
+    {
+      "epoch": 0.7910672507000995,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005545954239988042,
+      "loss": 1.1439,
+      "step": 9013
+    },
+    {
+      "epoch": 0.791155020283002,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005543475928978704,
+      "loss": 1.1348,
+      "step": 9014
+    },
+    {
+      "epoch": 0.7912427898659046,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005540998699334942,
+      "loss": 1.2028,
+      "step": 9015
+    },
+    {
+      "epoch": 0.791330559448807,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005538522551301247,
+      "loss": 1.1125,
+      "step": 9016
+    },
+    {
+      "epoch": 0.7914183290317095,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005536047485122001,
+      "loss": 1.0944,
+      "step": 9017
+    },
+    {
+      "epoch": 0.791506098614612,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005533573501041487,
+      "loss": 1.1213,
+      "step": 9018
+    },
+    {
+      "epoch": 0.7915938681975144,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005531100599303876,
+      "loss": 1.1804,
+      "step": 9019
+    },
+    {
+      "epoch": 0.791681637780417,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005528628780153232,
+      "loss": 1.1444,
+      "step": 9020
+    },
+    {
+      "epoch": 0.7917694073633195,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005526158043833518,
+      "loss": 1.1417,
+      "step": 9021
+    },
+    {
+      "epoch": 0.7918571769462219,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005523688390588582,
+      "loss": 1.1723,
+      "step": 9022
+    },
+    {
+      "epoch": 0.7919449465291244,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005521219820662174,
+      "loss": 1.1644,
+      "step": 9023
+    },
+    {
+      "epoch": 0.7920327161120269,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005518752334297929,
+      "loss": 1.1404,
+      "step": 9024
+    },
+    {
+      "epoch": 0.7921204856949294,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005516285931739377,
+      "loss": 1.149,
+      "step": 9025
+    },
+    {
+      "epoch": 0.7922082552778319,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005513820613229941,
+      "loss": 1.1354,
+      "step": 9026
+    },
+    {
+      "epoch": 0.7922960248607344,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005511356379012943,
+      "loss": 1.1133,
+      "step": 9027
+    },
+    {
+      "epoch": 0.7923837944436368,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005508893229331591,
+      "loss": 1.195,
+      "step": 9028
+    },
+    {
+      "epoch": 0.7924715640265393,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005506431164428987,
+      "loss": 1.1454,
+      "step": 9029
+    },
+    {
+      "epoch": 0.7925593336094419,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005503970184548122,
+      "loss": 1.0942,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7926471031923443,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005501510289931898,
+      "loss": 1.1718,
+      "step": 9031
+    },
+    {
+      "epoch": 0.7927348727752468,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005499051480823087,
+      "loss": 1.1282,
+      "step": 9032
+    },
+    {
+      "epoch": 0.7928226423581493,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005496593757464367,
+      "loss": 1.0991,
+      "step": 9033
+    },
+    {
+      "epoch": 0.7929104119410517,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005494137120098302,
+      "loss": 1.1465,
+      "step": 9034
+    },
+    {
+      "epoch": 0.7929981815239543,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005491681568967361,
+      "loss": 1.1394,
+      "step": 9035
+    },
+    {
+      "epoch": 0.7930859511068568,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005489227104313888,
+      "loss": 1.1009,
+      "step": 9036
+    },
+    {
+      "epoch": 0.7931737206897592,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005486773726380131,
+      "loss": 1.1412,
+      "step": 9037
+    },
+    {
+      "epoch": 0.7932614902726617,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005484321435408231,
+      "loss": 1.1032,
+      "step": 9038
+    },
+    {
+      "epoch": 0.7933492598555641,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005481870231640218,
+      "loss": 1.1568,
+      "step": 9039
+    },
+    {
+      "epoch": 0.7934370294384666,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005479420115318017,
+      "loss": 1.1301,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7935247990213692,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005476971086683444,
+      "loss": 1.1536,
+      "step": 9041
+    },
+    {
+      "epoch": 0.7936125686042717,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005474523145978209,
+      "loss": 1.112,
+      "step": 9042
+    },
+    {
+      "epoch": 0.7937003381871741,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005472076293443918,
+      "loss": 1.1446,
+      "step": 9043
+    },
+    {
+      "epoch": 0.7937881077700766,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005469630529322056,
+      "loss": 1.2182,
+      "step": 9044
+    },
+    {
+      "epoch": 0.793875877352979,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005467185853854019,
+      "loss": 1.2243,
+      "step": 9045
+    },
+    {
+      "epoch": 0.7939636469358816,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005464742267281082,
+      "loss": 1.1167,
+      "step": 9046
+    },
+    {
+      "epoch": 0.7940514165187841,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005462299769844422,
+      "loss": 1.1355,
+      "step": 9047
+    },
+    {
+      "epoch": 0.7941391861016865,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005459858361785097,
+      "loss": 1.2568,
+      "step": 9048
+    },
+    {
+      "epoch": 0.794226955684589,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005457418043344069,
+      "loss": 1.2059,
+      "step": 9049
+    },
+    {
+      "epoch": 0.7943147252674915,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005454978814762181,
+      "loss": 1.187,
+      "step": 9050
+    },
+    {
+      "epoch": 0.794402494850394,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005452540676280192,
+      "loss": 1.1619,
+      "step": 9051
+    },
+    {
+      "epoch": 0.7944902644332965,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005450103628138718,
+      "loss": 1.1557,
+      "step": 9052
+    },
+    {
+      "epoch": 0.794578034016199,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005447667670578296,
+      "loss": 1.1786,
+      "step": 9053
+    },
+    {
+      "epoch": 0.7946658035991014,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005445232803839344,
+      "loss": 1.1862,
+      "step": 9054
+    },
+    {
+      "epoch": 0.7947535731820039,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005442799028162176,
+      "loss": 1.1316,
+      "step": 9055
+    },
+    {
+      "epoch": 0.7948413427649065,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005440366343786987,
+      "loss": 1.1795,
+      "step": 9056
+    },
+    {
+      "epoch": 0.7949291123478089,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005437934750953879,
+      "loss": 1.1893,
+      "step": 9057
+    },
+    {
+      "epoch": 0.7950168819307114,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005435504249902842,
+      "loss": 1.1657,
+      "step": 9058
+    },
+    {
+      "epoch": 0.7951046515136139,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005433074840873759,
+      "loss": 1.1465,
+      "step": 9059
+    },
+    {
+      "epoch": 0.7951924210965163,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005430646524106393,
+      "loss": 1.1439,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7952801906794189,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005428219299840421,
+      "loss": 1.128,
+      "step": 9061
+    },
+    {
+      "epoch": 0.7953679602623214,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005425793168315392,
+      "loss": 1.1418,
+      "step": 9062
+    },
+    {
+      "epoch": 0.7954557298452238,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005423368129770763,
+      "loss": 1.2153,
+      "step": 9063
+    },
+    {
+      "epoch": 0.7955434994281263,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005420944184445869,
+      "loss": 1.1169,
+      "step": 9064
+    },
+    {
+      "epoch": 0.7956312690110288,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005418521332579946,
+      "loss": 1.205,
+      "step": 9065
+    },
+    {
+      "epoch": 0.7957190385939312,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.000541609957441212,
+      "loss": 1.145,
+      "step": 9066
+    },
+    {
+      "epoch": 0.7958068081768338,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005413678910181413,
+      "loss": 1.1264,
+      "step": 9067
+    },
+    {
+      "epoch": 0.7958945777597363,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005411259340126729,
+      "loss": 1.1575,
+      "step": 9068
+    },
+    {
+      "epoch": 0.7959823473426387,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.000540884086448687,
+      "loss": 1.2104,
+      "step": 9069
+    },
+    {
+      "epoch": 0.7960701169255412,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005406423483500534,
+      "loss": 1.1376,
+      "step": 9070
+    },
+    {
+      "epoch": 0.7961578865084437,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005404007197406303,
+      "loss": 1.1375,
+      "step": 9071
+    },
+    {
+      "epoch": 0.7962456560913462,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000540159200644266,
+      "loss": 1.1307,
+      "step": 9072
+    },
+    {
+      "epoch": 0.7963334256742487,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005399177910847968,
+      "loss": 1.142,
+      "step": 9073
+    },
+    {
+      "epoch": 0.7964211952571512,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005396764910860495,
+      "loss": 1.155,
+      "step": 9074
+    },
+    {
+      "epoch": 0.7965089648400536,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005394353006718394,
+      "loss": 1.1354,
+      "step": 9075
+    },
+    {
+      "epoch": 0.7965967344229561,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005391942198659707,
+      "loss": 1.1179,
+      "step": 9076
+    },
+    {
+      "epoch": 0.7966845040058587,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005389532486922372,
+      "loss": 1.1472,
+      "step": 9077
+    },
+    {
+      "epoch": 0.7967722735887611,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005387123871744218,
+      "loss": 1.1302,
+      "step": 9078
+    },
+    {
+      "epoch": 0.7968600431716636,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.000538471635336297,
+      "loss": 1.2193,
+      "step": 9079
+    },
+    {
+      "epoch": 0.796947812754566,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.000538230993201623,
+      "loss": 1.1597,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7970355823374685,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005379904607941512,
+      "loss": 1.1847,
+      "step": 9081
+    },
+    {
+      "epoch": 0.7971233519203711,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005377500381376213,
+      "loss": 1.1711,
+      "step": 9082
+    },
+    {
+      "epoch": 0.7972111215032736,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005375097252557614,
+      "loss": 1.1891,
+      "step": 9083
+    },
+    {
+      "epoch": 0.797298891086176,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005372695221722897,
+      "loss": 1.1764,
+      "step": 9084
+    },
+    {
+      "epoch": 0.7973866606690785,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005370294289109134,
+      "loss": 1.1586,
+      "step": 9085
+    },
+    {
+      "epoch": 0.797474430251981,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005367894454953289,
+      "loss": 1.1514,
+      "step": 9086
+    },
+    {
+      "epoch": 0.7975621998348835,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.000536549571949221,
+      "loss": 1.1282,
+      "step": 9087
+    },
+    {
+      "epoch": 0.797649969417786,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005363098082962646,
+      "loss": 1.1461,
+      "step": 9088
+    },
+    {
+      "epoch": 0.7977377390006885,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005360701545601235,
+      "loss": 1.191,
+      "step": 9089
+    },
+    {
+      "epoch": 0.7978255085835909,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005358306107644507,
+      "loss": 1.077,
+      "step": 9090
+    },
+    {
+      "epoch": 0.7979132781664934,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.000535591176932888,
+      "loss": 1.1277,
+      "step": 9091
+    },
+    {
+      "epoch": 0.798001047749396,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005353518530890665,
+      "loss": 1.0981,
+      "step": 9092
+    },
+    {
+      "epoch": 0.7980888173322984,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005351126392566069,
+      "loss": 1.0893,
+      "step": 9093
+    },
+    {
+      "epoch": 0.7981765869152009,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005348735354591186,
+      "loss": 1.1438,
+      "step": 9094
+    },
+    {
+      "epoch": 0.7982643564981033,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005346345417201997,
+      "loss": 1.1964,
+      "step": 9095
+    },
+    {
+      "epoch": 0.7983521260810058,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005343956580634383,
+      "loss": 1.1232,
+      "step": 9096
+    },
+    {
+      "epoch": 0.7984398956639083,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005341568845124112,
+      "loss": 1.1293,
+      "step": 9097
+    },
+    {
+      "epoch": 0.7985276652468108,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005339182210906846,
+      "loss": 1.1216,
+      "step": 9098
+    },
+    {
+      "epoch": 0.7986154348297133,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005336796678218134,
+      "loss": 1.1566,
+      "step": 9099
+    },
+    {
+      "epoch": 0.7987032044126158,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005334412247293414,
+      "loss": 1.18,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7987909739955182,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005332028918368031,
+      "loss": 1.1337,
+      "step": 9101
+    },
+    {
+      "epoch": 0.7988787435784207,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005329646691677205,
+      "loss": 1.1198,
+      "step": 9102
+    },
+    {
+      "epoch": 0.7989665131613233,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005327265567456049,
+      "loss": 1.1243,
+      "step": 9103
+    },
+    {
+      "epoch": 0.7990542827442257,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005324885545939574,
+      "loss": 1.1611,
+      "step": 9104
+    },
+    {
+      "epoch": 0.7991420523271282,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005322506627362676,
+      "loss": 1.1392,
+      "step": 9105
+    },
+    {
+      "epoch": 0.7992298219100307,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005320128811960152,
+      "loss": 1.1007,
+      "step": 9106
+    },
+    {
+      "epoch": 0.7993175914929331,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005317752099966675,
+      "loss": 1.1194,
+      "step": 9107
+    },
+    {
+      "epoch": 0.7994053610758357,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005315376491616819,
+      "loss": 1.1035,
+      "step": 9108
+    },
+    {
+      "epoch": 0.7994931306587382,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005313001987145047,
+      "loss": 1.158,
+      "step": 9109
+    },
+    {
+      "epoch": 0.7995809002416406,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005310628586785715,
+      "loss": 1.165,
+      "step": 9110
+    },
+    {
+      "epoch": 0.7996686698245431,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005308256290773066,
+      "loss": 1.2244,
+      "step": 9111
+    },
+    {
+      "epoch": 0.7997564394074456,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005305885099341239,
+      "loss": 1.1386,
+      "step": 9112
+    },
+    {
+      "epoch": 0.7998442089903481,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005303515012724261,
+      "loss": 1.1317,
+      "step": 9113
+    },
+    {
+      "epoch": 0.7999319785732506,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.000530114603115605,
+      "loss": 1.1596,
+      "step": 9114
+    },
+    {
+      "epoch": 0.8000197481561531,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005298778154870413,
+      "loss": 1.1014,
+      "step": 9115
+    },
+    {
+      "epoch": 0.8001075177390555,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.000529641138410105,
+      "loss": 1.112,
+      "step": 9116
+    },
+    {
+      "epoch": 0.800195287321958,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005294045719081551,
+      "loss": 1.1368,
+      "step": 9117
+    },
+    {
+      "epoch": 0.8002830569048606,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005291681160045405,
+      "loss": 1.169,
+      "step": 9118
+    },
+    {
+      "epoch": 0.800370826487763,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005289317707225977,
+      "loss": 1.1842,
+      "step": 9119
+    },
+    {
+      "epoch": 0.8004585960706655,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005286955360856528,
+      "loss": 1.1687,
+      "step": 9120
+    },
+    {
+      "epoch": 0.800546365653568,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005284594121170224,
+      "loss": 1.0765,
+      "step": 9121
+    },
+    {
+      "epoch": 0.8006341352364704,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00052822339884001,
+      "loss": 1.1148,
+      "step": 9122
+    },
+    {
+      "epoch": 0.8007219048193729,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005279874962779096,
+      "loss": 1.1028,
+      "step": 9123
+    },
+    {
+      "epoch": 0.8008096744022755,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005277517044540036,
+      "loss": 1.1567,
+      "step": 9124
+    },
+    {
+      "epoch": 0.8008974439851779,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005275160233915638,
+      "loss": 1.1724,
+      "step": 9125
+    },
+    {
+      "epoch": 0.8009852135680804,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005272804531138515,
+      "loss": 1.1334,
+      "step": 9126
+    },
+    {
+      "epoch": 0.8010729831509829,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005270449936441158,
+      "loss": 1.1379,
+      "step": 9127
+    },
+    {
+      "epoch": 0.8011607527338853,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005268096450055959,
+      "loss": 1.1433,
+      "step": 9128
+    },
+    {
+      "epoch": 0.8012485223167879,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005265744072215197,
+      "loss": 1.1502,
+      "step": 9129
+    },
+    {
+      "epoch": 0.8013362918996904,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005263392803151046,
+      "loss": 1.1294,
+      "step": 9130
+    },
+    {
+      "epoch": 0.8014240614825928,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005261042643095561,
+      "loss": 1.1032,
+      "step": 9131
+    },
+    {
+      "epoch": 0.8015118310654953,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005258693592280697,
+      "loss": 1.1825,
+      "step": 9132
+    },
+    {
+      "epoch": 0.8015996006483977,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00052563456509383,
+      "loss": 1.1561,
+      "step": 9133
+    },
+    {
+      "epoch": 0.8016873702313003,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005253998819300095,
+      "loss": 1.1731,
+      "step": 9134
+    },
+    {
+      "epoch": 0.8017751398142028,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005251653097597709,
+      "loss": 1.1745,
+      "step": 9135
+    },
+    {
+      "epoch": 0.8018629093971053,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005249308486062654,
+      "loss": 1.0354,
+      "step": 9136
+    },
+    {
+      "epoch": 0.8019506789800077,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005246964984926338,
