diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,28868 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 4118,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00024283632831471587,
+      "grad_norm": 2.3099043369293213,
+      "learning_rate": 0.0001,
+      "loss": 2.074,
+      "step": 1
+    },
+    {
+      "epoch": 0.00048567265662943174,
+      "grad_norm": 3.8508684635162354,
+      "learning_rate": 0.0001,
+      "loss": 2.0514,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007285089849441476,
+      "grad_norm": 1.006288766860962,
+      "learning_rate": 0.0001,
+      "loss": 1.9108,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009713453132588635,
+      "grad_norm": 1.1353191137313843,
+      "learning_rate": 0.0001,
+      "loss": 1.9215,
+      "step": 4
+    },
+    {
+      "epoch": 0.0012141816415735794,
+      "grad_norm": 0.7446403503417969,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 5
+    },
+    {
+      "epoch": 0.0014570179698882952,
+      "grad_norm": 0.8533063530921936,
+      "learning_rate": 0.0001,
+      "loss": 1.8683,
+      "step": 6
+    },
+    {
+      "epoch": 0.001699854298203011,
+      "grad_norm": 0.8186442852020264,
+      "learning_rate": 0.0001,
+      "loss": 2.0013,
+      "step": 7
+    },
+    {
+      "epoch": 0.001942690626517727,
+      "grad_norm": 0.744153618812561,
+      "learning_rate": 0.0001,
+      "loss": 2.0731,
+      "step": 8
+    },
+    {
+      "epoch": 0.002185526954832443,
+      "grad_norm": 0.7234266400337219,
+      "learning_rate": 0.0001,
+      "loss": 1.8693,
+      "step": 9
+    },
+    {
+      "epoch": 0.0024283632831471587,
+      "grad_norm": 0.6968668699264526,
+      "learning_rate": 0.0001,
+      "loss": 1.8265,
+      "step": 10
+    },
+    {
+      "epoch": 0.0026711996114618746,
+      "grad_norm": 0.6370717287063599,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 11
+    },
+    {
+      "epoch": 0.0029140359397765905,
+      "grad_norm": 0.6973312497138977,
+      "learning_rate": 0.0001,
+      "loss": 1.9457,
+      "step": 12
+    },
+    {
+      "epoch": 0.0031568722680913063,
+      "grad_norm": 0.658778965473175,
+      "learning_rate": 0.0001,
+      "loss": 1.8521,
+      "step": 13
+    },
+    {
+      "epoch": 0.003399708596406022,
+      "grad_norm": 0.6421470642089844,
+      "learning_rate": 0.0001,
+      "loss": 1.818,
+      "step": 14
+    },
+    {
+      "epoch": 0.003642544924720738,
+      "grad_norm": 0.6314427852630615,
+      "learning_rate": 0.0001,
+      "loss": 1.9109,
+      "step": 15
+    },
+    {
+      "epoch": 0.003885381253035454,
+      "grad_norm": 0.6784456372261047,
+      "learning_rate": 0.0001,
+      "loss": 1.9048,
+      "step": 16
+    },
+    {
+      "epoch": 0.00412821758135017,
+      "grad_norm": 0.6112809777259827,
+      "learning_rate": 0.0001,
+      "loss": 1.7607,
+      "step": 17
+    },
+    {
+      "epoch": 0.004371053909664886,
+      "grad_norm": 0.639802873134613,
+      "learning_rate": 0.0001,
+      "loss": 1.8273,
+      "step": 18
+    },
+    {
+      "epoch": 0.004613890237979602,
+      "grad_norm": 0.618982195854187,
+      "learning_rate": 0.0001,
+      "loss": 1.8666,
+      "step": 19
+    },
+    {
+      "epoch": 0.0048567265662943174,
+      "grad_norm": 0.667643666267395,
+      "learning_rate": 0.0001,
+      "loss": 1.8595,
+      "step": 20
+    },
+    {
+      "epoch": 0.005099562894609034,
+      "grad_norm": 0.6735926270484924,
+      "learning_rate": 0.0001,
+      "loss": 2.0393,
+      "step": 21
+    },
+    {
+      "epoch": 0.005342399222923749,
+      "grad_norm": 0.6211016774177551,
+      "learning_rate": 0.0001,
+      "loss": 1.9345,
+      "step": 22
+    },
+    {
+      "epoch": 0.0055852355512384655,
+      "grad_norm": 0.6592899560928345,
+      "learning_rate": 0.0001,
+      "loss": 1.887,
+      "step": 23
+    },
+    {
+      "epoch": 0.005828071879553181,
+      "grad_norm": 0.6509709358215332,
+      "learning_rate": 0.0001,
+      "loss": 1.6733,
+      "step": 24
+    },
+    {
+      "epoch": 0.006070908207867897,
+      "grad_norm": 0.6006460785865784,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 25
+    },
+    {
+      "epoch": 0.006313744536182613,
+      "grad_norm": 0.6942538022994995,
+      "learning_rate": 0.0001,
+      "loss": 1.875,
+      "step": 26
+    },
+    {
+      "epoch": 0.006556580864497329,
+      "grad_norm": 0.620377779006958,
+      "learning_rate": 0.0001,
+      "loss": 1.8469,
+      "step": 27
+    },
+    {
+      "epoch": 0.006799417192812044,
+      "grad_norm": 0.6397677659988403,
+      "learning_rate": 0.0001,
+      "loss": 1.9733,
+      "step": 28
+    },
+    {
+      "epoch": 0.007042253521126761,
+      "grad_norm": 0.592271089553833,
+      "learning_rate": 0.0001,
+      "loss": 1.8286,
+      "step": 29
+    },
+    {
+      "epoch": 0.007285089849441476,
+      "grad_norm": 0.6396237015724182,
+      "learning_rate": 0.0001,
+      "loss": 1.9415,
+      "step": 30
+    },
+    {
+      "epoch": 0.0075279261777561925,
+      "grad_norm": 0.6379568576812744,
+      "learning_rate": 0.0001,
+      "loss": 1.851,
+      "step": 31
+    },
+    {
+      "epoch": 0.007770762506070908,
+      "grad_norm": 0.5826771855354309,
+      "learning_rate": 0.0001,
+      "loss": 1.6987,
+      "step": 32
+    },
+    {
+      "epoch": 0.008013598834385623,
+      "grad_norm": 0.5559223890304565,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 33
+    },
+    {
+      "epoch": 0.00825643516270034,
+      "grad_norm": 0.6177660226821899,
+      "learning_rate": 0.0001,
+      "loss": 1.957,
+      "step": 34
+    },
+    {
+      "epoch": 0.008499271491015056,
+      "grad_norm": 0.6030952334403992,
+      "learning_rate": 0.0001,
+      "loss": 1.7835,
+      "step": 35
+    },
+    {
+      "epoch": 0.008742107819329771,
+      "grad_norm": 0.5825766921043396,
+      "learning_rate": 0.0001,
+      "loss": 1.8372,
+      "step": 36
+    },
+    {
+      "epoch": 0.008984944147644487,
+      "grad_norm": 0.5782750248908997,
+      "learning_rate": 0.0001,
+      "loss": 1.8552,
+      "step": 37
+    },
+    {
+      "epoch": 0.009227780475959204,
+      "grad_norm": 0.6544367074966431,
+      "learning_rate": 0.0001,
+      "loss": 2.0532,
+      "step": 38
+    },
+    {
+      "epoch": 0.00947061680427392,
+      "grad_norm": 0.5874798893928528,
+      "learning_rate": 0.0001,
+      "loss": 1.7794,
+      "step": 39
+    },
+    {
+      "epoch": 0.009713453132588635,
+      "grad_norm": 0.5625186562538147,
+      "learning_rate": 0.0001,
+      "loss": 1.8103,
+      "step": 40
+    },
+    {
+      "epoch": 0.00995628946090335,
+      "grad_norm": 0.5761361122131348,
+      "learning_rate": 0.0001,
+      "loss": 1.8301,
+      "step": 41
+    },
+    {
+      "epoch": 0.010199125789218067,
+      "grad_norm": 0.5836958885192871,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 42
+    },
+    {
+      "epoch": 0.010441962117532783,
+      "grad_norm": 0.5922273397445679,
+      "learning_rate": 0.0001,
+      "loss": 1.9372,
+      "step": 43
+    },
+    {
+      "epoch": 0.010684798445847498,
+      "grad_norm": 0.6721519827842712,
+      "learning_rate": 0.0001,
+      "loss": 1.9111,
+      "step": 44
+    },
+    {
+      "epoch": 0.010927634774162214,
+      "grad_norm": 0.537930428981781,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 45
+    },
+    {
+      "epoch": 0.011170471102476931,
+      "grad_norm": 0.575630247592926,
+      "learning_rate": 0.0001,
+      "loss": 1.8267,
+      "step": 46
+    },
+    {
+      "epoch": 0.011413307430791646,
+      "grad_norm": 0.6041532754898071,
+      "learning_rate": 0.0001,
+      "loss": 2.0025,
+      "step": 47
+    },
+    {
+      "epoch": 0.011656143759106362,
+      "grad_norm": 0.6168715357780457,
+      "learning_rate": 0.0001,
+      "loss": 1.9376,
+      "step": 48
+    },
+    {
+      "epoch": 0.011898980087421079,
+      "grad_norm": 0.5767819285392761,
+      "learning_rate": 0.0001,
+      "loss": 1.8109,
+      "step": 49
+    },
+    {
+      "epoch": 0.012141816415735794,
+      "grad_norm": 0.5658189058303833,
+      "learning_rate": 0.0001,
+      "loss": 1.8165,
+      "step": 50
+    },
+    {
+      "epoch": 0.01238465274405051,
+      "grad_norm": 0.5590968132019043,
+      "learning_rate": 0.0001,
+      "loss": 1.7482,
+      "step": 51
+    },
+    {
+      "epoch": 0.012627489072365225,
+      "grad_norm": 0.5551646947860718,
+      "learning_rate": 0.0001,
+      "loss": 1.7938,
+      "step": 52
+    },
+    {
+      "epoch": 0.012870325400679943,
+      "grad_norm": 0.5823336243629456,
+      "learning_rate": 0.0001,
+      "loss": 1.8563,
+      "step": 53
+    },
+    {
+      "epoch": 0.013113161728994658,
+      "grad_norm": 0.603715181350708,
+      "learning_rate": 0.0001,
+      "loss": 1.7271,
+      "step": 54
+    },
+    {
+      "epoch": 0.013355998057309373,
+      "grad_norm": 0.657590925693512,
+      "learning_rate": 0.0001,
+      "loss": 2.0677,
+      "step": 55
+    },
+    {
+      "epoch": 0.013598834385624089,
+      "grad_norm": 0.5561121702194214,
+      "learning_rate": 0.0001,
+      "loss": 1.7211,
+      "step": 56
+    },
+    {
+      "epoch": 0.013841670713938806,
+      "grad_norm": 0.591545581817627,
+      "learning_rate": 0.0001,
+      "loss": 1.9068,
+      "step": 57
+    },
+    {
+      "epoch": 0.014084507042253521,
+      "grad_norm": 0.604915201663971,
+      "learning_rate": 0.0001,
+      "loss": 1.8007,
+      "step": 58
+    },
+    {
+      "epoch": 0.014327343370568237,
+      "grad_norm": 0.5310607552528381,
+      "learning_rate": 0.0001,
+      "loss": 1.8028,
+      "step": 59
+    },
+    {
+      "epoch": 0.014570179698882952,
+      "grad_norm": 0.5268518328666687,
+      "learning_rate": 0.0001,
+      "loss": 1.7562,
+      "step": 60
+    },
+    {
+      "epoch": 0.01481301602719767,
+      "grad_norm": 0.5638241767883301,
+      "learning_rate": 0.0001,
+      "loss": 1.7515,
+      "step": 61
+    },
+    {
+      "epoch": 0.015055852355512385,
+      "grad_norm": 0.5837224125862122,
+      "learning_rate": 0.0001,
+      "loss": 1.8159,
+      "step": 62
+    },
+    {
+      "epoch": 0.0152986886838271,
+      "grad_norm": 0.5646676421165466,
+      "learning_rate": 0.0001,
+      "loss": 1.8585,
+      "step": 63
+    },
+    {
+      "epoch": 0.015541525012141816,
+      "grad_norm": 0.5547434687614441,
+      "learning_rate": 0.0001,
+      "loss": 1.745,
+      "step": 64
+    },
+    {
+      "epoch": 0.01578436134045653,
+      "grad_norm": 0.6039091944694519,
+      "learning_rate": 0.0001,
+      "loss": 1.8769,
+      "step": 65
+    },
+    {
+      "epoch": 0.016027197668771247,
+      "grad_norm": 0.5551029443740845,
+      "learning_rate": 0.0001,
+      "loss": 1.7518,
+      "step": 66
+    },
+    {
+      "epoch": 0.016270033997085966,
+      "grad_norm": 0.5494778752326965,
+      "learning_rate": 0.0001,
+      "loss": 1.8766,
+      "step": 67
+    },
+    {
+      "epoch": 0.01651287032540068,
+      "grad_norm": 0.5915890336036682,
+      "learning_rate": 0.0001,
+      "loss": 1.7991,
+      "step": 68
+    },
+    {
+      "epoch": 0.016755706653715396,
+      "grad_norm": 0.599169909954071,
+      "learning_rate": 0.0001,
+      "loss": 2.0,
+      "step": 69
+    },
+    {
+      "epoch": 0.016998542982030112,
+      "grad_norm": 0.587382972240448,
+      "learning_rate": 0.0001,
+      "loss": 1.8536,
+      "step": 70
+    },
+    {
+      "epoch": 0.017241379310344827,
+      "grad_norm": 0.5702838897705078,
+      "learning_rate": 0.0001,
+      "loss": 1.8945,
+      "step": 71
+    },
+    {
+      "epoch": 0.017484215638659543,
+      "grad_norm": 0.5253711342811584,
+      "learning_rate": 0.0001,
+      "loss": 1.7486,
+      "step": 72
+    },
+    {
+      "epoch": 0.017727051966974258,
+      "grad_norm": 0.5934901237487793,
+      "learning_rate": 0.0001,
+      "loss": 1.7831,
+      "step": 73
+    },
+    {
+      "epoch": 0.017969888295288974,
+      "grad_norm": 0.5517677068710327,
+      "learning_rate": 0.0001,
+      "loss": 1.9623,
+      "step": 74
+    },
+    {
+      "epoch": 0.018212724623603693,
+      "grad_norm": 0.6004043221473694,
+      "learning_rate": 0.0001,
+      "loss": 2.0,
+      "step": 75
+    },
+    {
+      "epoch": 0.018455560951918408,
+      "grad_norm": 0.5316723585128784,
+      "learning_rate": 0.0001,
+      "loss": 1.7527,
+      "step": 76
+    },
+    {
+      "epoch": 0.018698397280233123,
+      "grad_norm": 0.5355283617973328,
+      "learning_rate": 0.0001,
+      "loss": 1.7888,
+      "step": 77
+    },
+    {
+      "epoch": 0.01894123360854784,
+      "grad_norm": 0.5570546388626099,
+      "learning_rate": 0.0001,
+      "loss": 1.6755,
+      "step": 78
+    },
+    {
+      "epoch": 0.019184069936862554,
+      "grad_norm": 0.5318782925605774,
+      "learning_rate": 0.0001,
+      "loss": 1.7192,
+      "step": 79
+    },
+    {
+      "epoch": 0.01942690626517727,
+      "grad_norm": 0.5251316428184509,
+      "learning_rate": 0.0001,
+      "loss": 1.7928,
+      "step": 80
+    },
+    {
+      "epoch": 0.019669742593491985,
+      "grad_norm": 0.5667200088500977,
+      "learning_rate": 0.0001,
+      "loss": 1.8848,
+      "step": 81
+    },
+    {
+      "epoch": 0.0199125789218067,
+      "grad_norm": 0.5833965539932251,
+      "learning_rate": 0.0001,
+      "loss": 2.0491,
+      "step": 82
+    },
+    {
+      "epoch": 0.02015541525012142,
+      "grad_norm": 0.5401152968406677,
+      "learning_rate": 0.0001,
+      "loss": 1.7854,
+      "step": 83
+    },
+    {
+      "epoch": 0.020398251578436135,
+      "grad_norm": 0.5811483263969421,
+      "learning_rate": 0.0001,
+      "loss": 1.8976,
+      "step": 84
+    },
+    {
+      "epoch": 0.02064108790675085,
+      "grad_norm": 0.550742506980896,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 85
+    },
+    {
+      "epoch": 0.020883924235065566,
+      "grad_norm": 0.5710833668708801,
+      "learning_rate": 0.0001,
+      "loss": 1.8834,
+      "step": 86
+    },
+    {
+      "epoch": 0.02112676056338028,
+      "grad_norm": 0.5878604650497437,
+      "learning_rate": 0.0001,
+      "loss": 1.8921,
+      "step": 87
+    },
+    {
+      "epoch": 0.021369596891694997,
+      "grad_norm": 0.5854058861732483,
+      "learning_rate": 0.0001,
+      "loss": 1.8416,
+      "step": 88
+    },
+    {
+      "epoch": 0.021612433220009712,
+      "grad_norm": 0.552087664604187,
+      "learning_rate": 0.0001,
+      "loss": 1.8377,
+      "step": 89
+    },
+    {
+      "epoch": 0.021855269548324428,
+      "grad_norm": 0.5610334873199463,
+      "learning_rate": 0.0001,
+      "loss": 1.8038,
+      "step": 90
+    },
+    {
+      "epoch": 0.022098105876639147,
+      "grad_norm": 0.5685205459594727,
+      "learning_rate": 0.0001,
+      "loss": 1.763,
+      "step": 91
+    },
+    {
+      "epoch": 0.022340942204953862,
+      "grad_norm": 0.5359020233154297,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 92
+    },
+    {
+      "epoch": 0.022583778533268577,
+      "grad_norm": 0.5649325251579285,
+      "learning_rate": 0.0001,
+      "loss": 1.922,
+      "step": 93
+    },
+    {
+      "epoch": 0.022826614861583293,
+      "grad_norm": 0.6011147499084473,
+      "learning_rate": 0.0001,
+      "loss": 1.8993,
+      "step": 94
+    },
+    {
+      "epoch": 0.02306945118989801,
+      "grad_norm": 0.5482878088951111,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 95
+    },
+    {
+      "epoch": 0.023312287518212724,
+      "grad_norm": 0.5836300253868103,
+      "learning_rate": 0.0001,
+      "loss": 1.8026,
+      "step": 96
+    },
+    {
+      "epoch": 0.02355512384652744,
+      "grad_norm": 0.6109923124313354,
+      "learning_rate": 0.0001,
+      "loss": 1.9133,
+      "step": 97
+    },
+    {
+      "epoch": 0.023797960174842158,
+      "grad_norm": 0.5888505578041077,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 98
+    },
+    {
+      "epoch": 0.024040796503156873,
+      "grad_norm": 0.6011204719543457,
+      "learning_rate": 0.0001,
+      "loss": 1.8274,
+      "step": 99
+    },
+    {
+      "epoch": 0.02428363283147159,
+      "grad_norm": 0.6063917279243469,
+      "learning_rate": 0.0001,
+      "loss": 1.8999,
+      "step": 100
+    },
+    {
+      "epoch": 0.024526469159786304,
+      "grad_norm": 0.5995082259178162,
+      "learning_rate": 0.0001,
+      "loss": 1.8899,
+      "step": 101
+    },
+    {
+      "epoch": 0.02476930548810102,
+      "grad_norm": 0.6455526351928711,
+      "learning_rate": 0.0001,
+      "loss": 1.8617,
+      "step": 102
+    },
+    {
+      "epoch": 0.025012141816415735,
+      "grad_norm": 0.5625491142272949,
+      "learning_rate": 0.0001,
+      "loss": 1.8013,
+      "step": 103
+    },
+    {
+      "epoch": 0.02525497814473045,
+      "grad_norm": 0.5394976139068604,
+      "learning_rate": 0.0001,
+      "loss": 1.4962,
+      "step": 104
+    },
+    {
+      "epoch": 0.025497814473045166,
+      "grad_norm": 0.6297885179519653,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 105
+    },
+    {
+      "epoch": 0.025740650801359885,
+      "grad_norm": 0.5606225728988647,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 106
+    },
+    {
+      "epoch": 0.0259834871296746,
+      "grad_norm": 0.603681206703186,
+      "learning_rate": 0.0001,
+      "loss": 2.0088,
+      "step": 107
+    },
+    {
+      "epoch": 0.026226323457989316,
+      "grad_norm": 0.6018767952919006,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 108
+    },
+    {
+      "epoch": 0.02646915978630403,
+      "grad_norm": 0.601502001285553,
+      "learning_rate": 0.0001,
+      "loss": 1.9289,
+      "step": 109
+    },
+    {
+      "epoch": 0.026711996114618747,
+      "grad_norm": 0.5400576591491699,
+      "learning_rate": 0.0001,
+      "loss": 1.798,
+      "step": 110
+    },
+    {
+      "epoch": 0.026954832442933462,
+      "grad_norm": 0.5526056885719299,
+      "learning_rate": 0.0001,
+      "loss": 1.6961,
+      "step": 111
+    },
+    {
+      "epoch": 0.027197668771248178,
+      "grad_norm": 0.5839977264404297,
+      "learning_rate": 0.0001,
+      "loss": 1.8279,
+      "step": 112
+    },
+    {
+      "epoch": 0.027440505099562893,
+      "grad_norm": 0.5751993656158447,
+      "learning_rate": 0.0001,
+      "loss": 1.8064,
+      "step": 113
+    },
+    {
+      "epoch": 0.027683341427877612,
+      "grad_norm": 0.5579966306686401,
+      "learning_rate": 0.0001,
+      "loss": 1.6971,
+      "step": 114
+    },
+    {
+      "epoch": 0.027926177756192327,
+      "grad_norm": 0.5659162998199463,
+      "learning_rate": 0.0001,
+      "loss": 1.8134,
+      "step": 115
+    },
+    {
+      "epoch": 0.028169014084507043,
+      "grad_norm": 0.5924839377403259,
+      "learning_rate": 0.0001,
+      "loss": 1.7446,
+      "step": 116
+    },
+    {
+      "epoch": 0.02841185041282176,
+      "grad_norm": 0.5279017686843872,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 117
+    },
+    {
+      "epoch": 0.028654686741136474,
+      "grad_norm": 0.6021360158920288,
+      "learning_rate": 0.0001,
+      "loss": 1.8249,
+      "step": 118
+    },
+    {
+      "epoch": 0.02889752306945119,
+      "grad_norm": 0.552830696105957,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 119
+    },
+    {
+      "epoch": 0.029140359397765905,
+      "grad_norm": 0.5874643325805664,
+      "learning_rate": 0.0001,
+      "loss": 1.8309,
+      "step": 120
+    },
+    {
+      "epoch": 0.02938319572608062,
+      "grad_norm": 0.6058750152587891,
+      "learning_rate": 0.0001,
+      "loss": 1.8723,
+      "step": 121
+    },
+    {
+      "epoch": 0.02962603205439534,
+      "grad_norm": 0.5549308061599731,
+      "learning_rate": 0.0001,
+      "loss": 1.811,
+      "step": 122
+    },
+    {
+      "epoch": 0.029868868382710054,
+      "grad_norm": 0.5869531035423279,
+      "learning_rate": 0.0001,
+      "loss": 1.8061,
+      "step": 123
+    },
+    {
+      "epoch": 0.03011170471102477,
+      "grad_norm": 0.5605490207672119,
+      "learning_rate": 0.0001,
+      "loss": 1.862,
+      "step": 124
+    },
+    {
+      "epoch": 0.030354541039339485,
+      "grad_norm": 0.7070507407188416,
+      "learning_rate": 0.0001,
+      "loss": 1.9312,
+      "step": 125
+    },
+    {
+      "epoch": 0.0305973773676542,
+      "grad_norm": 0.560182511806488,
+      "learning_rate": 0.0001,
+      "loss": 1.7396,
+      "step": 126
+    },
+    {
+      "epoch": 0.030840213695968916,
+      "grad_norm": 0.5708908438682556,
+      "learning_rate": 0.0001,
+      "loss": 1.8544,
+      "step": 127
+    },
+    {
+      "epoch": 0.03108305002428363,
+      "grad_norm": 0.5821275115013123,
+      "learning_rate": 0.0001,
+      "loss": 1.8991,
+      "step": 128
+    },
+    {
+      "epoch": 0.03132588635259835,
+      "grad_norm": 0.5725240707397461,
+      "learning_rate": 0.0001,
+      "loss": 1.793,
+      "step": 129
+    },
+    {
+      "epoch": 0.03156872268091306,
+      "grad_norm": 0.5687121748924255,
+      "learning_rate": 0.0001,
+      "loss": 1.9397,
+      "step": 130
+    },
+    {
+      "epoch": 0.03181155900922778,
+      "grad_norm": 0.5817821621894836,
+      "learning_rate": 0.0001,
+      "loss": 1.9118,
+      "step": 131
+    },
+    {
+      "epoch": 0.03205439533754249,
+      "grad_norm": 0.5283122658729553,
+      "learning_rate": 0.0001,
+      "loss": 1.8213,
+      "step": 132
+    },
+    {
+      "epoch": 0.03229723166585721,
+      "grad_norm": 0.5369681715965271,
+      "learning_rate": 0.0001,
+      "loss": 1.7903,
+      "step": 133
+    },
+    {
+      "epoch": 0.03254006799417193,
+      "grad_norm": 0.5785312652587891,
+      "learning_rate": 0.0001,
+      "loss": 1.9701,
+      "step": 134
+    },
+    {
+      "epoch": 0.03278290432248664,
+      "grad_norm": 0.5557915568351746,
+      "learning_rate": 0.0001,
+      "loss": 1.7392,
+      "step": 135
+    },
+    {
+      "epoch": 0.03302574065080136,
+      "grad_norm": 0.5458675622940063,
+      "learning_rate": 0.0001,
+      "loss": 1.7289,
+      "step": 136
+    },
+    {
+      "epoch": 0.033268576979116074,
+      "grad_norm": 0.6089033484458923,
+      "learning_rate": 0.0001,
+      "loss": 1.9272,
+      "step": 137
+    },
+    {
+      "epoch": 0.03351141330743079,
+      "grad_norm": 0.5484150648117065,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 138
+    },
+    {
+      "epoch": 0.033754249635745505,
+      "grad_norm": 0.5457684397697449,
+      "learning_rate": 0.0001,
+      "loss": 1.7081,
+      "step": 139
+    },
+    {
+      "epoch": 0.033997085964060224,
+      "grad_norm": 0.5888879895210266,
+      "learning_rate": 0.0001,
+      "loss": 1.7877,
+      "step": 140
+    },
+    {
+      "epoch": 0.03423992229237494,
+      "grad_norm": 0.5889548063278198,
+      "learning_rate": 0.0001,
+      "loss": 1.9245,
+      "step": 141
+    },
+    {
+      "epoch": 0.034482758620689655,
+      "grad_norm": 0.5552160143852234,
+      "learning_rate": 0.0001,
+      "loss": 1.8471,
+      "step": 142
+    },
+    {
+      "epoch": 0.034725594949004374,
+      "grad_norm": 0.566940426826477,
+      "learning_rate": 0.0001,
+      "loss": 1.8551,
+      "step": 143
+    },
+    {
+      "epoch": 0.034968431277319086,
+      "grad_norm": 0.5903639197349548,
+      "learning_rate": 0.0001,
+      "loss": 1.7737,
+      "step": 144
+    },
+    {
+      "epoch": 0.035211267605633804,
+      "grad_norm": 0.5215669870376587,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 145
+    },
+    {
+      "epoch": 0.035454103933948516,
+      "grad_norm": 0.5972316861152649,
+      "learning_rate": 0.0001,
+      "loss": 1.9133,
+      "step": 146
+    },
+    {
+      "epoch": 0.035696940262263235,
+      "grad_norm": 0.5835139751434326,
+      "learning_rate": 0.0001,
+      "loss": 1.8421,
+      "step": 147
+    },
+    {
+      "epoch": 0.03593977659057795,
+      "grad_norm": 0.5471585988998413,
+      "learning_rate": 0.0001,
+      "loss": 1.7915,
+      "step": 148
+    },
+    {
+      "epoch": 0.036182612918892666,
+      "grad_norm": 0.5762492418289185,
+      "learning_rate": 0.0001,
+      "loss": 1.9026,
+      "step": 149
+    },
+    {
+      "epoch": 0.036425449247207385,
+      "grad_norm": 0.5917050838470459,
+      "learning_rate": 0.0001,
+      "loss": 1.9228,
+      "step": 150
+    },
+    {
+      "epoch": 0.0366682855755221,
+      "grad_norm": 0.5515230298042297,
+      "learning_rate": 0.0001,
+      "loss": 1.7392,
+      "step": 151
+    },
+    {
+      "epoch": 0.036911121903836816,
+      "grad_norm": 0.5346499085426331,
+      "learning_rate": 0.0001,
+      "loss": 1.6875,
+      "step": 152
+    },
+    {
+      "epoch": 0.03715395823215153,
+      "grad_norm": 0.5266364812850952,
+      "learning_rate": 0.0001,
+      "loss": 1.718,
+      "step": 153
+    },
+    {
+      "epoch": 0.03739679456046625,
+      "grad_norm": 0.5427219867706299,
+      "learning_rate": 0.0001,
+      "loss": 1.8072,
+      "step": 154
+    },
+    {
+      "epoch": 0.03763963088878096,
+      "grad_norm": 0.5867930054664612,
+      "learning_rate": 0.0001,
+      "loss": 1.7706,
+      "step": 155
+    },
+    {
+      "epoch": 0.03788246721709568,
+      "grad_norm": 0.5999542474746704,
+      "learning_rate": 0.0001,
+      "loss": 1.8921,
+      "step": 156
+    },
+    {
+      "epoch": 0.0381253035454104,
+      "grad_norm": 0.5528079271316528,
+      "learning_rate": 0.0001,
+      "loss": 1.6512,
+      "step": 157
+    },
+    {
+      "epoch": 0.03836813987372511,
+      "grad_norm": 0.580348014831543,
+      "learning_rate": 0.0001,
+      "loss": 1.9681,
+      "step": 158
+    },
+    {
+      "epoch": 0.03861097620203983,
+      "grad_norm": 0.6032450795173645,
+      "learning_rate": 0.0001,
+      "loss": 1.7902,
+      "step": 159
+    },
+    {
+      "epoch": 0.03885381253035454,
+      "grad_norm": 0.5768253803253174,
+      "learning_rate": 0.0001,
+      "loss": 1.8762,
+      "step": 160
+    },
+    {
+      "epoch": 0.03909664885866926,
+      "grad_norm": 0.5588753819465637,
+      "learning_rate": 0.0001,
+      "loss": 1.7458,
+      "step": 161
+    },
+    {
+      "epoch": 0.03933948518698397,
+      "grad_norm": 0.5667343139648438,
+      "learning_rate": 0.0001,
+      "loss": 1.7809,
+      "step": 162
+    },
+    {
+      "epoch": 0.03958232151529869,
+      "grad_norm": 0.5831610560417175,
+      "learning_rate": 0.0001,
+      "loss": 1.9095,
+      "step": 163
+    },
+    {
+      "epoch": 0.0398251578436134,
+      "grad_norm": 0.5786190032958984,
+      "learning_rate": 0.0001,
+      "loss": 1.7167,
+      "step": 164
+    },
+    {
+      "epoch": 0.04006799417192812,
+      "grad_norm": 0.5616288185119629,
+      "learning_rate": 0.0001,
+      "loss": 1.8568,
+      "step": 165
+    },
+    {
+      "epoch": 0.04031083050024284,
+      "grad_norm": 0.5448461771011353,
+      "learning_rate": 0.0001,
+      "loss": 1.8251,
+      "step": 166
+    },
+    {
+      "epoch": 0.04055366682855755,
+      "grad_norm": 0.5922189354896545,
+      "learning_rate": 0.0001,
+      "loss": 1.8343,
+      "step": 167
+    },
+    {
+      "epoch": 0.04079650315687227,
+      "grad_norm": 0.5826832056045532,
+      "learning_rate": 0.0001,
+      "loss": 1.8375,
+      "step": 168
+    },
+    {
+      "epoch": 0.04103933948518698,
+      "grad_norm": 0.546203076839447,
+      "learning_rate": 0.0001,
+      "loss": 1.8048,
+      "step": 169
+    },
+    {
+      "epoch": 0.0412821758135017,
+      "grad_norm": 0.5722417831420898,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 170
+    },
+    {
+      "epoch": 0.04152501214181641,
+      "grad_norm": 0.5691736936569214,
+      "learning_rate": 0.0001,
+      "loss": 1.8391,
+      "step": 171
+    },
+    {
+      "epoch": 0.04176784847013113,
+      "grad_norm": 0.5580527186393738,
+      "learning_rate": 0.0001,
+      "loss": 1.7919,
+      "step": 172
+    },
+    {
+      "epoch": 0.04201068479844585,
+      "grad_norm": 0.5881775617599487,
+      "learning_rate": 0.0001,
+      "loss": 1.9728,
+      "step": 173
+    },
+    {
+      "epoch": 0.04225352112676056,
+      "grad_norm": 0.5363543033599854,
+      "learning_rate": 0.0001,
+      "loss": 1.6842,
+      "step": 174
+    },
+    {
+      "epoch": 0.04249635745507528,
+      "grad_norm": 0.5504969954490662,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 175
+    },
+    {
+      "epoch": 0.042739193783389993,
+      "grad_norm": 0.5678769946098328,
+      "learning_rate": 0.0001,
+      "loss": 1.8638,
+      "step": 176
+    },
+    {
+      "epoch": 0.04298203011170471,
+      "grad_norm": 0.5707623958587646,
+      "learning_rate": 0.0001,
+      "loss": 1.7592,
+      "step": 177
+    },
+    {
+      "epoch": 0.043224866440019424,
+      "grad_norm": 0.5393390655517578,
+      "learning_rate": 0.0001,
+      "loss": 1.7732,
+      "step": 178
+    },
+    {
+      "epoch": 0.04346770276833414,
+      "grad_norm": 0.5405838489532471,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 179
+    },
+    {
+      "epoch": 0.043710539096648855,
+      "grad_norm": 0.5595617294311523,
+      "learning_rate": 0.0001,
+      "loss": 1.7885,
+      "step": 180
+    },
+    {
+      "epoch": 0.043953375424963574,
+      "grad_norm": 0.5614048838615417,
+      "learning_rate": 0.0001,
+      "loss": 1.8507,
+      "step": 181
+    },
+    {
+      "epoch": 0.04419621175327829,
+      "grad_norm": 0.5554660558700562,
+      "learning_rate": 0.0001,
+      "loss": 1.8694,
+      "step": 182
+    },
+    {
+      "epoch": 0.044439048081593005,
+      "grad_norm": 0.5462179780006409,
+      "learning_rate": 0.0001,
+      "loss": 1.828,
+      "step": 183
+    },
+    {
+      "epoch": 0.044681884409907724,
+      "grad_norm": 0.5049857497215271,
+      "learning_rate": 0.0001,
+      "loss": 1.494,
+      "step": 184
+    },
+    {
+      "epoch": 0.044924720738222436,
+      "grad_norm": 0.5400852560997009,
+      "learning_rate": 0.0001,
+      "loss": 1.8124,
+      "step": 185
+    },
+    {
+      "epoch": 0.045167557066537155,
+      "grad_norm": 0.5604100227355957,
+      "learning_rate": 0.0001,
+      "loss": 1.9107,
+      "step": 186
+    },
+    {
+      "epoch": 0.04541039339485187,
+      "grad_norm": 0.5613305568695068,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 187
+    },
+    {
+      "epoch": 0.045653229723166586,
+      "grad_norm": 0.5521450042724609,
+      "learning_rate": 0.0001,
+      "loss": 1.7774,
+      "step": 188
+    },
+    {
+      "epoch": 0.045896066051481305,
+      "grad_norm": 0.558240532875061,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 189
+    },
+    {
+      "epoch": 0.04613890237979602,
+      "grad_norm": 0.5644837021827698,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 190
+    },
+    {
+      "epoch": 0.046381738708110735,
+      "grad_norm": 0.5644278526306152,
+      "learning_rate": 0.0001,
+      "loss": 1.8844,
+      "step": 191
+    },
+    {
+      "epoch": 0.04662457503642545,
+      "grad_norm": 0.529413640499115,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 192
+    },
+    {
+      "epoch": 0.046867411364740166,
+      "grad_norm": 0.5736052989959717,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 193
+    },
+    {
+      "epoch": 0.04711024769305488,
+      "grad_norm": 0.5963498950004578,
+      "learning_rate": 0.0001,
+      "loss": 1.8204,
+      "step": 194
+    },
+    {
+      "epoch": 0.0473530840213696,
+      "grad_norm": 0.5892695188522339,
+      "learning_rate": 0.0001,
+      "loss": 1.6753,
+      "step": 195
+    },
+    {
+      "epoch": 0.047595920349684316,
+      "grad_norm": 0.570740282535553,
+      "learning_rate": 0.0001,
+      "loss": 1.8047,
+      "step": 196
+    },
+    {
+      "epoch": 0.04783875667799903,
+      "grad_norm": 0.5566459894180298,
+      "learning_rate": 0.0001,
+      "loss": 1.7479,
+      "step": 197
+    },
+    {
+      "epoch": 0.04808159300631375,
+      "grad_norm": 0.5618220567703247,
+      "learning_rate": 0.0001,
+      "loss": 1.7769,
+      "step": 198
+    },
+    {
+      "epoch": 0.04832442933462846,
+      "grad_norm": 0.5499823093414307,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 199
+    },
+    {
+      "epoch": 0.04856726566294318,
+      "grad_norm": 0.5597728490829468,
+      "learning_rate": 0.0001,
+      "loss": 1.8234,
+      "step": 200
+    },
+    {
+      "epoch": 0.04881010199125789,
+      "grad_norm": 0.5423206090927124,
+      "learning_rate": 0.0001,
+      "loss": 1.7442,
+      "step": 201
+    },
+    {
+      "epoch": 0.04905293831957261,
+      "grad_norm": 0.59138023853302,
+      "learning_rate": 0.0001,
+      "loss": 1.932,
+      "step": 202
+    },
+    {
+      "epoch": 0.04929577464788732,
+      "grad_norm": 0.57453852891922,
+      "learning_rate": 0.0001,
+      "loss": 1.7707,
+      "step": 203
+    },
+    {
+      "epoch": 0.04953861097620204,
+      "grad_norm": 0.5949112176895142,
+      "learning_rate": 0.0001,
+      "loss": 2.0603,
+      "step": 204
+    },
+    {
+      "epoch": 0.04978144730451676,
+      "grad_norm": 0.5683001279830933,
+      "learning_rate": 0.0001,
+      "loss": 1.8341,
+      "step": 205
+    },
+    {
+      "epoch": 0.05002428363283147,
+      "grad_norm": 0.511799156665802,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 206
+    },
+    {
+      "epoch": 0.05026711996114619,
+      "grad_norm": 0.5489233732223511,
+      "learning_rate": 0.0001,
+      "loss": 1.7677,
+      "step": 207
+    },
+    {
+      "epoch": 0.0505099562894609,
+      "grad_norm": 0.5438036322593689,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 208
+    },
+    {
+      "epoch": 0.05075279261777562,
+      "grad_norm": 0.5517812967300415,
+      "learning_rate": 0.0001,
+      "loss": 1.7707,
+      "step": 209
+    },
+    {
+      "epoch": 0.05099562894609033,
+      "grad_norm": 0.5675941109657288,
+      "learning_rate": 0.0001,
+      "loss": 1.776,
+      "step": 210
+    },
+    {
+      "epoch": 0.05123846527440505,
+      "grad_norm": 0.5526789426803589,
+      "learning_rate": 0.0001,
+      "loss": 1.9368,
+      "step": 211
+    },
+    {
+      "epoch": 0.05148130160271977,
+      "grad_norm": 0.5648570656776428,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 212
+    },
+    {
+      "epoch": 0.05172413793103448,
+      "grad_norm": 0.5832803249359131,
+      "learning_rate": 0.0001,
+      "loss": 1.8035,
+      "step": 213
+    },
+    {
+      "epoch": 0.0519669742593492,
+      "grad_norm": 0.5673675537109375,
+      "learning_rate": 0.0001,
+      "loss": 1.7283,
+      "step": 214
+    },
+    {
+      "epoch": 0.05220981058766391,
+      "grad_norm": 0.5454904437065125,
+      "learning_rate": 0.0001,
+      "loss": 1.8284,
+      "step": 215
+    },
+    {
+      "epoch": 0.05245264691597863,
+      "grad_norm": 0.6127611398696899,
+      "learning_rate": 0.0001,
+      "loss": 1.9195,
+      "step": 216
+    },
+    {
+      "epoch": 0.052695483244293344,
+      "grad_norm": 0.5642631649971008,
+      "learning_rate": 0.0001,
+      "loss": 1.848,
+      "step": 217
+    },
+    {
+      "epoch": 0.05293831957260806,
+      "grad_norm": 0.5837909579277039,
+      "learning_rate": 0.0001,
+      "loss": 1.9077,
+      "step": 218
+    },
+    {
+      "epoch": 0.053181155900922775,
+      "grad_norm": 0.5727541446685791,
+      "learning_rate": 0.0001,
+      "loss": 1.7965,
+      "step": 219
+    },
+    {
+      "epoch": 0.053423992229237494,
+      "grad_norm": 0.5601940751075745,
+      "learning_rate": 0.0001,
+      "loss": 1.9039,
+      "step": 220
+    },
+    {
+      "epoch": 0.05366682855755221,
+      "grad_norm": 0.5595248341560364,
+      "learning_rate": 0.0001,
+      "loss": 1.8851,
+      "step": 221
+    },
+    {
+      "epoch": 0.053909664885866924,
+      "grad_norm": 0.49235615134239197,
+      "learning_rate": 0.0001,
+      "loss": 1.6141,
+      "step": 222
+    },
+    {
+      "epoch": 0.05415250121418164,
+      "grad_norm": 0.5064696073532104,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 223
+    },
+    {
+      "epoch": 0.054395337542496355,
+      "grad_norm": 0.5605242848396301,
+      "learning_rate": 0.0001,
+      "loss": 1.8623,
+      "step": 224
+    },
+    {
+      "epoch": 0.054638173870811074,
+      "grad_norm": 0.5589006543159485,
+      "learning_rate": 0.0001,
+      "loss": 1.6309,
+      "step": 225
+    },
+    {
+      "epoch": 0.054881010199125786,
+      "grad_norm": 0.5964434146881104,
+      "learning_rate": 0.0001,
+      "loss": 1.8167,
+      "step": 226
+    },
+    {
+      "epoch": 0.055123846527440505,
+      "grad_norm": 0.5265910625457764,
+      "learning_rate": 0.0001,
+      "loss": 1.7182,
+      "step": 227
+    },
+    {
+      "epoch": 0.055366682855755224,
+      "grad_norm": 0.5511795878410339,
+      "learning_rate": 0.0001,
+      "loss": 1.6657,
+      "step": 228
+    },
+    {
+      "epoch": 0.055609519184069936,
+      "grad_norm": 0.5586657524108887,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 229
+    },
+    {
+      "epoch": 0.055852355512384655,
+      "grad_norm": 0.5953067541122437,
+      "learning_rate": 0.0001,
+      "loss": 1.8374,
+      "step": 230
+    },
+    {
+      "epoch": 0.05609519184069937,
+      "grad_norm": 0.5542361736297607,
+      "learning_rate": 0.0001,
+      "loss": 1.7509,
+      "step": 231
+    },
+    {
+      "epoch": 0.056338028169014086,
+      "grad_norm": 0.5865808129310608,
+      "learning_rate": 0.0001,
+      "loss": 1.9627,
+      "step": 232
+    },
+    {
+      "epoch": 0.0565808644973288,
+      "grad_norm": 0.514796793460846,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 233
+    },
+    {
+      "epoch": 0.05682370082564352,
+      "grad_norm": 0.5481152534484863,
+      "learning_rate": 0.0001,
+      "loss": 1.8672,
+      "step": 234
+    },
+    {
+      "epoch": 0.057066537153958236,
+      "grad_norm": 0.49685022234916687,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 235
+    },
+    {
+      "epoch": 0.05730937348227295,
+      "grad_norm": 0.6000615954399109,
+      "learning_rate": 0.0001,
+      "loss": 1.9551,
+      "step": 236
+    },
+    {
+      "epoch": 0.057552209810587666,
+      "grad_norm": 0.5260239839553833,
+      "learning_rate": 0.0001,
+      "loss": 1.8325,
+      "step": 237
+    },
+    {
+      "epoch": 0.05779504613890238,
+      "grad_norm": 0.5536266565322876,
+      "learning_rate": 0.0001,
+      "loss": 1.884,
+      "step": 238
+    },
+    {
+      "epoch": 0.0580378824672171,
+      "grad_norm": 0.5461103916168213,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 239
+    },
+    {
+      "epoch": 0.05828071879553181,
+      "grad_norm": 0.5518085360527039,
+      "learning_rate": 0.0001,
+      "loss": 1.8653,
+      "step": 240
+    },
+    {
+      "epoch": 0.05852355512384653,
+      "grad_norm": 0.5721222162246704,
+      "learning_rate": 0.0001,
+      "loss": 1.874,
+      "step": 241
+    },
+    {
+      "epoch": 0.05876639145216124,
+      "grad_norm": 0.5601527690887451,
+      "learning_rate": 0.0001,
+      "loss": 1.8091,
+      "step": 242
+    },
+    {
+      "epoch": 0.05900922778047596,
+      "grad_norm": 0.5886834263801575,
+      "learning_rate": 0.0001,
+      "loss": 1.9075,
+      "step": 243
+    },
+    {
+      "epoch": 0.05925206410879068,
+      "grad_norm": 0.5505769848823547,
+      "learning_rate": 0.0001,
+      "loss": 1.7758,
+      "step": 244
+    },
+    {
+      "epoch": 0.05949490043710539,
+      "grad_norm": 0.5973044037818909,
+      "learning_rate": 0.0001,
+      "loss": 1.8534,
+      "step": 245
+    },
+    {
+      "epoch": 0.05973773676542011,
+      "grad_norm": 0.5689461827278137,
+      "learning_rate": 0.0001,
+      "loss": 1.8422,
+      "step": 246
+    },
+    {
+      "epoch": 0.05998057309373482,
+      "grad_norm": 0.5527995824813843,
+      "learning_rate": 0.0001,
+      "loss": 1.7791,
+      "step": 247
+    },
+    {
+      "epoch": 0.06022340942204954,
+      "grad_norm": 0.5316581130027771,
+      "learning_rate": 0.0001,
+      "loss": 1.7325,
+      "step": 248
+    },
+    {
+      "epoch": 0.06046624575036425,
+      "grad_norm": 0.5318779349327087,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 249
+    },
+    {
+      "epoch": 0.06070908207867897,
+      "grad_norm": 0.5586565136909485,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 250
+    },
+    {
+      "epoch": 0.06095191840699369,
+      "grad_norm": 0.58368980884552,
+      "learning_rate": 0.0001,
+      "loss": 1.8094,
+      "step": 251
+    },
+    {
+      "epoch": 0.0611947547353084,
+      "grad_norm": 0.5369036793708801,
+      "learning_rate": 0.0001,
+      "loss": 1.8643,
+      "step": 252
+    },
+    {
+      "epoch": 0.06143759106362312,
+      "grad_norm": 0.5384863615036011,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 253
+    },
+    {
+      "epoch": 0.06168042739193783,
+      "grad_norm": 0.5588333606719971,
+      "learning_rate": 0.0001,
+      "loss": 1.8526,
+      "step": 254
+    },
+    {
+      "epoch": 0.06192326372025255,
+      "grad_norm": 0.5602843761444092,
+      "learning_rate": 0.0001,
+      "loss": 1.9226,
+      "step": 255
+    },
+    {
+      "epoch": 0.06216610004856726,
+      "grad_norm": 0.5340338349342346,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 256
+    },
+    {
+      "epoch": 0.06240893637688198,
+      "grad_norm": 0.5598003268241882,
+      "learning_rate": 0.0001,
+      "loss": 1.7225,
+      "step": 257
+    },
+    {
+      "epoch": 0.0626517727051967,
+      "grad_norm": 0.5247579216957092,
+      "learning_rate": 0.0001,
+      "loss": 1.7159,
+      "step": 258
+    },
+    {
+      "epoch": 0.06289460903351142,
+      "grad_norm": 0.5494512319564819,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 259
+    },
+    {
+      "epoch": 0.06313744536182612,
+      "grad_norm": 0.5570547580718994,
+      "learning_rate": 0.0001,
+      "loss": 1.8625,
+      "step": 260
+    },
+    {
+      "epoch": 0.06338028169014084,
+      "grad_norm": 0.5337409973144531,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 261
+    },
+    {
+      "epoch": 0.06362311801845556,
+      "grad_norm": 0.5944165587425232,
+      "learning_rate": 0.0001,
+      "loss": 1.9068,
+      "step": 262
+    },
+    {
+      "epoch": 0.06386595434677028,
+      "grad_norm": 0.5736029148101807,
+      "learning_rate": 0.0001,
+      "loss": 1.7135,
+      "step": 263
+    },
+    {
+      "epoch": 0.06410879067508499,
+      "grad_norm": 0.5729188323020935,
+      "learning_rate": 0.0001,
+      "loss": 1.8619,
+      "step": 264
+    },
+    {
+      "epoch": 0.0643516270033997,
+      "grad_norm": 0.5908741354942322,
+      "learning_rate": 0.0001,
+      "loss": 1.7791,
+      "step": 265
+    },
+    {
+      "epoch": 0.06459446333171442,
+      "grad_norm": 0.550342857837677,
+      "learning_rate": 0.0001,
+      "loss": 1.722,
+      "step": 266
+    },
+    {
+      "epoch": 0.06483729966002914,
+      "grad_norm": 0.5404797792434692,
+      "learning_rate": 0.0001,
+      "loss": 1.8185,
+      "step": 267
+    },
+    {
+      "epoch": 0.06508013598834386,
+      "grad_norm": 0.6320937275886536,
+      "learning_rate": 0.0001,
+      "loss": 1.8696,
+      "step": 268
+    },
+    {
+      "epoch": 0.06532297231665857,
+      "grad_norm": 0.5689648389816284,
+      "learning_rate": 0.0001,
+      "loss": 1.9712,
+      "step": 269
+    },
+    {
+      "epoch": 0.06556580864497329,
+      "grad_norm": 0.5600985288619995,
+      "learning_rate": 0.0001,
+      "loss": 1.8229,
+      "step": 270
+    },
+    {
+      "epoch": 0.065808644973288,
+      "grad_norm": 0.6124951243400574,
+      "learning_rate": 0.0001,
+      "loss": 1.8365,
+      "step": 271
+    },
+    {
+      "epoch": 0.06605148130160272,
+      "grad_norm": 0.6249804496765137,
+      "learning_rate": 0.0001,
+      "loss": 1.8418,
+      "step": 272
+    },
+    {
+      "epoch": 0.06629431762991743,
+      "grad_norm": 0.5479178428649902,
+      "learning_rate": 0.0001,
+      "loss": 1.7385,
+      "step": 273
+    },
+    {
+      "epoch": 0.06653715395823215,
+      "grad_norm": 0.562118411064148,
+      "learning_rate": 0.0001,
+      "loss": 1.6261,
+      "step": 274
+    },
+    {
+      "epoch": 0.06677999028654687,
+      "grad_norm": 0.5987095832824707,
+      "learning_rate": 0.0001,
+      "loss": 2.0492,
+      "step": 275
+    },
+    {
+      "epoch": 0.06702282661486159,
+      "grad_norm": 0.5354698896408081,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 276
+    },
+    {
+      "epoch": 0.0672656629431763,
+      "grad_norm": 0.5242313742637634,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 277
+    },
+    {
+      "epoch": 0.06750849927149101,
+      "grad_norm": 0.6085278391838074,
+      "learning_rate": 0.0001,
+      "loss": 1.8389,
+      "step": 278
+    },
+    {
+      "epoch": 0.06775133559980573,
+      "grad_norm": 0.592673659324646,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 279
+    },
+    {
+      "epoch": 0.06799417192812045,
+      "grad_norm": 0.6025213599205017,
+      "learning_rate": 0.0001,
+      "loss": 1.9998,
+      "step": 280
+    },
+    {
+      "epoch": 0.06823700825643517,
+      "grad_norm": 0.5785611867904663,
+      "learning_rate": 0.0001,
+      "loss": 1.8031,
+      "step": 281
+    },
+    {
+      "epoch": 0.06847984458474989,
+      "grad_norm": 0.5544794797897339,
+      "learning_rate": 0.0001,
+      "loss": 1.7671,
+      "step": 282
+    },
+    {
+      "epoch": 0.06872268091306459,
+      "grad_norm": 0.5426813364028931,
+      "learning_rate": 0.0001,
+      "loss": 1.7511,
+      "step": 283
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 0.6183847784996033,
+      "learning_rate": 0.0001,
+      "loss": 1.805,
+      "step": 284
+    },
+    {
+      "epoch": 0.06920835356969403,
+      "grad_norm": 0.5462707877159119,
+      "learning_rate": 0.0001,
+      "loss": 1.7911,
+      "step": 285
+    },
+    {
+      "epoch": 0.06945118989800875,
+      "grad_norm": 0.5737053155899048,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 286
+    },
+    {
+      "epoch": 0.06969402622632345,
+      "grad_norm": 0.6043452620506287,
+      "learning_rate": 0.0001,
+      "loss": 1.8558,
+      "step": 287
+    },
+    {
+      "epoch": 0.06993686255463817,
+      "grad_norm": 0.606366753578186,
+      "learning_rate": 0.0001,
+      "loss": 2.0038,
+      "step": 288
+    },
+    {
+      "epoch": 0.07017969888295289,
+      "grad_norm": 0.5447037220001221,
+      "learning_rate": 0.0001,
+      "loss": 1.8431,
+      "step": 289
+    },
+    {
+      "epoch": 0.07042253521126761,