+      "loss": 1.1606,
+      "step": 9137
+    },
+    {
+      "epoch": 0.8020384485629102,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.000524462259442005,
+      "loss": 1.1721,
+      "step": 9138
+    },
+    {
+      "epoch": 0.8021262181458128,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005242281314774976,
+      "loss": 1.225,
+      "step": 9139
+    },
+    {
+      "epoch": 0.8022139877287152,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.000523994114622219,
+      "loss": 1.1347,
+      "step": 9140
+    },
+    {
+      "epoch": 0.8023017573116177,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005237602088992661,
+      "loss": 1.1451,
+      "step": 9141
+    },
+    {
+      "epoch": 0.8023895268945201,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005235264143317242,
+      "loss": 1.1548,
+      "step": 9142
+    },
+    {
+      "epoch": 0.8024772964774226,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005232927309426679,
+      "loss": 1.1654,
+      "step": 9143
+    },
+    {
+      "epoch": 0.8025650660603252,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005230591587551611,
+      "loss": 1.1732,
+      "step": 9144
+    },
+    {
+      "epoch": 0.8026528356432276,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005228256977922565,
+      "loss": 1.1644,
+      "step": 9145
+    },
+    {
+      "epoch": 0.8027406052261301,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005225923480769952,
+      "loss": 1.1104,
+      "step": 9146
+    },
+    {
+      "epoch": 0.8028283748090326,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005223591096324082,
+      "loss": 1.1375,
+      "step": 9147
+    },
+    {
+      "epoch": 0.802916144391935,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005221259824815151,
+      "loss": 1.1243,
+      "step": 9148
+    },
+    {
+      "epoch": 0.8030039139748376,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005218929666473253,
+      "loss": 1.1871,
+      "step": 9149
+    },
+    {
+      "epoch": 0.8030916835577401,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005216600621528355,
+      "loss": 1.1607,
+      "step": 9150
+    },
+    {
+      "epoch": 0.8031794531406425,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005214272690210328,
+      "loss": 1.1671,
+      "step": 9151
+    },
+    {
+      "epoch": 0.803267222723545,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005211945872748931,
+      "loss": 1.0958,
+      "step": 9152
+    },
+    {
+      "epoch": 0.8033549923064475,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005209620169373819,
+      "loss": 1.055,
+      "step": 9153
+    },
+    {
+      "epoch": 0.8034427618893499,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005207295580314516,
+      "loss": 1.1125,
+      "step": 9154
+    },
+    {
+      "epoch": 0.8035305314722525,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005204972105800458,
+      "loss": 1.1623,
+      "step": 9155
+    },
+    {
+      "epoch": 0.803618301055155,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005202649746060961,
+      "loss": 1.1364,
+      "step": 9156
+    },
+    {
+      "epoch": 0.8037060706380574,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005200328501325235,
+      "loss": 1.0546,
+      "step": 9157
+    },
+    {
+      "epoch": 0.8037938402209599,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005198008371822374,
+      "loss": 1.1576,
+      "step": 9158
+    },
+    {
+      "epoch": 0.8038816098038624,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005195689357781366,
+      "loss": 1.1762,
+      "step": 9159
+    },
+    {
+      "epoch": 0.8039693793867649,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005193371459431088,
+      "loss": 1.1866,
+      "step": 9160
+    },
+    {
+      "epoch": 0.8040571489696674,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.000519105467700031,
+      "loss": 1.1453,
+      "step": 9161
+    },
+    {
+      "epoch": 0.8041449185525699,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005188739010717688,
+      "loss": 1.1806,
+      "step": 9162
+    },
+    {
+      "epoch": 0.8042326881354723,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.000518642446081177,
+      "loss": 1.1276,
+      "step": 9163
+    },
+    {
+      "epoch": 0.8043204577183748,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005184111027510992,
+      "loss": 1.167,
+      "step": 9164
+    },
+    {
+      "epoch": 0.8044082273012774,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005181798711043683,
+      "loss": 1.1271,
+      "step": 9165
+    },
+    {
+      "epoch": 0.8044959968841798,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005179487511638055,
+      "loss": 1.1251,
+      "step": 9166
+    },
+    {
+      "epoch": 0.8045837664670823,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005177177429522216,
+      "loss": 1.126,
+      "step": 9167
+    },
+    {
+      "epoch": 0.8046715360499848,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005174868464924164,
+      "loss": 1.0961,
+      "step": 9168
+    },
+    {
+      "epoch": 0.8047593056328872,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005172560618071784,
+      "loss": 1.13,
+      "step": 9169
+    },
+    {
+      "epoch": 0.8048470752157898,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000517025388919285,
+      "loss": 1.0778,
+      "step": 9170
+    },
+    {
+      "epoch": 0.8049348447986923,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005167948278515022,
+      "loss": 1.1504,
+      "step": 9171
+    },
+    {
+      "epoch": 0.8050226143815947,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005165643786265867,
+      "loss": 1.1051,
+      "step": 9172
+    },
+    {
+      "epoch": 0.8051103839644972,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005163340412672821,
+      "loss": 1.1957,
+      "step": 9173
+    },
+    {
+      "epoch": 0.8051981535473997,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005161038157963221,
+      "loss": 1.1336,
+      "step": 9174
+    },
+    {
+      "epoch": 0.8052859231303022,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005158737022364284,
+      "loss": 1.1472,
+      "step": 9175
+    },
+    {
+      "epoch": 0.8053736927132047,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005156437006103133,
+      "loss": 1.1258,
+      "step": 9176
+    },
+    {
+      "epoch": 0.8054614622961072,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005154138109406763,
+      "loss": 1.1292,
+      "step": 9177
+    },
+    {
+      "epoch": 0.8055492318790096,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005151840332502068,
+      "loss": 1.1918,
+      "step": 9178
+    },
+    {
+      "epoch": 0.8056370014619121,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.000514954367561583,
+      "loss": 1.1488,
+      "step": 9179
+    },
+    {
+      "epoch": 0.8057247710448145,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005147248138974719,
+      "loss": 1.1525,
+      "step": 9180
+    },
+    {
+      "epoch": 0.8058125406277171,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005144953722805296,
+      "loss": 1.1173,
+      "step": 9181
+    },
+    {
+      "epoch": 0.8059003102106196,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005142660427334011,
+      "loss": 1.1576,
+      "step": 9182
+    },
+    {
+      "epoch": 0.805988079793522,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005140368252787203,
+      "loss": 1.1719,
+      "step": 9183
+    },
+    {
+      "epoch": 0.8060758493764245,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005138077199391103,
+      "loss": 1.0829,
+      "step": 9184
+    },
+    {
+      "epoch": 0.806163618959327,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005135787267371826,
+      "loss": 1.2137,
+      "step": 9185
+    },
+    {
+      "epoch": 0.8062513885422296,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005133498456955378,
+      "loss": 1.1431,
+      "step": 9186
+    },
+    {
+      "epoch": 0.806339158125132,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005131210768367657,
+      "loss": 1.1098,
+      "step": 9187
+    },
+    {
+      "epoch": 0.8064269277080345,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005128924201834453,
+      "loss": 1.1274,
+      "step": 9188
+    },
+    {
+      "epoch": 0.8065146972909369,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005126638757581433,
+      "loss": 1.0879,
+      "step": 9189
+    },
+    {
+      "epoch": 0.8066024668738394,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005124354435834167,
+      "loss": 1.1796,
+      "step": 9190
+    },
+    {
+      "epoch": 0.806690236456742,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005122071236818102,
+      "loss": 1.1503,
+      "step": 9191
+    },
+    {
+      "epoch": 0.8067780060396444,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005119789160758595,
+      "loss": 1.1695,
+      "step": 9192
+    },
+    {
+      "epoch": 0.8068657756225469,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005117508207880863,
+      "loss": 1.1298,
+      "step": 9193
+    },
+    {
+      "epoch": 0.8069535452054494,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005115228378410033,
+      "loss": 1.1224,
+      "step": 9194
+    },
+    {
+      "epoch": 0.8070413147883518,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005112949672571115,
+      "loss": 1.1256,
+      "step": 9195
+    },
+    {
+      "epoch": 0.8071290843712544,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005110672090589011,
+      "loss": 1.1734,
+      "step": 9196
+    },
+    {
+      "epoch": 0.8072168539541569,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005108395632688504,
+      "loss": 1.1133,
+      "step": 9197
+    },
+    {
+      "epoch": 0.8073046235370593,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005106120299094274,
+      "loss": 1.1568,
+      "step": 9198
+    },
+    {
+      "epoch": 0.8073923931199618,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005103846090030886,
+      "loss": 1.1556,
+      "step": 9199
+    },
+    {
+      "epoch": 0.8074801627028643,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.00051015730057228,
+      "loss": 1.1108,
+      "step": 9200
+    },
+    {
+      "epoch": 0.8075679322857668,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005099301046394352,
+      "loss": 1.1654,
+      "step": 9201
+    },
+    {
+      "epoch": 0.8076557018686693,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005097030212269781,
+      "loss": 1.1495,
+      "step": 9202
+    },
+    {
+      "epoch": 0.8077434714515718,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005094760503573211,
+      "loss": 1.1387,
+      "step": 9203
+    },
+    {
+      "epoch": 0.8078312410344742,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005092491920528654,
+      "loss": 1.1227,
+      "step": 9204
+    },
+    {
+      "epoch": 0.8079190106173767,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005090224463360003,
+      "loss": 1.16,
+      "step": 9205
+    },
+    {
+      "epoch": 0.8080067802002792,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005087958132291051,
+      "loss": 1.1642,
+      "step": 9206
+    },
+    {
+      "epoch": 0.8080945497831817,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005085692927545478,
+      "loss": 1.136,
+      "step": 9207
+    },
+    {
+      "epoch": 0.8081823193660842,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.000508342884934685,
+      "loss": 1.115,
+      "step": 9208
+    },
+    {
+      "epoch": 0.8082700889489867,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005081165897918619,
+      "loss": 1.185,
+      "step": 9209
+    },
+    {
+      "epoch": 0.8083578585318891,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005078904073484132,
+      "loss": 1.1221,
+      "step": 9210
+    },
+    {
+      "epoch": 0.8084456281147916,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005076643376266623,
+      "loss": 1.1809,
+      "step": 9211
+    },
+    {
+      "epoch": 0.8085333976976942,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005074383806489212,
+      "loss": 1.1315,
+      "step": 9212
+    },
+    {
+      "epoch": 0.8086211672805966,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005072125364374908,
+      "loss": 1.1556,
+      "step": 9213
+    },
+    {
+      "epoch": 0.8087089368634991,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005069868050146618,
+      "loss": 1.1457,
+      "step": 9214
+    },
+    {
+      "epoch": 0.8087967064464016,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.000506761186402712,
+      "loss": 1.0863,
+      "step": 9215
+    },
+    {
+      "epoch": 0.808884476029304,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00050653568062391,
+      "loss": 1.1771,
+      "step": 9216
+    },
+    {
+      "epoch": 0.8089722456122066,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005063102877005117,
+      "loss": 1.1611,
+      "step": 9217
+    },
+    {
+      "epoch": 0.8090600151951091,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005060850076547623,
+      "loss": 1.1667,
+      "step": 9218
+    },
+    {
+      "epoch": 0.8091477847780115,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005058598405088969,
+      "loss": 1.2018,
+      "step": 9219
+    },
+    {
+      "epoch": 0.809235554360914,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005056347862851379,
+      "loss": 1.1314,
+      "step": 9220
+    },
+    {
+      "epoch": 0.8093233239438165,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005054098450056971,
+      "loss": 1.1171,
+      "step": 9221
+    },
+    {
+      "epoch": 0.809411093526719,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005051850166927759,
+      "loss": 1.1424,
+      "step": 9222
+    },
+    {
+      "epoch": 0.8094988631096215,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005049603013685642,
+      "loss": 1.1792,
+      "step": 9223
+    },
+    {
+      "epoch": 0.809586632692524,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005047356990552397,
+      "loss": 1.1418,
+      "step": 9224
+    },
+    {
+      "epoch": 0.8096744022754264,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005045112097749702,
+      "loss": 1.1035,
+      "step": 9225
+    },
+    {
+      "epoch": 0.8097621718583289,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005042868335499118,
+      "loss": 1.1591,
+      "step": 9226
+    },
+    {
+      "epoch": 0.8098499414412315,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005040625704022096,
+      "loss": 1.1358,
+      "step": 9227
+    },
+    {
+      "epoch": 0.8099377110241339,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005038384203539974,
+      "loss": 1.1534,
+      "step": 9228
+    },
+    {
+      "epoch": 0.8100254806070364,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000503614383427398,
+      "loss": 1.1313,
+      "step": 9229
+    },
+    {
+      "epoch": 0.8101132501899388,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005033904596445227,
+      "loss": 1.1608,
+      "step": 9230
+    },
+    {
+      "epoch": 0.8102010197728413,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005031666490274724,
+      "loss": 1.1194,
+      "step": 9231
+    },
+    {
+      "epoch": 0.8102887893557439,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005029429515983359,
+      "loss": 1.1638,
+      "step": 9232
+    },
+    {
+      "epoch": 0.8103765589386464,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005027193673791915,
+      "loss": 1.144,
+      "step": 9233
+    },
+    {
+      "epoch": 0.8104643285215488,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.000502495896392106,
+      "loss": 1.199,
+      "step": 9234
+    },
+    {
+      "epoch": 0.8105520981044513,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005022725386591352,
+      "loss": 1.1318,
+      "step": 9235
+    },
+    {
+      "epoch": 0.8106398676873537,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005020492942023234,
+      "loss": 1.1008,
+      "step": 9236
+    },
+    {
+      "epoch": 0.8107276372702562,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005018261630437039,
+      "loss": 1.1348,
+      "step": 9237
+    },
+    {
+      "epoch": 0.8108154068531588,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.000501603145205299,
+      "loss": 1.1459,
+      "step": 9238
+    },
+    {
+      "epoch": 0.8109031764360612,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005013802407091201,
+      "loss": 1.1858,
+      "step": 9239
+    },
+    {
+      "epoch": 0.8109909460189637,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005011574495771662,
+      "loss": 1.2639,
+      "step": 9240
+    },
+    {
+      "epoch": 0.8110787156018662,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005009347718314261,
+      "loss": 1.1292,
+      "step": 9241
+    },
+    {
+      "epoch": 0.8111664851847686,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005007122074938777,
+      "loss": 1.1613,
+      "step": 9242
+    },
+    {
+      "epoch": 0.8112542547676712,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000500489756586487,
+      "loss": 1.1701,
+      "step": 9243
+    },
+    {
+      "epoch": 0.8113420243505737,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005002674191312088,
+      "loss": 1.1788,
+      "step": 9244
+    },
+    {
+      "epoch": 0.8114297939334761,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005000451951499871,
+      "loss": 1.1164,
+      "step": 9245
+    },
+    {
+      "epoch": 0.8115175635163786,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0004998230846647545,
+      "loss": 1.1005,
+      "step": 9246
+    },
+    {
+      "epoch": 0.8116053330992811,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004996010876974328,
+      "loss": 1.1737,
+      "step": 9247
+    },
+    {
+      "epoch": 0.8116931026821836,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0004993792042699314,
+      "loss": 1.1325,
+      "step": 9248
+    },
+    {
+      "epoch": 0.8117808722650861,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004991574344041497,
+      "loss": 1.1588,
+      "step": 9249
+    },
+    {
+      "epoch": 0.8118686418479886,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004989357781219756,
+      "loss": 1.1691,
+      "step": 9250
+    },
+    {
+      "epoch": 0.811956411430891,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004987142354452861,
+      "loss": 1.1891,
+      "step": 9251
+    },
+    {
+      "epoch": 0.8120441810137935,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004984928063959453,
+      "loss": 1.0909,
+      "step": 9252
+    },
+    {
+      "epoch": 0.8121319505966961,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004982714909958089,
+      "loss": 1.1129,
+      "step": 9253
+    },
+    {
+      "epoch": 0.8122197201795985,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004980502892667189,
+      "loss": 1.1848,
+      "step": 9254
+    },
+    {
+      "epoch": 0.812307489762501,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004978292012305077,
+      "loss": 1.0914,
+      "step": 9255
+    },
+    {
+      "epoch": 0.8123952593454035,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.000497608226908995,
+      "loss": 1.0911,
+      "step": 9256
+    },
+    {
+      "epoch": 0.8124830289283059,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004973873663239906,
+      "loss": 1.1424,
+      "step": 9257
+    },
+    {
+      "epoch": 0.8125707985112085,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0004971666194972926,
+      "loss": 1.1301,
+      "step": 9258
+    },
+    {
+      "epoch": 0.812658568094111,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.000496945986450688,
+      "loss": 1.1518,
+      "step": 9259
+    },
+    {
+      "epoch": 0.8127463376770134,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0004967254672059517,
+      "loss": 1.1348,
+      "step": 9260
+    },
+    {
+      "epoch": 0.8128341072599159,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004965050617848487,
+      "loss": 1.1305,
+      "step": 9261
+    },
+    {