+      "grad_norm": 0.5541792511940002,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 290
+    },
+    {
+      "epoch": 0.07066537153958233,
+      "grad_norm": 0.5450261831283569,
+      "learning_rate": 0.0001,
+      "loss": 1.8592,
+      "step": 291
+    },
+    {
+      "epoch": 0.07090820786789703,
+      "grad_norm": 0.5191848874092102,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 292
+    },
+    {
+      "epoch": 0.07115104419621175,
+      "grad_norm": 0.5771126747131348,
+      "learning_rate": 0.0001,
+      "loss": 1.7581,
+      "step": 293
+    },
+    {
+      "epoch": 0.07139388052452647,
+      "grad_norm": 0.550542414188385,
+      "learning_rate": 0.0001,
+      "loss": 1.7619,
+      "step": 294
+    },
+    {
+      "epoch": 0.07163671685284119,
+      "grad_norm": 0.5441504716873169,
+      "learning_rate": 0.0001,
+      "loss": 1.8268,
+      "step": 295
+    },
+    {
+      "epoch": 0.0718795531811559,
+      "grad_norm": 0.5738163590431213,
+      "learning_rate": 0.0001,
+      "loss": 1.7749,
+      "step": 296
+    },
+    {
+      "epoch": 0.07212238950947061,
+      "grad_norm": 0.5475561022758484,
+      "learning_rate": 0.0001,
+      "loss": 1.7159,
+      "step": 297
+    },
+    {
+      "epoch": 0.07236522583778533,
+      "grad_norm": 0.5339605808258057,
+      "learning_rate": 0.0001,
+      "loss": 1.7179,
+      "step": 298
+    },
+    {
+      "epoch": 0.07260806216610005,
+      "grad_norm": 0.5403867363929749,
+      "learning_rate": 0.0001,
+      "loss": 1.7337,
+      "step": 299
+    },
+    {
+      "epoch": 0.07285089849441477,
+      "grad_norm": 0.546708345413208,
+      "learning_rate": 0.0001,
+      "loss": 1.7792,
+      "step": 300
+    },
+    {
+      "epoch": 0.07309373482272948,
+      "grad_norm": 0.5604628920555115,
+      "learning_rate": 0.0001,
+      "loss": 1.8536,
+      "step": 301
+    },
+    {
+      "epoch": 0.0733365711510442,
+      "grad_norm": 0.5335520505905151,
+      "learning_rate": 0.0001,
+      "loss": 1.6597,
+      "step": 302
+    },
+    {
+      "epoch": 0.07357940747935891,
+      "grad_norm": 0.5464479327201843,
+      "learning_rate": 0.0001,
+      "loss": 1.7763,
+      "step": 303
+    },
+    {
+      "epoch": 0.07382224380767363,
+      "grad_norm": 0.5682089328765869,
+      "learning_rate": 0.0001,
+      "loss": 1.8342,
+      "step": 304
+    },
+    {
+      "epoch": 0.07406508013598834,
+      "grad_norm": 0.5149857401847839,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 305
+    },
+    {
+      "epoch": 0.07430791646430306,
+      "grad_norm": 0.5400729179382324,
+      "learning_rate": 0.0001,
+      "loss": 1.8352,
+      "step": 306
+    },
+    {
+      "epoch": 0.07455075279261777,
+      "grad_norm": 0.5587572455406189,
+      "learning_rate": 0.0001,
+      "loss": 1.8093,
+      "step": 307
+    },
+    {
+      "epoch": 0.0747935891209325,
+      "grad_norm": 0.5496796369552612,
+      "learning_rate": 0.0001,
+      "loss": 1.8077,
+      "step": 308
+    },
+    {
+      "epoch": 0.07503642544924721,
+      "grad_norm": 0.5394822359085083,
+      "learning_rate": 0.0001,
+      "loss": 1.7357,
+      "step": 309
+    },
+    {
+      "epoch": 0.07527926177756192,
+      "grad_norm": 0.5829740166664124,
+      "learning_rate": 0.0001,
+      "loss": 1.8976,
+      "step": 310
+    },
+    {
+      "epoch": 0.07552209810587664,
+      "grad_norm": 0.5242211818695068,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 311
+    },
+    {
+      "epoch": 0.07576493443419136,
+      "grad_norm": 0.5243309140205383,
+      "learning_rate": 0.0001,
+      "loss": 1.6823,
+      "step": 312
+    },
+    {
+      "epoch": 0.07600777076250607,
+      "grad_norm": 0.5284510254859924,
+      "learning_rate": 0.0001,
+      "loss": 1.6511,
+      "step": 313
+    },
+    {
+      "epoch": 0.0762506070908208,
+      "grad_norm": 0.5880502462387085,
+      "learning_rate": 0.0001,
+      "loss": 1.8045,
+      "step": 314
+    },
+    {
+      "epoch": 0.0764934434191355,
+      "grad_norm": 0.5681189298629761,
+      "learning_rate": 0.0001,
+      "loss": 1.8092,
+      "step": 315
+    },
+    {
+      "epoch": 0.07673627974745022,
+      "grad_norm": 0.5528023838996887,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 316
+    },
+    {
+      "epoch": 0.07697911607576494,
+      "grad_norm": 0.545785665512085,
+      "learning_rate": 0.0001,
+      "loss": 1.817,
+      "step": 317
+    },
+    {
+      "epoch": 0.07722195240407966,
+      "grad_norm": 0.5269185304641724,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 318
+    },
+    {
+      "epoch": 0.07746478873239436,
+      "grad_norm": 0.5679982304573059,
+      "learning_rate": 0.0001,
+      "loss": 1.8103,
+      "step": 319
+    },
+    {
+      "epoch": 0.07770762506070908,
+      "grad_norm": 0.5480614900588989,
+      "learning_rate": 0.0001,
+      "loss": 1.6871,
+      "step": 320
+    },
+    {
+      "epoch": 0.0779504613890238,
+      "grad_norm": 0.5294215679168701,
+      "learning_rate": 0.0001,
+      "loss": 1.7389,
+      "step": 321
+    },
+    {
+      "epoch": 0.07819329771733852,
+      "grad_norm": 0.5534301400184631,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 322
+    },
+    {
+      "epoch": 0.07843613404565324,
+      "grad_norm": 0.5598313808441162,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 323
+    },
+    {
+      "epoch": 0.07867897037396794,
+      "grad_norm": 0.5626444816589355,
+      "learning_rate": 0.0001,
+      "loss": 1.8142,
+      "step": 324
+    },
+    {
+      "epoch": 0.07892180670228266,
+      "grad_norm": 0.528076708316803,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 325
+    },
+    {
+      "epoch": 0.07916464303059738,
+      "grad_norm": 0.5425437092781067,
+      "learning_rate": 0.0001,
+      "loss": 1.7188,
+      "step": 326
+    },
+    {
+      "epoch": 0.0794074793589121,
+      "grad_norm": 0.5973831415176392,
+      "learning_rate": 0.0001,
+      "loss": 1.8852,
+      "step": 327
+    },
+    {
+      "epoch": 0.0796503156872268,
+      "grad_norm": 0.5841144919395447,
+      "learning_rate": 0.0001,
+      "loss": 1.8122,
+      "step": 328
+    },
+    {
+      "epoch": 0.07989315201554152,
+      "grad_norm": 0.5223991274833679,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 329
+    },
+    {
+      "epoch": 0.08013598834385624,
+      "grad_norm": 0.573898434638977,
+      "learning_rate": 0.0001,
+      "loss": 1.8858,
+      "step": 330
+    },
+    {
+      "epoch": 0.08037882467217096,
+      "grad_norm": 0.5554970502853394,
+      "learning_rate": 0.0001,
+      "loss": 1.7687,
+      "step": 331
+    },
+    {
+      "epoch": 0.08062166100048568,
+      "grad_norm": 0.5309914350509644,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 332
+    },
+    {
+      "epoch": 0.08086449732880038,
+      "grad_norm": 0.5667017698287964,
+      "learning_rate": 0.0001,
+      "loss": 1.8222,
+      "step": 333
+    },
+    {
+      "epoch": 0.0811073336571151,
+      "grad_norm": 0.5442120432853699,
+      "learning_rate": 0.0001,
+      "loss": 1.7532,
+      "step": 334
+    },
+    {
+      "epoch": 0.08135016998542982,
+      "grad_norm": 0.48812708258628845,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 335
+    },
+    {
+      "epoch": 0.08159300631374454,
+      "grad_norm": 0.523797333240509,
+      "learning_rate": 0.0001,
+      "loss": 1.7718,
+      "step": 336
+    },
+    {
+      "epoch": 0.08183584264205926,
+      "grad_norm": 0.5603326559066772,
+      "learning_rate": 0.0001,
+      "loss": 1.835,
+      "step": 337
+    },
+    {
+      "epoch": 0.08207867897037396,
+      "grad_norm": 0.5247459411621094,
+      "learning_rate": 0.0001,
+      "loss": 1.7186,
+      "step": 338
+    },
+    {
+      "epoch": 0.08232151529868868,
+      "grad_norm": 0.5363231301307678,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 339
+    },
+    {
+      "epoch": 0.0825643516270034,
+      "grad_norm": 0.5604921579360962,
+      "learning_rate": 0.0001,
+      "loss": 1.745,
+      "step": 340
+    },
+    {
+      "epoch": 0.08280718795531812,
+      "grad_norm": 0.5731708407402039,
+      "learning_rate": 0.0001,
+      "loss": 1.5748,
+      "step": 341
+    },
+    {
+      "epoch": 0.08305002428363283,
+      "grad_norm": 0.6106705069541931,
+      "learning_rate": 0.0001,
+      "loss": 1.8616,
+      "step": 342
+    },
+    {
+      "epoch": 0.08329286061194754,
+      "grad_norm": 0.5862675309181213,
+      "learning_rate": 0.0001,
+      "loss": 1.8843,
+      "step": 343
+    },
+    {
+      "epoch": 0.08353569694026226,
+      "grad_norm": 0.5550743937492371,
+      "learning_rate": 0.0001,
+      "loss": 1.8349,
+      "step": 344
+    },
+    {
+      "epoch": 0.08377853326857698,
+      "grad_norm": 0.5295913219451904,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 345
+    },
+    {
+      "epoch": 0.0840213695968917,
+      "grad_norm": 0.5335150361061096,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 346
+    },
+    {
+      "epoch": 0.0842642059252064,
+      "grad_norm": 0.5739315748214722,
+      "learning_rate": 0.0001,
+      "loss": 1.7884,
+      "step": 347
+    },
+    {
+      "epoch": 0.08450704225352113,
+      "grad_norm": 0.49940741062164307,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 348
+    },
+    {
+      "epoch": 0.08474987858183584,
+      "grad_norm": 0.5813695788383484,
+      "learning_rate": 0.0001,
+      "loss": 1.7981,
+      "step": 349
+    },
+    {
+      "epoch": 0.08499271491015056,
+      "grad_norm": 0.5490919351577759,
+      "learning_rate": 0.0001,
+      "loss": 1.7776,
+      "step": 350
+    },
+    {
+      "epoch": 0.08523555123846527,
+      "grad_norm": 0.5911244750022888,
+      "learning_rate": 0.0001,
+      "loss": 1.8169,
+      "step": 351
+    },
+    {
+      "epoch": 0.08547838756677999,
+      "grad_norm": 0.5575651526451111,
+      "learning_rate": 0.0001,
+      "loss": 1.7757,
+      "step": 352
+    },
+    {
+      "epoch": 0.0857212238950947,
+      "grad_norm": 0.510899007320404,
+      "learning_rate": 0.0001,
+      "loss": 1.7117,
+      "step": 353
+    },
+    {
+      "epoch": 0.08596406022340942,
+      "grad_norm": 0.5488718152046204,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 354
+    },
+    {
+      "epoch": 0.08620689655172414,
+      "grad_norm": 0.5245040059089661,
+      "learning_rate": 0.0001,
+      "loss": 1.8676,
+      "step": 355
+    },
+    {
+      "epoch": 0.08644973288003885,
+      "grad_norm": 0.543440043926239,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 356
+    },
+    {
+      "epoch": 0.08669256920835357,
+      "grad_norm": 0.5688736438751221,
+      "learning_rate": 0.0001,
+      "loss": 1.7556,
+      "step": 357
+    },
+    {
+      "epoch": 0.08693540553666829,
+      "grad_norm": 0.5552835464477539,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 358
+    },
+    {
+      "epoch": 0.087178241864983,
+      "grad_norm": 0.5715907216072083,
+      "learning_rate": 0.0001,
+      "loss": 1.9301,
+      "step": 359
+    },
+    {
+      "epoch": 0.08742107819329771,
+      "grad_norm": 0.5471308827400208,
+      "learning_rate": 0.0001,
+      "loss": 1.7324,
+      "step": 360
+    },
+    {
+      "epoch": 0.08766391452161243,
+      "grad_norm": 0.5335566997528076,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 361
+    },
+    {
+      "epoch": 0.08790675084992715,
+      "grad_norm": 0.5608660578727722,
+      "learning_rate": 0.0001,
+      "loss": 1.8098,
+      "step": 362
+    },
+    {
+      "epoch": 0.08814958717824187,
+      "grad_norm": 0.5688713788986206,
+      "learning_rate": 0.0001,
+      "loss": 1.8482,
+      "step": 363
+    },
+    {
+      "epoch": 0.08839242350655659,
+      "grad_norm": 0.5294666886329651,
+      "learning_rate": 0.0001,
+      "loss": 1.7772,
+      "step": 364
+    },
+    {
+      "epoch": 0.08863525983487129,
+      "grad_norm": 0.5550897717475891,
+      "learning_rate": 0.0001,
+      "loss": 1.766,
+      "step": 365
+    },
+    {
+      "epoch": 0.08887809616318601,
+      "grad_norm": 0.6038578152656555,
+      "learning_rate": 0.0001,
+      "loss": 1.951,
+      "step": 366
+    },
+    {
+      "epoch": 0.08912093249150073,
+      "grad_norm": 0.5179253220558167,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 367
+    },
+    {
+      "epoch": 0.08936376881981545,
+      "grad_norm": 0.54139643907547,
+      "learning_rate": 0.0001,
+      "loss": 1.8326,
+      "step": 368
+    },
+    {
+      "epoch": 0.08960660514813017,
+      "grad_norm": 0.5586341619491577,
+      "learning_rate": 0.0001,
+      "loss": 1.858,
+      "step": 369
+    },
+    {
+      "epoch": 0.08984944147644487,
+      "grad_norm": 0.5631245374679565,
+      "learning_rate": 0.0001,
+      "loss": 1.8237,
+      "step": 370
+    },
+    {
+      "epoch": 0.09009227780475959,
+      "grad_norm": 0.536080539226532,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 371
+    },
+    {
+      "epoch": 0.09033511413307431,
+      "grad_norm": 0.5671922564506531,
+      "learning_rate": 0.0001,
+      "loss": 1.7635,
+      "step": 372
+    },
+    {
+      "epoch": 0.09057795046138903,
+      "grad_norm": 0.5378475785255432,
+      "learning_rate": 0.0001,
+      "loss": 1.754,
+      "step": 373
+    },
+    {
+      "epoch": 0.09082078678970373,
+      "grad_norm": 0.5762247443199158,
+      "learning_rate": 0.0001,
+      "loss": 1.9381,
+      "step": 374
+    },
+    {
+      "epoch": 0.09106362311801845,
+      "grad_norm": 0.4861982762813568,
+      "learning_rate": 0.0001,
+      "loss": 1.7204,
+      "step": 375
+    },
+    {
+      "epoch": 0.09130645944633317,
+      "grad_norm": 0.5585159063339233,
+      "learning_rate": 0.0001,
+      "loss": 1.8194,
+      "step": 376
+    },
+    {
+      "epoch": 0.09154929577464789,
+      "grad_norm": 0.49415215849876404,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 377
+    },
+    {
+      "epoch": 0.09179213210296261,
+      "grad_norm": 0.5581274032592773,
+      "learning_rate": 0.0001,
+      "loss": 1.8006,
+      "step": 378
+    },
+    {
+      "epoch": 0.09203496843127731,
+      "grad_norm": 0.5406607389450073,
+      "learning_rate": 0.0001,
+      "loss": 1.7876,
+      "step": 379
+    },
+    {
+      "epoch": 0.09227780475959203,
+      "grad_norm": 0.5319144129753113,
+      "learning_rate": 0.0001,
+      "loss": 1.726,
+      "step": 380
+    },
+    {
+      "epoch": 0.09252064108790675,
+      "grad_norm": 0.5684472918510437,
+      "learning_rate": 0.0001,
+      "loss": 1.7701,
+      "step": 381
+    },
+    {
+      "epoch": 0.09276347741622147,
+      "grad_norm": 0.5876220464706421,
+      "learning_rate": 0.0001,
+      "loss": 1.8187,
+      "step": 382
+    },
+    {
+      "epoch": 0.09300631374453618,
+      "grad_norm": 0.5094630122184753,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 383
+    },
+    {
+      "epoch": 0.0932491500728509,
+      "grad_norm": 0.5458014607429504,
+      "learning_rate": 0.0001,
+      "loss": 1.8323,
+      "step": 384
+    },
+    {
+      "epoch": 0.09349198640116561,
+      "grad_norm": 0.5321098566055298,
+      "learning_rate": 0.0001,
+      "loss": 1.7619,
+      "step": 385
+    },
+    {
+      "epoch": 0.09373482272948033,
+      "grad_norm": 0.542500376701355,
+      "learning_rate": 0.0001,
+      "loss": 1.6506,
+      "step": 386
+    },
+    {
+      "epoch": 0.09397765905779505,
+      "grad_norm": 0.5316786170005798,
+      "learning_rate": 0.0001,
+      "loss": 1.8175,
+      "step": 387
+    },
+    {
+      "epoch": 0.09422049538610976,
+      "grad_norm": 0.5709975957870483,
+      "learning_rate": 0.0001,
+      "loss": 1.8854,
+      "step": 388
+    },
+    {
+      "epoch": 0.09446333171442448,
+      "grad_norm": 0.5439728498458862,
+      "learning_rate": 0.0001,
+      "loss": 1.8345,
+      "step": 389
+    },
+    {
+      "epoch": 0.0947061680427392,
+      "grad_norm": 0.5377857089042664,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 390
+    },
+    {
+      "epoch": 0.09494900437105391,
+      "grad_norm": 0.5467667579650879,
+      "learning_rate": 0.0001,
+      "loss": 1.7572,
+      "step": 391
+    },
+    {
+      "epoch": 0.09519184069936863,
+      "grad_norm": 0.5458954572677612,
+      "learning_rate": 0.0001,
+      "loss": 1.7995,
+      "step": 392
+    },
+    {
+      "epoch": 0.09543467702768334,
+      "grad_norm": 0.5365225672721863,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 393
+    },
+    {
+      "epoch": 0.09567751335599806,
+      "grad_norm": 0.5744127631187439,
+      "learning_rate": 0.0001,
+      "loss": 1.8339,
+      "step": 394
+    },
+    {
+      "epoch": 0.09592034968431278,
+      "grad_norm": 0.585131049156189,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 395
+    },
+    {
+      "epoch": 0.0961631860126275,
+      "grad_norm": 0.5712559223175049,
+      "learning_rate": 0.0001,
+      "loss": 1.9143,
+      "step": 396
+    },
+    {
+      "epoch": 0.0964060223409422,
+      "grad_norm": 0.5720378160476685,
+      "learning_rate": 0.0001,
+      "loss": 1.8163,
+      "step": 397
+    },
+    {
+      "epoch": 0.09664885866925692,
+      "grad_norm": 0.5640956163406372,
+      "learning_rate": 0.0001,
+      "loss": 1.8803,
+      "step": 398
+    },
+    {
+      "epoch": 0.09689169499757164,
+      "grad_norm": 0.5292410254478455,
+      "learning_rate": 0.0001,
+      "loss": 1.6497,
+      "step": 399
+    },
+    {
+      "epoch": 0.09713453132588636,
+      "grad_norm": 0.5474023818969727,
+      "learning_rate": 0.0001,
+      "loss": 1.8508,
+      "step": 400
+    },
+    {
+      "epoch": 0.09737736765420107,
+      "grad_norm": 0.5421199202537537,
+      "learning_rate": 0.0001,
+      "loss": 1.9272,
+      "step": 401
+    },
+    {
+      "epoch": 0.09762020398251578,
+      "grad_norm": 0.5888630151748657,
+      "learning_rate": 0.0001,
+      "loss": 1.7785,
+      "step": 402
+    },
+    {
+      "epoch": 0.0978630403108305,
+      "grad_norm": 0.5838062167167664,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 403
+    },
+    {
+      "epoch": 0.09810587663914522,
+      "grad_norm": 0.5154374241828918,
+      "learning_rate": 0.0001,
+      "loss": 1.5593,
+      "step": 404
+    },
+    {
+      "epoch": 0.09834871296745994,
+      "grad_norm": 0.5997213125228882,
+      "learning_rate": 0.0001,
+      "loss": 1.8053,
+      "step": 405
+    },
+    {
+      "epoch": 0.09859154929577464,
+      "grad_norm": 0.572761058807373,
+      "learning_rate": 0.0001,
+      "loss": 1.8878,
+      "step": 406
+    },
+    {
+      "epoch": 0.09883438562408936,
+      "grad_norm": 0.5701769590377808,
+      "learning_rate": 0.0001,
+      "loss": 1.9937,
+      "step": 407
+    },
+    {
+      "epoch": 0.09907722195240408,
+      "grad_norm": 0.5593425035476685,
+      "learning_rate": 0.0001,
+      "loss": 1.8615,
+      "step": 408
+    },
+    {
+      "epoch": 0.0993200582807188,
+      "grad_norm": 0.5380374193191528,
+      "learning_rate": 0.0001,
+      "loss": 1.8052,
+      "step": 409
+    },
+    {
+      "epoch": 0.09956289460903352,
+      "grad_norm": 0.5586409568786621,
+      "learning_rate": 0.0001,
+      "loss": 1.9118,
+      "step": 410
+    },
+    {
+      "epoch": 0.09980573093734822,
+      "grad_norm": 0.5825559496879578,
+      "learning_rate": 0.0001,
+      "loss": 1.9286,
+      "step": 411
+    },
+    {
+      "epoch": 0.10004856726566294,
+      "grad_norm": 0.5659518837928772,
+      "learning_rate": 0.0001,
+      "loss": 1.9078,
+      "step": 412
+    },
+    {
+      "epoch": 0.10029140359397766,
+      "grad_norm": 0.5435416102409363,
+      "learning_rate": 0.0001,
+      "loss": 1.7668,
+      "step": 413
+    },
+    {
+      "epoch": 0.10053423992229238,
+      "grad_norm": 0.5503891706466675,
+      "learning_rate": 0.0001,
+      "loss": 1.9453,
+      "step": 414
+    },
+    {
+      "epoch": 0.1007770762506071,
+      "grad_norm": 0.5533131957054138,
+      "learning_rate": 0.0001,
+      "loss": 1.8397,
+      "step": 415
+    },
+    {
+      "epoch": 0.1010199125789218,
+      "grad_norm": 0.5576867461204529,
+      "learning_rate": 0.0001,
+      "loss": 1.8323,
+      "step": 416
+    },
+    {
+      "epoch": 0.10126274890723652,
+      "grad_norm": 0.5436316132545471,
+      "learning_rate": 0.0001,
+      "loss": 1.8363,
+      "step": 417
+    },
+    {
+      "epoch": 0.10150558523555124,
+      "grad_norm": 0.5690177083015442,
+      "learning_rate": 0.0001,
+      "loss": 1.9228,
+      "step": 418
+    },
+    {
+      "epoch": 0.10174842156386596,
+      "grad_norm": 0.5737699270248413,
+      "learning_rate": 0.0001,
+      "loss": 1.7261,
+      "step": 419
+    },
+    {
+      "epoch": 0.10199125789218066,
+      "grad_norm": 0.6116786599159241,
+      "learning_rate": 0.0001,
+      "loss": 1.738,
+      "step": 420
+    },
+    {
+      "epoch": 0.10223409422049538,
+      "grad_norm": 0.5913645625114441,
+      "learning_rate": 0.0001,
+      "loss": 1.9034,
+      "step": 421
+    },
+    {
+      "epoch": 0.1024769305488101,
+      "grad_norm": 0.537891685962677,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 422
+    },
+    {
+      "epoch": 0.10271976687712482,
+      "grad_norm": 0.5630314350128174,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 423
+    },
+    {
+      "epoch": 0.10296260320543954,
+      "grad_norm": 0.5510636568069458,
+      "learning_rate": 0.0001,
+      "loss": 1.7351,
+      "step": 424
+    },
+    {
+      "epoch": 0.10320543953375425,
+      "grad_norm": 0.5301394462585449,
+      "learning_rate": 0.0001,
+      "loss": 1.8174,
+      "step": 425
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 0.5566518902778625,
+      "learning_rate": 0.0001,
+      "loss": 1.7677,
+      "step": 426
+    },
+    {
+      "epoch": 0.10369111219038368,
+      "grad_norm": 0.5097194314002991,
+      "learning_rate": 0.0001,
+      "loss": 1.5447,
+      "step": 427
+    },
+    {
+      "epoch": 0.1039339485186984,
+      "grad_norm": 0.5431993007659912,
+      "learning_rate": 0.0001,
+      "loss": 1.828,
+      "step": 428
+    },
+    {
+      "epoch": 0.1041767848470131,
+      "grad_norm": 0.5613725781440735,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 429
+    },
+    {
+      "epoch": 0.10441962117532783,
+      "grad_norm": 0.5759568214416504,
+      "learning_rate": 0.0001,
+      "loss": 1.8452,
+      "step": 430
+    },
+    {
+      "epoch": 0.10466245750364254,
+      "grad_norm": 0.544442892074585,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 431
+    },
+    {
+      "epoch": 0.10490529383195726,
+      "grad_norm": 0.5675405263900757,
+      "learning_rate": 0.0001,
+      "loss": 1.8674,
+      "step": 432
+    },
+    {
+      "epoch": 0.10514813016027198,
+      "grad_norm": 0.5533949732780457,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 433
+    },
+    {
+      "epoch": 0.10539096648858669,
+      "grad_norm": 0.5979740023612976,
+      "learning_rate": 0.0001,
+      "loss": 1.736,
+      "step": 434
+    },
+    {
+      "epoch": 0.1056338028169014,
+      "grad_norm": 0.5390839576721191,
+      "learning_rate": 0.0001,
+      "loss": 1.7611,
+      "step": 435
+    },
+    {
+      "epoch": 0.10587663914521613,
+      "grad_norm": 0.5322712659835815,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 436
+    },
+    {
+      "epoch": 0.10611947547353084,
+      "grad_norm": 0.5771969556808472,
+      "learning_rate": 0.0001,
+      "loss": 1.8153,
+      "step": 437
+    },
+    {
+      "epoch": 0.10636231180184555,
+      "grad_norm": 0.568136990070343,
+      "learning_rate": 0.0001,
+      "loss": 1.8945,
+      "step": 438
+    },
+    {
+      "epoch": 0.10660514813016027,
+      "grad_norm": 0.5258873105049133,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 439
+    },
+    {
+      "epoch": 0.10684798445847499,
+      "grad_norm": 0.5741325616836548,
+      "learning_rate": 0.0001,
+      "loss": 1.9262,
+      "step": 440
+    },
+    {
+      "epoch": 0.1070908207867897,
+      "grad_norm": 0.5695716738700867,
+      "learning_rate": 0.0001,
+      "loss": 1.7516,
+      "step": 441
+    },
+    {
+      "epoch": 0.10733365711510442,
+      "grad_norm": 0.5673108696937561,
+      "learning_rate": 0.0001,
+      "loss": 1.787,
+      "step": 442
+    },
+    {
+      "epoch": 0.10757649344341913,
+      "grad_norm": 0.5834952592849731,
+      "learning_rate": 0.0001,
+      "loss": 1.8915,
+      "step": 443
+    },
+    {
+      "epoch": 0.10781932977173385,
+      "grad_norm": 0.5457293391227722,
+      "learning_rate": 0.0001,
+      "loss": 1.8129,
+      "step": 444
+    },
+    {
+      "epoch": 0.10806216610004857,
+      "grad_norm": 0.5748411417007446,
+      "learning_rate": 0.0001,
+      "loss": 1.79,
+      "step": 445
+    },
+    {
+      "epoch": 0.10830500242836329,
+      "grad_norm": 0.5327291488647461,
+      "learning_rate": 0.0001,
+      "loss": 1.6976,
+      "step": 446
+    },
+    {
+      "epoch": 0.108547838756678,
+      "grad_norm": 0.5944225192070007,
+      "learning_rate": 0.0001,
+      "loss": 1.9803,
+      "step": 447
+    },
+    {
+      "epoch": 0.10879067508499271,
+      "grad_norm": 0.5857844352722168,
+      "learning_rate": 0.0001,
+      "loss": 1.8585,
+      "step": 448
+    },
+    {
+      "epoch": 0.10903351141330743,
+      "grad_norm": 0.5551210641860962,
+      "learning_rate": 0.0001,
+      "loss": 1.7326,
+      "step": 449
+    },
+    {
+      "epoch": 0.10927634774162215,
+      "grad_norm": 0.5699621438980103,
+      "learning_rate": 0.0001,
+      "loss": 1.8009,
+      "step": 450
+    },
+    {
+      "epoch": 0.10951918406993687,
+      "grad_norm": 0.5252717733383179,
+      "learning_rate": 0.0001,
+      "loss": 1.7593,
+      "step": 451
+    },
+    {
+      "epoch": 0.10976202039825157,
+      "grad_norm": 0.5586408376693726,
+      "learning_rate": 0.0001,
+      "loss": 1.9124,
+      "step": 452
+    },
+    {
+      "epoch": 0.11000485672656629,
+      "grad_norm": 0.5575873255729675,
+      "learning_rate": 0.0001,
+      "loss": 1.8058,
+      "step": 453
+    },
+    {
+      "epoch": 0.11024769305488101,
+      "grad_norm": 0.5508807897567749,
+      "learning_rate": 0.0001,
+      "loss": 1.7938,
+      "step": 454
+    },
+    {
+      "epoch": 0.11049052938319573,
+      "grad_norm": 0.5738120675086975,
+      "learning_rate": 0.0001,
+      "loss": 1.8046,
+      "step": 455
+    },
+    {
+      "epoch": 0.11073336571151045,
+      "grad_norm": 0.5300111770629883,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 456
+    },
+    {
+      "epoch": 0.11097620203982515,
+      "grad_norm": 0.5261697173118591,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 457
+    },
+    {
+      "epoch": 0.11121903836813987,
+      "grad_norm": 0.5182989835739136,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 458
+    },
+    {
+      "epoch": 0.11146187469645459,
+      "grad_norm": 0.5611332058906555,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 459
+    },
+    {
+      "epoch": 0.11170471102476931,
+      "grad_norm": 0.5654703974723816,
+      "learning_rate": 0.0001,
+      "loss": 1.811,
+      "step": 460
+    },
+    {
+      "epoch": 0.11194754735308401,
+      "grad_norm": 0.5420575737953186,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 461
+    },
+    {
+      "epoch": 0.11219038368139873,
+      "grad_norm": 0.5482935309410095,
+      "learning_rate": 0.0001,
+      "loss": 1.7207,
+      "step": 462
+    },
+    {
+      "epoch": 0.11243322000971345,
+      "grad_norm": 0.5594455599784851,
+      "learning_rate": 0.0001,
+      "loss": 1.7714,
+      "step": 463
+    },
+    {
+      "epoch": 0.11267605633802817,
+      "grad_norm": 0.5463600158691406,
+      "learning_rate": 0.0001,
+      "loss": 1.6993,
+      "step": 464
+    },
+    {
+      "epoch": 0.11291889266634289,
+      "grad_norm": 0.5660488605499268,
+      "learning_rate": 0.0001,
+      "loss": 1.7934,
+      "step": 465
+    },
+    {
+      "epoch": 0.1131617289946576,
+      "grad_norm": 0.5747204422950745,
+      "learning_rate": 0.0001,
+      "loss": 1.7009,
+      "step": 466
+    },
+    {
+      "epoch": 0.11340456532297231,
+      "grad_norm": 0.5203810930252075,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 467
+    },
+    {
+      "epoch": 0.11364740165128703,
+      "grad_norm": 0.5617368221282959,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 468
+    },
+    {
+      "epoch": 0.11389023797960175,
+      "grad_norm": 0.5709205269813538,
+      "learning_rate": 0.0001,
+      "loss": 1.8537,
+      "step": 469
+    },
+    {
+      "epoch": 0.11413307430791647,
+      "grad_norm": 0.5530130863189697,
+      "learning_rate": 0.0001,
+      "loss": 1.7883,
+      "step": 470
+    },
+    {
+      "epoch": 0.11437591063623118,
+      "grad_norm": 0.4973085820674896,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 471
+    },
+    {
+      "epoch": 0.1146187469645459,
+      "grad_norm": 0.5351885557174683,
+      "learning_rate": 0.0001,
+      "loss": 1.7357,
+      "step": 472
+    },
+    {
+      "epoch": 0.11486158329286061,
+      "grad_norm": 0.5383778214454651,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 473
+    },
+    {
+      "epoch": 0.11510441962117533,
+      "grad_norm": 0.5456533432006836,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 474
+    },
+    {
+      "epoch": 0.11534725594949004,
+      "grad_norm": 0.5580303072929382,
+      "learning_rate": 0.0001,
+      "loss": 1.8015,
+      "step": 475
+    },
+    {
+      "epoch": 0.11559009227780476,
+      "grad_norm": 0.5299339890480042,
+      "learning_rate": 0.0001,
+      "loss": 1.7562,
+      "step": 476
+    },
+    {
+      "epoch": 0.11583292860611948,
+      "grad_norm": 0.5573864579200745,
+      "learning_rate": 0.0001,
+      "loss": 1.8338,
+      "step": 477
+    },
+    {
+      "epoch": 0.1160757649344342,
+      "grad_norm": 0.5550556182861328,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 478
+    },
+    {
+      "epoch": 0.11631860126274891,
+      "grad_norm": 0.6052720546722412,
+      "learning_rate": 0.0001,
+      "loss": 1.9028,
+      "step": 479
+    },
+    {
+      "epoch": 0.11656143759106362,
+      "grad_norm": 0.5638125538825989,
+      "learning_rate": 0.0001,
+      "loss": 1.7026,
+      "step": 480
+    },
+    {
+      "epoch": 0.11680427391937834,
+      "grad_norm": 0.5557655096054077,
+      "learning_rate": 0.0001,
+      "loss": 1.7897,
+      "step": 481
+    },
+    {
+      "epoch": 0.11704711024769306,
+      "grad_norm": 0.5520438551902771,
+      "learning_rate": 0.0001,
+      "loss": 1.6795,
+      "step": 482
+    },
+    {
+      "epoch": 0.11728994657600778,
+      "grad_norm": 0.5554108619689941,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 483
+    },
+    {
+      "epoch": 0.11753278290432248,
+      "grad_norm": 0.5558226108551025,
+      "learning_rate": 0.0001,
+      "loss": 1.8267,
+      "step": 484
+    },
+    {
+      "epoch": 0.1177756192326372,
+      "grad_norm": 0.5317284464836121,
+      "learning_rate": 0.0001,
+      "loss": 1.6754,
+      "step": 485
+    },
+    {
+      "epoch": 0.11801845556095192,
+      "grad_norm": 0.4989995062351227,
+      "learning_rate": 0.0001,
+      "loss": 1.6589,
+      "step": 486
+    },
+    {
+      "epoch": 0.11826129188926664,
+      "grad_norm": 0.5741257071495056,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 487
+    },
+    {
+      "epoch": 0.11850412821758136,
+      "grad_norm": 0.5618738532066345,
+      "learning_rate": 0.0001,
+      "loss": 1.7277,
+      "step": 488
+    },
+    {
+      "epoch": 0.11874696454589606,
+      "grad_norm": 0.5853535532951355,
+      "learning_rate": 0.0001,
+      "loss": 1.7813,
+      "step": 489
+    },
+    {
+      "epoch": 0.11898980087421078,
+      "grad_norm": 0.6120092868804932,
+      "learning_rate": 0.0001,
+      "loss": 1.9246,
+      "step": 490
+    },
+    {
+      "epoch": 0.1192326372025255,
+      "grad_norm": 0.5622900724411011,
+      "learning_rate": 0.0001,
+      "loss": 1.8058,
+      "step": 491
+    },
+    {
+      "epoch": 0.11947547353084022,
+      "grad_norm": 0.5811873078346252,
+      "learning_rate": 0.0001,
+      "loss": 1.8664,
+      "step": 492
+    },
+    {
+      "epoch": 0.11971830985915492,
+      "grad_norm": 0.5615583062171936,
+      "learning_rate": 0.0001,
+      "loss": 1.7442,
+      "step": 493
+    },
+    {
+      "epoch": 0.11996114618746964,
+      "grad_norm": 0.5460301041603088,
+      "learning_rate": 0.0001,
+      "loss": 1.7299,
+      "step": 494
+    },
+    {
+      "epoch": 0.12020398251578436,
+      "grad_norm": 0.5650546550750732,
+      "learning_rate": 0.0001,
+      "loss": 1.6994,
+      "step": 495
+    },
+    {
+      "epoch": 0.12044681884409908,
+      "grad_norm": 0.6062639951705933,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 496
+    },
+    {
+      "epoch": 0.1206896551724138,
+      "grad_norm": 0.5884907841682434,
+      "learning_rate": 0.0001,
+      "loss": 1.881,
+      "step": 497
+    },
+    {
+      "epoch": 0.1209324915007285,
+      "grad_norm": 0.5676560997962952,
+      "learning_rate": 0.0001,
+      "loss": 1.9511,
+      "step": 498
+    },
+    {
+      "epoch": 0.12117532782904322,
+      "grad_norm": 0.5525282025337219,
+      "learning_rate": 0.0001,
+      "loss": 1.8009,
+      "step": 499
+    },
+    {
+      "epoch": 0.12141816415735794,
+      "grad_norm": 0.5765119194984436,
+      "learning_rate": 0.0001,
+      "loss": 1.8173,
+      "step": 500
+    },
+    {
+      "epoch": 0.12166100048567266,
+      "grad_norm": 0.5602638125419617,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 501
+    },
+    {
+      "epoch": 0.12190383681398738,
+      "grad_norm": 0.5428099036216736,
+      "learning_rate": 0.0001,
+      "loss": 1.6568,
+      "step": 502
+    },
+    {
+      "epoch": 0.12214667314230208,
+      "grad_norm": 0.5926585793495178,
+      "learning_rate": 0.0001,
+      "loss": 1.9099,
+      "step": 503
+    },
+    {
+      "epoch": 0.1223895094706168,
+      "grad_norm": 0.5717451572418213,
+      "learning_rate": 0.0001,
+      "loss": 1.6819,
+      "step": 504
+    },
+    {
+      "epoch": 0.12263234579893152,
+      "grad_norm": 0.5724567770957947,
+      "learning_rate": 0.0001,
+      "loss": 1.7439,
+      "step": 505
+    },
+    {
+      "epoch": 0.12287518212724624,
+      "grad_norm": 0.5597602725028992,
+      "learning_rate": 0.0001,
+      "loss": 1.7791,
+      "step": 506
+    },
+    {
+      "epoch": 0.12311801845556095,
+      "grad_norm": 0.5225638151168823,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 507
+    },
+    {
+      "epoch": 0.12336085478387566,
+      "grad_norm": 0.5752341747283936,
+      "learning_rate": 0.0001,
+      "loss": 1.8222,
+      "step": 508
+    },
+    {
+      "epoch": 0.12360369111219038,
+      "grad_norm": 0.5637819170951843,
+      "learning_rate": 0.0001,
+      "loss": 1.8684,
+      "step": 509
+    },
+    {
+      "epoch": 0.1238465274405051,
+      "grad_norm": 0.5527138113975525,
+      "learning_rate": 0.0001,
+      "loss": 1.8502,
+      "step": 510
+    },
+    {
+      "epoch": 0.12408936376881982,
+      "grad_norm": 0.5274344086647034,
+      "learning_rate": 0.0001,
+      "loss": 1.6784,
+      "step": 511
+    },
+    {
+      "epoch": 0.12433220009713453,
+      "grad_norm": 0.500634491443634,
+      "learning_rate": 0.0001,
+      "loss": 1.6773,
+      "step": 512
+    },
+    {
+      "epoch": 0.12457503642544925,
+      "grad_norm": 0.5613259077072144,
+      "learning_rate": 0.0001,
+      "loss": 1.7038,
+      "step": 513
+    },
+    {
+      "epoch": 0.12481787275376396,
+      "grad_norm": 0.5362138152122498,
+      "learning_rate": 0.0001,
+      "loss": 1.6937,
+      "step": 514
+    },
+    {
+      "epoch": 0.12506070908207867,
+      "grad_norm": 0.6005094647407532,
+      "learning_rate": 0.0001,
+      "loss": 1.9453,
+      "step": 515
+    },
+    {
+      "epoch": 0.1253035454103934,
+      "grad_norm": 0.5446475148200989,
+      "learning_rate": 0.0001,
+      "loss": 1.7107,
+      "step": 516
+    },
+    {
+      "epoch": 0.1255463817387081,
+      "grad_norm": 0.5308226346969604,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 517
+    },
+    {
+      "epoch": 0.12578921806702284,
+      "grad_norm": 0.5435710549354553,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 518
+    },
+    {
+      "epoch": 0.12603205439533754,
+      "grad_norm": 0.5712341666221619,
+      "learning_rate": 0.0001,
+      "loss": 1.971,
+      "step": 519
+    },
+    {
+      "epoch": 0.12627489072365225,
+      "grad_norm": 0.5840036869049072,
+      "learning_rate": 0.0001,
+      "loss": 1.8336,
+      "step": 520
+    },
+    {
+      "epoch": 0.12651772705196698,
+      "grad_norm": 0.5734742283821106,
+      "learning_rate": 0.0001,
+      "loss": 1.8283,
+      "step": 521
+    },
+    {
+      "epoch": 0.1267605633802817,
+      "grad_norm": 0.5368530750274658,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 522
+    },
+    {
+      "epoch": 0.1270033997085964,
+      "grad_norm": 0.6023390293121338,
+      "learning_rate": 0.0001,
+      "loss": 1.8824,
+      "step": 523
+    },
+    {
+      "epoch": 0.12724623603691113,
+      "grad_norm": 0.5599238872528076,
+      "learning_rate": 0.0001,
+      "loss": 1.8165,
+      "step": 524
+    },
+    {
+      "epoch": 0.12748907236522583,
+      "grad_norm": 0.5743386745452881,
+      "learning_rate": 0.0001,
+      "loss": 1.7759,
+      "step": 525
+    },
+    {
+      "epoch": 0.12773190869354056,
+      "grad_norm": 0.5817521810531616,
+      "learning_rate": 0.0001,
+      "loss": 1.933,
+      "step": 526
+    },
+    {
+      "epoch": 0.12797474502185527,
+      "grad_norm": 0.5742323994636536,
+      "learning_rate": 0.0001,
+      "loss": 1.7725,
+      "step": 527
+    },
+    {
+      "epoch": 0.12821758135016997,
+      "grad_norm": 0.5679665207862854,
+      "learning_rate": 0.0001,
+      "loss": 1.8016,
+      "step": 528
+    },
+    {
+      "epoch": 0.1284604176784847,
+      "grad_norm": 0.5591177940368652,
+      "learning_rate": 0.0001,
+      "loss": 1.8947,
+      "step": 529
+    },
+    {
+      "epoch": 0.1287032540067994,
+      "grad_norm": 0.5632833242416382,
+      "learning_rate": 0.0001,
+      "loss": 1.8154,
+      "step": 530
+    },
+    {
+      "epoch": 0.12894609033511414,
+      "grad_norm": 0.5830283761024475,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 531
+    },
+    {
+      "epoch": 0.12918892666342885,
+      "grad_norm": 0.5660063624382019,
+      "learning_rate": 0.0001,
+      "loss": 1.8233,
+      "step": 532
+    },
+    {
+      "epoch": 0.12943176299174355,
+      "grad_norm": 0.5650138258934021,
+      "learning_rate": 0.0001,
+      "loss": 1.7795,
+      "step": 533
+    },
+    {
+      "epoch": 0.1296745993200583,
+      "grad_norm": 0.6010497808456421,
+      "learning_rate": 0.0001,
+      "loss": 1.812,
+      "step": 534
+    },
+    {
+      "epoch": 0.129917435648373,
+      "grad_norm": 0.5512215495109558,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 535
+    },
+    {
+      "epoch": 0.13016027197668772,
+      "grad_norm": 0.5600783228874207,
+      "learning_rate": 0.0001,
+      "loss": 1.7371,
+      "step": 536
+    },
+    {
+      "epoch": 0.13040310830500243,
+      "grad_norm": 0.5541916489601135,
+      "learning_rate": 0.0001,
+      "loss": 1.7264,
+      "step": 537
+    },
+    {
+      "epoch": 0.13064594463331713,
+      "grad_norm": 0.5739514827728271,
+      "learning_rate": 0.0001,
+      "loss": 1.7232,
+      "step": 538
+    },
+    {
+      "epoch": 0.13088878096163187,
+      "grad_norm": 0.5634263157844543,
+      "learning_rate": 0.0001,
+      "loss": 1.7582,
+      "step": 539
+    },
+    {
+      "epoch": 0.13113161728994657,
+      "grad_norm": 0.5988732576370239,
+      "learning_rate": 0.0001,
+      "loss": 1.7761,
+      "step": 540
+    },
+    {
+      "epoch": 0.1313744536182613,
+      "grad_norm": 0.5475578308105469,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 541
+    },
+    {
+      "epoch": 0.131617289946576,
+      "grad_norm": 0.5896528363227844,
+      "learning_rate": 0.0001,
+      "loss": 1.828,
+      "step": 542
+    },
+    {
+      "epoch": 0.13186012627489072,
+      "grad_norm": 0.5900474786758423,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 543
+    },
+    {
+      "epoch": 0.13210296260320545,
+      "grad_norm": 0.583486795425415,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 544
+    },
+    {
+      "epoch": 0.13234579893152015,
+      "grad_norm": 0.5893517732620239,
+      "learning_rate": 0.0001,
+      "loss": 1.8338,
+      "step": 545
+    },
+    {
+      "epoch": 0.13258863525983486,
+      "grad_norm": 0.5527064204216003,
+      "learning_rate": 0.0001,
+      "loss": 1.8413,
+      "step": 546
+    },
+    {
+      "epoch": 0.1328314715881496,
+      "grad_norm": 0.6100444793701172,
+      "learning_rate": 0.0001,
+      "loss": 1.8663,
+      "step": 547
+    },
+    {
+      "epoch": 0.1330743079164643,
+      "grad_norm": 0.5398809313774109,
+      "learning_rate": 0.0001,
+      "loss": 1.7586,
+      "step": 548
+    },
+    {
+      "epoch": 0.13331714424477903,
+      "grad_norm": 0.5335167050361633,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 549
+    },
+    {
+      "epoch": 0.13355998057309373,
+      "grad_norm": 0.5480557680130005,
+      "learning_rate": 0.0001,
+      "loss": 1.8034,
+      "step": 550
+    },
+    {
+      "epoch": 0.13380281690140844,
+      "grad_norm": 0.5606412291526794,
+      "learning_rate": 0.0001,
+      "loss": 1.7259,
+      "step": 551
+    },
+    {
+      "epoch": 0.13404565322972317,
+      "grad_norm": 0.5881384015083313,
+      "learning_rate": 0.0001,
+      "loss": 1.9435,
+      "step": 552
+    },
+    {
+      "epoch": 0.13428848955803788,
+      "grad_norm": 0.5396336317062378,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 553
+    },
+    {
+      "epoch": 0.1345313258863526,
+      "grad_norm": 0.5790011286735535,
+      "learning_rate": 0.0001,
+      "loss": 1.9107,
+      "step": 554
+    },
+    {
+      "epoch": 0.13477416221466731,
+      "grad_norm": 0.5559581518173218,
+      "learning_rate": 0.0001,
+      "loss": 1.8296,
+      "step": 555
+    },
+    {
+      "epoch": 0.13501699854298202,
+      "grad_norm": 0.5403438210487366,
+      "learning_rate": 0.0001,
+      "loss": 1.6947,
+      "step": 556
+    },
+    {
+      "epoch": 0.13525983487129675,
+      "grad_norm": 0.5910641551017761,
+      "learning_rate": 0.0001,
+      "loss": 1.9624,
+      "step": 557
+    },
+    {
+      "epoch": 0.13550267119961146,
+      "grad_norm": 0.5610319375991821,
+      "learning_rate": 0.0001,
+      "loss": 1.8825,
+      "step": 558
+    },
+    {
+      "epoch": 0.1357455075279262,
+      "grad_norm": 0.5739595293998718,
+      "learning_rate": 0.0001,
+      "loss": 1.9115,
+      "step": 559
+    },
+    {
+      "epoch": 0.1359883438562409,
+      "grad_norm": 0.5951991081237793,
+      "learning_rate": 0.0001,
+      "loss": 1.8239,
+      "step": 560
+    },
+    {
+      "epoch": 0.1362311801845556,
+      "grad_norm": 0.5959383845329285,
+      "learning_rate": 0.0001,
+      "loss": 1.7641,
+      "step": 561
+    },
+    {
+      "epoch": 0.13647401651287033,
+      "grad_norm": 0.5285969376564026,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 562
+    },
+    {
+      "epoch": 0.13671685284118504,
+      "grad_norm": 0.6021578311920166,
+      "learning_rate": 0.0001,
+      "loss": 1.7826,
+      "step": 563
+    },
+    {
+      "epoch": 0.13695968916949977,
+      "grad_norm": 0.5408231616020203,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 564
+    },
+    {
+      "epoch": 0.13720252549781448,
+      "grad_norm": 0.5222688317298889,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 565
+    },
+    {
+      "epoch": 0.13744536182612918,
+      "grad_norm": 0.5697501301765442,
+      "learning_rate": 0.0001,
+      "loss": 1.7813,
+      "step": 566
+    },
+    {
+      "epoch": 0.1376881981544439,
+      "grad_norm": 0.5833176970481873,
+      "learning_rate": 0.0001,
+      "loss": 1.8235,
+      "step": 567
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 0.5586931705474854,
+      "learning_rate": 0.0001,
+      "loss": 1.8319,
+      "step": 568
+    },
+    {
+      "epoch": 0.13817387081107332,
+      "grad_norm": 0.5566940903663635,
+      "learning_rate": 0.0001,
+      "loss": 1.8003,
+      "step": 569
+    },
+    {
+      "epoch": 0.13841670713938806,
+      "grad_norm": 0.5803408622741699,
+      "learning_rate": 0.0001,
+      "loss": 1.8709,
+      "step": 570
+    },
+    {
+      "epoch": 0.13865954346770276,
+      "grad_norm": 0.5673038363456726,
+      "learning_rate": 0.0001,
+      "loss": 1.8451,
+      "step": 571
+    },
+    {
+      "epoch": 0.1389023797960175,
+      "grad_norm": 0.537318766117096,
+      "learning_rate": 0.0001,
+      "loss": 1.7931,
+      "step": 572
+    },
+    {
+      "epoch": 0.1391452161243322,
+      "grad_norm": 0.5993166565895081,
+      "learning_rate": 0.0001,
+      "loss": 1.8135,
+      "step": 573
+    },
+    {
+      "epoch": 0.1393880524526469,
+      "grad_norm": 0.6365535259246826,
+      "learning_rate": 0.0001,
+      "loss": 1.8214,
+      "step": 574
+    },
+    {
+      "epoch": 0.13963088878096164,
+      "grad_norm": 0.5775412917137146,
+      "learning_rate": 0.0001,
+      "loss": 1.7692,
+      "step": 575
+    },
+    {
+      "epoch": 0.13987372510927634,
+      "grad_norm": 0.5516401529312134,
+      "learning_rate": 0.0001,
+      "loss": 1.8644,
+      "step": 576
+    },
+    {
+      "epoch": 0.14011656143759108,
+      "grad_norm": 0.5734835863113403,
+      "learning_rate": 0.0001,
+      "loss": 1.7655,
+      "step": 577
+    },
+    {
+      "epoch": 0.14035939776590578,
+      "grad_norm": 0.5712597966194153,
+      "learning_rate": 0.0001,
+      "loss": 1.7431,
+      "step": 578
+    },
+    {
+      "epoch": 0.14060223409422049,
+      "grad_norm": 0.5707169771194458,
+      "learning_rate": 0.0001,
+      "loss": 1.8038,
+      "step": 579
+    },
+    {
+      "epoch": 0.14084507042253522,
+      "grad_norm": 0.611261785030365,
+      "learning_rate": 0.0001,
+      "loss": 1.7464,
+      "step": 580
+    },
+    {
+      "epoch": 0.14108790675084992,
+      "grad_norm": 0.5502864718437195,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 581
+    },
+    {
+      "epoch": 0.14133074307916466,
+      "grad_norm": 0.5708633065223694,
+      "learning_rate": 0.0001,
+      "loss": 1.7671,
+      "step": 582