+      "epoch": 0.8129218768428184,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0004962847702091317,
+      "loss": 1.2002,
+      "step": 9262
+    },
+    {
+      "epoch": 0.8130096464257208,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.000496064592500543,
+      "loss": 1.0933,
+      "step": 9263
+    },
+    {
+      "epoch": 0.8130974160086234,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0004958445286808132,
+      "loss": 1.1787,
+      "step": 9264
+    },
+    {
+      "epoch": 0.8131851855915259,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004956245787716616,
+      "loss": 1.1523,
+      "step": 9265
+    },
+    {
+      "epoch": 0.8132729551744283,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0004954047427947963,
+      "loss": 1.0948,
+      "step": 9266
+    },
+    {
+      "epoch": 0.8133607247573308,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0004951850207719149,
+      "loss": 1.1293,
+      "step": 9267
+    },
+    {
+      "epoch": 0.8134484943402333,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004949654127247022,
+      "loss": 1.1337,
+      "step": 9268
+    },
+    {
+      "epoch": 0.8135362639231358,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.000494745918674833,
+      "loss": 1.19,
+      "step": 9269
+    },
+    {
+      "epoch": 0.8136240335060383,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0004945265386439706,
+      "loss": 1.1657,
+      "step": 9270
+    },
+    {
+      "epoch": 0.8137118030889408,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004943072726537671,
+      "loss": 1.1703,
+      "step": 9271
+    },
+    {
+      "epoch": 0.8137995726718432,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0004940881207258622,
+      "loss": 1.1373,
+      "step": 9272
+    },
+    {
+      "epoch": 0.8138873422547457,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0004938690828818865,
+      "loss": 1.1332,
+      "step": 9273
+    },
+    {
+      "epoch": 0.8139751118376483,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0004936501591434581,
+      "loss": 1.111,
+      "step": 9274
+    },
+    {
+      "epoch": 0.8140628814205507,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004934313495321832,
+      "loss": 1.1197,
+      "step": 9275
+    },
+    {
+      "epoch": 0.8141506510034532,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004932126540696577,
+      "loss": 1.1471,
+      "step": 9276
+    },
+    {
+      "epoch": 0.8142384205863556,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0004929940727774661,
+      "loss": 1.1183,
+      "step": 9277
+    },
+    {
+      "epoch": 0.8143261901692581,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0004927756056771818,
+      "loss": 1.1524,
+      "step": 9278
+    },
+    {
+      "epoch": 0.8144139597521607,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.000492557252790366,
+      "loss": 1.1279,
+      "step": 9279
+    },
+    {
+      "epoch": 0.8145017293350632,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0004923390141385697,
+      "loss": 1.1465,
+      "step": 9280
+    },
+    {
+      "epoch": 0.8145894989179656,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004921208897433319,
+      "loss": 1.0831,
+      "step": 9281
+    },
+    {
+      "epoch": 0.8146772685008681,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0004919028796261809,
+      "loss": 1.1769,
+      "step": 9282
+    },
+    {
+      "epoch": 0.8147650380837705,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0004916849838086332,
+      "loss": 1.085,
+      "step": 9283
+    },
+    {
+      "epoch": 0.8148528076666731,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0004914672023121945,
+      "loss": 1.1509,
+      "step": 9284
+    },
+    {
+      "epoch": 0.8149405772495756,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004912495351583588,
+      "loss": 1.1553,
+      "step": 9285
+    },
+    {
+      "epoch": 0.815028346832478,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004910319823686094,
+      "loss": 1.1426,
+      "step": 9286
+    },
+    {
+      "epoch": 0.8151161164153805,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.000490814543964417,
+      "loss": 1.1452,
+      "step": 9287
+    },
+    {
+      "epoch": 0.815203885998283,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004905972199672427,
+      "loss": 1.1796,
+      "step": 9288
+    },
+    {
+      "epoch": 0.8152916555811855,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0004903800103985352,
+      "loss": 1.1054,
+      "step": 9289
+    },
+    {
+      "epoch": 0.815379425164088,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004901629152797327,
+      "loss": 1.1386,
+      "step": 9290
+    },
+    {
+      "epoch": 0.8154671947469905,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0004899459346322607,
+      "loss": 1.1624,
+      "step": 9291
+    },
+    {
+      "epoch": 0.8155549643298929,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0004897290684775346,
+      "loss": 1.1791,
+      "step": 9292
+    },
+    {
+      "epoch": 0.8156427339127954,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0004895123168369589,
+      "loss": 1.1799,
+      "step": 9293
+    },
+    {
+      "epoch": 0.8157305034956979,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004892956797319259,
+      "loss": 1.1177,
+      "step": 9294
+    },
+    {
+      "epoch": 0.8158182730786004,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004890791571838163,
+      "loss": 1.1116,
+      "step": 9295
+    },
+    {
+      "epoch": 0.8159060426615029,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004888627492140005,
+      "loss": 1.126,
+      "step": 9296
+    },
+    {
+      "epoch": 0.8159938122444054,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004886464558438369,
+      "loss": 1.143,
+      "step": 9297
+    },
+    {
+      "epoch": 0.8160815818273078,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000488430277094673,
+      "loss": 1.1287,
+      "step": 9298
+    },
+    {
+      "epoch": 0.8161693514102103,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00048821421298784454,
+      "loss": 1.1346,
+      "step": 9299
+    },
+    {
+      "epoch": 0.8162571209931129,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004879982635446763,
+      "loss": 1.0995,
+      "step": 9300
+    },
+    {
+      "epoch": 0.8163448905760153,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00048778242878648155,
+      "loss": 1.0986,
+      "step": 9301
+    },
+    {
+      "epoch": 0.8164326601589178,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00048756670873456256,
+      "loss": 1.1638,
+      "step": 9302
+    },
+    {
+      "epoch": 0.8165204297418203,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004873511034102097,
+      "loss": 1.1004,
+      "step": 9303
+    },
+    {
+      "epoch": 0.8166081993247227,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0004871356128347027,
+      "loss": 1.0876,
+      "step": 9304
+    },
+    {
+      "epoch": 0.8166959689076253,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004869202370293094,
+      "loss": 1.1024,
+      "step": 9305
+    },
+    {
+      "epoch": 0.8167837384905278,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.000486704976015287,
+      "loss": 1.17,
+      "step": 9306
+    },
+    {
+      "epoch": 0.8168715080734302,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004864898298138802,
+      "loss": 1.1716,
+      "step": 9307
+    },
+    {
+      "epoch": 0.8169592776563327,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0004862747984463234,
+      "loss": 1.1753,
+      "step": 9308
+    },
+    {
+      "epoch": 0.8170470472392352,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0004860598819338394,
+      "loss": 1.1196,
+      "step": 9309
+    },
+    {
+      "epoch": 0.8171348168221377,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.00048584508029763995,
+      "loss": 1.1767,
+      "step": 9310
+    },
+    {
+      "epoch": 0.8172225864050402,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0004856303935589244,
+      "loss": 1.1554,
+      "step": 9311
+    },
+    {
+      "epoch": 0.8173103559879427,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0004854158217388815,
+      "loss": 1.184,
+      "step": 9312
+    },
+    {
+      "epoch": 0.8173981255708451,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004852013648586896,
+      "loss": 1.1078,
+      "step": 9313
+    },
+    {
+      "epoch": 0.8174858951537476,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0004849870229395137,
+      "loss": 1.1678,
+      "step": 9314
+    },
+    {
+      "epoch": 0.8175736647366502,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.000484772796002509,
+      "loss": 1.1619,
+      "step": 9315
+    },
+    {
+      "epoch": 0.8176614343195526,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0004845586840688188,
+      "loss": 1.1484,
+      "step": 9316
+    },
+    {
+      "epoch": 0.8177492039024551,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00048434468715957495,
+      "loss": 1.1716,
+      "step": 9317
+    },
+    {
+      "epoch": 0.8178369734853576,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00048413080529589864,
+      "loss": 1.1005,
+      "step": 9318
+    },
+    {
+      "epoch": 0.81792474306826,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0004839170384988983,
+      "loss": 1.0901,
+      "step": 9319
+    },
+    {
+      "epoch": 0.8180125126511625,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0004837033867896724,
+      "loss": 1.1537,
+      "step": 9320
+    },
+    {
+      "epoch": 0.8181002822340651,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0004834898501893074,
+      "loss": 1.1119,
+      "step": 9321
+    },
+    {
+      "epoch": 0.8181880518169675,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00048327642871887863,
+      "loss": 1.113,
+      "step": 9322
+    },
+    {
+      "epoch": 0.81827582139987,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00048306312239944967,
+      "loss": 1.2455,
+      "step": 9323
+    },
+    {
+      "epoch": 0.8183635909827724,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004828499312520734,
+      "loss": 1.1536,
+      "step": 9324
+    },
+    {
+      "epoch": 0.8184513605656749,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00048263685529779075,
+      "loss": 1.1409,
+      "step": 9325
+    },
+    {
+      "epoch": 0.8185391301485775,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00048242389455763137,
+      "loss": 1.1356,
+      "step": 9326
+    },
+    {
+      "epoch": 0.81862689973148,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0004822110490526137,
+      "loss": 1.1511,
+      "step": 9327
+    },
+    {
+      "epoch": 0.8187146693143824,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0004819983188037448,
+      "loss": 1.1238,
+      "step": 9328
+    },
+    {
+      "epoch": 0.8188024388972849,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00048178570383202064,
+      "loss": 1.1728,
+      "step": 9329
+    },
+    {
+      "epoch": 0.8188902084801873,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00048157320415842465,
+      "loss": 1.1115,
+      "step": 9330
+    },
+    {
+      "epoch": 0.8189779780630899,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00048136081980393027,
+      "loss": 1.1469,
+      "step": 9331
+    },
+    {
+      "epoch": 0.8190657476459924,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0004811485507894985,
+      "loss": 1.1119,
+      "step": 9332
+    },
+    {
+      "epoch": 0.8191535172288948,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00048093639713608064,
+      "loss": 1.1426,
+      "step": 9333
+    },
+    {
+      "epoch": 0.8192412868117973,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00048072435886461416,
+      "loss": 1.1439,
+      "step": 9334
+    },
+    {
+      "epoch": 0.8193290563946998,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.000480512435996027,
+      "loss": 1.1627,
+      "step": 9335
+    },
+    {
+      "epoch": 0.8194168259776023,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0004803006285512349,
+      "loss": 1.1393,
+      "step": 9336
+    },
+    {
+      "epoch": 0.8195045955605048,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0004800889365511428,
+      "loss": 1.1539,
+      "step": 9337
+    },
+    {
+      "epoch": 0.8195923651434073,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004798773600166433,
+      "loss": 1.1256,
+      "step": 9338
+    },
+    {
+      "epoch": 0.8196801347263097,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00047966589896861845,
+      "loss": 1.1581,
+      "step": 9339
+    },
+    {
+      "epoch": 0.8197679043092122,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00047945455342793855,
+      "loss": 1.1495,
+      "step": 9340
+    },
+    {
+      "epoch": 0.8198556738921148,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0004792433234154629,
+      "loss": 1.1052,
+      "step": 9341
+    },
+    {
+      "epoch": 0.8199434434750172,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0004790322089520382,
+      "loss": 1.1272,
+      "step": 9342
+    },
+    {
+      "epoch": 0.8200312130579197,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00047882121005850145,
+      "loss": 1.1583,
+      "step": 9343
+    },
+    {
+      "epoch": 0.8201189826408222,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00047861032675567707,
+      "loss": 1.2194,
+      "step": 9344
+    },
+    {
+      "epoch": 0.8202067522237246,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.000478399559064379,
+      "loss": 1.1778,
+      "step": 9345
+    },
+    {
+      "epoch": 0.8202945218066271,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0004781889070054082,
+      "loss": 1.1187,
+      "step": 9346
+    },
+    {
+      "epoch": 0.8203822913895297,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0004779783705995556,
+      "loss": 1.1798,
+      "step": 9347
+    },
+    {
+      "epoch": 0.8204700609724321,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00047776794986760057,
+      "loss": 1.2195,
+      "step": 9348
+    },
+    {
+      "epoch": 0.8205578305553346,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00047755764483031077,
+      "loss": 1.1954,
+      "step": 9349
+    },
+    {
+      "epoch": 0.8206456001382371,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0004773474555084421,
+      "loss": 1.1612,
+      "step": 9350
+    },
+    {
+      "epoch": 0.8207333697211395,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.00047713738192273946,
+      "loss": 1.1229,
+      "step": 9351
+    },
+    {
+      "epoch": 0.8208211393040421,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00047692742409393655,
+      "loss": 1.1118,
+      "step": 9352
+    },
+    {
+      "epoch": 0.8209089088869446,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0004767175820427552,
+      "loss": 1.2038,
+      "step": 9353
+    },
+    {
+      "epoch": 0.820996678469847,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00047650785578990615,
+      "loss": 1.1538,
+      "step": 9354
+    },
+    {
+      "epoch": 0.8210844480527495,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00047629824535608836,
+      "loss": 1.1044,
+      "step": 9355
+    },
+    {
+      "epoch": 0.821172217635652,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00047608875076198974,
+      "loss": 1.1103,
+      "step": 9356
+    },
+    {
+      "epoch": 0.8212599872185545,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004758793720282868,
+      "loss": 1.1381,
+      "step": 9357
+    },
+    {
+      "epoch": 0.821347756801457,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00047567010917564376,
+      "loss": 1.1365,
+      "step": 9358
+    },
+    {
+      "epoch": 0.8214355263843595,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004754609622247142,
+      "loss": 1.0849,
+      "step": 9359
+    },
+    {
+      "epoch": 0.8215232959672619,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004752519311961404,
+      "loss": 1.1069,
+      "step": 9360
+    },
+    {
+      "epoch": 0.8216110655501644,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0004750430161105531,
+      "loss": 1.1356,
+      "step": 9361
+    },
+    {
+      "epoch": 0.821698835133067,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004748342169885703,
+      "loss": 1.1191,
+      "step": 9362
+    },
+    {
+      "epoch": 0.8217866047159694,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00047462553385080075,
+      "loss": 1.1051,
+      "step": 9363
+    },
+    {
+      "epoch": 0.8218743742988719,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00047441696671784056,
+      "loss": 1.1741,
+      "step": 9364
+    },
+    {
+      "epoch": 0.8219621438817744,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00047420851561027393,
+      "loss": 1.1361,
+      "step": 9365
+    },
+    {
+      "epoch": 0.8220499134646768,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00047400018054867435,
+      "loss": 1.1817,
+      "step": 9366
+    },
+    {
+      "epoch": 0.8221376830475794,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00047379196155360393,
+      "loss": 1.1199,
+      "step": 9367
+    },
+    {
+      "epoch": 0.8222254526304819,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004735838586456128,
+      "loss": 1.1059,
+      "step": 9368
+    },
+    {
+      "epoch": 0.8223132222133843,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00047337587184524015,
+      "loss": 1.1253,
+      "step": 9369
+    },
+    {
+      "epoch": 0.8224009917962868,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00047316800117301323,
+      "loss": 1.1424,
+      "step": 9370
+    },
+    {
+      "epoch": 0.8224887613791892,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0004729602466494479,
+      "loss": 1.1796,
+      "step": 9371
+    },
+    {
+      "epoch": 0.8225765309620918,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.000472752608295049,
+      "loss": 1.1283,
+      "step": 9372
+    },
+    {
+      "epoch": 0.8226643005449943,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004725450861303096,
+      "loss": 1.1749,
+      "step": 9373
+    },
+    {
+      "epoch": 0.8227520701278968,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00047233768017571114,
+      "loss": 1.16,
+      "step": 9374
+    },
+    {
+      "epoch": 0.8228398397107992,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.00047213039045172387,
+      "loss": 1.1019,
+      "step": 9375
+    },
+    {
+      "epoch": 0.8229276092937017,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0004719232169788068,
+      "loss": 1.1412,
+      "step": 9376
+    },
+    {
+      "epoch": 0.8230153788766041,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0004717161597774065,
+      "loss": 1.1687,
+      "step": 9377
+    },
+    {
+      "epoch": 0.8231031484595067,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004715092188679589,
+      "loss": 1.0412,
+      "step": 9378
+    },
+    {
+      "epoch": 0.8231909180424092,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00047130239427088845,
+      "loss": 1.1693,
+      "step": 9379
+    },
+    {
+      "epoch": 0.8232786876253116,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00047109568600660817,
+      "loss": 1.1064,
+      "step": 9380
+    },
+    {
+      "epoch": 0.8233664572082141,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00047088909409551854,
+      "loss": 1.1454,
+      "step": 9381
+    },
+    {
+      "epoch": 0.8234542267911166,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0004706826185580099,
+      "loss": 1.1541,
+      "step": 9382
+    },
+    {
+      "epoch": 0.8235419963740191,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00047047625941446013,
+      "loss": 1.2286,
+      "step": 9383
+    },
+    {
+      "epoch": 0.8236297659569216,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0004702700166852371,
+      "loss": 1.1457,
+      "step": 9384
+    },
+    {
+      "epoch": 0.8237175355398241,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0004700638903906952,
+      "loss": 1.1991,
+      "step": 9385
+    },
+    {
+      "epoch": 0.8238053051227265,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0004698578805511783,
+      "loss": 1.1624,
+      "step": 9386
+    },
+    {
+      "epoch": 0.823893074705629,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.00046965198718701927,
+      "loss": 1.1899,
+      "step": 9387
+    },
+    {
+      "epoch": 0.8239808442885316,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.00046944621031853894,
+      "loss": 1.1198,
+      "step": 9388
+    },
+    {
+      "epoch": 0.824068613871434,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0004692405499660461,
+      "loss": 1.1432,
+      "step": 9389
+    },
+    {
+      "epoch": 0.8241563834543365,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00046903500614983886,
+      "loss": 1.1497,
+      "step": 9390
+    },
+    {
+      "epoch": 0.824244153037239,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00046882957889020376,
+      "loss": 1.2337,
+      "step": 9391
+    },
+    {
+      "epoch": 0.8243319226201414,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0004686242682074159,
+      "loss": 1.1307,
+      "step": 9392
+    },
+    {
+      "epoch": 0.824419692203044,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00046841907412173764,
+      "loss": 1.1619,
+      "step": 9393
+    },
+    {
+      "epoch": 0.8245074617859465,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00046821399665342194,
+      "loss": 1.1852,
+      "step": 9394
+    },
+    {
+      "epoch": 0.8245952313688489,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0004680090358227084,
+      "loss": 1.1414,
+      "step": 9395
+    },
+    {
+      "epoch": 0.8246830009517514,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00046780419164982654,
+      "loss": 1.0955,
+      "step": 9396
+    },
+    {
+      "epoch": 0.8247707705346539,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00046759946415499294,
+      "loss": 1.1488,
+      "step": 9397
+    },
+    {
+      "epoch": 0.8248585401175564,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0004673948533584138,
+      "loss": 1.1821,
+      "step": 9398
+    },
+    {
+      "epoch": 0.8249463097004589,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0004671903592802833,
+      "loss": 1.1903,
+      "step": 9399
+    },
+    {
+      "epoch": 0.8250340792833614,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00046698598194078444,
+      "loss": 1.1385,
+      "step": 9400
+    },
+    {
+      "epoch": 0.8251218488662638,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004667817213600879,
+      "loss": 1.1571,
+      "step": 9401
+    },
+    {
+      "epoch": 0.8252096184491663,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00046657757755835376,
+      "loss": 1.1323,
+      "step": 9402
+    },
+    {
+      "epoch": 0.8252973880320688,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0004663735505557301,
+      "loss": 1.1569,
+      "step": 9403
+    },
+    {
+      "epoch": 0.8253851576149713,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004661696403723536,
+      "loss": 1.1458,
+      "step": 9404
+    },
+    {
+      "epoch": 0.8254729271978738,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.00046596584702834936,
+      "loss": 1.136,
+      "step": 9405
+    },
+    {
+      "epoch": 0.8255606967807763,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0004657621705438312,
+      "loss": 1.1251,
+      "step": 9406
+    },
+    {
+      "epoch": 0.8256484663636787,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00046555861093890095,
+      "loss": 1.1269,
+      "step": 9407
+    },
+    {
+      "epoch": 0.8257362359465812,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00046535516823364943,
+      "loss": 1.1497,
+      "step": 9408
+    },
+    {
+      "epoch": 0.8258240055294838,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004651518424481551,
+      "loss": 1.2032,
+      "step": 9409
+    },
+    {
+      "epoch": 0.8259117751123862,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004649486336024859,
+      "loss": 1.1792,
+      "step": 9410
+    },
+    {
+      "epoch": 0.8259995446952887,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0004647455417166974,
+      "loss": 1.1555,
+      "step": 9411
+    },
+    {
+      "epoch": 0.8260873142781912,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00046454256681083446,
+      "loss": 1.1305,
+      "step": 9412
+    },
+    {
+      "epoch": 0.8261750838610936,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0004643397089049289,
+      "loss": 1.1378,
+      "step": 9413
+    },
+    {
+      "epoch": 0.8262628534439962,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00046413696801900304,
+      "loss": 1.1685,
+      "step": 9414
+    },
+    {
+      "epoch": 0.8263506230268987,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00046393434417306634,
+      "loss": 1.1149,
+      "step": 9415
+    },
+    {
+      "epoch": 0.8264383926098011,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0004637318373871166,
+      "loss": 1.1799,
+      "step": 9416
+    },
+    {
+      "epoch": 0.8265261621927036,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004635294476811408,
+      "loss": 1.1454,
+      "step": 9417
+    },
+    {
+      "epoch": 0.826613931775606,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0004633271750751139,
+      "loss": 1.1528,
+      "step": 9418
+    },
+    {
+      "epoch": 0.8267017013585086,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0004631250195889993,
+      "loss": 1.1508,
+      "step": 9419
+    },
+    {
+      "epoch": 0.8267894709414111,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00046292298124274927,
+      "loss": 1.1538,
+      "step": 9420
+    },
+    {
+      "epoch": 0.8268772405243136,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004627210600563038,
+      "loss": 1.0882,
+      "step": 9421
+    },
+    {
+      "epoch": 0.826965010107216,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00046251925604959173,
+      "loss": 1.105,
+      "step": 9422
+    },
+    {
+      "epoch": 0.8270527796901185,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004623175692425306,
+      "loss": 1.1177,
+      "step": 9423
+    },
+    {
+      "epoch": 0.827140549273021,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00046211599965502594,
+      "loss": 1.2445,
+      "step": 9424
+    },
+    {
+      "epoch": 0.8272283188559235,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004619145473069719,
+      "loss": 1.1468,
+      "step": 9425
+    },
+    {
+      "epoch": 0.827316088438826,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.000461713212218251,
+      "loss": 1.182,
+      "step": 9426
+    },
+    {
+      "epoch": 0.8274038580217284,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00046151199440873456,
+      "loss": 1.1593,
+      "step": 9427
+    },
+    {
+      "epoch": 0.8274916276046309,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004613108938982813,
+      "loss": 1.0936,
+      "step": 9428
+    },
+    {
+      "epoch": 0.8275793971875335,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004611099107067395,
+      "loss": 1.1584,
+      "step": 9429
+    },
+    {
+      "epoch": 0.827667166770436,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00046090904485394517,
+      "loss": 1.0985,
+      "step": 9430
+    },
+    {
+      "epoch": 0.8277549363533384,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00046070829635972354,
+      "loss": 1.1076,
+      "step": 9431
+    },
+    {
+      "epoch": 0.8278427059362409,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00046050766524388686,
+      "loss": 1.1398,
+      "step": 9432
+    },
+    {
+      "epoch": 0.8279304755191433,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000460307151526237,
+      "loss": 1.14,
+      "step": 9433
+    },
+    {
+      "epoch": 0.8280182451020458,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0004601067552265639,
+      "loss": 1.1787,
+      "step": 9434
+    },
+    {
+      "epoch": 0.8281060146849484,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0004599064763646464,
+      "loss": 1.1504,
+      "step": 9435
+    },
+    {
+      "epoch": 0.8281937842678508,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0004597063149602502,
+      "loss": 1.2183,
+      "step": 9436
+    },
+    {
+      "epoch": 0.8282815538507533,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00045950627103313115,
+      "loss": 1.1221,
+      "step": 9437
+    },
+    {
+      "epoch": 0.8283693234336558,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004593063446030325,
+      "loss": 1.0786,
+      "step": 9438
+    },
+    {
+      "epoch": 0.8284570930165582,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0004591065356896867,
+      "loss": 1.1546,
+      "step": 9439
+    },
+    {
+      "epoch": 0.8285448625994608,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00045890684431281327,
+      "loss": 1.1136,
+      "step": 9440
+    },
+    {
+      "epoch": 0.8286326321823633,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00045870727049212147,
+      "loss": 1.16,
+      "step": 9441
+    },
+    {
+      "epoch": 0.8287204017652657,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0004585078142473083,
+      "loss": 1.1298,
+      "step": 9442
+    },
+    {
+      "epoch": 0.8288081713481682,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0004583084755980594,
+      "loss": 1.16,
+      "step": 9443
+    },
+    {
+      "epoch": 0.8288959409310707,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00045810925456404877,
+      "loss": 1.206,
+      "step": 9444
+    },
+    {
+      "epoch": 0.8289837105139732,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0004579101511649385,
+      "loss": 1.1888,
+      "step": 9445
+    },
+    {
+      "epoch": 0.8290714800968757,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004577111654203795,
+      "loss": 1.2021,
+      "step": 9446
+    },
+    {
+      "epoch": 0.8291592496797782,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00045751229735001113,
+      "loss": 1.075,
+      "step": 9447
+    },
+    {
+      "epoch": 0.8292470192626806,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0004573135469734603,
+      "loss": 1.1437,
+      "step": 9448
+    },
+    {
+      "epoch": 0.8293347888455831,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0004571149143103431,
+      "loss": 1.1174,
+      "step": 9449
+    },
+    {
+      "epoch": 0.8294225584284857,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0004569163993802638,
+      "loss": 1.1832,
+      "step": 9450
+    },
+    {
+      "epoch": 0.8295103280113881,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004567180022028156,
+      "loss": 1.1571,
+      "step": 9451
+    },
+    {
+      "epoch": 0.8295980975942906,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0004565197227975785,
+      "loss": 1.1225,
+      "step": 9452
+    },
+    {
+      "epoch": 0.8296858671771931,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00045632156118412227,
+      "loss": 1.1656,
+      "step": 9453
+    },
+    {
+      "epoch": 0.8297736367600955,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004561235173820052,
+      "loss": 1.1598,
+      "step": 9454
+    },
+    {
+      "epoch": 0.8298614063429981,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00045592559141077293,
+      "loss": 1.1415,
+      "step": 9455
+    },
+    {
+      "epoch": 0.8299491759259006,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00045572778328995987,
+      "loss": 1.1484,
+      "step": 9456
+    },
+    {
+      "epoch": 0.830036945508803,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004555300930390893,
+      "loss": 1.1808,
+      "step": 9457
+    },
+    {
+      "epoch": 0.8301247150917055,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0004553325206776723,
+      "loss": 1.1848,
+      "step": 9458
+    },
+    {
+      "epoch": 0.830212484674608,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0004551350662252084,
+      "loss": 1.1633,
+      "step": 9459
+    },
+    {
+      "epoch": 0.8303002542575104,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004549377297011856,
+      "loss": 1.1793,
+      "step": 9460
+    },
+    {
+      "epoch": 0.830388023840413,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.00045474051112508017,
+      "loss": 1.1732,
+      "step": 9461
+    },
+    {
+      "epoch": 0.8304757934233155,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.000454543410516357,
+      "loss": 1.0652,
+      "step": 9462
+    },
+    {
+      "epoch": 0.8305635630062179,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000454346427894469,
+      "loss": 1.1463,
+      "step": 9463
+    },
+    {
+      "epoch": 0.8306513325891204,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00045414956327885747,
+      "loss": 1.164,
+      "step": 9464
+    },
+    {
+      "epoch": 0.8307391021720228,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0004539528166889524,
+      "loss": 1.1562,
+      "step": 9465
+    },
+    {
+      "epoch": 0.8308268717549254,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000453756188144172,
+      "loss": 1.186,
+      "step": 9466
+    },
+    {
+      "epoch": 0.8309146413378279,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004535596776639224,
+      "loss": 1.1028,
+      "step": 9467
+    },
+    {
+      "epoch": 0.8310024109207303,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00045336328526759847,
+      "loss": 1.1901,
+      "step": 9468
+    },
+    {
+      "epoch": 0.8310901805036328,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00045316701097458336,
+      "loss": 1.1502,
+      "step": 9469
+    },
+    {
+      "epoch": 0.8311779500865353,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0004529708548042487,
+      "loss": 1.161,
+      "step": 9470
+    },
+    {
+      "epoch": 0.8312657196694379,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00045277481677595457,
+      "loss": 1.1026,
+      "step": 9471
+    },
+    {
+      "epoch": 0.8313534892523403,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00045257889690904863,
+      "loss": 1.1435,
+      "step": 9472
+    },
+    {
+      "epoch": 0.8314412588352428,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00045238309522286713,
+      "loss": 1.1375,
+      "step": 9473
+    },
+    {
+      "epoch": 0.8315290284181452,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004521874117367361,
+      "loss": 1.1023,
+      "step": 9474
+    },
+    {
+      "epoch": 0.8316167980010477,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0004519918464699678,
+      "loss": 1.1544,
+      "step": 9475
+    },
+    {
+      "epoch": 0.8317045675839503,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0004517963994418639,
+      "loss": 1.2107,
+      "step": 9476
+    },
+    {
+      "epoch": 0.8317923371668527,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00045160107067171413,
+      "loss": 1.176,
+      "step": 9477
+    },
+    {
+      "epoch": 0.8318801067497552,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0004514058601787973,
+      "loss": 1.1176,
+      "step": 9478
+    },
+    {
+      "epoch": 0.8319678763326577,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0004512107679823793,
+      "loss": 1.1566,
+      "step": 9479
+    },
+    {
+      "epoch": 0.8320556459155601,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00045101579410171497,
+      "loss": 1.1697,
+      "step": 9480
+    },
+    {
+      "epoch": 0.8321434154984627,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00045082093855604776,
+      "loss": 1.1717,
+      "step": 9481
+    },
+    {
+      "epoch": 0.8322311850813652,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00045062620136460916,
+      "loss": 1.1271,
+      "step": 9482
+    },
+    {
+      "epoch": 0.8323189546642676,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0004504315825466183,
+      "loss": 1.1453,
+      "step": 9483
+    },
+    {
+      "epoch": 0.8324067242471701,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004502370821212841,
+      "loss": 1.1126,
+      "step": 9484
+    },
+    {
+      "epoch": 0.8324944938300726,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0004500427001078026,
+      "loss": 1.0803,
+      "step": 9485
+    },
+    {
+      "epoch": 0.832582263412975,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.000449848436525359,
+      "loss": 1.174,
+      "step": 9486
+    },
+    {
+      "epoch": 0.8326700329958776,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.00044965429139312546,
+      "loss": 1.1416,
+      "step": 9487
+    },
+    {
+      "epoch": 0.8327578025787801,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.000449460264730264,
+      "loss": 1.1207,
+      "step": 9488
+    },
+    {
+      "epoch": 0.8328455721616825,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0004492663565559241,
+      "loss": 1.1123,
+      "step": 9489
+    },
+    {
+      "epoch": 0.832933341744585,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00044907256688924395,
+      "loss": 1.1491,
+      "step": 9490
+    },
+    {
+      "epoch": 0.8330211113274875,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004488788957493494,
+      "loss": 1.0891,
+      "step": 9491
+    },
+    {
+      "epoch": 0.83310888091039,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00044868534315535536,
+      "loss": 1.1756,
+      "step": 9492
+    },
+    {
+      "epoch": 0.8331966504932925,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00044849190912636433,
+      "loss": 1.1445,
+      "step": 9493
+    },
+    {
+      "epoch": 0.833284420076195,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.000448298593681468,
+      "loss": 1.1605,
+      "step": 9494
+    },
+    {
+      "epoch": 0.8333721896590974,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004481053968397454,
+      "loss": 1.1369,
+      "step": 9495
+    },
+    {
+      "epoch": 0.8334599592419999,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0004479123186202644,
+      "loss": 1.1376,
+      "step": 9496
+    },
+    {
+      "epoch": 0.8335477288249025,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0004477193590420814,
+      "loss": 1.1301,
+      "step": 9497
+    },
+    {
+      "epoch": 0.8336354984078049,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0004475265181242405,
+      "loss": 1.1133,
+      "step": 9498
+    },
+    {
+      "epoch": 0.8337232679907074,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0004473337958857742,
+      "loss": 1.1859,
+      "step": 9499
+    },
+    {
+      "epoch": 0.8338110375736099,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0004471411923457035,
+      "loss": 1.1379,
+      "step": 9500
+    },
+    {
+      "epoch": 0.8338110375736099,
+      "eval_loss": 1.1350057125091553,
+      "eval_runtime": 437.4513,
+      "eval_samples_per_second": 33.695,
+      "eval_steps_per_second": 8.424,
+      "step": 9500
+    },
+    {
+      "epoch": 0.8338988071565123,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0004469487075230376,
+      "loss": 1.1654,
+      "step": 9501
+    },
+    {
+      "epoch": 0.8339865767394149,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0004467563414367743,
+      "loss": 1.1386,
+      "step": 9502
+    },
+    {
+      "epoch": 0.8340743463223174,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0004465640941058989,
+      "loss": 1.1563,
+      "step": 9503
+    },
+    {
+      "epoch": 0.8341621159052198,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0004463719655493854,
+      "loss": 1.1215,
+      "step": 9504
+    },
+    {
+      "epoch": 0.8342498854881223,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0004461799557861968,
+      "loss": 1.1728,
+      "step": 9505
+    },
+    {
+      "epoch": 0.8343376550710248,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00044598806483528317,
+      "loss": 1.1515,
+      "step": 9506
+    },
+    {
+      "epoch": 0.8344254246539273,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0004457962927155834,
+      "loss": 1.0876,
+      "step": 9507
+    },
+    {
+      "epoch": 0.8345131942368298,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00044560463944602483,