+    },
+    {
+      "epoch": 0.14157357940747936,
+      "grad_norm": 0.5802814960479736,
+      "learning_rate": 0.0001,
+      "loss": 1.7979,
+      "step": 583
+    },
+    {
+      "epoch": 0.14181641573579407,
+      "grad_norm": 0.5676042437553406,
+      "learning_rate": 0.0001,
+      "loss": 1.7693,
+      "step": 584
+    },
+    {
+      "epoch": 0.1420592520641088,
+      "grad_norm": 0.5899032950401306,
+      "learning_rate": 0.0001,
+      "loss": 1.7929,
+      "step": 585
+    },
+    {
+      "epoch": 0.1423020883924235,
+      "grad_norm": 0.5601672530174255,
+      "learning_rate": 0.0001,
+      "loss": 1.8005,
+      "step": 586
+    },
+    {
+      "epoch": 0.1425449247207382,
+      "grad_norm": 0.5701844096183777,
+      "learning_rate": 0.0001,
+      "loss": 1.8274,
+      "step": 587
+    },
+    {
+      "epoch": 0.14278776104905294,
+      "grad_norm": 0.5746443271636963,
+      "learning_rate": 0.0001,
+      "loss": 1.8188,
+      "step": 588
+    },
+    {
+      "epoch": 0.14303059737736765,
+      "grad_norm": 0.520227313041687,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 589
+    },
+    {
+      "epoch": 0.14327343370568238,
+      "grad_norm": 0.5633252263069153,
+      "learning_rate": 0.0001,
+      "loss": 1.9063,
+      "step": 590
+    },
+    {
+      "epoch": 0.14351627003399708,
+      "grad_norm": 0.5445085763931274,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 591
+    },
+    {
+      "epoch": 0.1437591063623118,
+      "grad_norm": 0.5793913006782532,
+      "learning_rate": 0.0001,
+      "loss": 1.7959,
+      "step": 592
+    },
+    {
+      "epoch": 0.14400194269062652,
+      "grad_norm": 0.561595618724823,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 593
+    },
+    {
+      "epoch": 0.14424477901894123,
+      "grad_norm": 0.5914781093597412,
+      "learning_rate": 0.0001,
+      "loss": 1.9152,
+      "step": 594
+    },
+    {
+      "epoch": 0.14448761534725596,
+      "grad_norm": 0.5659077763557434,
+      "learning_rate": 0.0001,
+      "loss": 1.8534,
+      "step": 595
+    },
+    {
+      "epoch": 0.14473045167557066,
+      "grad_norm": 0.5098112225532532,
+      "learning_rate": 0.0001,
+      "loss": 1.5326,
+      "step": 596
+    },
+    {
+      "epoch": 0.14497328800388537,
+      "grad_norm": 0.5428852438926697,
+      "learning_rate": 0.0001,
+      "loss": 1.8016,
+      "step": 597
+    },
+    {
+      "epoch": 0.1452161243322001,
+      "grad_norm": 0.549568235874176,
+      "learning_rate": 0.0001,
+      "loss": 1.7529,
+      "step": 598
+    },
+    {
+      "epoch": 0.1454589606605148,
+      "grad_norm": 0.5077096223831177,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 599
+    },
+    {
+      "epoch": 0.14570179698882954,
+      "grad_norm": 0.5143014788627625,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 600
+    },
+    {
+      "epoch": 0.14594463331714425,
+      "grad_norm": 0.588847279548645,
+      "learning_rate": 0.0001,
+      "loss": 1.84,
+      "step": 601
+    },
+    {
+      "epoch": 0.14618746964545895,
+      "grad_norm": 0.5263761281967163,
+      "learning_rate": 0.0001,
+      "loss": 1.7258,
+      "step": 602
+    },
+    {
+      "epoch": 0.14643030597377368,
+      "grad_norm": 0.5702909827232361,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 603
+    },
+    {
+      "epoch": 0.1466731423020884,
+      "grad_norm": 0.5721048712730408,
+      "learning_rate": 0.0001,
+      "loss": 1.8626,
+      "step": 604
+    },
+    {
+      "epoch": 0.14691597863040312,
+      "grad_norm": 0.5193774700164795,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 605
+    },
+    {
+      "epoch": 0.14715881495871783,
+      "grad_norm": 0.5777069926261902,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 606
+    },
+    {
+      "epoch": 0.14740165128703253,
+      "grad_norm": 0.5141522884368896,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 607
+    },
+    {
+      "epoch": 0.14764448761534726,
+      "grad_norm": 0.5801394581794739,
+      "learning_rate": 0.0001,
+      "loss": 1.7541,
+      "step": 608
+    },
+    {
+      "epoch": 0.14788732394366197,
+      "grad_norm": 0.5814424157142639,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 609
+    },
+    {
+      "epoch": 0.14813016027197667,
+      "grad_norm": 0.5588964223861694,
+      "learning_rate": 0.0001,
+      "loss": 1.8527,
+      "step": 610
+    },
+    {
+      "epoch": 0.1483729966002914,
+      "grad_norm": 0.5577553510665894,
+      "learning_rate": 0.0001,
+      "loss": 1.8548,
+      "step": 611
+    },
+    {
+      "epoch": 0.1486158329286061,
+      "grad_norm": 0.5552173852920532,
+      "learning_rate": 0.0001,
+      "loss": 1.8248,
+      "step": 612
+    },
+    {
+      "epoch": 0.14885866925692084,
+      "grad_norm": 0.5496284365653992,
+      "learning_rate": 0.0001,
+      "loss": 1.9658,
+      "step": 613
+    },
+    {
+      "epoch": 0.14910150558523555,
+      "grad_norm": 0.5887159705162048,
+      "learning_rate": 0.0001,
+      "loss": 1.7976,
+      "step": 614
+    },
+    {
+      "epoch": 0.14934434191355025,
+      "grad_norm": 0.5603215098381042,
+      "learning_rate": 0.0001,
+      "loss": 1.8203,
+      "step": 615
+    },
+    {
+      "epoch": 0.149587178241865,
+      "grad_norm": 0.5411328077316284,
+      "learning_rate": 0.0001,
+      "loss": 1.807,
+      "step": 616
+    },
+    {
+      "epoch": 0.1498300145701797,
+      "grad_norm": 0.5373426675796509,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 617
+    },
+    {
+      "epoch": 0.15007285089849443,
+      "grad_norm": 0.5616528987884521,
+      "learning_rate": 0.0001,
+      "loss": 1.8262,
+      "step": 618
+    },
+    {
+      "epoch": 0.15031568722680913,
+      "grad_norm": 0.5507047772407532,
+      "learning_rate": 0.0001,
+      "loss": 1.8568,
+      "step": 619
+    },
+    {
+      "epoch": 0.15055852355512384,
+      "grad_norm": 0.5572707056999207,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 620
+    },
+    {
+      "epoch": 0.15080135988343857,
+      "grad_norm": 0.6481385231018066,
+      "learning_rate": 0.0001,
+      "loss": 1.7724,
+      "step": 621
+    },
+    {
+      "epoch": 0.15104419621175327,
+      "grad_norm": 0.5856874585151672,
+      "learning_rate": 0.0001,
+      "loss": 1.7932,
+      "step": 622
+    },
+    {
+      "epoch": 0.151287032540068,
+      "grad_norm": 0.5337699055671692,
+      "learning_rate": 0.0001,
+      "loss": 1.7423,
+      "step": 623
+    },
+    {
+      "epoch": 0.1515298688683827,
+      "grad_norm": 0.5539548993110657,
+      "learning_rate": 0.0001,
+      "loss": 1.8657,
+      "step": 624
+    },
+    {
+      "epoch": 0.15177270519669742,
+      "grad_norm": 0.5827417373657227,
+      "learning_rate": 0.0001,
+      "loss": 1.9231,
+      "step": 625
+    },
+    {
+      "epoch": 0.15201554152501215,
+      "grad_norm": 0.5438136458396912,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 626
+    },
+    {
+      "epoch": 0.15225837785332685,
+      "grad_norm": 0.525128185749054,
+      "learning_rate": 0.0001,
+      "loss": 1.6657,
+      "step": 627
+    },
+    {
+      "epoch": 0.1525012141816416,
+      "grad_norm": 0.5475497841835022,
+      "learning_rate": 0.0001,
+      "loss": 1.7372,
+      "step": 628
+    },
+    {
+      "epoch": 0.1527440505099563,
+      "grad_norm": 0.5585061311721802,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 629
+    },
+    {
+      "epoch": 0.152986886838271,
+      "grad_norm": 0.5918571949005127,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 630
+    },
+    {
+      "epoch": 0.15322972316658573,
+      "grad_norm": 0.5802246332168579,
+      "learning_rate": 0.0001,
+      "loss": 1.814,
+      "step": 631
+    },
+    {
+      "epoch": 0.15347255949490043,
+      "grad_norm": 0.5600754022598267,
+      "learning_rate": 0.0001,
+      "loss": 1.7924,
+      "step": 632
+    },
+    {
+      "epoch": 0.15371539582321514,
+      "grad_norm": 0.5247310996055603,
+      "learning_rate": 0.0001,
+      "loss": 1.7274,
+      "step": 633
+    },
+    {
+      "epoch": 0.15395823215152987,
+      "grad_norm": 0.5142088532447815,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 634
+    },
+    {
+      "epoch": 0.15420106847984458,
+      "grad_norm": 0.5472610592842102,
+      "learning_rate": 0.0001,
+      "loss": 1.8009,
+      "step": 635
+    },
+    {
+      "epoch": 0.1544439048081593,
+      "grad_norm": 0.5571399331092834,
+      "learning_rate": 0.0001,
+      "loss": 1.9098,
+      "step": 636
+    },
+    {
+      "epoch": 0.15468674113647402,
+      "grad_norm": 0.534423291683197,
+      "learning_rate": 0.0001,
+      "loss": 1.7576,
+      "step": 637
+    },
+    {
+      "epoch": 0.15492957746478872,
+      "grad_norm": 0.5361340641975403,
+      "learning_rate": 0.0001,
+      "loss": 1.7545,
+      "step": 638
+    },
+    {
+      "epoch": 0.15517241379310345,
+      "grad_norm": 0.5327625274658203,
+      "learning_rate": 0.0001,
+      "loss": 1.7251,
+      "step": 639
+    },
+    {
+      "epoch": 0.15541525012141816,
+      "grad_norm": 0.6097555160522461,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 640
+    },
+    {
+      "epoch": 0.1556580864497329,
+      "grad_norm": 0.5422442555427551,
+      "learning_rate": 0.0001,
+      "loss": 1.7612,
+      "step": 641
+    },
+    {
+      "epoch": 0.1559009227780476,
+      "grad_norm": 0.5585924983024597,
+      "learning_rate": 0.0001,
+      "loss": 1.7741,
+      "step": 642
+    },
+    {
+      "epoch": 0.1561437591063623,
+      "grad_norm": 0.583871603012085,
+      "learning_rate": 0.0001,
+      "loss": 1.7839,
+      "step": 643
+    },
+    {
+      "epoch": 0.15638659543467703,
+      "grad_norm": 0.5999270677566528,
+      "learning_rate": 0.0001,
+      "loss": 1.7371,
+      "step": 644
+    },
+    {
+      "epoch": 0.15662943176299174,
+      "grad_norm": 0.5860697031021118,
+      "learning_rate": 0.0001,
+      "loss": 1.9128,
+      "step": 645
+    },
+    {
+      "epoch": 0.15687226809130647,
+      "grad_norm": 0.5972449779510498,
+      "learning_rate": 0.0001,
+      "loss": 1.983,
+      "step": 646
+    },
+    {
+      "epoch": 0.15711510441962118,
+      "grad_norm": 0.5599138140678406,
+      "learning_rate": 0.0001,
+      "loss": 1.7721,
+      "step": 647
+    },
+    {
+      "epoch": 0.15735794074793588,
+      "grad_norm": 0.588979184627533,
+      "learning_rate": 0.0001,
+      "loss": 1.8579,
+      "step": 648
+    },
+    {
+      "epoch": 0.15760077707625061,
+      "grad_norm": 0.5571702718734741,
+      "learning_rate": 0.0001,
+      "loss": 1.9163,
+      "step": 649
+    },
+    {
+      "epoch": 0.15784361340456532,
+      "grad_norm": 0.536984920501709,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 650
+    },
+    {
+      "epoch": 0.15808644973288005,
+      "grad_norm": 0.5638613700866699,
+      "learning_rate": 0.0001,
+      "loss": 1.7381,
+      "step": 651
+    },
+    {
+      "epoch": 0.15832928606119476,
+      "grad_norm": 0.5769402980804443,
+      "learning_rate": 0.0001,
+      "loss": 1.9559,
+      "step": 652
+    },
+    {
+      "epoch": 0.15857212238950946,
+      "grad_norm": 0.5716903805732727,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 653
+    },
+    {
+      "epoch": 0.1588149587178242,
+      "grad_norm": 0.5627726316452026,
+      "learning_rate": 0.0001,
+      "loss": 1.8505,
+      "step": 654
+    },
+    {
+      "epoch": 0.1590577950461389,
+      "grad_norm": 0.5364795923233032,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 655
+    },
+    {
+      "epoch": 0.1593006313744536,
+      "grad_norm": 0.5715546607971191,
+      "learning_rate": 0.0001,
+      "loss": 1.8778,
+      "step": 656
+    },
+    {
+      "epoch": 0.15954346770276834,
+      "grad_norm": 0.5425900220870972,
+      "learning_rate": 0.0001,
+      "loss": 1.7544,
+      "step": 657
+    },
+    {
+      "epoch": 0.15978630403108304,
+      "grad_norm": 0.5716409683227539,
+      "learning_rate": 0.0001,
+      "loss": 1.8701,
+      "step": 658
+    },
+    {
+      "epoch": 0.16002914035939778,
+      "grad_norm": 0.566852331161499,
+      "learning_rate": 0.0001,
+      "loss": 1.8293,
+      "step": 659
+    },
+    {
+      "epoch": 0.16027197668771248,
+      "grad_norm": 0.5446115732192993,
+      "learning_rate": 0.0001,
+      "loss": 1.7935,
+      "step": 660
+    },
+    {
+      "epoch": 0.16051481301602719,
+      "grad_norm": 0.5492187738418579,
+      "learning_rate": 0.0001,
+      "loss": 1.7554,
+      "step": 661
+    },
+    {
+      "epoch": 0.16075764934434192,
+      "grad_norm": 0.5912359952926636,
+      "learning_rate": 0.0001,
+      "loss": 1.8386,
+      "step": 662
+    },
+    {
+      "epoch": 0.16100048567265662,
+      "grad_norm": 0.5367072224617004,
+      "learning_rate": 0.0001,
+      "loss": 1.7206,
+      "step": 663
+    },
+    {
+      "epoch": 0.16124332200097136,
+      "grad_norm": 0.5620670914649963,
+      "learning_rate": 0.0001,
+      "loss": 1.8007,
+      "step": 664
+    },
+    {
+      "epoch": 0.16148615832928606,
+      "grad_norm": 0.5639497637748718,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 665
+    },
+    {
+      "epoch": 0.16172899465760077,
+      "grad_norm": 0.5727313756942749,
+      "learning_rate": 0.0001,
+      "loss": 1.7677,
+      "step": 666
+    },
+    {
+      "epoch": 0.1619718309859155,
+      "grad_norm": 0.5115989446640015,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 667
+    },
+    {
+      "epoch": 0.1622146673142302,
+      "grad_norm": 0.5412555932998657,
+      "learning_rate": 0.0001,
+      "loss": 1.7562,
+      "step": 668
+    },
+    {
+      "epoch": 0.16245750364254494,
+      "grad_norm": 0.61517733335495,
+      "learning_rate": 0.0001,
+      "loss": 1.8744,
+      "step": 669
+    },
+    {
+      "epoch": 0.16270033997085964,
+      "grad_norm": 0.5591433048248291,
+      "learning_rate": 0.0001,
+      "loss": 1.7625,
+      "step": 670
+    },
+    {
+      "epoch": 0.16294317629917435,
+      "grad_norm": 0.5341270565986633,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 671
+    },
+    {
+      "epoch": 0.16318601262748908,
+      "grad_norm": 0.5355855822563171,
+      "learning_rate": 0.0001,
+      "loss": 1.7055,
+      "step": 672
+    },
+    {
+      "epoch": 0.16342884895580378,
+      "grad_norm": 0.5489002466201782,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 673
+    },
+    {
+      "epoch": 0.16367168528411852,
+      "grad_norm": 0.5780882835388184,
+      "learning_rate": 0.0001,
+      "loss": 1.883,
+      "step": 674
+    },
+    {
+      "epoch": 0.16391452161243322,
+      "grad_norm": 0.6101503372192383,
+      "learning_rate": 0.0001,
+      "loss": 2.0329,
+      "step": 675
+    },
+    {
+      "epoch": 0.16415735794074793,
+      "grad_norm": 0.5983835458755493,
+      "learning_rate": 0.0001,
+      "loss": 1.883,
+      "step": 676
+    },
+    {
+      "epoch": 0.16440019426906266,
+      "grad_norm": 0.5713289380073547,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 677
+    },
+    {
+      "epoch": 0.16464303059737737,
+      "grad_norm": 0.5344922542572021,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 678
+    },
+    {
+      "epoch": 0.16488586692569207,
+      "grad_norm": 0.5629361867904663,
+      "learning_rate": 0.0001,
+      "loss": 1.8222,
+      "step": 679
+    },
+    {
+      "epoch": 0.1651287032540068,
+      "grad_norm": 0.572493314743042,
+      "learning_rate": 0.0001,
+      "loss": 1.7996,
+      "step": 680
+    },
+    {
+      "epoch": 0.1653715395823215,
+      "grad_norm": 0.545283317565918,
+      "learning_rate": 0.0001,
+      "loss": 1.7302,
+      "step": 681
+    },
+    {
+      "epoch": 0.16561437591063624,
+      "grad_norm": 0.5692222714424133,
+      "learning_rate": 0.0001,
+      "loss": 1.7713,
+      "step": 682
+    },
+    {
+      "epoch": 0.16585721223895095,
+      "grad_norm": 0.5323293209075928,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 683
+    },
+    {
+      "epoch": 0.16610004856726565,
+      "grad_norm": 0.546804666519165,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 684
+    },
+    {
+      "epoch": 0.16634288489558038,
+      "grad_norm": 0.5531345009803772,
+      "learning_rate": 0.0001,
+      "loss": 1.878,
+      "step": 685
+    },
+    {
+      "epoch": 0.1665857212238951,
+      "grad_norm": 0.5590841770172119,
+      "learning_rate": 0.0001,
+      "loss": 1.8109,
+      "step": 686
+    },
+    {
+      "epoch": 0.16682855755220982,
+      "grad_norm": 0.5341556668281555,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 687
+    },
+    {
+      "epoch": 0.16707139388052453,
+      "grad_norm": 0.5584942698478699,
+      "learning_rate": 0.0001,
+      "loss": 1.8527,
+      "step": 688
+    },
+    {
+      "epoch": 0.16731423020883923,
+      "grad_norm": 0.57704097032547,
+      "learning_rate": 0.0001,
+      "loss": 1.8635,
+      "step": 689
+    },
+    {
+      "epoch": 0.16755706653715396,
+      "grad_norm": 0.5278995633125305,
+      "learning_rate": 0.0001,
+      "loss": 1.8115,
+      "step": 690
+    },
+    {
+      "epoch": 0.16779990286546867,
+      "grad_norm": 0.559356153011322,
+      "learning_rate": 0.0001,
+      "loss": 1.7171,
+      "step": 691
+    },
+    {
+      "epoch": 0.1680427391937834,
+      "grad_norm": 0.5431156754493713,
+      "learning_rate": 0.0001,
+      "loss": 1.7339,
+      "step": 692
+    },
+    {
+      "epoch": 0.1682855755220981,
+      "grad_norm": 0.5194575190544128,
+      "learning_rate": 0.0001,
+      "loss": 1.4545,
+      "step": 693
+    },
+    {
+      "epoch": 0.1685284118504128,
+      "grad_norm": 0.604436993598938,
+      "learning_rate": 0.0001,
+      "loss": 1.922,
+      "step": 694
+    },
+    {
+      "epoch": 0.16877124817872755,
+      "grad_norm": 0.5614901185035706,
+      "learning_rate": 0.0001,
+      "loss": 1.725,
+      "step": 695
+    },
+    {
+      "epoch": 0.16901408450704225,
+      "grad_norm": 0.592366635799408,
+      "learning_rate": 0.0001,
+      "loss": 1.8416,
+      "step": 696
+    },
+    {
+      "epoch": 0.16925692083535698,
+      "grad_norm": 0.5448355674743652,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 697
+    },
+    {
+      "epoch": 0.1694997571636717,
+      "grad_norm": 0.5555615425109863,
+      "learning_rate": 0.0001,
+      "loss": 1.7578,
+      "step": 698
+    },
+    {
+      "epoch": 0.1697425934919864,
+      "grad_norm": 0.5792731642723083,
+      "learning_rate": 0.0001,
+      "loss": 1.8133,
+      "step": 699
+    },
+    {
+      "epoch": 0.16998542982030113,
+      "grad_norm": 0.5417819619178772,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 700
+    },
+    {
+      "epoch": 0.17022826614861583,
+      "grad_norm": 0.5344753265380859,
+      "learning_rate": 0.0001,
+      "loss": 1.7461,
+      "step": 701
+    },
+    {
+      "epoch": 0.17047110247693054,
+      "grad_norm": 0.565560519695282,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 702
+    },
+    {
+      "epoch": 0.17071393880524527,
+      "grad_norm": 0.5037214159965515,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 703
+    },
+    {
+      "epoch": 0.17095677513355997,
+      "grad_norm": 0.5803849101066589,
+      "learning_rate": 0.0001,
+      "loss": 1.8091,
+      "step": 704
+    },
+    {
+      "epoch": 0.1711996114618747,
+      "grad_norm": 0.5579473376274109,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 705
+    },
+    {
+      "epoch": 0.1714424477901894,
+      "grad_norm": 0.5527069568634033,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 706
+    },
+    {
+      "epoch": 0.17168528411850412,
+      "grad_norm": 0.5678402185440063,
+      "learning_rate": 0.0001,
+      "loss": 1.9185,
+      "step": 707
+    },
+    {
+      "epoch": 0.17192812044681885,
+      "grad_norm": 0.5627220273017883,
+      "learning_rate": 0.0001,
+      "loss": 1.7465,
+      "step": 708
+    },
+    {
+      "epoch": 0.17217095677513355,
+      "grad_norm": 0.604686975479126,
+      "learning_rate": 0.0001,
+      "loss": 1.8794,
+      "step": 709
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 0.5554637908935547,
+      "learning_rate": 0.0001,
+      "loss": 1.8323,
+      "step": 710
+    },
+    {
+      "epoch": 0.172656629431763,
+      "grad_norm": 0.561477541923523,
+      "learning_rate": 0.0001,
+      "loss": 1.7666,
+      "step": 711
+    },
+    {
+      "epoch": 0.1728994657600777,
+      "grad_norm": 0.5362895727157593,
+      "learning_rate": 0.0001,
+      "loss": 1.7603,
+      "step": 712
+    },
+    {
+      "epoch": 0.17314230208839243,
+      "grad_norm": 0.541498064994812,
+      "learning_rate": 0.0001,
+      "loss": 1.6341,
+      "step": 713
+    },
+    {
+      "epoch": 0.17338513841670714,
+      "grad_norm": 0.557447612285614,
+      "learning_rate": 0.0001,
+      "loss": 1.7325,
+      "step": 714
+    },
+    {
+      "epoch": 0.17362797474502187,
+      "grad_norm": 0.5641613006591797,
+      "learning_rate": 0.0001,
+      "loss": 1.8528,
+      "step": 715
+    },
+    {
+      "epoch": 0.17387081107333657,
+      "grad_norm": 0.5534732341766357,
+      "learning_rate": 0.0001,
+      "loss": 1.7265,
+      "step": 716
+    },
+    {
+      "epoch": 0.17411364740165128,
+      "grad_norm": 0.576434314250946,
+      "learning_rate": 0.0001,
+      "loss": 1.8017,
+      "step": 717
+    },
+    {
+      "epoch": 0.174356483729966,
+      "grad_norm": 0.5623506307601929,
+      "learning_rate": 0.0001,
+      "loss": 1.8295,
+      "step": 718
+    },
+    {
+      "epoch": 0.17459932005828072,
+      "grad_norm": 0.5556817650794983,
+      "learning_rate": 0.0001,
+      "loss": 1.8157,
+      "step": 719
+    },
+    {
+      "epoch": 0.17484215638659542,
+      "grad_norm": 0.5228356122970581,
+      "learning_rate": 0.0001,
+      "loss": 1.3937,
+      "step": 720
+    },
+    {
+      "epoch": 0.17508499271491015,
+      "grad_norm": 0.5386799573898315,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 721
+    },
+    {
+      "epoch": 0.17532782904322486,
+      "grad_norm": 0.6024092435836792,
+      "learning_rate": 0.0001,
+      "loss": 1.997,
+      "step": 722
+    },
+    {
+      "epoch": 0.1755706653715396,
+      "grad_norm": 0.5360486507415771,
+      "learning_rate": 0.0001,
+      "loss": 1.6886,
+      "step": 723
+    },
+    {
+      "epoch": 0.1758135016998543,
+      "grad_norm": 0.5964798927307129,
+      "learning_rate": 0.0001,
+      "loss": 1.8914,
+      "step": 724
+    },
+    {
+      "epoch": 0.176056338028169,
+      "grad_norm": 0.6057879328727722,
+      "learning_rate": 0.0001,
+      "loss": 1.946,
+      "step": 725
+    },
+    {
+      "epoch": 0.17629917435648373,
+      "grad_norm": 0.5556308627128601,
+      "learning_rate": 0.0001,
+      "loss": 1.7441,
+      "step": 726
+    },
+    {
+      "epoch": 0.17654201068479844,
+      "grad_norm": 0.5340707302093506,
+      "learning_rate": 0.0001,
+      "loss": 1.6801,
+      "step": 727
+    },
+    {
+      "epoch": 0.17678484701311317,
+      "grad_norm": 0.5532394647598267,
+      "learning_rate": 0.0001,
+      "loss": 1.8096,
+      "step": 728
+    },
+    {
+      "epoch": 0.17702768334142788,
+      "grad_norm": 0.5861799716949463,
+      "learning_rate": 0.0001,
+      "loss": 1.9008,