+      "loss": 1.1347,
+      "step": 9508
+    },
+    {
+      "epoch": 0.8346009638197323,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0004454131050455227,
+      "loss": 1.159,
+      "step": 9509
+    },
+    {
+      "epoch": 0.8346887334026347,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.000445221689532981,
+      "loss": 1.1365,
+      "step": 9510
+    },
+    {
+      "epoch": 0.8347765029855372,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00044503039292729147,
+      "loss": 1.1857,
+      "step": 9511
+    },
+    {
+      "epoch": 0.8348642725684398,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00044483921524733405,
+      "loss": 1.1607,
+      "step": 9512
+    },
+    {
+      "epoch": 0.8349520421513422,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0004446481565119775,
+      "loss": 1.1179,
+      "step": 9513
+    },
+    {
+      "epoch": 0.8350398117342447,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0004444572167400789,
+      "loss": 1.1439,
+      "step": 9514
+    },
+    {
+      "epoch": 0.8351275813171471,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0004442663959504825,
+      "loss": 1.1219,
+      "step": 9515
+    },
+    {
+      "epoch": 0.8352153509000496,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00044407569416202234,
+      "loss": 1.1428,
+      "step": 9516
+    },
+    {
+      "epoch": 0.8353031204829521,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00044388511139351966,
+      "loss": 1.2045,
+      "step": 9517
+    },
+    {
+      "epoch": 0.8353908900658547,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004436946476637838,
+      "loss": 1.1238,
+      "step": 9518
+    },
+    {
+      "epoch": 0.8354786596487571,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00044350430299161314,
+      "loss": 1.1734,
+      "step": 9519
+    },
+    {
+      "epoch": 0.8355664292316596,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000443314077395794,
+      "loss": 1.1489,
+      "step": 9520
+    },
+    {
+      "epoch": 0.835654198814562,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0004431239708951007,
+      "loss": 1.1613,
+      "step": 9521
+    },
+    {
+      "epoch": 0.8357419683974645,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00044293398350829646,
+      "loss": 1.1368,
+      "step": 9522
+    },
+    {
+      "epoch": 0.8358297379803671,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004427441152541316,
+      "loss": 1.1457,
+      "step": 9523
+    },
+    {
+      "epoch": 0.8359175075632695,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00044255436615134536,
+      "loss": 1.1207,
+      "step": 9524
+    },
+    {
+      "epoch": 0.836005277146172,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00044236473621866627,
+      "loss": 1.1667,
+      "step": 9525
+    },
+    {
+      "epoch": 0.8360930467290745,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004421752254748089,
+      "loss": 1.1462,
+      "step": 9526
+    },
+    {
+      "epoch": 0.8361808163119769,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0004419858339384778,
+      "loss": 1.1592,
+      "step": 9527
+    },
+    {
+      "epoch": 0.8362685858948795,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00044179656162836496,
+      "loss": 1.1738,
+      "step": 9528
+    },
+    {
+      "epoch": 0.836356355477782,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00044160740856315114,
+      "loss": 1.1201,
+      "step": 9529
+    },
+    {
+      "epoch": 0.8364441250606844,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0004414183747615044,
+      "loss": 1.0869,
+      "step": 9530
+    },
+    {
+      "epoch": 0.8365318946435869,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00044122946024208203,
+      "loss": 1.1686,
+      "step": 9531
+    },
+    {
+      "epoch": 0.8366196642264894,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00044104066502352905,
+      "loss": 1.1534,
+      "step": 9532
+    },
+    {
+      "epoch": 0.8367074338093919,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004408519891244791,
+      "loss": 1.1992,
+      "step": 9533
+    },
+    {
+      "epoch": 0.8367952033922944,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.000440663432563553,
+      "loss": 1.1154,
+      "step": 9534
+    },
+    {
+      "epoch": 0.8368829729751969,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.000440474995359361,
+      "loss": 1.1071,
+      "step": 9535
+    },
+    {
+      "epoch": 0.8369707425580993,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0004402866775305014,
+      "loss": 1.1482,
+      "step": 9536
+    },
+    {
+      "epoch": 0.8370585121410018,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0004400984790955603,
+      "loss": 1.1341,
+      "step": 9537
+    },
+    {
+      "epoch": 0.8371462817239044,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00043991040007311175,
+      "loss": 1.1941,
+      "step": 9538
+    },
+    {
+      "epoch": 0.8372340513068068,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00043972244048171846,
+      "loss": 1.1915,
+      "step": 9539
+    },
+    {
+      "epoch": 0.8373218208897093,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00043953460033993157,
+      "loss": 1.1745,
+      "step": 9540
+    },
+    {
+      "epoch": 0.8374095904726118,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004393468796662904,
+      "loss": 1.1155,
+      "step": 9541
+    },
+    {
+      "epoch": 0.8374973600555142,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0004391592784793215,
+      "loss": 1.1273,
+      "step": 9542
+    },
+    {
+      "epoch": 0.8375851296384167,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00043897179679754084,
+      "loss": 1.1085,
+      "step": 9543
+    },
+    {
+      "epoch": 0.8376728992213193,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0004387844346394521,
+      "loss": 1.1469,
+      "step": 9544
+    },
+    {
+      "epoch": 0.8377606688042217,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00043859719202354695,
+      "loss": 1.1038,
+      "step": 9545
+    },
+    {
+      "epoch": 0.8378484383871242,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0004384100689683058,
+      "loss": 1.1395,
+      "step": 9546
+    },
+    {
+      "epoch": 0.8379362079700267,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00043822306549219674,
+      "loss": 1.1296,
+      "step": 9547
+    },
+    {
+      "epoch": 0.8380239775529291,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00043803618161367635,
+      "loss": 1.2539,
+      "step": 9548
+    },
+    {
+      "epoch": 0.8381117471358317,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004378494173511897,
+      "loss": 1.1518,
+      "step": 9549
+    },
+    {
+      "epoch": 0.8381995167187342,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00043766277272316905,
+      "loss": 1.1359,
+      "step": 9550
+    },
+    {
+      "epoch": 0.8382872863016366,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0004374762477480357,
+      "loss": 1.1382,
+      "step": 9551
+    },
+    {
+      "epoch": 0.8383750558845391,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0004372898424441991,
+      "loss": 1.1369,
+      "step": 9552
+    },
+    {
+      "epoch": 0.8384628254674416,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0004371035568300567,
+      "loss": 1.0816,
+      "step": 9553
+    },
+    {
+      "epoch": 0.8385505950503441,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00043691739092399347,
+      "loss": 1.1678,
+      "step": 9554
+    },
+    {
+      "epoch": 0.8386383646332466,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00043673134474438423,
+      "loss": 1.1539,
+      "step": 9555
+    },
+    {
+      "epoch": 0.838726134216149,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004365454183095909,
+      "loss": 1.1243,
+      "step": 9556
+    },
+    {
+      "epoch": 0.8388139037990515,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.000436359611637963,
+      "loss": 1.1623,
+      "step": 9557
+    },
+    {
+      "epoch": 0.838901673381954,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00043617392474783943,
+      "loss": 1.0973,
+      "step": 9558
+    },
+    {
+      "epoch": 0.8389894429648566,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00043598835765754654,
+      "loss": 1.1768,
+      "step": 9559
+    },
+    {
+      "epoch": 0.839077212547759,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00043580291038539906,
+      "loss": 1.128,
+      "step": 9560
+    },
+    {
+      "epoch": 0.8391649821306615,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004356175829497004,
+      "loss": 1.0904,
+      "step": 9561
+    },
+    {
+      "epoch": 0.839252751713564,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0004354323753687409,
+      "loss": 1.121,
+      "step": 9562
+    },
+    {
+      "epoch": 0.8393405212964664,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0004352472876608001,
+      "loss": 1.1574,
+      "step": 9563
+    },
+    {
+      "epoch": 0.839428290879369,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0004350623198441456,
+      "loss": 1.1589,
+      "step": 9564
+    },
+    {
+      "epoch": 0.8395160604622715,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0004348774719370328,
+      "loss": 1.1202,
+      "step": 9565
+    },
+    {
+      "epoch": 0.8396038300451739,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0004346927439577055,
+      "loss": 1.1633,
+      "step": 9566
+    },
+    {
+      "epoch": 0.8396915996280764,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00043450813592439566,
+      "loss": 1.146,
+      "step": 9567
+    },
+    {
+      "epoch": 0.8397793692109788,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00043432364785532356,
+      "loss": 1.1165,
+      "step": 9568
+    },
+    {
+      "epoch": 0.8398671387938814,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0004341392797686971,
+      "loss": 1.1181,
+      "step": 9569
+    },
+    {
+      "epoch": 0.8399549083767839,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004339550316827127,
+      "loss": 1.1052,
+      "step": 9570
+    },
+    {
+      "epoch": 0.8400426779596863,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00043377090361555504,
+      "loss": 1.1346,
+      "step": 9571
+    },
+    {
+      "epoch": 0.8401304475425888,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0004335868955853971,
+      "loss": 1.0916,
+      "step": 9572
+    },
+    {
+      "epoch": 0.8402182171254913,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0004334030076103992,
+      "loss": 1.1606,
+      "step": 9573
+    },
+    {
+      "epoch": 0.8403059867083937,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004332192397087104,
+      "loss": 1.1797,
+      "step": 9574
+    },
+    {
+      "epoch": 0.8403937562912963,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00043303559189846824,
+      "loss": 1.1261,
+      "step": 9575
+    },
+    {
+      "epoch": 0.8404815258741988,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0004328520641977982,
+      "loss": 1.1418,
+      "step": 9576
+    },
+    {
+      "epoch": 0.8405692954571012,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00043266865662481305,
+      "loss": 1.218,
+      "step": 9577
+    },
+    {
+      "epoch": 0.8406570650400037,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00043248536919761495,
+      "loss": 1.2162,
+      "step": 9578
+    },
+    {
+      "epoch": 0.8407448346229062,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00043230220193429326,
+      "loss": 1.2278,
+      "step": 9579
+    },
+    {
+      "epoch": 0.8408326042058087,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00043211915485292634,
+      "loss": 1.1255,
+      "step": 9580
+    },
+    {
+      "epoch": 0.8409203737887112,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0004319362279715797,
+      "loss": 1.1381,
+      "step": 9581
+    },
+    {
+      "epoch": 0.8410081433716137,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00043175342130830785,
+      "loss": 1.1095,
+      "step": 9582
+    },
+    {
+      "epoch": 0.8410959129545161,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00043157073488115283,
+      "loss": 1.1511,
+      "step": 9583
+    },
+    {
+      "epoch": 0.8411836825374186,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00043138816870814516,
+      "loss": 1.137,
+      "step": 9584
+    },
+    {
+      "epoch": 0.8412714521203212,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00043120572280730353,
+      "loss": 1.2037,
+      "step": 9585
+    },
+    {
+      "epoch": 0.8413592217032236,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00043102339719663447,
+      "loss": 1.1621,
+      "step": 9586
+    },
+    {
+      "epoch": 0.8414469912861261,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00043084119189413286,
+      "loss": 1.1378,
+      "step": 9587
+    },
+    {
+      "epoch": 0.8415347608690286,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00043065910691778193,
+      "loss": 1.1506,
+      "step": 9588
+    },
+    {
+      "epoch": 0.841622530451931,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0004304771422855522,
+      "loss": 1.1027,
+      "step": 9589
+    },
+    {
+      "epoch": 0.8417103000348336,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00043029529801540296,
+      "loss": 1.1099,
+      "step": 9590
+    },
+    {
+      "epoch": 0.8417980696177361,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00043011357412528196,
+      "loss": 1.1757,
+      "step": 9591
+    },
+    {
+      "epoch": 0.8418858392006385,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0004299319706331244,
+      "loss": 1.2074,
+      "step": 9592
+    },
+    {
+      "epoch": 0.841973608783541,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0004297504875568535,
+      "loss": 1.1249,
+      "step": 9593
+    },
+    {
+      "epoch": 0.8420613783664435,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00042956912491438106,
+      "loss": 1.0762,
+      "step": 9594
+    },
+    {
+      "epoch": 0.842149147949346,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00042938788272360747,
+      "loss": 1.1121,
+      "step": 9595
+    },
+    {
+      "epoch": 0.8422369175322485,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00042920676100241994,
+      "loss": 1.1766,
+      "step": 9596
+    },
+    {
+      "epoch": 0.842324687115151,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0004290257597686946,
+      "loss": 1.185,
+      "step": 9597
+    },
+    {
+      "epoch": 0.8424124566980534,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00042884487904029563,
+      "loss": 1.1066,
+      "step": 9598
+    },
+    {
+      "epoch": 0.8425002262809559,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.000428664118835075,
+      "loss": 1.0952,
+      "step": 9599
+    },
+    {
+      "epoch": 0.8425879958638584,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0004284834791708739,
+      "loss": 1.1589,
+      "step": 9600
+    },
+    {
+      "epoch": 0.8426757654467609,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0004283029600655197,
+      "loss": 1.1642,
+      "step": 9601
+    },
+    {
+      "epoch": 0.8427635350296634,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004281225615368292,
+      "loss": 1.136,
+      "step": 9602
+    },
+    {
+      "epoch": 0.8428513046125659,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0004279422836026073,
+      "loss": 1.1187,
+      "step": 9603
+    },
+    {
+      "epoch": 0.8429390741954683,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00042776212628064646,
+      "loss": 1.1319,
+      "step": 9604
+    },
+    {
+      "epoch": 0.8430268437783708,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00042758208958872774,
+      "loss": 1.0956,
+      "step": 9605
+    },
+    {
+      "epoch": 0.8431146133612734,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004274021735446199,
+      "loss": 1.1665,
+      "step": 9606
+    },
+    {
+      "epoch": 0.8432023829441758,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00042722237816608004,
+      "loss": 1.1218,
+      "step": 9607
+    },
+    {
+      "epoch": 0.8432901525270783,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.00042704270347085306,
+      "loss": 1.1025,
+      "step": 9608
+    },
+    {
+      "epoch": 0.8433779221099807,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004268631494766722,
+      "loss": 1.146,
+      "step": 9609
+    },
+    {
+      "epoch": 0.8434656916928832,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004266837162012588,
+      "loss": 1.1386,
+      "step": 9610
+    },
+    {
+      "epoch": 0.8435534612757858,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0004265044036623221,
+      "loss": 1.1241,
+      "step": 9611
+    },
+    {
+      "epoch": 0.8436412308586883,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004263252118775601,
+      "loss": 1.1457,
+      "step": 9612
+    },
+    {
+      "epoch": 0.8437290004415907,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0004261461408646573,
+      "loss": 1.1557,
+      "step": 9613
+    },
+    {
+      "epoch": 0.8438167700244932,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0004259671906412879,
+      "loss": 1.1449,
+      "step": 9614
+    },
+    {
+      "epoch": 0.8439045396073956,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.00042578836122511386,
+      "loss": 1.0978,
+      "step": 9615
+    },
+    {
+      "epoch": 0.8439923091902982,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00042560965263378444,
+      "loss": 1.1763,
+      "step": 9616
+    },
+    {
+      "epoch": 0.8440800787732007,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0004254310648849375,
+      "loss": 1.1806,
+      "step": 9617
+    },
+    {
+      "epoch": 0.8441678483561031,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004252525979961992,
+      "loss": 1.1016,
+      "step": 9618
+    },
+    {
+      "epoch": 0.8442556179390056,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00042507425198518364,
+      "loss": 1.1329,
+      "step": 9619
+    },
+    {
+      "epoch": 0.8443433875219081,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0004248960268694923,
+      "loss": 1.1158,
+      "step": 9620
+    },
+    {
+      "epoch": 0.8444311571048106,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004247179226667157,
+      "loss": 1.1605,
+      "step": 9621
+    },
+    {
+      "epoch": 0.8445189266877131,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00042453993939443185,
+      "loss": 1.2055,
+      "step": 9622
+    },
+    {
+      "epoch": 0.8446066962706156,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0004243620770702074,
+      "loss": 1.1339,
+      "step": 9623
+    },
+    {
+      "epoch": 0.844694465853518,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00042418433571159587,
+      "loss": 1.112,
+      "step": 9624
+    },
+    {
+      "epoch": 0.8447822354364205,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004240067153361402,
+      "loss": 1.2494,
+      "step": 9625
+    },
+    {
+      "epoch": 0.844870005019323,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0004238292159613708,
+      "loss": 1.1444,
+      "step": 9626
+    },
+    {
+      "epoch": 0.8449577746022255,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0004236518376048064,
+      "loss": 1.1882,
+      "step": 9627
+    },
+    {
+      "epoch": 0.845045544185128,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0004234745802839529,
+      "loss": 1.1626,
+      "step": 9628
+    },
+    {
+      "epoch": 0.8451333137680305,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00042329744401630506,
+      "loss": 1.1524,
+      "step": 9629
+    },
+    {
+      "epoch": 0.8452210833509329,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0004231204288193459,
+      "loss": 1.1448,
+      "step": 9630
+    },
+    {
+      "epoch": 0.8453088529338354,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0004229435347105461,
+      "loss": 1.1478,
+      "step": 9631
+    },
+    {