+      "step": 729
+    },
+    {
+      "epoch": 0.17727051966974258,
+      "grad_norm": 0.5458225011825562,
+      "learning_rate": 0.0001,
+      "loss": 1.6961,
+      "step": 730
+    },
+    {
+      "epoch": 0.17751335599805732,
+      "grad_norm": 0.5501299500465393,
+      "learning_rate": 0.0001,
+      "loss": 1.7769,
+      "step": 731
+    },
+    {
+      "epoch": 0.17775619232637202,
+      "grad_norm": 0.5216866135597229,
+      "learning_rate": 0.0001,
+      "loss": 1.6921,
+      "step": 732
+    },
+    {
+      "epoch": 0.17799902865468675,
+      "grad_norm": 0.5326796174049377,
+      "learning_rate": 0.0001,
+      "loss": 1.6341,
+      "step": 733
+    },
+    {
+      "epoch": 0.17824186498300146,
+      "grad_norm": 0.5195381045341492,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 734
+    },
+    {
+      "epoch": 0.17848470131131616,
+      "grad_norm": 0.5662425756454468,
+      "learning_rate": 0.0001,
+      "loss": 1.7098,
+      "step": 735
+    },
+    {
+      "epoch": 0.1787275376396309,
+      "grad_norm": 0.5749154090881348,
+      "learning_rate": 0.0001,
+      "loss": 1.747,
+      "step": 736
+    },
+    {
+      "epoch": 0.1789703739679456,
+      "grad_norm": 0.6091803908348083,
+      "learning_rate": 0.0001,
+      "loss": 1.937,
+      "step": 737
+    },
+    {
+      "epoch": 0.17921321029626033,
+      "grad_norm": 0.5344775319099426,
+      "learning_rate": 0.0001,
+      "loss": 1.7386,
+      "step": 738
+    },
+    {
+      "epoch": 0.17945604662457504,
+      "grad_norm": 0.5442355871200562,
+      "learning_rate": 0.0001,
+      "loss": 1.7379,
+      "step": 739
+    },
+    {
+      "epoch": 0.17969888295288974,
+      "grad_norm": 0.5308052897453308,
+      "learning_rate": 0.0001,
+      "loss": 1.5692,
+      "step": 740
+    },
+    {
+      "epoch": 0.17994171928120448,
+      "grad_norm": 0.5485922694206238,
+      "learning_rate": 0.0001,
+      "loss": 1.8535,
+      "step": 741
+    },
+    {
+      "epoch": 0.18018455560951918,
+      "grad_norm": 0.5594990849494934,
+      "learning_rate": 0.0001,
+      "loss": 1.9047,
+      "step": 742
+    },
+    {
+      "epoch": 0.1804273919378339,
+      "grad_norm": 0.5496913194656372,
+      "learning_rate": 0.0001,
+      "loss": 1.8267,
+      "step": 743
+    },
+    {
+      "epoch": 0.18067022826614862,
+      "grad_norm": 0.608910858631134,
+      "learning_rate": 0.0001,
+      "loss": 1.8546,
+      "step": 744
+    },
+    {
+      "epoch": 0.18091306459446332,
+      "grad_norm": 0.6221304535865784,
+      "learning_rate": 0.0001,
+      "loss": 1.9559,
+      "step": 745
+    },
+    {
+      "epoch": 0.18115590092277806,
+      "grad_norm": 0.5525220036506653,
+      "learning_rate": 0.0001,
+      "loss": 1.8696,
+      "step": 746
+    },
+    {
+      "epoch": 0.18139873725109276,
+      "grad_norm": 0.5693623423576355,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 747
+    },
+    {
+      "epoch": 0.18164157357940747,
+      "grad_norm": 0.5546811819076538,
+      "learning_rate": 0.0001,
+      "loss": 1.7519,
+      "step": 748
+    },
+    {
+      "epoch": 0.1818844099077222,
+      "grad_norm": 0.5642764568328857,
+      "learning_rate": 0.0001,
+      "loss": 1.7697,
+      "step": 749
+    },
+    {
+      "epoch": 0.1821272462360369,
+      "grad_norm": 0.5557659268379211,
+      "learning_rate": 0.0001,
+      "loss": 1.8393,
+      "step": 750
+    },
+    {
+      "epoch": 0.18237008256435164,
+      "grad_norm": 0.5520908236503601,
+      "learning_rate": 0.0001,
+      "loss": 1.6867,
+      "step": 751
+    },
+    {
+      "epoch": 0.18261291889266634,
+      "grad_norm": 0.5650697946548462,
+      "learning_rate": 0.0001,
+      "loss": 1.7904,
+      "step": 752
+    },
+    {
+      "epoch": 0.18285575522098105,
+      "grad_norm": 0.5700987577438354,
+      "learning_rate": 0.0001,
+      "loss": 1.8719,
+      "step": 753
+    },
+    {
+      "epoch": 0.18309859154929578,
+      "grad_norm": 0.5345878601074219,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 754
+    },
+    {
+      "epoch": 0.18334142787761049,
+      "grad_norm": 0.5695345401763916,
+      "learning_rate": 0.0001,
+      "loss": 1.8629,
+      "step": 755
+    },
+    {
+      "epoch": 0.18358426420592522,
+      "grad_norm": 0.5671497583389282,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 756
+    },
+    {
+      "epoch": 0.18382710053423992,
+      "grad_norm": 0.5771461129188538,
+      "learning_rate": 0.0001,
+      "loss": 1.8345,
+      "step": 757
+    },
+    {
+      "epoch": 0.18406993686255463,
+      "grad_norm": 0.5537222623825073,
+      "learning_rate": 0.0001,
+      "loss": 1.7433,
+      "step": 758
+    },
+    {
+      "epoch": 0.18431277319086936,
+      "grad_norm": 0.555461049079895,
+      "learning_rate": 0.0001,
+      "loss": 1.767,
+      "step": 759
+    },
+    {
+      "epoch": 0.18455560951918407,
+      "grad_norm": 0.5350934267044067,
+      "learning_rate": 0.0001,
+      "loss": 1.6869,
+      "step": 760
+    },
+    {
+      "epoch": 0.1847984458474988,
+      "grad_norm": 0.5174617767333984,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 761
+    },
+    {
+      "epoch": 0.1850412821758135,
+      "grad_norm": 0.560255765914917,
+      "learning_rate": 0.0001,
+      "loss": 1.8531,
+      "step": 762
+    },
+    {
+      "epoch": 0.1852841185041282,
+      "grad_norm": 0.5425716042518616,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 763
+    },
+    {
+      "epoch": 0.18552695483244294,
+      "grad_norm": 0.5555941462516785,
+      "learning_rate": 0.0001,
+      "loss": 1.6561,
+      "step": 764
+    },
+    {
+      "epoch": 0.18576979116075765,
+      "grad_norm": 0.5448790192604065,
+      "learning_rate": 0.0001,
+      "loss": 1.7301,
+      "step": 765
+    },
+    {
+      "epoch": 0.18601262748907235,
+      "grad_norm": 0.5178407430648804,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 766
+    },
+    {
+      "epoch": 0.18625546381738708,
+      "grad_norm": 0.5792193412780762,
+      "learning_rate": 0.0001,
+      "loss": 1.9282,
+      "step": 767
+    },
+    {
+      "epoch": 0.1864983001457018,
+      "grad_norm": 0.5438252091407776,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 768
+    },
+    {
+      "epoch": 0.18674113647401652,
+      "grad_norm": 0.5777905583381653,
+      "learning_rate": 0.0001,
+      "loss": 1.8311,
+      "step": 769
+    },
+    {
+      "epoch": 0.18698397280233123,
+      "grad_norm": 0.5501772165298462,
+      "learning_rate": 0.0001,
+      "loss": 1.8035,
+      "step": 770
+    },
+    {
+      "epoch": 0.18722680913064593,
+      "grad_norm": 0.540941059589386,
+      "learning_rate": 0.0001,
+      "loss": 1.8926,
+      "step": 771
+    },
+    {
+      "epoch": 0.18746964545896067,
+      "grad_norm": 0.5447785258293152,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 772
+    },
+    {
+      "epoch": 0.18771248178727537,
+      "grad_norm": 0.5447104573249817,
+      "learning_rate": 0.0001,
+      "loss": 1.7891,
+      "step": 773
+    },
+    {
+      "epoch": 0.1879553181155901,
+      "grad_norm": 0.5534484386444092,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 774
+    },
+    {
+      "epoch": 0.1881981544439048,
+      "grad_norm": 0.56410151720047,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 775
+    },
+    {
+      "epoch": 0.1884409907722195,
+      "grad_norm": 0.5520610809326172,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 776
+    },
+    {
+      "epoch": 0.18868382710053425,
+      "grad_norm": 0.5323498249053955,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 777
+    },
+    {
+      "epoch": 0.18892666342884895,
+      "grad_norm": 0.5767523050308228,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 778
+    },
+    {
+      "epoch": 0.18916949975716368,
+      "grad_norm": 0.5612403154373169,
+      "learning_rate": 0.0001,
+      "loss": 1.8959,
+      "step": 779
+    },
+    {
+      "epoch": 0.1894123360854784,
+      "grad_norm": 0.5474254488945007,
+      "learning_rate": 0.0001,
+      "loss": 1.718,
+      "step": 780
+    },
+    {
+      "epoch": 0.1896551724137931,
+      "grad_norm": 0.5918302536010742,
+      "learning_rate": 0.0001,
+      "loss": 1.8335,
+      "step": 781
+    },
+    {
+      "epoch": 0.18989800874210783,
+      "grad_norm": 0.5537752509117126,
+      "learning_rate": 0.0001,
+      "loss": 1.8291,
+      "step": 782
+    },
+    {
+      "epoch": 0.19014084507042253,
+      "grad_norm": 0.5441523194313049,
+      "learning_rate": 0.0001,
+      "loss": 1.8237,
+      "step": 783
+    },
+    {
+      "epoch": 0.19038368139873726,
+      "grad_norm": 0.5317227244377136,
+      "learning_rate": 0.0001,
+      "loss": 1.6781,
+      "step": 784
+    },
+    {
+      "epoch": 0.19062651772705197,
+      "grad_norm": 0.5500926971435547,
+      "learning_rate": 0.0001,
+      "loss": 1.6763,
+      "step": 785
+    },
+    {
+      "epoch": 0.19086935405536667,
+      "grad_norm": 0.5860820412635803,
+      "learning_rate": 0.0001,
+      "loss": 1.9524,
+      "step": 786
+    },
+    {
+      "epoch": 0.1911121903836814,
+      "grad_norm": 0.5804112553596497,
+      "learning_rate": 0.0001,
+      "loss": 1.6927,
+      "step": 787
+    },
+    {
+      "epoch": 0.1913550267119961,
+      "grad_norm": 0.5534296035766602,
+      "learning_rate": 0.0001,
+      "loss": 1.7464,
+      "step": 788
+    },
+    {
+      "epoch": 0.19159786304031082,
+      "grad_norm": 0.5835431814193726,
+      "learning_rate": 0.0001,
+      "loss": 1.8327,
+      "step": 789
+    },
+    {
+      "epoch": 0.19184069936862555,
+      "grad_norm": 0.6011634469032288,
+      "learning_rate": 0.0001,
+      "loss": 1.961,
+      "step": 790
+    },
+    {
+      "epoch": 0.19208353569694026,
+      "grad_norm": 0.5693417191505432,
+      "learning_rate": 0.0001,
+      "loss": 1.7782,
+      "step": 791
+    },
+    {
+      "epoch": 0.192326372025255,
+      "grad_norm": 0.6142314672470093,
+      "learning_rate": 0.0001,
+      "loss": 1.7706,
+      "step": 792
+    },
+    {
+      "epoch": 0.1925692083535697,
+      "grad_norm": 0.548800528049469,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 793
+    },
+    {
+      "epoch": 0.1928120446818844,
+      "grad_norm": 0.5322363376617432,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 794
+    },
+    {
+      "epoch": 0.19305488101019913,
+      "grad_norm": 0.5774746537208557,
+      "learning_rate": 0.0001,
+      "loss": 1.7622,
+      "step": 795
+    },
+    {
+      "epoch": 0.19329771733851384,
+      "grad_norm": 0.6003028750419617,
+      "learning_rate": 0.0001,
+      "loss": 1.896,
+      "step": 796
+    },
+    {
+      "epoch": 0.19354055366682857,
+      "grad_norm": 0.5864676833152771,
+      "learning_rate": 0.0001,
+      "loss": 1.9316,
+      "step": 797
+    },
+    {
+      "epoch": 0.19378338999514327,
+      "grad_norm": 0.5733742713928223,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 798
+    },
+    {
+      "epoch": 0.19402622632345798,
+      "grad_norm": 0.5421572327613831,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 799
+    },
+    {
+      "epoch": 0.1942690626517727,
+      "grad_norm": 0.5882205963134766,
+      "learning_rate": 0.0001,
+      "loss": 1.9301,
+      "step": 800
+    },
+    {
+      "epoch": 0.19451189898008742,
+      "grad_norm": 0.5595299601554871,
+      "learning_rate": 0.0001,
+      "loss": 1.7524,
+      "step": 801
+    },
+    {
+      "epoch": 0.19475473530840215,
+      "grad_norm": 0.6007490754127502,
+      "learning_rate": 0.0001,
+      "loss": 1.8614,
+      "step": 802
+    },
+    {
+      "epoch": 0.19499757163671685,
+      "grad_norm": 0.60100257396698,
+      "learning_rate": 0.0001,
+      "loss": 1.9322,
+      "step": 803
+    },
+    {
+      "epoch": 0.19524040796503156,
+      "grad_norm": 0.5655602216720581,
+      "learning_rate": 0.0001,
+      "loss": 1.8182,
+      "step": 804
+    },
+    {
+      "epoch": 0.1954832442933463,
+      "grad_norm": 0.5522935390472412,
+      "learning_rate": 0.0001,
+      "loss": 1.8435,
+      "step": 805
+    },
+    {
+      "epoch": 0.195726080621661,
+      "grad_norm": 0.5583087801933289,
+      "learning_rate": 0.0001,
+      "loss": 1.6269,
+      "step": 806
+    },
+    {
+      "epoch": 0.19596891694997573,
+      "grad_norm": 0.5243350267410278,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 807
+    },
+    {
+      "epoch": 0.19621175327829043,
+      "grad_norm": 0.5580244660377502,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 808
+    },
+    {
+      "epoch": 0.19645458960660514,
+      "grad_norm": 0.5781289339065552,
+      "learning_rate": 0.0001,
+      "loss": 1.7999,
+      "step": 809
+    },
+    {
+      "epoch": 0.19669742593491987,
+      "grad_norm": 0.5763341188430786,
+      "learning_rate": 0.0001,
+      "loss": 1.8655,
+      "step": 810
+    },
+    {
+      "epoch": 0.19694026226323458,
+      "grad_norm": 0.5569537281990051,
+      "learning_rate": 0.0001,
+      "loss": 1.8148,
+      "step": 811
+    },
+    {
+      "epoch": 0.19718309859154928,
+      "grad_norm": 0.5850344896316528,
+      "learning_rate": 0.0001,
+      "loss": 1.7526,
+      "step": 812
+    },
+    {
+      "epoch": 0.19742593491986402,
+      "grad_norm": 0.5795556902885437,
+      "learning_rate": 0.0001,
+      "loss": 1.8331,
+      "step": 813
+    },
+    {
+      "epoch": 0.19766877124817872,
+      "grad_norm": 0.5692217350006104,
+      "learning_rate": 0.0001,
+      "loss": 1.8168,
+      "step": 814
+    },
+    {
+      "epoch": 0.19791160757649345,
+      "grad_norm": 0.5934564471244812,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 815
+    },
+    {
+      "epoch": 0.19815444390480816,
+      "grad_norm": 0.5410759449005127,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 816
+    },
+    {
+      "epoch": 0.19839728023312286,
+      "grad_norm": 0.5464872121810913,
+      "learning_rate": 0.0001,
+      "loss": 1.7449,
+      "step": 817
+    },
+    {
+      "epoch": 0.1986401165614376,
+      "grad_norm": 0.5795077085494995,
+      "learning_rate": 0.0001,
+      "loss": 1.9202,
+      "step": 818
+    },
+    {
+      "epoch": 0.1988829528897523,
+      "grad_norm": 0.6018663048744202,
+      "learning_rate": 0.0001,
+      "loss": 1.8138,
+      "step": 819
+    },
+    {
+      "epoch": 0.19912578921806703,
+      "grad_norm": 0.5785474181175232,
+      "learning_rate": 0.0001,
+      "loss": 1.7411,
+      "step": 820
+    },
+    {
+      "epoch": 0.19936862554638174,
+      "grad_norm": 0.61650550365448,
+      "learning_rate": 0.0001,
+      "loss": 1.8478,
+      "step": 821
+    },
+    {
+      "epoch": 0.19961146187469644,
+      "grad_norm": 0.5855715274810791,
+      "learning_rate": 0.0001,
+      "loss": 1.7753,
+      "step": 822
+    },
+    {
+      "epoch": 0.19985429820301118,
+      "grad_norm": 0.5476219654083252,
+      "learning_rate": 0.0001,
+      "loss": 1.7139,
+      "step": 823
+    },
+    {
+      "epoch": 0.20009713453132588,
+      "grad_norm": 0.5218155384063721,
+      "learning_rate": 0.0001,
+      "loss": 1.6086,
+      "step": 824
+    },
+    {
+      "epoch": 0.20033997085964061,
+      "grad_norm": 0.5409383773803711,
+      "learning_rate": 0.0001,
+      "loss": 1.785,
+      "step": 825
+    },
+    {
+      "epoch": 0.20058280718795532,
+      "grad_norm": 0.585970938205719,
+      "learning_rate": 0.0001,
+      "loss": 1.9259,
+      "step": 826
+    },
+    {
+      "epoch": 0.20082564351627002,
+      "grad_norm": 0.5450654029846191,
+      "learning_rate": 0.0001,
+      "loss": 1.7319,
+      "step": 827
+    },
+    {
+      "epoch": 0.20106847984458476,
+      "grad_norm": 0.5078698396682739,
+      "learning_rate": 0.0001,
+      "loss": 1.6365,
+      "step": 828
+    },
+    {
+      "epoch": 0.20131131617289946,
+      "grad_norm": 0.5683661103248596,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 829
+    },
+    {
+      "epoch": 0.2015541525012142,
+      "grad_norm": 0.5884541869163513,
+      "learning_rate": 0.0001,
+      "loss": 1.8233,
+      "step": 830
+    },
+    {
+      "epoch": 0.2017969888295289,
+      "grad_norm": 0.5697531700134277,
+      "learning_rate": 0.0001,
+      "loss": 1.8651,
+      "step": 831
+    },
+    {
+      "epoch": 0.2020398251578436,
+      "grad_norm": 0.5456650853157043,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 832
+    },
+    {
+      "epoch": 0.20228266148615834,
+      "grad_norm": 0.5575149655342102,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 833
+    },
+    {
+      "epoch": 0.20252549781447304,
+      "grad_norm": 0.5606904029846191,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 834
+    },
+    {
+      "epoch": 0.20276833414278775,
+      "grad_norm": 0.5217230319976807,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 835
+    },
+    {
+      "epoch": 0.20301117047110248,
+      "grad_norm": 0.5786769986152649,
+      "learning_rate": 0.0001,
+      "loss": 1.7364,
+      "step": 836
+    },
+    {
+      "epoch": 0.2032540067994172,
+      "grad_norm": 0.536147952079773,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 837
+    },
+    {
+      "epoch": 0.20349684312773192,
+      "grad_norm": 0.5949231386184692,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 838
+    },
+    {
+      "epoch": 0.20373967945604662,
+      "grad_norm": 0.5834288001060486,
+      "learning_rate": 0.0001,
+      "loss": 1.8865,
+      "step": 839
+    },
+    {
+      "epoch": 0.20398251578436133,
+      "grad_norm": 0.6145623326301575,
+      "learning_rate": 0.0001,
+      "loss": 1.8882,
+      "step": 840
+    },
+    {
+      "epoch": 0.20422535211267606,
+      "grad_norm": 0.5596982836723328,
+      "learning_rate": 0.0001,
+      "loss": 1.8797,
+      "step": 841
+    },
+    {
+      "epoch": 0.20446818844099077,
+      "grad_norm": 0.5657252073287964,
+      "learning_rate": 0.0001,
+      "loss": 1.7358,
+      "step": 842
+    },
+    {
+      "epoch": 0.2047110247693055,
+      "grad_norm": 0.5526270270347595,
+      "learning_rate": 0.0001,
+      "loss": 1.6714,
+      "step": 843
+    },
+    {
+      "epoch": 0.2049538610976202,
+      "grad_norm": 0.5630422830581665,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 844
+    },
+    {
+      "epoch": 0.2051966974259349,
+      "grad_norm": 0.5791303515434265,
+      "learning_rate": 0.0001,
+      "loss": 1.7582,
+      "step": 845
+    },
+    {
+      "epoch": 0.20543953375424964,
+      "grad_norm": 0.5686047673225403,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 846
+    },
+    {
+      "epoch": 0.20568237008256435,
+      "grad_norm": 0.5482386946678162,
+      "learning_rate": 0.0001,
+      "loss": 1.7276,
+      "step": 847
+    },
+    {
+      "epoch": 0.20592520641087908,
+      "grad_norm": 0.5421972870826721,
+      "learning_rate": 0.0001,
+      "loss": 1.7984,
+      "step": 848
+    },
+    {
+      "epoch": 0.20616804273919379,
+      "grad_norm": 0.5663183331489563,
+      "learning_rate": 0.0001,
+      "loss": 1.7296,
+      "step": 849
+    },
+    {
+      "epoch": 0.2064108790675085,
+      "grad_norm": 0.5500679016113281,
+      "learning_rate": 0.0001,
+      "loss": 1.8638,
+      "step": 850
+    },
+    {
+      "epoch": 0.20665371539582322,
+      "grad_norm": 0.5626316666603088,
+      "learning_rate": 0.0001,
+      "loss": 1.7687,
+      "step": 851
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 0.5308297872543335,
+      "learning_rate": 0.0001,
+      "loss": 1.7183,
+      "step": 852
+    },
+    {
+      "epoch": 0.20713938805245263,
+      "grad_norm": 0.5546897053718567,
+      "learning_rate": 0.0001,
+      "loss": 1.6975,
+      "step": 853
+    },
+    {
+      "epoch": 0.20738222438076737,
+      "grad_norm": 0.5396243929862976,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 854
+    },
+    {
+      "epoch": 0.20762506070908207,
+      "grad_norm": 0.5337682366371155,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 855
+    },
+    {
+      "epoch": 0.2078678970373968,
+      "grad_norm": 0.5745347142219543,
+      "learning_rate": 0.0001,
+      "loss": 1.8399,
+      "step": 856
+    },
+    {
+      "epoch": 0.2081107333657115,
+      "grad_norm": 0.5393202900886536,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 857
+    },
+    {
+      "epoch": 0.2083535696940262,
+      "grad_norm": 0.5366864204406738,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 858
+    },
+    {
+      "epoch": 0.20859640602234095,
+      "grad_norm": 0.55235755443573,
+      "learning_rate": 0.0001,
+      "loss": 1.8314,
+      "step": 859
+    },
+    {
+      "epoch": 0.20883924235065565,
+      "grad_norm": 0.5640939474105835,
+      "learning_rate": 0.0001,
+      "loss": 1.8421,
+      "step": 860
+    },
+    {
+      "epoch": 0.20908207867897038,
+      "grad_norm": 0.5542709231376648,
+      "learning_rate": 0.0001,
+      "loss": 1.7897,
+      "step": 861
+    },
+    {
+      "epoch": 0.2093249150072851,
+      "grad_norm": 0.5564791560173035,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 862
+    },
+    {
+      "epoch": 0.2095677513355998,
+      "grad_norm": 0.560176432132721,
+      "learning_rate": 0.0001,
+      "loss": 1.7619,
+      "step": 863
+    },
+    {
+      "epoch": 0.20981058766391453,
+      "grad_norm": 0.6058011054992676,
+      "learning_rate": 0.0001,
+      "loss": 1.8572,
+      "step": 864
+    },
+    {
+      "epoch": 0.21005342399222923,
+      "grad_norm": 0.5366670489311218,
+      "learning_rate": 0.0001,
+      "loss": 1.7734,
+      "step": 865
+    },
+    {
+      "epoch": 0.21029626032054397,
+      "grad_norm": 0.5931606888771057,
+      "learning_rate": 0.0001,
+      "loss": 1.7994,
+      "step": 866
+    },
+    {
+      "epoch": 0.21053909664885867,
+      "grad_norm": 0.5367823243141174,
+      "learning_rate": 0.0001,
+      "loss": 1.6912,
+      "step": 867
+    },
+    {
+      "epoch": 0.21078193297717338,
+      "grad_norm": 0.5488685369491577,
+      "learning_rate": 0.0001,
+      "loss": 1.7204,
+      "step": 868
+    },
+    {
+      "epoch": 0.2110247693054881,
+      "grad_norm": 0.5571208000183105,
+      "learning_rate": 0.0001,
+      "loss": 1.7688,
+      "step": 869
+    },
+    {
+      "epoch": 0.2112676056338028,
+      "grad_norm": 0.5351872444152832,
+      "learning_rate": 0.0001,
+      "loss": 1.6713,
+      "step": 870
+    },
+    {
+      "epoch": 0.21151044196211755,
+      "grad_norm": 0.5415477752685547,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 871
+    },
+    {
+      "epoch": 0.21175327829043225,
+      "grad_norm": 0.5760199427604675,
+      "learning_rate": 0.0001,
+      "loss": 1.8462,
+      "step": 872
+    },
+    {
+      "epoch": 0.21199611461874696,
+      "grad_norm": 0.5803876519203186,
+      "learning_rate": 0.0001,
+      "loss": 1.6798,
+      "step": 873
+    },
+    {
+      "epoch": 0.2122389509470617,
+      "grad_norm": 0.5319856405258179,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 874
+    },
+    {
+      "epoch": 0.2124817872753764,
+      "grad_norm": 0.5371922254562378,
+      "learning_rate": 0.0001,
+      "loss": 1.5163,
+      "step": 875
+    },
+    {
+      "epoch": 0.2127246236036911,
+      "grad_norm": 0.57842618227005,
+      "learning_rate": 0.0001,
+      "loss": 1.8315,
+      "step": 876