+      "epoch": 0.845396622516738,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00042276676170736383,
+      "loss": 1.1319,
+      "step": 9632
+    },
+    {
+      "epoch": 0.8454843920996404,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0004225901098272464,
+      "loss": 1.1704,
+      "step": 9633
+    },
+    {
+      "epoch": 0.8455721616825429,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00042241357908762855,
+      "loss": 1.1539,
+      "step": 9634
+    },
+    {
+      "epoch": 0.8456599312654454,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00042223716950593295,
+      "loss": 1.1799,
+      "step": 9635
+    },
+    {
+      "epoch": 0.8457477008483478,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00042206088109957083,
+      "loss": 1.166,
+      "step": 9636
+    },
+    {
+      "epoch": 0.8458354704312504,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00042188471388594095,
+      "loss": 1.1075,
+      "step": 9637
+    },
+    {
+      "epoch": 0.8459232400141529,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004217086678824304,
+      "loss": 1.1453,
+      "step": 9638
+    },
+    {
+      "epoch": 0.8460110095970553,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0004215327431064147,
+      "loss": 1.117,
+      "step": 9639
+    },
+    {
+      "epoch": 0.8460987791799578,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00042135693957525593,
+      "loss": 1.1434,
+      "step": 9640
+    },
+    {
+      "epoch": 0.8461865487628603,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00042118125730630595,
+      "loss": 1.1504,
+      "step": 9641
+    },
+    {
+      "epoch": 0.8462743183457628,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00042100569631690345,
+      "loss": 1.1546,
+      "step": 9642
+    },
+    {
+      "epoch": 0.8463620879286653,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.00042083025662437623,
+      "loss": 1.1257,
+      "step": 9643
+    },
+    {
+      "epoch": 0.8464498575115678,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0004206549382460388,
+      "loss": 1.0835,
+      "step": 9644
+    },
+    {
+      "epoch": 0.8465376270944702,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0004204797411991945,
+      "loss": 1.1674,
+      "step": 9645
+    },
+    {
+      "epoch": 0.8466253966773727,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004203046655011352,
+      "loss": 1.1332,
+      "step": 9646
+    },
+    {
+      "epoch": 0.8467131662602753,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0004201297111691397,
+      "loss": 1.1232,
+      "step": 9647
+    },
+    {
+      "epoch": 0.8468009358431777,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004199548782204751,
+      "loss": 1.1242,
+      "step": 9648
+    },
+    {
+      "epoch": 0.8468887054260802,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00041978016667239715,
+      "loss": 1.131,
+      "step": 9649
+    },
+    {
+      "epoch": 0.8469764750089827,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0004196055765421489,
+      "loss": 1.2042,
+      "step": 9650
+    },
+    {
+      "epoch": 0.8470642445918851,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00041943110784696205,
+      "loss": 1.1285,
+      "step": 9651
+    },
+    {
+      "epoch": 0.8471520141747877,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00041925676060405555,
+      "loss": 1.1546,
+      "step": 9652
+    },
+    {
+      "epoch": 0.8472397837576902,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00041908253483063706,
+      "loss": 1.1338,
+      "step": 9653
+    },
+    {
+      "epoch": 0.8473275533405926,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004189084305439019,
+      "loss": 1.1703,
+      "step": 9654
+    },
+    {
+      "epoch": 0.8474153229234951,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0004187344477610333,
+      "loss": 1.143,
+      "step": 9655
+    },
+    {
+      "epoch": 0.8475030925063975,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00041856058649920283,
+      "loss": 1.1176,
+      "step": 9656
+    },
+    {
+      "epoch": 0.8475908620893,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.00041838684677557,
+      "loss": 1.1515,
+      "step": 9657
+    },
+    {
+      "epoch": 0.8476786316722026,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0004182132286072824,
+      "loss": 1.0822,
+      "step": 9658
+    },
+    {
+      "epoch": 0.847766401255105,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0004180397320114749,
+      "loss": 1.1598,
+      "step": 9659
+    },
+    {
+      "epoch": 0.8478541708380075,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004178663570052713,
+      "loss": 1.1167,
+      "step": 9660
+    },
+    {
+      "epoch": 0.84794194042091,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.00041769310360578293,
+      "loss": 1.1354,
+      "step": 9661
+    },
+    {
+      "epoch": 0.8480297100038124,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0004175199718301092,
+      "loss": 1.1138,
+      "step": 9662
+    },
+    {
+      "epoch": 0.848117479586715,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0004173469616953379,
+      "loss": 1.1331,
+      "step": 9663
+    },
+    {
+      "epoch": 0.8482052491696175,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0004171740732185438,
+      "loss": 1.1111,
+      "step": 9664
+    },
+    {
+      "epoch": 0.84829301875252,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004170013064167904,
+      "loss": 1.1484,
+      "step": 9665
+    },
+    {
+      "epoch": 0.8483807883354224,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00041682866130712977,
+      "loss": 1.1501,
+      "step": 9666
+    },
+    {
+      "epoch": 0.8484685579183249,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00041665613790660064,
+      "loss": 1.1116,
+      "step": 9667
+    },
+    {
+      "epoch": 0.8485563275012274,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.00041648373623223066,
+      "loss": 1.2232,
+      "step": 9668
+    },
+    {
+      "epoch": 0.8486440970841299,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00041631145630103514,
+      "loss": 1.1565,
+      "step": 9669
+    },
+    {
+      "epoch": 0.8487318666670324,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00041613929813001763,
+      "loss": 1.1636,
+      "step": 9670
+    },
+    {
+      "epoch": 0.8488196362499348,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00041596726173616903,
+      "loss": 1.1711,
+      "step": 9671
+    },
+    {
+      "epoch": 0.8489074058328373,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0004157953471364687,
+      "loss": 1.1473,
+      "step": 9672
+    },
+    {
+      "epoch": 0.8489951754157399,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00041562355434788433,
+      "loss": 1.1565,
+      "step": 9673
+    },
+    {
+      "epoch": 0.8490829449986423,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0004154518833873711,
+      "loss": 1.2,
+      "step": 9674
+    },
+    {
+      "epoch": 0.8491707145815448,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0004152803342718716,
+      "loss": 1.135,
+      "step": 9675
+    },
+    {
+      "epoch": 0.8492584841644473,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004151089070183179,
+      "loss": 1.1714,
+      "step": 9676
+    },
+    {
+      "epoch": 0.8493462537473497,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00041493760164362895,
+      "loss": 1.1627,
+      "step": 9677
+    },
+    {
+      "epoch": 0.8494340233302523,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0004147664181647119,
+      "loss": 1.1647,
+      "step": 9678
+    },
+    {
+      "epoch": 0.8495217929131548,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0004145953565984616,
+      "loss": 1.1418,
+      "step": 9679
+    },
+    {
+      "epoch": 0.8496095624960572,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0004144244169617614,
+      "loss": 1.1405,
+      "step": 9680
+    },
+    {
+      "epoch": 0.8496973320789597,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00041425359927148227,
+      "loss": 1.1223,
+      "step": 9681
+    },
+    {
+      "epoch": 0.8497851016618622,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00041408290354448356,
+      "loss": 1.1526,
+      "step": 9682
+    },
+    {
+      "epoch": 0.8498728712447646,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0004139123297976117,
+      "loss": 1.1734,
+      "step": 9683
+    },
+    {
+      "epoch": 0.8499606408276672,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00041374187804770195,
+      "loss": 1.0994,
+      "step": 9684
+    },
+    {
+      "epoch": 0.8500484104105697,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0004135715483115772,
+      "loss": 1.148,
+      "step": 9685
+    },
+    {
+      "epoch": 0.8501361799934721,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00041340134060604845,
+      "loss": 1.1772,
+      "step": 9686
+    },
+    {
+      "epoch": 0.8502239495763746,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00041323125494791436,
+      "loss": 1.1098,
+      "step": 9687
+    },
+    {
+      "epoch": 0.8503117191592771,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0004130612913539616,
+      "loss": 1.1726,
+      "step": 9688
+    },
+    {
+      "epoch": 0.8503994887421796,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00041289144984096523,
+      "loss": 1.1255,
+      "step": 9689
+    },
+    {
+      "epoch": 0.8504872583250821,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004127217304256879,
+      "loss": 1.206,
+      "step": 9690
+    },
+    {
+      "epoch": 0.8505750279079846,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00041255213312488,
+      "loss": 1.1122,
+      "step": 9691
+    },
+    {
+      "epoch": 0.850662797490887,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00041238265795528003,
+      "loss": 1.1273,
+      "step": 9692
+    },
+    {
+      "epoch": 0.8507505670737895,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00041221330493361485,
+      "loss": 1.1549,
+      "step": 9693
+    },
+    {
+      "epoch": 0.8508383366566921,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0004120440740765991,
+      "loss": 1.1528,
+      "step": 9694
+    },
+    {
+      "epoch": 0.8509261062395945,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00041187496540093433,
+      "loss": 1.1006,
+      "step": 9695
+    },
+    {
+      "epoch": 0.851013875822497,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0004117059789233117,
+      "loss": 1.1029,
+      "step": 9696
+    },
+    {
+      "epoch": 0.8511016454053995,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0004115371146604096,
+      "loss": 1.0974,
+      "step": 9697
+    },
+    {
+      "epoch": 0.8511894149883019,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00041136837262889357,
+      "loss": 1.1491,
+      "step": 9698
+    },
+    {
+      "epoch": 0.8512771845712045,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004111997528454182,
+      "loss": 1.1651,
+      "step": 9699
+    },
+    {
+      "epoch": 0.851364954154107,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00041103125532662547,
+      "loss": 1.1555,
+      "step": 9700
+    },
+    {
+      "epoch": 0.8514527237370094,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00041086288008914545,
+      "loss": 1.1593,
+      "step": 9701
+    },
+    {
+      "epoch": 0.8515404933199119,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00041069462714959665,
+      "loss": 1.1407,
+      "step": 9702
+    },
+    {
+      "epoch": 0.8516282629028143,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.000410526496524584,
+      "loss": 1.1883,
+      "step": 9703
+    },
+    {
+      "epoch": 0.8517160324857169,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0004103584882307018,
+      "loss": 1.1351,
+      "step": 9704
+    },
+    {
+      "epoch": 0.8518038020686194,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.000410190602284532,
+      "loss": 1.1059,
+      "step": 9705
+    },
+    {
+      "epoch": 0.8518915716515219,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00041002283870264375,
+      "loss": 1.2351,
+      "step": 9706
+    },
+    {
+      "epoch": 0.8519793412344243,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004098551975015952,
+      "loss": 1.1519,
+      "step": 9707
+    },
+    {
+      "epoch": 0.8520671108173268,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00040968767869793156,
+      "loss": 1.101,
+      "step": 9708
+    },
+    {
+      "epoch": 0.8521548804002294,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00040952028230818667,
+      "loss": 1.1989,
+      "step": 9709
+    },
+    {
+      "epoch": 0.8522426499831318,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00040935300834888137,
+      "loss": 1.1165,
+      "step": 9710
+    },
+    {
+      "epoch": 0.8523304195660343,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.00040918585683652497,
+      "loss": 1.118,
+      "step": 9711
+    },
+    {
+      "epoch": 0.8524181891489367,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00040901882778761496,
+      "loss": 1.1265,
+      "step": 9712
+    },
+    {
+      "epoch": 0.8525059587318392,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0004088519212186365,
+      "loss": 1.147,
+      "step": 9713
+    },
+    {
+      "epoch": 0.8525937283147417,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.00040868513714606246,
+      "loss": 1.1572,
+      "step": 9714
+    },
+    {
+      "epoch": 0.8526814978976442,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0004085184755863533,
+      "loss": 1.114,
+      "step": 9715
+    },
+    {
+      "epoch": 0.8527692674805467,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0004083519365559587,
+      "loss": 1.1633,
+      "step": 9716
+    },
+    {
+      "epoch": 0.8528570370634492,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0004081855200713151,
+      "loss": 1.1615,
+      "step": 9717
+    },
+    {
+      "epoch": 0.8529448066463516,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00040801922614884674,
+      "loss": 1.1467,
+      "step": 9718
+    },
+    {
+      "epoch": 0.8530325762292541,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00040785305480496673,
+      "loss": 1.1752,
+      "step": 9719
+    },
+    {
+      "epoch": 0.8531203458121567,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0004076870060560751,
+      "loss": 1.1422,
+      "step": 9720
+    },
+    {
+      "epoch": 0.8532081153950591,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.00040752107991856053,
+      "loss": 1.0989,
+      "step": 9721
+    },
+    {
+      "epoch": 0.8532958849779616,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00040735527640879915,
+      "loss": 1.1325,
+      "step": 9722
+    },
+    {
+      "epoch": 0.8533836545608641,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00040718959554315507,
+      "loss": 1.1403,
+      "step": 9723
+    },
+    {
+      "epoch": 0.8534714241437665,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0004070240373379802,
+      "loss": 1.1822,
+      "step": 9724
+    },
+    {
+      "epoch": 0.8535591937266691,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004068586018096147,
+      "loss": 1.1511,
+      "step": 9725
+    },
+    {
+      "epoch": 0.8536469633095716,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.00040669328897438627,
+      "loss": 1.0843,
+      "step": 9726
+    },
+    {
+      "epoch": 0.853734732892474,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00040652809884861074,
+      "loss": 1.1378,
+      "step": 9727
+    },
+    {
+      "epoch": 0.8538225024753765,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0004063630314485917,
+      "loss": 1.0872,
+      "step": 9728
+    },
+    {
+      "epoch": 0.853910272058279,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.000406198086790621,
+      "loss": 1.1444,
+      "step": 9729
+    },
+    {
+      "epoch": 0.8539980416411815,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0004060332648909772,
+      "loss": 1.0751,
+      "step": 9730
+    },
+    {
+      "epoch": 0.854085811224084,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00040586856576592807,
+      "loss": 1.1012,
+      "step": 9731
+    },
+    {
+      "epoch": 0.8541735808069865,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00040570398943172887,
+      "loss": 1.1045,
+      "step": 9732
+    },
+    {
+      "epoch": 0.8542613503898889,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0004055395359046226,
+      "loss": 1.187,
+      "step": 9733
+    },
+    {
+      "epoch": 0.8543491199727914,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0004053752052008398,
+      "loss": 1.1406,
+      "step": 9734
+    },
+    {
+      "epoch": 0.854436889555694,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0004052109973365993,
+      "loss": 1.1567,
+      "step": 9735
+    },
+    {
+      "epoch": 0.8545246591385964,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0004050469123281085,
+      "loss": 1.1348,
+      "step": 9736
+    },
+    {
+      "epoch": 0.8546124287214989,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00040488295019156114,
+      "loss": 1.1389,
+      "step": 9737
+    },
+    {
+      "epoch": 0.8547001983044014,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00040471911094313993,
+      "loss": 1.171,
+      "step": 9738
+    },
+    {
+      "epoch": 0.8547879678873038,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00040455539459901516,
+      "loss": 1.1337,
+      "step": 9739
+    },
+    {
+      "epoch": 0.8548757374702063,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0004043918011753449,
+      "loss": 1.1162,
+      "step": 9740
+    },
+    {
+      "epoch": 0.8549635070531089,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00040422833068827556,
+      "loss": 1.1309,
+      "step": 9741
+    },
+    {
+      "epoch": 0.8550512766360113,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00040406498315394046,
+      "loss": 1.1017,
+      "step": 9742
+    },
+    {
+      "epoch": 0.8551390462189138,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0004039017585884615,
+      "loss": 1.1218,
+      "step": 9743
+    },
+    {
+      "epoch": 0.8552268158018163,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00040373865700794865,
+      "loss": 1.1559,
+      "step": 9744
+    },
+    {
+      "epoch": 0.8553145853847187,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0004035756784284989,
+      "loss": 1.163,
+      "step": 9745
+    },
+    {
+      "epoch": 0.8554023549676213,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0004034128228661979,
+      "loss": 1.1836,
+      "step": 9746
+    },
+    {
+      "epoch": 0.8554901245505238,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0004032500903371188,
+      "loss": 1.1171,
+      "step": 9747
+    },
+    {
+      "epoch": 0.8555778941334262,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00040308748085732287,
+      "loss": 1.1633,
+      "step": 9748
+    },
+    {
+      "epoch": 0.8556656637163287,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.00040292499444285866,
+      "loss": 1.1629,
+      "step": 9749
+    },
+    {
+      "epoch": 0.8557534332992311,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0004027626311097629,
+      "loss": 1.1306,
+      "step": 9750
+    },
+    {
+      "epoch": 0.8558412028821337,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0004026003908740605,
+      "loss": 1.1308,
+      "step": 9751
+    },
+    {
+      "epoch": 0.8559289724650362,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00040243827375176374,
+      "loss": 1.2031,
+      "step": 9752
+    },
+    {
+      "epoch": 0.8560167420479387,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00040227627975887337,
+      "loss": 1.1065,
+      "step": 9753
+    },
+    {
+      "epoch": 0.8561045116308411,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.00040211440891137684,
+      "loss": 1.0742,
+      "step": 9754
+    },
+    {
+      "epoch": 0.8561922812137436,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004019526612252505,
+      "loss": 1.1438,