+    },
+    {
+      "epoch": 0.21296745993200583,
+      "grad_norm": 0.5532588362693787,
+      "learning_rate": 0.0001,
+      "loss": 1.7318,
+      "step": 877
+    },
+    {
+      "epoch": 0.21321029626032054,
+      "grad_norm": 0.5485522150993347,
+      "learning_rate": 0.0001,
+      "loss": 1.7128,
+      "step": 878
+    },
+    {
+      "epoch": 0.21345313258863527,
+      "grad_norm": 0.5748867392539978,
+      "learning_rate": 0.0001,
+      "loss": 1.8404,
+      "step": 879
+    },
+    {
+      "epoch": 0.21369596891694997,
+      "grad_norm": 0.5486942529678345,
+      "learning_rate": 0.0001,
+      "loss": 1.7819,
+      "step": 880
+    },
+    {
+      "epoch": 0.21393880524526468,
+      "grad_norm": 0.5755213499069214,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 881
+    },
+    {
+      "epoch": 0.2141816415735794,
+      "grad_norm": 0.5474473237991333,
+      "learning_rate": 0.0001,
+      "loss": 1.7719,
+      "step": 882
+    },
+    {
+      "epoch": 0.21442447790189412,
+      "grad_norm": 0.5740113854408264,
+      "learning_rate": 0.0001,
+      "loss": 1.9424,
+      "step": 883
+    },
+    {
+      "epoch": 0.21466731423020885,
+      "grad_norm": 0.5920238494873047,
+      "learning_rate": 0.0001,
+      "loss": 1.6995,
+      "step": 884
+    },
+    {
+      "epoch": 0.21491015055852355,
+      "grad_norm": 0.5916893482208252,
+      "learning_rate": 0.0001,
+      "loss": 1.8407,
+      "step": 885
+    },
+    {
+      "epoch": 0.21515298688683826,
+      "grad_norm": 0.5308418273925781,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 886
+    },
+    {
+      "epoch": 0.215395823215153,
+      "grad_norm": 0.5787625312805176,
+      "learning_rate": 0.0001,
+      "loss": 1.7945,
+      "step": 887
+    },
+    {
+      "epoch": 0.2156386595434677,
+      "grad_norm": 0.5507543087005615,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 888
+    },
+    {
+      "epoch": 0.21588149587178243,
+      "grad_norm": 0.5690258741378784,
+      "learning_rate": 0.0001,
+      "loss": 1.8161,
+      "step": 889
+    },
+    {
+      "epoch": 0.21612433220009714,
+      "grad_norm": 0.561421275138855,
+      "learning_rate": 0.0001,
+      "loss": 1.7688,
+      "step": 890
+    },
+    {
+      "epoch": 0.21636716852841184,
+      "grad_norm": 0.5234627723693848,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 891
+    },
+    {
+      "epoch": 0.21661000485672657,
+      "grad_norm": 0.5844734311103821,
+      "learning_rate": 0.0001,
+      "loss": 1.6985,
+      "step": 892
+    },
+    {
+      "epoch": 0.21685284118504128,
+      "grad_norm": 0.5320757031440735,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 893
+    },
+    {
+      "epoch": 0.217095677513356,
+      "grad_norm": 0.5599240660667419,
+      "learning_rate": 0.0001,
+      "loss": 1.7955,
+      "step": 894
+    },
+    {
+      "epoch": 0.21733851384167072,
+      "grad_norm": 0.6245287656784058,
+      "learning_rate": 0.0001,
+      "loss": 1.691,
+      "step": 895
+    },
+    {
+      "epoch": 0.21758135016998542,
+      "grad_norm": 0.5418979525566101,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 896
+    },
+    {
+      "epoch": 0.21782418649830015,
+      "grad_norm": 0.6089024543762207,
+      "learning_rate": 0.0001,
+      "loss": 1.9166,
+      "step": 897
+    },
+    {
+      "epoch": 0.21806702282661486,
+      "grad_norm": 0.5672260522842407,
+      "learning_rate": 0.0001,
+      "loss": 1.7904,
+      "step": 898
+    },
+    {
+      "epoch": 0.21830985915492956,
+      "grad_norm": 0.5514928698539734,
+      "learning_rate": 0.0001,
+      "loss": 1.7884,
+      "step": 899
+    },
+    {
+      "epoch": 0.2185526954832443,
+      "grad_norm": 0.5628061890602112,
+      "learning_rate": 0.0001,
+      "loss": 1.751,
+      "step": 900
+    },
+    {
+      "epoch": 0.218795531811559,
+      "grad_norm": 0.5473241209983826,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 901
+    },
+    {
+      "epoch": 0.21903836813987373,
+      "grad_norm": 0.545527994632721,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 902
+    },
+    {
+      "epoch": 0.21928120446818844,
+      "grad_norm": 0.5235995054244995,
+      "learning_rate": 0.0001,
+      "loss": 1.7536,
+      "step": 903
+    },
+    {
+      "epoch": 0.21952404079650314,
+      "grad_norm": 0.5851334929466248,
+      "learning_rate": 0.0001,
+      "loss": 1.985,
+      "step": 904
+    },
+    {
+      "epoch": 0.21976687712481788,
+      "grad_norm": 0.5480401515960693,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 905
+    },
+    {
+      "epoch": 0.22000971345313258,
+      "grad_norm": 0.5195808410644531,
+      "learning_rate": 0.0001,
+      "loss": 1.7583,
+      "step": 906
+    },
+    {
+      "epoch": 0.22025254978144732,
+      "grad_norm": 0.5582727193832397,
+      "learning_rate": 0.0001,
+      "loss": 1.9367,
+      "step": 907
+    },
+    {
+      "epoch": 0.22049538610976202,
+      "grad_norm": 0.5689213871955872,
+      "learning_rate": 0.0001,
+      "loss": 1.9294,
+      "step": 908
+    },
+    {
+      "epoch": 0.22073822243807673,
+      "grad_norm": 0.5949645638465881,
+      "learning_rate": 0.0001,
+      "loss": 1.7627,
+      "step": 909
+    },
+    {
+      "epoch": 0.22098105876639146,
+      "grad_norm": 0.5724778175354004,
+      "learning_rate": 0.0001,
+      "loss": 1.7718,
+      "step": 910
+    },
+    {
+      "epoch": 0.22122389509470616,
+      "grad_norm": 0.5361983180046082,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 911
+    },
+    {
+      "epoch": 0.2214667314230209,
+      "grad_norm": 0.5725252032279968,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 912
+    },
+    {
+      "epoch": 0.2217095677513356,
+      "grad_norm": 0.5657257437705994,
+      "learning_rate": 0.0001,
+      "loss": 1.7613,
+      "step": 913
+    },
+    {
+      "epoch": 0.2219524040796503,
+      "grad_norm": 0.5461857318878174,
+      "learning_rate": 0.0001,
+      "loss": 1.6713,
+      "step": 914
+    },
+    {
+      "epoch": 0.22219524040796504,
+      "grad_norm": 0.5253598690032959,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 915
+    },
+    {
+      "epoch": 0.22243807673627974,
+      "grad_norm": 0.5413147211074829,
+      "learning_rate": 0.0001,
+      "loss": 1.7325,
+      "step": 916
+    },
+    {
+      "epoch": 0.22268091306459448,
+      "grad_norm": 0.5368465185165405,
+      "learning_rate": 0.0001,
+      "loss": 1.7434,
+      "step": 917
+    },
+    {
+      "epoch": 0.22292374939290918,
+      "grad_norm": 0.5531226992607117,
+      "learning_rate": 0.0001,
+      "loss": 1.6896,
+      "step": 918
+    },
+    {
+      "epoch": 0.2231665857212239,
+      "grad_norm": 0.5544846653938293,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 919
+    },
+    {
+      "epoch": 0.22340942204953862,
+      "grad_norm": 0.5697352290153503,
+      "learning_rate": 0.0001,
+      "loss": 1.7903,
+      "step": 920
+    },
+    {
+      "epoch": 0.22365225837785332,
+      "grad_norm": 0.5445935130119324,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 921
+    },
+    {
+      "epoch": 0.22389509470616803,
+      "grad_norm": 0.5548669695854187,
+      "learning_rate": 0.0001,
+      "loss": 1.7802,
+      "step": 922
+    },
+    {
+      "epoch": 0.22413793103448276,
+      "grad_norm": 0.5918106436729431,
+      "learning_rate": 0.0001,
+      "loss": 1.813,
+      "step": 923
+    },
+    {
+      "epoch": 0.22438076736279747,
+      "grad_norm": 0.5879506468772888,
+      "learning_rate": 0.0001,
+      "loss": 1.9609,
+      "step": 924
+    },
+    {
+      "epoch": 0.2246236036911122,
+      "grad_norm": 0.5621781945228577,
+      "learning_rate": 0.0001,
+      "loss": 1.9015,
+      "step": 925
+    },
+    {
+      "epoch": 0.2248664400194269,
+      "grad_norm": 0.5593553781509399,
+      "learning_rate": 0.0001,
+      "loss": 1.7965,
+      "step": 926
+    },
+    {
+      "epoch": 0.2251092763477416,
+      "grad_norm": 0.6174554824829102,
+      "learning_rate": 0.0001,
+      "loss": 2.0494,
+      "step": 927
+    },
+    {
+      "epoch": 0.22535211267605634,
+      "grad_norm": 0.5466719269752502,
+      "learning_rate": 0.0001,
+      "loss": 1.8213,
+      "step": 928
+    },
+    {
+      "epoch": 0.22559494900437105,
+      "grad_norm": 0.5589328408241272,
+      "learning_rate": 0.0001,
+      "loss": 1.6947,
+      "step": 929
+    },
+    {
+      "epoch": 0.22583778533268578,
+      "grad_norm": 0.5526649355888367,
+      "learning_rate": 0.0001,
+      "loss": 1.8647,
+      "step": 930
+    },
+    {
+      "epoch": 0.22608062166100049,
+      "grad_norm": 0.5893845558166504,
+      "learning_rate": 0.0001,
+      "loss": 1.898,
+      "step": 931
+    },
+    {
+      "epoch": 0.2263234579893152,
+      "grad_norm": 0.6088446378707886,
+      "learning_rate": 0.0001,
+      "loss": 1.8883,
+      "step": 932
+    },
+    {
+      "epoch": 0.22656629431762992,
+      "grad_norm": 0.5552610754966736,
+      "learning_rate": 0.0001,
+      "loss": 1.6933,
+      "step": 933
+    },
+    {
+      "epoch": 0.22680913064594463,
+      "grad_norm": 0.5152078866958618,
+      "learning_rate": 0.0001,
+      "loss": 1.7231,
+      "step": 934
+    },
+    {
+      "epoch": 0.22705196697425936,
+      "grad_norm": 0.5598558783531189,
+      "learning_rate": 0.0001,
+      "loss": 1.7015,
+      "step": 935
+    },
+    {
+      "epoch": 0.22729480330257407,
+      "grad_norm": 0.608623743057251,
+      "learning_rate": 0.0001,
+      "loss": 1.8497,
+      "step": 936
+    },
+    {
+      "epoch": 0.22753763963088877,
+      "grad_norm": 0.570016086101532,
+      "learning_rate": 0.0001,
+      "loss": 1.8382,
+      "step": 937
+    },
+    {
+      "epoch": 0.2277804759592035,
+      "grad_norm": 0.5475924015045166,
+      "learning_rate": 0.0001,
+      "loss": 1.798,
+      "step": 938
+    },
+    {
+      "epoch": 0.2280233122875182,
+      "grad_norm": 0.5976389050483704,
+      "learning_rate": 0.0001,
+      "loss": 1.9021,
+      "step": 939
+    },
+    {
+      "epoch": 0.22826614861583294,
+      "grad_norm": 0.5569254159927368,
+      "learning_rate": 0.0001,
+      "loss": 1.8965,
+      "step": 940
+    },
+    {
+      "epoch": 0.22850898494414765,
+      "grad_norm": 0.5021308064460754,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 941
+    },
+    {
+      "epoch": 0.22875182127246235,
+      "grad_norm": 0.515293300151825,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 942
+    },
+    {
+      "epoch": 0.22899465760077709,
+      "grad_norm": 0.515910804271698,
+      "learning_rate": 0.0001,
+      "loss": 1.6818,
+      "step": 943
+    },
+    {
+      "epoch": 0.2292374939290918,
+      "grad_norm": 0.5279413461685181,
+      "learning_rate": 0.0001,
+      "loss": 1.6615,
+      "step": 944
+    },
+    {
+      "epoch": 0.2294803302574065,
+      "grad_norm": 0.5485278367996216,
+      "learning_rate": 0.0001,
+      "loss": 1.7407,
+      "step": 945
+    },
+    {
+      "epoch": 0.22972316658572123,
+      "grad_norm": 0.5664181709289551,
+      "learning_rate": 0.0001,
+      "loss": 1.7865,
+      "step": 946
+    },
+    {
+      "epoch": 0.22996600291403593,
+      "grad_norm": 0.5307700037956238,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 947
+    },
+    {
+      "epoch": 0.23020883924235067,
+      "grad_norm": 0.5400621294975281,
+      "learning_rate": 0.0001,
+      "loss": 1.7756,
+      "step": 948
+    },
+    {
+      "epoch": 0.23045167557066537,
+      "grad_norm": 0.5360437631607056,
+      "learning_rate": 0.0001,
+      "loss": 1.8162,
+      "step": 949
+    },
+    {
+      "epoch": 0.23069451189898008,
+      "grad_norm": 0.5398291945457458,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 950
+    },
+    {
+      "epoch": 0.2309373482272948,
+      "grad_norm": 0.5587419271469116,
+      "learning_rate": 0.0001,
+      "loss": 1.945,
+      "step": 951
+    },
+    {
+      "epoch": 0.2311801845556095,
+      "grad_norm": 0.5683267116546631,
+      "learning_rate": 0.0001,
+      "loss": 1.8404,
+      "step": 952
+    },
+    {
+      "epoch": 0.23142302088392425,
+      "grad_norm": 0.5431987643241882,
+      "learning_rate": 0.0001,
+      "loss": 1.8186,
+      "step": 953
+    },
+    {
+      "epoch": 0.23166585721223895,
+      "grad_norm": 0.5460397005081177,
+      "learning_rate": 0.0001,
+      "loss": 1.7055,
+      "step": 954
+    },
+    {
+      "epoch": 0.23190869354055366,
+      "grad_norm": 0.6160047054290771,
+      "learning_rate": 0.0001,
+      "loss": 1.9158,
+      "step": 955
+    },
+    {
+      "epoch": 0.2321515298688684,
+      "grad_norm": 0.5519976615905762,
+      "learning_rate": 0.0001,
+      "loss": 1.9105,
+      "step": 956
+    },
+    {
+      "epoch": 0.2323943661971831,
+      "grad_norm": 0.5553170442581177,
+      "learning_rate": 0.0001,
+      "loss": 1.7359,
+      "step": 957
+    },
+    {
+      "epoch": 0.23263720252549783,
+      "grad_norm": 0.5691775679588318,
+      "learning_rate": 0.0001,
+      "loss": 1.853,
+      "step": 958
+    },
+    {
+      "epoch": 0.23288003885381253,
+      "grad_norm": 0.5368521809577942,
+      "learning_rate": 0.0001,
+      "loss": 1.7095,
+      "step": 959
+    },
+    {
+      "epoch": 0.23312287518212724,
+      "grad_norm": 0.5350582003593445,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 960
+    },
+    {
+      "epoch": 0.23336571151044197,
+      "grad_norm": 0.5377877354621887,
+      "learning_rate": 0.0001,
+      "loss": 1.6835,
+      "step": 961
+    },
+    {
+      "epoch": 0.23360854783875667,
+      "grad_norm": 0.6030954122543335,
+      "learning_rate": 0.0001,
+      "loss": 1.792,
+      "step": 962
+    },
+    {
+      "epoch": 0.2338513841670714,
+      "grad_norm": 0.5422897338867188,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 963
+    },
+    {
+      "epoch": 0.2340942204953861,
+      "grad_norm": 0.5576704144477844,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 964
+    },
+    {
+      "epoch": 0.23433705682370082,
+      "grad_norm": 0.5503125190734863,
+      "learning_rate": 0.0001,
+      "loss": 1.701,
+      "step": 965
+    },
+    {
+      "epoch": 0.23457989315201555,
+      "grad_norm": 0.5446116328239441,
+      "learning_rate": 0.0001,
+      "loss": 1.7575,
+      "step": 966
+    },
+    {
+      "epoch": 0.23482272948033026,
+      "grad_norm": 0.5460963845252991,
+      "learning_rate": 0.0001,
+      "loss": 1.7283,
+      "step": 967
+    },
+    {
+      "epoch": 0.23506556580864496,
+      "grad_norm": 0.5579452514648438,
+      "learning_rate": 0.0001,
+      "loss": 1.8132,
+      "step": 968
+    },
+    {
+      "epoch": 0.2353084021369597,
+      "grad_norm": 0.5324477553367615,
+      "learning_rate": 0.0001,
+      "loss": 1.6876,
+      "step": 969
+    },
+    {
+      "epoch": 0.2355512384652744,
+      "grad_norm": 0.5197933316230774,
+      "learning_rate": 0.0001,
+      "loss": 1.6824,
+      "step": 970
+    },
+    {
+      "epoch": 0.23579407479358913,
+      "grad_norm": 0.5394790172576904,
+      "learning_rate": 0.0001,
+      "loss": 1.8378,
+      "step": 971
+    },
+    {
+      "epoch": 0.23603691112190384,
+      "grad_norm": 0.5891541242599487,
+      "learning_rate": 0.0001,
+      "loss": 1.8853,
+      "step": 972
+    },
+    {
+      "epoch": 0.23627974745021854,
+      "grad_norm": 0.539494514465332,
+      "learning_rate": 0.0001,
+      "loss": 1.7937,
+      "step": 973
+    },
+    {
+      "epoch": 0.23652258377853327,
+      "grad_norm": 0.5574384927749634,
+      "learning_rate": 0.0001,
+      "loss": 1.781,
+      "step": 974
+    },
+    {
+      "epoch": 0.23676542010684798,
+      "grad_norm": 0.5909311175346375,
+      "learning_rate": 0.0001,
+      "loss": 1.7712,
+      "step": 975
+    },
+    {
+      "epoch": 0.2370082564351627,
+      "grad_norm": 0.6242678761482239,
+      "learning_rate": 0.0001,
+      "loss": 1.978,
+      "step": 976
+    },
+    {
+      "epoch": 0.23725109276347742,
+      "grad_norm": 0.5489070415496826,
+      "learning_rate": 0.0001,
+      "loss": 1.8114,
+      "step": 977
+    },
+    {
+      "epoch": 0.23749392909179212,
+      "grad_norm": 0.5614756345748901,
+      "learning_rate": 0.0001,
+      "loss": 1.8186,
+      "step": 978
+    },
+    {
+      "epoch": 0.23773676542010685,
+      "grad_norm": 0.5582174062728882,
+      "learning_rate": 0.0001,
+      "loss": 1.8372,
+      "step": 979
+    },
+    {
+      "epoch": 0.23797960174842156,
+      "grad_norm": 0.5511270761489868,
+      "learning_rate": 0.0001,
+      "loss": 1.6409,
+      "step": 980
+    },
+    {
+      "epoch": 0.2382224380767363,
+      "grad_norm": 0.5524898767471313,
+      "learning_rate": 0.0001,
+      "loss": 1.705,
+      "step": 981
+    },
+    {
+      "epoch": 0.238465274405051,
+      "grad_norm": 0.542084813117981,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 982
+    },
+    {
+      "epoch": 0.2387081107333657,
+      "grad_norm": 0.5258582234382629,
+      "learning_rate": 0.0001,
+      "loss": 1.4242,
+      "step": 983
+    },
+    {
+      "epoch": 0.23895094706168044,
+      "grad_norm": 0.5367980599403381,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 984
+    },
+    {
+      "epoch": 0.23919378338999514,
+      "grad_norm": 0.5734711289405823,
+      "learning_rate": 0.0001,
+      "loss": 1.8304,
+      "step": 985
+    },
+    {
+      "epoch": 0.23943661971830985,
+      "grad_norm": 0.532623827457428,
+      "learning_rate": 0.0001,
+      "loss": 1.7613,
+      "step": 986
+    },
+    {
+      "epoch": 0.23967945604662458,
+      "grad_norm": 0.5503738522529602,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 987
+    },
+    {
+      "epoch": 0.23992229237493928,
+      "grad_norm": 0.5924004316329956,
+      "learning_rate": 0.0001,
+      "loss": 1.7314,
+      "step": 988
+    },
+    {
+      "epoch": 0.24016512870325402,
+      "grad_norm": 0.6099340319633484,
+      "learning_rate": 0.0001,
+      "loss": 1.907,
+      "step": 989
+    },
+    {
+      "epoch": 0.24040796503156872,
+      "grad_norm": 0.5783790349960327,
+      "learning_rate": 0.0001,
+      "loss": 1.8731,
+      "step": 990
+    },
+    {
+      "epoch": 0.24065080135988343,
+      "grad_norm": 0.5989636182785034,
+      "learning_rate": 0.0001,
+      "loss": 1.9282,
+      "step": 991
+    },
+    {
+      "epoch": 0.24089363768819816,
+      "grad_norm": 0.5222843289375305,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 992
+    },
+    {
+      "epoch": 0.24113647401651286,
+      "grad_norm": 0.5943722724914551,
+      "learning_rate": 0.0001,
+      "loss": 1.7093,
+      "step": 993
+    },
+    {
+      "epoch": 0.2413793103448276,
+      "grad_norm": 0.6033304333686829,
+      "learning_rate": 0.0001,
+      "loss": 1.8118,
+      "step": 994
+    },
+    {
+      "epoch": 0.2416221466731423,
+      "grad_norm": 0.5955982804298401,
+      "learning_rate": 0.0001,
+      "loss": 1.851,
+      "step": 995
+    },
+    {
+      "epoch": 0.241864983001457,
+      "grad_norm": 0.5730704069137573,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 996
+    },
+    {
+      "epoch": 0.24210781932977174,
+      "grad_norm": 0.5521137118339539,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 997
+    },
+    {
+      "epoch": 0.24235065565808644,
+      "grad_norm": 0.5170513391494751,
+      "learning_rate": 0.0001,
+      "loss": 1.7792,
+      "step": 998
+    },
+    {
+      "epoch": 0.24259349198640118,
+      "grad_norm": 0.5559035539627075,
+      "learning_rate": 0.0001,
+      "loss": 1.7803,
+      "step": 999
+    },
+    {
+      "epoch": 0.24283632831471588,
+      "grad_norm": 0.5826027393341064,
+      "learning_rate": 0.0001,
+      "loss": 1.8425,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2430791646430306,
+      "grad_norm": 1.0654746294021606,
+      "learning_rate": 0.0001,
+      "loss": 1.7884,
+      "step": 1001
+    },
+    {
+      "epoch": 0.24332200097134532,
+      "grad_norm": 0.5563173890113831,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 1002
+    },
+    {
+      "epoch": 0.24356483729966003,
+      "grad_norm": 0.5636867880821228,
+      "learning_rate": 0.0001,
+      "loss": 1.8316,
+      "step": 1003
+    },
+    {
+      "epoch": 0.24380767362797476,
+      "grad_norm": 0.5363386869430542,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 1004
+    },
+    {
+      "epoch": 0.24405050995628946,
+      "grad_norm": 0.5657698512077332,
+      "learning_rate": 0.0001,
+      "loss": 1.8426,
+      "step": 1005
+    },
+    {
+      "epoch": 0.24429334628460417,
+      "grad_norm": 0.548893392086029,
+      "learning_rate": 0.0001,
+      "loss": 1.7432,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2445361826129189,
+      "grad_norm": 0.5803833603858948,
+      "learning_rate": 0.0001,
+      "loss": 1.8235,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2447790189412336,
+      "grad_norm": 0.5895869731903076,
+      "learning_rate": 0.0001,
+      "loss": 1.787,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2450218552695483,
+      "grad_norm": 0.5639642477035522,
+      "learning_rate": 0.0001,
+      "loss": 1.6861,
+      "step": 1009
+    },
+    {
+      "epoch": 0.24526469159786304,
+      "grad_norm": 0.5599086880683899,
+      "learning_rate": 0.0001,
+      "loss": 1.7854,
+      "step": 1010
+    },
+    {
+      "epoch": 0.24550752792617775,
+      "grad_norm": 0.5383694171905518,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 1011
+    },
+    {
+      "epoch": 0.24575036425449248,
+      "grad_norm": 0.5369172096252441,
+      "learning_rate": 0.0001,
+      "loss": 1.5798,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2459932005828072,
+      "grad_norm": 0.5793865919113159,
+      "learning_rate": 0.0001,
+      "loss": 1.8873,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2462360369111219,
+      "grad_norm": 0.5521877408027649,
+      "learning_rate": 0.0001,
+      "loss": 1.743,
+      "step": 1014
+    },
+    {
+      "epoch": 0.24647887323943662,
+      "grad_norm": 0.5447569489479065,
+      "learning_rate": 0.0001,
+      "loss": 1.8138,
+      "step": 1015
+    },
+    {
+      "epoch": 0.24672170956775133,
+      "grad_norm": 0.5787389278411865,
+      "learning_rate": 0.0001,
+      "loss": 1.8368,
+      "step": 1016
+    },
+    {
+      "epoch": 0.24696454589606606,
+      "grad_norm": 0.5416997075080872,
+      "learning_rate": 0.0001,
+      "loss": 1.6505,
+      "step": 1017
+    },
+    {
+      "epoch": 0.24720738222438077,
+      "grad_norm": 0.5685380697250366,
+      "learning_rate": 0.0001,
+      "loss": 1.7282,
+      "step": 1018
+    },
+    {
+      "epoch": 0.24745021855269547,
+      "grad_norm": 0.562731921672821,
+      "learning_rate": 0.0001,
+      "loss": 1.7308,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2476930548810102,
+      "grad_norm": 0.5102028846740723,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2479358912093249,
+      "grad_norm": 0.5441479682922363,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 1021
+    },
+    {
+      "epoch": 0.24817872753763964,
+      "grad_norm": 0.5577821135520935,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 1022
+    },
+    {
+      "epoch": 0.24842156386595435,
+      "grad_norm": 0.5477888584136963,