+      "step": 9755
+    },
+    {
+      "epoch": 0.8562800507966462,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00040179103671645835,
+      "loss": 1.1648,
+      "step": 9756
+    },
+    {
+      "epoch": 0.8563678203795486,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0004016295354009519,
+      "loss": 1.135,
+      "step": 9757
+    },
+    {
+      "epoch": 0.8564555899624511,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0004014681572946707,
+      "loss": 1.1191,
+      "step": 9758
+    },
+    {
+      "epoch": 0.8565433595453535,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00040130690241354215,
+      "loss": 1.1168,
+      "step": 9759
+    },
+    {
+      "epoch": 0.856631129128256,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0004011457707734818,
+      "loss": 1.1147,
+      "step": 9760
+    },
+    {
+      "epoch": 0.8567188987111586,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0004009847623903922,
+      "loss": 1.1237,
+      "step": 9761
+    },
+    {
+      "epoch": 0.856806668294061,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0004008238772801645,
+      "loss": 1.147,
+      "step": 9762
+    },
+    {
+      "epoch": 0.8568944378769635,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0004006631154586771,
+      "loss": 1.1555,
+      "step": 9763
+    },
+    {
+      "epoch": 0.856982207459866,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0004005024769417971,
+      "loss": 1.1799,
+      "step": 9764
+    },
+    {
+      "epoch": 0.8570699770427684,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0004003419617453788,
+      "loss": 1.1782,
+      "step": 9765
+    },
+    {
+      "epoch": 0.8571577466256709,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0004001815698852637,
+      "loss": 1.1383,
+      "step": 9766
+    },
+    {
+      "epoch": 0.8572455162085735,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0004000213013772826,
+      "loss": 1.1146,
+      "step": 9767
+    },
+    {
+      "epoch": 0.8573332857914759,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00039986115623725355,
+      "loss": 1.1155,
+      "step": 9768
+    },
+    {
+      "epoch": 0.8574210553743784,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0003997011344809813,
+      "loss": 1.1592,
+      "step": 9769
+    },
+    {
+      "epoch": 0.8575088249572809,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0003995412361242601,
+      "loss": 1.1519,
+      "step": 9770
+    },
+    {
+      "epoch": 0.8575965945401833,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00039938146118287106,
+      "loss": 1.1014,
+      "step": 9771
+    },
+    {
+      "epoch": 0.8576843641230859,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0003992218096725835,
+      "loss": 1.1021,
+      "step": 9772
+    },
+    {
+      "epoch": 0.8577721337059884,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.000399062281609154,
+      "loss": 1.2096,
+      "step": 9773
+    },
+    {
+      "epoch": 0.8578599032888908,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0003989028770083277,
+      "loss": 1.1754,
+      "step": 9774
+    },
+    {
+      "epoch": 0.8579476728717933,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00039874359588583694,
+      "loss": 1.1583,
+      "step": 9775
+    },
+    {
+      "epoch": 0.8580354424546958,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00039858443825740243,
+      "loss": 1.148,
+      "step": 9776
+    },
+    {
+      "epoch": 0.8581232120375983,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00039842540413873224,
+      "loss": 1.1239,
+      "step": 9777
+    },
+    {
+      "epoch": 0.8582109816205008,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00039826649354552247,
+      "loss": 1.1699,
+      "step": 9778
+    },
+    {
+      "epoch": 0.8582987512034033,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0003981077064934569,
+      "loss": 1.224,
+      "step": 9779
+    },
+    {
+      "epoch": 0.8583865207863057,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00039794904299820735,
+      "loss": 1.1282,
+      "step": 9780
+    },
+    {
+      "epoch": 0.8584742903692082,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00039779050307543306,
+      "loss": 1.129,
+      "step": 9781
+    },
+    {
+      "epoch": 0.8585620599521108,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0003976320867407812,
+      "loss": 1.1876,
+      "step": 9782
+    },
+    {
+      "epoch": 0.8586498295350132,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00039747379400988716,
+      "loss": 1.1544,
+      "step": 9783
+    },
+    {
+      "epoch": 0.8587375991179157,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0003973156248983738,
+      "loss": 1.1118,
+      "step": 9784
+    },
+    {
+      "epoch": 0.8588253687008182,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00039715757942185146,
+      "loss": 1.1961,
+      "step": 9785
+    },
+    {
+      "epoch": 0.8589131382837206,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0003969996575959185,
+      "loss": 1.1813,
+      "step": 9786
+    },
+    {
+      "epoch": 0.8590009078666232,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00039684185943616193,
+      "loss": 1.1197,
+      "step": 9787
+    },
+    {
+      "epoch": 0.8590886774495257,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0003966841849581552,
+      "loss": 1.1282,
+      "step": 9788
+    },
+    {
+      "epoch": 0.8591764470324281,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00039652663417746013,
+      "loss": 1.1363,
+      "step": 9789
+    },
+    {
+      "epoch": 0.8592642166153306,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0003963692071096266,
+      "loss": 1.1603,
+      "step": 9790
+    },
+    {
+      "epoch": 0.859351986198233,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0003962119037701921,
+      "loss": 1.091,
+      "step": 9791
+    },
+    {
+      "epoch": 0.8594397557811356,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0003960547241746817,
+      "loss": 1.0957,
+      "step": 9792
+    },
+    {
+      "epoch": 0.8595275253640381,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00039589766833860834,
+      "loss": 1.2,
+      "step": 9793
+    },
+    {
+      "epoch": 0.8596152949469406,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0003957407362774728,
+      "loss": 1.124,
+      "step": 9794
+    },
+    {
+      "epoch": 0.859703064529843,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00039558392800676385,
+      "loss": 1.1312,
+      "step": 9795
+    },
+    {
+      "epoch": 0.8597908341127455,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00039542724354195755,
+      "loss": 1.1543,
+      "step": 9796
+    },
+    {
+      "epoch": 0.859878603695648,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0003952706828985183,
+      "loss": 1.154,
+      "step": 9797
+    },
+    {
+      "epoch": 0.8599663732785505,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.000395114246091898,
+      "loss": 1.1432,
+      "step": 9798
+    },
+    {
+      "epoch": 0.860054142861453,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00039495793313753635,
+      "loss": 1.1728,
+      "step": 9799
+    },
+    {
+      "epoch": 0.8601419124443554,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0003948017440508607,
+      "loss": 1.1807,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8602296820272579,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00039464567884728627,
+      "loss": 1.1476,
+      "step": 9801
+    },
+    {
+      "epoch": 0.8603174516101604,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0003944897375422162,
+      "loss": 1.1364,
+      "step": 9802
+    },
+    {
+      "epoch": 0.860405221193063,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0003943339201510411,
+      "loss": 1.1502,
+      "step": 9803
+    },
+    {
+      "epoch": 0.8604929907759654,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0003941782266891399,
+      "loss": 1.1092,
+      "step": 9804
+    },
+    {
+      "epoch": 0.8605807603588679,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00039402265717187877,
+      "loss": 1.1427,
+      "step": 9805
+    },
+    {
+      "epoch": 0.8606685299417703,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00039386721161461125,
+      "loss": 1.1597,
+      "step": 9806
+    },
+    {
+      "epoch": 0.8607562995246728,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.00039371189003268014,
+      "loss": 1.1643,
+      "step": 9807
+    },
+    {
+      "epoch": 0.8608440691075754,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00039355669244141425,
+      "loss": 1.1377,
+      "step": 9808
+    },
+    {
+      "epoch": 0.8609318386904778,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00039340161885613147,
+      "loss": 1.0785,
+      "step": 9809
+    },
+    {
+      "epoch": 0.8610196082733803,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0003932466692921367,
+      "loss": 1.1626,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8611073778562828,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00039309184376472333,
+      "loss": 1.0934,
+      "step": 9811
+    },
+    {
+      "epoch": 0.8611951474391852,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0003929371422891713,
+      "loss": 1.1034,
+      "step": 9812
+    },
+    {
+      "epoch": 0.8612829170220878,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00039278256488074943,
+      "loss": 1.1235,
+      "step": 9813
+    },
+    {
+      "epoch": 0.8613706866049903,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00039262811155471377,
+      "loss": 1.1675,
+      "step": 9814
+    },
+    {
+      "epoch": 0.8614584561878927,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00039247378232630845,
+      "loss": 1.1094,
+      "step": 9815
+    },
+    {
+      "epoch": 0.8615462257707952,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00039231957721076505,
+      "loss": 1.1157,
+      "step": 9816
+    },
+    {
+      "epoch": 0.8616339953536977,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.00039216549622330283,
+      "loss": 1.1869,
+      "step": 9817
+    },
+    {
+      "epoch": 0.8617217649366002,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0003920115393791293,
+      "loss": 1.1278,
+      "step": 9818
+    },
+    {
+      "epoch": 0.8618095345195027,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0003918577066934393,
+      "loss": 1.1371,
+      "step": 9819
+    },
+    {
+      "epoch": 0.8618973041024052,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00039170399818141526,
+      "loss": 1.1693,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8619850736853076,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00039155041385822786,
+      "loss": 1.1107,
+      "step": 9821
+    },
+    {
+      "epoch": 0.8620728432682101,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.000391396953739035,
+      "loss": 1.103,
+      "step": 9822
+    },
+    {
+      "epoch": 0.8621606128511126,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00039124361783898315,
+      "loss": 1.1765,
+      "step": 9823
+    },
+    {
+      "epoch": 0.8622483824340151,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00039109040617320534,
+      "loss": 1.1602,
+      "step": 9824
+    },
+    {
+      "epoch": 0.8623361520169176,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00039093731875682313,
+      "loss": 1.1121,
+      "step": 9825
+    },
+    {
+      "epoch": 0.8624239215998201,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00039078435560494537,
+      "loss": 1.1673,
+      "step": 9826
+    },
+    {
+      "epoch": 0.8625116911827225,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0003906315167326698,
+      "loss": 1.1765,
+      "step": 9827
+    },
+    {
+      "epoch": 0.862599460765625,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00039047880215508035,
+      "loss": 1.098,
+      "step": 9828
+    },
+    {
+      "epoch": 0.8626872303485276,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0003903262118872493,
+      "loss": 1.0526,
+      "step": 9829
+    },
+    {
+      "epoch": 0.86277499993143,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00039017374594423683,
+      "loss": 1.1724,
+      "step": 9830
+    },
+    {
+      "epoch": 0.8628627695143325,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00039002140434109115,
+      "loss": 1.1691,
+      "step": 9831
+    },
+    {
+      "epoch": 0.862950539097235,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00038986918709284706,
+      "loss": 1.1355,
+      "step": 9832
+    },
+    {
+      "epoch": 0.8630383086801374,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0003897170942145279,
+      "loss": 1.1531,
+      "step": 9833
+    },
+    {
+      "epoch": 0.86312607826304,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00038956512572114517,
+      "loss": 1.1456,
+      "step": 9834
+    },
+    {
+      "epoch": 0.8632138478459425,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0003894132816276974,
+      "loss": 1.1688,
+      "step": 9835
+    },
+    {
+      "epoch": 0.8633016174288449,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0003892615619491704,
+      "loss": 1.1386,
+      "step": 9836
+    },
+    {
+      "epoch": 0.8633893870117474,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0003891099667005389,
+      "loss": 1.1209,
+      "step": 9837
+    },
+    {
+      "epoch": 0.8634771565946499,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00038895849589676474,
+      "loss": 1.0883,
+      "step": 9838
+    },
+    {
+      "epoch": 0.8635649261775524,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0003888071495527973,
+      "loss": 1.1014,
+      "step": 9839
+    },
+    {
+      "epoch": 0.8636526957604549,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00038865592768357385,
+      "loss": 1.1076,
+      "step": 9840
+    },
+    {
+      "epoch": 0.8637404653433574,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00038850483030401935,
+      "loss": 1.1548,
+      "step": 9841
+    },
+    {
+      "epoch": 0.8638282349262598,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00038835385742904663,
+      "loss": 1.121,
+      "step": 9842
+    },
+    {
+      "epoch": 0.8639160045091623,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00038820300907355646,
+      "loss": 1.1375,
+      "step": 9843
+    },
+    {
+      "epoch": 0.8640037740920649,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00038805228525243637,
+      "loss": 1.1382,
+      "step": 9844
+    },
+    {
+      "epoch": 0.8640915436749673,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00038790168598056235,
+      "loss": 1.1999,
+      "step": 9845
+    },
+    {
+      "epoch": 0.8641793132578698,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0003877512112727982,
+      "loss": 1.1439,
+      "step": 9846
+    },
+    {
+      "epoch": 0.8642670828407722,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0003876008611439948,
+      "loss": 1.0809,
+      "step": 9847
+    },
+    {
+      "epoch": 0.8643548524236747,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0003874506356089915,
+      "loss": 1.0693,
+      "step": 9848
+    },
+    {
+      "epoch": 0.8644426220065773,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00038730053468261484,
+      "loss": 1.1688,
+      "step": 9849
+    },
+    {
+      "epoch": 0.8645303915894798,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00038715055837967925,
+      "loss": 1.1314,
+      "step": 9850
+    },
+    {
+      "epoch": 0.8646181611723822,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.00038700070671498653,
+      "loss": 1.1156,
+      "step": 9851
+    },
+    {
+      "epoch": 0.8647059307552847,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0003868509797033266,
+      "loss": 1.2144,
+      "step": 9852
+    },
+    {
+      "epoch": 0.8647937003381871,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0003867013773594771,
+      "loss": 1.1408,
+      "step": 9853
+    },
+    {
+      "epoch": 0.8648814699210896,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0003865518996982029,
+      "loss": 1.2124,
+      "step": 9854
+    },
+    {
+      "epoch": 0.8649692395039922,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0003864025467342573,
+      "loss": 1.1662,
+      "step": 9855
+    },
+    {
+      "epoch": 0.8650570090868946,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00038625331848238,
+      "loss": 1.1644,
+      "step": 9856
+    },
+    {
+      "epoch": 0.8651447786697971,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00038610421495729997,
+      "loss": 1.1365,
+      "step": 9857
+    },
+    {
+      "epoch": 0.8652325482526996,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00038595523617373313,
+      "loss": 1.1616,
+      "step": 9858
+    },
+    {
+      "epoch": 0.865320317835602,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0003858063821463829,
+      "loss": 1.1559,
+      "step": 9859
+    },
+    {
+      "epoch": 0.8654080874185046,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00038565765288994043,
+      "loss": 1.1517,
+      "step": 9860
+    },
+    {
+      "epoch": 0.8654958570014071,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00038550904841908496,
+      "loss": 1.1731,
+      "step": 9861
+    },
+    {
+      "epoch": 0.8655836265843095,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00038536056874848305,
+      "loss": 1.1627,
+      "step": 9862
+    },
+    {
+      "epoch": 0.865671396167212,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0003852122138927891,
+      "loss": 1.1719,
+      "step": 9863
+    },
+    {
+      "epoch": 0.8657591657501145,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.00038506398386664494,
+      "loss": 1.173,
+      "step": 9864
+    },
+    {
+      "epoch": 0.865846935333017,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0003849158786846807,
+      "loss": 1.1164,
+      "step": 9865
+    },
+    {
+      "epoch": 0.8659347049159195,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00038476789836151343,
+      "loss": 1.152,
+      "step": 9866
+    },
+    {
+      "epoch": 0.866022474498822,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00038462004291174833,
+      "loss": 1.1,
+      "step": 9867
+    },
+    {
+      "epoch": 0.8661102440817244,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0003844723123499781,
+      "loss": 1.1117,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8661980136646269,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00038432470669078333,
+      "loss": 1.1363,
+      "step": 9869
+    },
+    {
+      "epoch": 0.8662857832475295,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0003841772259487321,
+      "loss": 1.123,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8663735528304319,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0003840298701383797,
+      "loss": 1.1536,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8664613224133344,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0003838826392742702,
+      "loss": 1.193,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8665490919962369,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0003837355333709343,
+      "loss": 1.2064,
+      "step": 9873
+    },
+    {
+      "epoch": 0.8666368615791393,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.000383588552442891,
+      "loss": 1.1672,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8667246311620419,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00038344169650464654,
+      "loss": 1.1275,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8668124007449444,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00038329496557069513,
+      "loss": 1.1471,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8669001703278468,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0003831483596555183,
+      "loss": 1.2019,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8669879399107493,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00038300187877358624,
+      "loss": 1.1238,
+      "step": 9878
+    },
+    {
+      "epoch": 0.8670757094936518,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.000382855522939355,
+      "loss": 1.1119,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8671634790765542,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00038270929216727004,
+      "loss": 1.1416,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8672512486594568,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0003825631864717634,
+      "loss": 1.1716,
+      "step": 9881
+    },
+    {
+      "epoch": 0.8673390182423593,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0003824172058672554,
+      "loss": 1.1687,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8674267878252617,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00038227135036815354,
+      "loss": 1.1712,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8675145574081642,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0003821256199888533,
+      "loss": 1.1382,
+      "step": 9884
+    },
+    {
+      "epoch": 0.8676023269910667,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0003819800147437375,
+      "loss": 1.1416,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8676900965739692,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00038183453464717723,
+      "loss": 1.1258,
+      "step": 9886
+    },
+    {
+      "epoch": 0.8677778661568717,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00038168917971353013,
+      "loss": 1.2058,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8678656357397742,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00038154394995714265,
+      "loss": 1.1157,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8679534053226766,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00038139884539234854,
+      "loss": 1.1964,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8680411749055791,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00038125386603346853,
+      "loss": 1.1316,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8681289444884817,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0003811090118948119,
+      "loss": 1.1649,
+      "step": 9891
+    },
+    {
+      "epoch": 0.8682167140713841,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0003809642829906748,
+      "loss": 1.2118,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8683044836542866,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00038081967933534185,
+      "loss": 1.0604,
+      "step": 9893
+    },
+    {
+      "epoch": 0.868392253237189,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0003806752009430849,
+      "loss": 1.1295,
+      "step": 9894
+    },
+    {
+      "epoch": 0.8684800228200915,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0003805308478281628,
+      "loss": 1.1527,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8685677924029941,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.000380386620004823,
+      "loss": 1.1826,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8686555619858966,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0003802425174873002,
+      "loss": 1.1031,
+      "step": 9897
+    },
+    {
+      "epoch": 0.868743331568799,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0003800985402898169,
+      "loss": 1.1868,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8688311011517015,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0003799546884265828,
+      "loss": 1.1689,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8689188707346039,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00037981096191179567,
+      "loss": 1.1208,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8690066403175065,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0003796673607596408,
+      "loss": 1.1898,
+      "step": 9901
+    },
+    {
+      "epoch": 0.869094409900409,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0003795238849842908,
+      "loss": 1.1445,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8691821794833114,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00037938053459990646,
+      "loss": 1.1533,
+      "step": 9903
+    },
+    {
+      "epoch": 0.8692699490662139,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0003792373096206357,
+      "loss": 1.1398,
+      "step": 9904
+    },
+    {
+      "epoch": 0.8693577186491164,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00037909421006061437,
+      "loss": 1.1296,
+      "step": 9905
+    },
+    {
+      "epoch": 0.8694454882320188,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0003789512359339659,
+      "loss": 1.1569,
+      "step": 9906
+    },
+    {
+      "epoch": 0.8695332578149214,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00037880838725480077,
+      "loss": 1.1752,
+      "step": 9907
+    },
+    {
+      "epoch": 0.8696210273978239,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00037866566403721826,
+      "loss": 1.1446,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8697087969807263,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00037852306629530445,
+      "loss": 1.2035,
+      "step": 9909
+    },
+    {
+      "epoch": 0.8697965665636288,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00037838059404313276,
+      "loss": 1.1532,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8698843361465313,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0003782382472947649,
+      "loss": 1.1863,
+      "step": 9911
+    },
+    {
+      "epoch": 0.8699721057294338,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00037809602606425,
+      "loss": 1.1602,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8700598753123363,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0003779539303656247,
+      "loss": 1.128,
+      "step": 9913
+    },
+    {
+      "epoch": 0.8701476448952388,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00037781196021291323,
+      "loss": 1.1514,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8702354144781412,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0003776701156201274,
+      "loss": 1.0894,
+      "step": 9915
+    },
+    {
+      "epoch": 0.8703231840610437,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0003775283966012667,
+      "loss": 1.1332,
+      "step": 9916
+    },
+    {
+      "epoch": 0.8704109536439463,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00037738680317031856,
+      "loss": 1.1748,
+      "step": 9917
+    },
+    {
+      "epoch": 0.8704987232268487,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0003772453353412573,
+      "loss": 1.1544,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8705864928097512,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0003771039931280455,
+      "loss": 1.1327,
+      "step": 9919
+    },
+    {
+      "epoch": 0.8706742623926537,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0003769627765446329,
+      "loss": 1.1258,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8707620319755561,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0003768216856049574,
+      "loss": 1.1451,
+      "step": 9921
+    },
+    {
+      "epoch": 0.8708498015584587,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.00037668072032294355,
+      "loss": 1.2127,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8709375711413612,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0003765398807125043,
+      "loss": 1.1144,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8710253407242636,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0003763991667875401,
+      "loss": 1.136,
+      "step": 9924
+    },
+    {
+      "epoch": 0.8711131103071661,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00037625857856193884,
+      "loss": 1.1095,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8712008798900686,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00037611811604957587,
+      "loss": 1.0732,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8712886494729711,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.000375977779264314,
+      "loss": 1.1608,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8713764190558736,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00037583756822000454,
+      "loss": 1.1846,
+      "step": 9928
+    },
+    {
+      "epoch": 0.8714641886387761,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00037569748293048567,
+      "loss": 1.1692,
+      "step": 9929
+    },
+    {
+      "epoch": 0.8715519582216785,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00037555752340958303,
+      "loss": 1.1975,
+      "step": 9930
+    },
+    {
+      "epoch": 0.871639727804581,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0003754176896711099,
+      "loss": 1.1741,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8717274973874836,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00037527798172886756,
+      "loss": 1.144,
+      "step": 9932
+    },
+    {
+      "epoch": 0.871815266970386,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.00037513839959664474,
+      "loss": 1.0925,
+      "step": 9933
+    },
+    {
+      "epoch": 0.8719030365532885,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0003749989432882175,
+      "loss": 1.1693,
+      "step": 9934
+    },
+    {
+      "epoch": 0.871990806136191,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0003748596128173496,
+      "loss": 1.1317,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8720785757190934,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00037472040819779234,
+      "loss": 1.0846,
+      "step": 9936
+    },
+    {
+      "epoch": 0.8721663453019959,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00037458132944328474,
+      "loss": 1.123,
+      "step": 9937
+    },
+    {
+      "epoch": 0.8722541148848985,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0003744423765675535,
+      "loss": 1.1483,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8723418844678009,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0003743035495843124,
+      "loss": 1.1288,
+      "step": 9939
+    },
+    {
+      "epoch": 0.8724296540507034,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00037416484850726366,
+      "loss": 1.1846,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8725174236336058,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0003740262733500957,
+      "loss": 1.2011,
+      "step": 9941
+    },
+    {
+      "epoch": 0.8726051932165083,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.000373887824126486,
+      "loss": 1.1778,
+      "step": 9942
+    },
+    {
+      "epoch": 0.8726929627994109,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00037374950085009863,
+      "loss": 1.2031,
+      "step": 9943
+    },
+    {
+      "epoch": 0.8727807323823134,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00037361130353458574,
+      "loss": 1.1577,
+      "step": 9944
+    },
+    {
+      "epoch": 0.8728685019652158,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00037347323219358693,
+      "loss": 1.1761,
+      "step": 9945
+    },
+    {
+      "epoch": 0.8729562715481183,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.000373335286840729,
+      "loss": 1.1342,
+      "step": 9946
+    },
+    {
+      "epoch": 0.8730440411310207,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00037319746748962646,
+      "loss": 1.1175,
+      "step": 9947
+    },
+    {
+      "epoch": 0.8731318107139233,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0003730597741538822,
+      "loss": 1.1262,
+      "step": 9948
+    },
+    {
+      "epoch": 0.8732195802968258,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00037292220684708555,
+      "loss": 1.1144,
+      "step": 9949
+    },
+    {
+      "epoch": 0.8733073498797282,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00037278476558281386,
+      "loss": 1.1561,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8733951194626307,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00037264745037463205,
+      "loss": 1.1134,
+      "step": 9951
+    },
+    {
+      "epoch": 0.8734828890455332,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00037251026123609303,
+      "loss": 1.2114,
+      "step": 9952
+    },
+    {
+      "epoch": 0.8735706586284357,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0003723731981807362,
+      "loss": 1.1304,
+      "step": 9953
+    },
+    {
+      "epoch": 0.8736584282113382,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00037223626122208933,
+      "loss": 1.1971,
+      "step": 9954
+    },
+    {
+      "epoch": 0.8737461977942407,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0003720994503736676,
+      "loss": 1.0968,
+      "step": 9955
+    },
+    {
+      "epoch": 0.8738339673771431,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00037196276564897384,
+      "loss": 1.1628,
+      "step": 9956
+    },
+    {
+      "epoch": 0.8739217369600456,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.000371826207061498,
+      "loss": 1.1351,
+      "step": 9957
+    },
+    {
+      "epoch": 0.8740095065429482,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0003716897746247183,
+      "loss": 1.1524,
+      "step": 9958
+    },
+    {
+      "epoch": 0.8740972761258506,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0003715534683520997,
+      "loss": 1.1715,
+      "step": 9959
+    },
+    {
+      "epoch": 0.8741850457087531,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0003714172882570953,
+      "loss": 1.1306,
+      "step": 9960
+    },
+    {
+      "epoch": 0.8742728152916556,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0003712812343531454,
+      "loss": 1.2196,
+      "step": 9961
+    },
+    {
+      "epoch": 0.874360584874558,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00037114530665367795,
+      "loss": 1.1467,
+      "step": 9962
+    },
+    {
+      "epoch": 0.8744483544574605,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0003710095051721086,
+      "loss": 1.116,
+      "step": 9963
+    },
+    {
+      "epoch": 0.8745361240403631,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.00037087382992184053,
+      "loss": 1.1046,
+      "step": 9964
+    },
+    {
+      "epoch": 0.8746238936232655,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0003707382809162641,
+      "loss": 1.1149,
+      "step": 9965
+    },
+    {
+      "epoch": 0.874711663206168,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0003706028581687575,
+      "loss": 1.1651,
+      "step": 9966
+    },
+    {
+      "epoch": 0.8747994327890705,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0003704675616926862,
+      "loss": 1.1624,
+      "step": 9967
+    },
+    {
+      "epoch": 0.8748872023719729,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0003703323915014043,
+      "loss": 1.1245,
+      "step": 9968
+    },
+    {
+      "epoch": 0.8749749719548755,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00037019734760825154,
+      "loss": 1.114,
+      "step": 9969
+    },
+    {
+      "epoch": 0.875062741537778,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0003700624300265568,
+      "loss": 1.1565,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8751505111206804,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0003699276387696357,
+      "loss": 1.1222,
+      "step": 9971
+    },
+    {
+      "epoch": 0.8752382807035829,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.00036979297385079186,
+      "loss": 1.1451,
+      "step": 9972
+    },
+    {
+      "epoch": 0.8753260502864854,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0003696584352833158,
+      "loss": 1.162,
+      "step": 9973
+    },
+    {
+      "epoch": 0.8754138198693879,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0003695240230804862,
+      "loss": 1.1655,
+      "step": 9974
+    },
+    {
+      "epoch": 0.8755015894522904,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0003693897372555688,
+      "loss": 1.125,
+      "step": 9975
+    },
+    {
+      "epoch": 0.8755893590351929,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0003692555778218175,
+      "loss": 1.1609,
+      "step": 9976
+    },
+    {
+      "epoch": 0.8756771286180953,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0003691215447924726,
+      "loss": 1.1157,
+      "step": 9977
+    },
+    {
+      "epoch": 0.8757648982009978,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00036898763818076315,
+      "loss": 1.1267,
+      "step": 9978
+    },
+    {
+      "epoch": 0.8758526677839004,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0003688538579999053,
+      "loss": 1.099,
+      "step": 9979
+    },
+    {
+      "epoch": 0.8759404373668028,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0003687202042631025,
+      "loss": 1.1694,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8760282069497053,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0003685866769835455,
+      "loss": 1.1847,
+      "step": 9981
+    },
+    {
+      "epoch": 0.8761159765326078,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0003684532761744132,
+      "loss": 1.1332,
+      "step": 9982
+    },
+    {
+      "epoch": 0.8762037461155102,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00036832000184887175,
+      "loss": 1.2232,
+      "step": 9983
+    },
+    {
+      "epoch": 0.8762915156984128,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0003681868540200751,
+      "loss": 1.1645,
+      "step": 9984
+    },
+    {
+      "epoch": 0.8763792852813153,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00036805383270116374,
+      "loss": 1.1652,
+      "step": 9985
+    },
+    {
+      "epoch": 0.8764670548642177,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00036792093790526673,
+      "loss": 1.1228,
+      "step": 9986
+    },
+    {
+      "epoch": 0.8765548244471202,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0003677881696455001,
+      "loss": 1.0418,
+      "step": 9987
+    },
+    {
+      "epoch": 0.8766425940300226,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00036765552793496787,
+      "loss": 1.165,
+      "step": 9988
+    },
+    {
+      "epoch": 0.8767303636129252,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0003675230127867609,
+      "loss": 1.1381,
+      "step": 9989
+    },
+    {
+      "epoch": 0.8768181331958277,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0003673906242139581,
+      "loss": 1.1614,
+      "step": 9990
+    },
+    {
+      "epoch": 0.8769059027787302,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0003672583622296259,
+      "loss": 1.129,
+      "step": 9991
+    },
+    {
+      "epoch": 0.8769936723616326,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0003671262268468177,
+      "loss": 1.1133,
+      "step": 9992
+    },
+    {
+      "epoch": 0.8770814419445351,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0003669942180785746,
+      "loss": 1.123,
+      "step": 9993
+    },
+    {
+      "epoch": 0.8771692115274375,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00036686233593792567,
+      "loss": 1.1622,
+      "step": 9994
+    },
+    {
+      "epoch": 0.8772569811103401,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00036673058043788705,
+      "loss": 1.134,
+      "step": 9995
+    },
+    {
+      "epoch": 0.8773447506932426,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00036659895159146273,
+      "loss": 1.1804,
+      "step": 9996
+    },
+    {
+      "epoch": 0.877432520276145,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00036646744941164336,
+      "loss": 1.1165,
+      "step": 9997
+    },
+    {
+      "epoch": 0.8775202898590475,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00036633607391140823,
+      "loss": 1.0902,
+      "step": 9998
+    },
+    {
+      "epoch": 0.87760805944195,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00036620482510372354,
+      "loss": 1.1392,
+      "step": 9999
+    },
+    {
+      "epoch": 0.8776958290248525,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0003660737030015428,
+      "loss": 1.1464,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8776958290248525,
+      "eval_loss": 1.1333048343658447,
+      "eval_runtime": 437.564,
+      "eval_samples_per_second": 33.687,
+      "eval_steps_per_second": 8.422,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 11000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.948783792128e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}