+      "learning_rate": 0.0001,
+      "loss": 1.7947,
+      "step": 1023
+    },
+    {
+      "epoch": 0.24866440019426905,
+      "grad_norm": 0.5860451459884644,
+      "learning_rate": 0.0001,
+      "loss": 1.9354,
+      "step": 1024
+    },
+    {
+      "epoch": 0.24890723652258379,
+      "grad_norm": 0.5447387099266052,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2491500728508985,
+      "grad_norm": 0.5881152749061584,
+      "learning_rate": 0.0001,
+      "loss": 1.8635,
+      "step": 1026
+    },
+    {
+      "epoch": 0.24939290917921322,
+      "grad_norm": 0.5507918000221252,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 1027
+    },
+    {
+      "epoch": 0.24963574550752793,
+      "grad_norm": 0.5391066670417786,
+      "learning_rate": 0.0001,
+      "loss": 1.6698,
+      "step": 1028
+    },
+    {
+      "epoch": 0.24987858183584263,
+      "grad_norm": 0.5557683706283569,
+      "learning_rate": 0.0001,
+      "loss": 1.901,
+      "step": 1029
+    },
+    {
+      "epoch": 0.25012141816415734,
+      "grad_norm": 0.5551528930664062,
+      "learning_rate": 0.0001,
+      "loss": 1.7396,
+      "step": 1030
+    },
+    {
+      "epoch": 0.25036425449247207,
+      "grad_norm": 0.5728172659873962,
+      "learning_rate": 0.0001,
+      "loss": 1.9029,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2506070908207868,
+      "grad_norm": 0.5627360343933105,
+      "learning_rate": 0.0001,
+      "loss": 1.6951,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2508499271491015,
+      "grad_norm": 0.5626469850540161,
+      "learning_rate": 0.0001,
+      "loss": 1.8209,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2510927634774162,
+      "grad_norm": 0.5618252754211426,
+      "learning_rate": 0.0001,
+      "loss": 1.8646,
+      "step": 1034
+    },
+    {
+      "epoch": 0.25133559980573095,
+      "grad_norm": 0.5571959018707275,
+      "learning_rate": 0.0001,
+      "loss": 1.8346,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2515784361340457,
+      "grad_norm": 0.5539636611938477,
+      "learning_rate": 0.0001,
+      "loss": 1.7712,
+      "step": 1036
+    },
+    {
+      "epoch": 0.25182127246236036,
+      "grad_norm": 0.559863269329071,
+      "learning_rate": 0.0001,
+      "loss": 1.8712,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2520641087906751,
+      "grad_norm": 0.5371668934822083,
+      "learning_rate": 0.0001,
+      "loss": 1.8333,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2523069451189898,
+      "grad_norm": 0.5461929440498352,
+      "learning_rate": 0.0001,
+      "loss": 1.7055,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2525497814473045,
+      "grad_norm": 0.6017919778823853,
+      "learning_rate": 0.0001,
+      "loss": 1.7399,
+      "step": 1040
+    },
+    {
+      "epoch": 0.25279261777561923,
+      "grad_norm": 0.5876109600067139,
+      "learning_rate": 0.0001,
+      "loss": 1.9032,
+      "step": 1041
+    },
+    {
+      "epoch": 0.25303545410393397,
+      "grad_norm": 0.5787210464477539,
+      "learning_rate": 0.0001,
+      "loss": 1.753,
+      "step": 1042
+    },
+    {
+      "epoch": 0.25327829043224864,
+      "grad_norm": 0.5309500694274902,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2535211267605634,
+      "grad_norm": 0.539039134979248,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2537639630888781,
+      "grad_norm": 0.5269420742988586,
+      "learning_rate": 0.0001,
+      "loss": 1.6942,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2540067994171928,
+      "grad_norm": 0.5767653584480286,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2542496357455075,
+      "grad_norm": 0.5234490633010864,
+      "learning_rate": 0.0001,
+      "loss": 1.7217,
+      "step": 1047
+    },
+    {
+      "epoch": 0.25449247207382225,
+      "grad_norm": 0.5650292634963989,
+      "learning_rate": 0.0001,
+      "loss": 1.8731,
+      "step": 1048
+    },
+    {
+      "epoch": 0.254735308402137,
+      "grad_norm": 0.5449150800704956,
+      "learning_rate": 0.0001,
+      "loss": 1.8406,
+      "step": 1049
+    },
+    {
+      "epoch": 0.25497814473045166,
+      "grad_norm": 0.5538469552993774,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2552209810587664,
+      "grad_norm": 0.5362374186515808,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2554638173870811,
+      "grad_norm": 0.5948169231414795,
+      "learning_rate": 0.0001,
+      "loss": 1.7964,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2557066537153958,
+      "grad_norm": 0.5787668228149414,
+      "learning_rate": 0.0001,
+      "loss": 1.812,
+      "step": 1053
+    },
+    {
+      "epoch": 0.25594949004371054,
+      "grad_norm": 0.589809238910675,
+      "learning_rate": 0.0001,
+      "loss": 1.7048,
+      "step": 1054
+    },
+    {
+      "epoch": 0.25619232637202527,
+      "grad_norm": 0.5409502387046814,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 1055
+    },
+    {
+      "epoch": 0.25643516270033995,
+      "grad_norm": 0.6000292897224426,
+      "learning_rate": 0.0001,
+      "loss": 1.8869,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2566779990286547,
+      "grad_norm": 0.5915602445602417,
+      "learning_rate": 0.0001,
+      "loss": 1.9479,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2569208353569694,
+      "grad_norm": 0.5632345080375671,
+      "learning_rate": 0.0001,
+      "loss": 1.7527,
+      "step": 1058
+    },
+    {
+      "epoch": 0.25716367168528415,
+      "grad_norm": 0.6115126013755798,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2574065080135988,
+      "grad_norm": 0.5524129867553711,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 1060
+    },
+    {
+      "epoch": 0.25764934434191356,
+      "grad_norm": 0.5773287415504456,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2578921806702283,
+      "grad_norm": 0.5275243520736694,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 1062
+    },
+    {
+      "epoch": 0.25813501699854297,
+      "grad_norm": 0.5649922490119934,
+      "learning_rate": 0.0001,
+      "loss": 1.738,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2583778533268577,
+      "grad_norm": 0.5758867263793945,
+      "learning_rate": 0.0001,
+      "loss": 1.796,
+      "step": 1064
+    },
+    {
+      "epoch": 0.25862068965517243,
+      "grad_norm": 0.529830276966095,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2588635259834871,
+      "grad_norm": 0.6376872062683105,
+      "learning_rate": 0.0001,
+      "loss": 2.0468,
+      "step": 1066
+    },
+    {
+      "epoch": 0.25910636231180184,
+      "grad_norm": 0.5491669774055481,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2593491986401166,
+      "grad_norm": 0.5449734926223755,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 1068
+    },
+    {
+      "epoch": 0.25959203496843125,
+      "grad_norm": 0.6246454119682312,
+      "learning_rate": 0.0001,
+      "loss": 1.814,
+      "step": 1069
+    },
+    {
+      "epoch": 0.259834871296746,
+      "grad_norm": 0.5419249534606934,
+      "learning_rate": 0.0001,
+      "loss": 1.8207,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2600777076250607,
+      "grad_norm": 0.5403563976287842,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 1071
+    },
+    {
+      "epoch": 0.26032054395337545,
+      "grad_norm": 0.5629746913909912,
+      "learning_rate": 0.0001,
+      "loss": 1.8187,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2605633802816901,
+      "grad_norm": 0.5278500318527222,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 1073
+    },
+    {
+      "epoch": 0.26080621661000486,
+      "grad_norm": 0.5906748175621033,
+      "learning_rate": 0.0001,
+      "loss": 1.9218,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2610490529383196,
+      "grad_norm": 0.5823220014572144,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 1075
+    },
+    {
+      "epoch": 0.26129188926663427,
+      "grad_norm": 0.5446853041648865,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 1076
+    },
+    {
+      "epoch": 0.261534725594949,
+      "grad_norm": 0.5715393424034119,
+      "learning_rate": 0.0001,
+      "loss": 1.8105,
+      "step": 1077
+    },
+    {
+      "epoch": 0.26177756192326374,
+      "grad_norm": 0.5600488781929016,
+      "learning_rate": 0.0001,
+      "loss": 1.8232,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2620203982515784,
+      "grad_norm": 0.5303180813789368,
+      "learning_rate": 0.0001,
+      "loss": 1.8029,
+      "step": 1079
+    },
+    {
+      "epoch": 0.26226323457989315,
+      "grad_norm": 0.632108211517334,
+      "learning_rate": 0.0001,
+      "loss": 1.9495,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2625060709082079,
+      "grad_norm": 0.5606825947761536,
+      "learning_rate": 0.0001,
+      "loss": 1.7417,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2627489072365226,
+      "grad_norm": 0.5599796772003174,
+      "learning_rate": 0.0001,
+      "loss": 1.7526,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2629917435648373,
+      "grad_norm": 0.5996298789978027,
+      "learning_rate": 0.0001,
+      "loss": 1.8755,
+      "step": 1083
+    },
+    {
+      "epoch": 0.263234579893152,
+      "grad_norm": 0.5917564630508423,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 1084
+    },
+    {
+      "epoch": 0.26347741622146675,
+      "grad_norm": 0.5896962285041809,
+      "learning_rate": 0.0001,
+      "loss": 1.8442,
+      "step": 1085
+    },
+    {
+      "epoch": 0.26372025254978143,
+      "grad_norm": 0.5371586084365845,
+      "learning_rate": 0.0001,
+      "loss": 1.7769,
+      "step": 1086
+    },
+    {
+      "epoch": 0.26396308887809616,
+      "grad_norm": 0.5033586621284485,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2642059252064109,
+      "grad_norm": 0.595294713973999,
+      "learning_rate": 0.0001,
+      "loss": 1.9274,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2644487615347256,
+      "grad_norm": 0.5519765615463257,
+      "learning_rate": 0.0001,
+      "loss": 1.8435,
+      "step": 1089
+    },
+    {
+      "epoch": 0.2646915978630403,
+      "grad_norm": 0.5802309513092041,
+      "learning_rate": 0.0001,
+      "loss": 1.97,
+      "step": 1090
+    },
+    {
+      "epoch": 0.26493443419135504,
+      "grad_norm": 0.5462941527366638,
+      "learning_rate": 0.0001,
+      "loss": 1.8116,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2651772705196697,
+      "grad_norm": 0.5389935970306396,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 1092
+    },
+    {
+      "epoch": 0.26542010684798445,
+      "grad_norm": 0.5654126405715942,
+      "learning_rate": 0.0001,
+      "loss": 1.8127,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2656629431762992,
+      "grad_norm": 0.5769006013870239,
+      "learning_rate": 0.0001,
+      "loss": 1.8799,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2659057795046139,
+      "grad_norm": 0.5631387829780579,
+      "learning_rate": 0.0001,
+      "loss": 1.6312,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2661486158329286,
+      "grad_norm": 0.5755802392959595,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2663914521612433,
+      "grad_norm": 0.573484480381012,
+      "learning_rate": 0.0001,
+      "loss": 1.8436,
+      "step": 1097
+    },
+    {
+      "epoch": 0.26663428848955806,
+      "grad_norm": 0.5359703302383423,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 1098
+    },
+    {
+      "epoch": 0.26687712481787274,
+      "grad_norm": 0.516389012336731,
+      "learning_rate": 0.0001,
+      "loss": 1.5106,
+      "step": 1099
+    },
+    {
+      "epoch": 0.26711996114618747,
+      "grad_norm": 0.5863427519798279,
+      "learning_rate": 0.0001,
+      "loss": 1.7809,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2673627974745022,
+      "grad_norm": 0.5418011546134949,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2676056338028169,
+      "grad_norm": 0.5960160493850708,
+      "learning_rate": 0.0001,
+      "loss": 1.8422,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2678484701311316,
+      "grad_norm": 0.6008249521255493,
+      "learning_rate": 0.0001,
+      "loss": 1.7556,
+      "step": 1103
+    },
+    {
+      "epoch": 0.26809130645944634,
+      "grad_norm": 0.5573046803474426,
+      "learning_rate": 0.0001,
+      "loss": 1.7899,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2683341427877611,
+      "grad_norm": 0.529295027256012,
+      "learning_rate": 0.0001,
+      "loss": 1.7662,
+      "step": 1105
+    },
+    {
+      "epoch": 0.26857697911607575,
+      "grad_norm": 0.5498364567756653,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2688198154443905,
+      "grad_norm": 0.5485572814941406,
+      "learning_rate": 0.0001,
+      "loss": 1.8122,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2690626517727052,
+      "grad_norm": 0.5335333943367004,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2693054881010199,
+      "grad_norm": 0.5681390762329102,
+      "learning_rate": 0.0001,
+      "loss": 1.871,
+      "step": 1109
+    },
+    {
+      "epoch": 0.26954832442933463,
+      "grad_norm": 0.5439334511756897,
+      "learning_rate": 0.0001,
+      "loss": 1.8251,
+      "step": 1110
+    },
+    {
+      "epoch": 0.26979116075764936,
+      "grad_norm": 0.560706377029419,
+      "learning_rate": 0.0001,
+      "loss": 1.7744,
+      "step": 1111
+    },
+    {
+      "epoch": 0.27003399708596404,
+      "grad_norm": 0.5368844270706177,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 1112
+    },
+    {
+      "epoch": 0.27027683341427877,
+      "grad_norm": 0.5631430149078369,
+      "learning_rate": 0.0001,
+      "loss": 1.7116,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2705196697425935,
+      "grad_norm": 0.5687524676322937,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2707625060709082,
+      "grad_norm": 0.5624735951423645,
+      "learning_rate": 0.0001,
+      "loss": 1.724,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2710053423992229,
+      "grad_norm": 0.586024284362793,
+      "learning_rate": 0.0001,
+      "loss": 1.8873,
+      "step": 1116
+    },
+    {
+      "epoch": 0.27124817872753765,
+      "grad_norm": 0.5499001145362854,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2714910150558524,
+      "grad_norm": 0.5544242858886719,
+      "learning_rate": 0.0001,
+      "loss": 1.8069,
+      "step": 1118
+    },
+    {
+      "epoch": 0.27173385138416706,
+      "grad_norm": 0.5592672228813171,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2719766877124818,
+      "grad_norm": 0.524213433265686,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2722195240407965,
+      "grad_norm": 0.5297691822052002,
+      "learning_rate": 0.0001,
+      "loss": 1.8047,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2724623603691112,
+      "grad_norm": 0.5715131759643555,
+      "learning_rate": 0.0001,
+      "loss": 1.7379,
+      "step": 1122
+    },
+    {
+      "epoch": 0.27270519669742593,
+      "grad_norm": 0.5559701323509216,
+      "learning_rate": 0.0001,
+      "loss": 1.9112,
+      "step": 1123
+    },
+    {
+      "epoch": 0.27294803302574067,
+      "grad_norm": 0.5795487761497498,
+      "learning_rate": 0.0001,
+      "loss": 1.7956,
+      "step": 1124
+    },
+    {
+      "epoch": 0.27319086935405534,
+      "grad_norm": 0.5590142607688904,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2734337056823701,
+      "grad_norm": 0.5858544707298279,
+      "learning_rate": 0.0001,
+      "loss": 1.9377,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2736765420106848,
+      "grad_norm": 0.5422412753105164,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 1127
+    },
+    {
+      "epoch": 0.27391937833899954,
+      "grad_norm": 0.5824558138847351,
+      "learning_rate": 0.0001,
+      "loss": 1.8094,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2741622146673142,
+      "grad_norm": 0.5579871535301208,
+      "learning_rate": 0.0001,
+      "loss": 1.6865,
+      "step": 1129
+    },
+    {
+      "epoch": 0.27440505099562895,
+      "grad_norm": 0.5918609499931335,
+      "learning_rate": 0.0001,
+      "loss": 1.7295,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2746478873239437,
+      "grad_norm": 0.5762107968330383,
+      "learning_rate": 0.0001,
+      "loss": 1.7305,
+      "step": 1131
+    },
+    {
+      "epoch": 0.27489072365225836,
+      "grad_norm": 0.5555712580680847,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2751335599805731,
+      "grad_norm": 0.5766242146492004,
+      "learning_rate": 0.0001,
+      "loss": 1.9184,
+      "step": 1133
+    },
+    {
+      "epoch": 0.2753763963088878,
+      "grad_norm": 0.5332757234573364,
+      "learning_rate": 0.0001,
+      "loss": 1.7171,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2756192326372025,
+      "grad_norm": 0.6191022992134094,
+      "learning_rate": 0.0001,
+      "loss": 1.9584,
+      "step": 1135
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 0.603369414806366,
+      "learning_rate": 0.0001,
+      "loss": 1.7676,
+      "step": 1136
+    },
+    {
+      "epoch": 0.27610490529383197,
+      "grad_norm": 0.5369250178337097,
+      "learning_rate": 0.0001,
+      "loss": 1.7546,
+      "step": 1137
+    },
+    {
+      "epoch": 0.27634774162214665,
+      "grad_norm": 0.5725906491279602,
+      "learning_rate": 0.0001,
+      "loss": 1.923,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2765905779504614,
+      "grad_norm": 0.5639214515686035,
+      "learning_rate": 0.0001,
+      "loss": 1.7745,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2768334142787761,
+      "grad_norm": 0.5801315903663635,
+      "learning_rate": 0.0001,
+      "loss": 1.8606,
+      "step": 1140
+    },
+    {
+      "epoch": 0.27707625060709085,
+      "grad_norm": 0.5294753313064575,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2773190869354055,
+      "grad_norm": 0.5789963603019714,
+      "learning_rate": 0.0001,
+      "loss": 1.877,
+      "step": 1142
+    },
+    {
+      "epoch": 0.27756192326372026,
+      "grad_norm": 0.5986112356185913,
+      "learning_rate": 0.0001,
+      "loss": 1.7511,
+      "step": 1143
+    },
+    {
+      "epoch": 0.277804759592035,
+      "grad_norm": 0.611415684223175,
+      "learning_rate": 0.0001,
+      "loss": 1.772,
+      "step": 1144
+    },
+    {
+      "epoch": 0.27804759592034967,
+      "grad_norm": 0.5708376169204712,
+      "learning_rate": 0.0001,
+      "loss": 1.5659,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2782904322486644,
+      "grad_norm": 0.5890704989433289,
+      "learning_rate": 0.0001,
+      "loss": 1.7244,
+      "step": 1146
+    },
+    {
+      "epoch": 0.27853326857697913,
+      "grad_norm": 0.5593096613883972,
+      "learning_rate": 0.0001,
+      "loss": 1.796,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2787761049052938,
+      "grad_norm": 0.6066704392433167,
+      "learning_rate": 0.0001,
+      "loss": 1.8155,
+      "step": 1148
+    },
+    {
+      "epoch": 0.27901894123360854,
+      "grad_norm": 0.551701009273529,
+      "learning_rate": 0.0001,
+      "loss": 1.7212,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2792617775619233,
+      "grad_norm": 0.5450371503829956,
+      "learning_rate": 0.0001,
+      "loss": 1.6942,
+      "step": 1150
+    },
+    {
+      "epoch": 0.279504613890238,
+      "grad_norm": 0.5442568063735962,
+      "learning_rate": 0.0001,
+      "loss": 1.823,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2797474502185527,
+      "grad_norm": 0.5607459545135498,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2799902865468674,
+      "grad_norm": 0.5624642372131348,
+      "learning_rate": 0.0001,
+      "loss": 1.84,
+      "step": 1153
+    },
+    {
+      "epoch": 0.28023312287518215,
+      "grad_norm": 0.6023097038269043,
+      "learning_rate": 0.0001,
+      "loss": 1.695,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2804759592034968,
+      "grad_norm": 0.5366615653038025,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 1155
+    },
+    {
+      "epoch": 0.28071879553181156,
+      "grad_norm": 0.5833327174186707,
+      "learning_rate": 0.0001,
+      "loss": 1.8255,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2809616318601263,
+      "grad_norm": 0.5773898363113403,
+      "learning_rate": 0.0001,
+      "loss": 1.9525,
+      "step": 1157
+    },
+    {
+      "epoch": 0.28120446818844097,
+      "grad_norm": 0.569523274898529,
+      "learning_rate": 0.0001,
+      "loss": 1.7754,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2814473045167557,
+      "grad_norm": 0.5390274524688721,
+      "learning_rate": 0.0001,
+      "loss": 1.4668,
+      "step": 1159
+    },
+    {
+      "epoch": 0.28169014084507044,
+      "grad_norm": 0.5608516931533813,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2819329771733851,
+      "grad_norm": 0.5500361919403076,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 1161
+    },
+    {
+      "epoch": 0.28217581350169985,
+      "grad_norm": 0.5973489284515381,
+      "learning_rate": 0.0001,
+      "loss": 1.8089,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2824186498300146,
+      "grad_norm": 0.5561203360557556,
+      "learning_rate": 0.0001,
+      "loss": 1.6301,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2826614861583293,
+      "grad_norm": 0.6018040776252747,
+      "learning_rate": 0.0001,
+      "loss": 1.8193,
+      "step": 1164
+    },
+    {
+      "epoch": 0.282904322486644,
+      "grad_norm": 0.5456592440605164,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2831471588149587,
+      "grad_norm": 0.5560385584831238,
+      "learning_rate": 0.0001,
+      "loss": 1.8017,
+      "step": 1166
+    },
+    {
+      "epoch": 0.28338999514327345,
+      "grad_norm": 0.5754451751708984,
+      "learning_rate": 0.0001,
+      "loss": 1.9064,
+      "step": 1167
+    },
+    {
+      "epoch": 0.28363283147158813,
+      "grad_norm": 0.5309687256813049,
+      "learning_rate": 0.0001,
+      "loss": 1.7043,
+      "step": 1168
+    },
+    {
+      "epoch": 0.28387566779990286,
+      "grad_norm": 1.5462857484817505,
+      "learning_rate": 0.0001,
+      "loss": 1.7804,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2841185041282176,
+      "grad_norm": 0.5605103373527527,
+      "learning_rate": 0.0001,
+      "loss": 1.7713,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2843613404565323,
+      "grad_norm": 0.5687307715415955,
+      "learning_rate": 0.0001,
+      "loss": 1.7273,
+      "step": 1171
+    },
+    {
+      "epoch": 0.284604176784847,
+      "grad_norm": 0.5158697366714478,
+      "learning_rate": 0.0001,
+      "loss": 1.4238,
+      "step": 1172
+    },
+    {
+      "epoch": 0.28484701311316174,
+      "grad_norm": 0.5550355315208435,
+      "learning_rate": 0.0001,
+      "loss": 1.7404,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2850898494414764,
+      "grad_norm": 0.5797061920166016,
+      "learning_rate": 0.0001,
+      "loss": 1.7718,
+      "step": 1174
+    },
+    {
+      "epoch": 0.28533268576979115,
+      "grad_norm": 0.5334068536758423,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2855755220981059,
+      "grad_norm": 0.5494942665100098,
+      "learning_rate": 0.0001,
+      "loss": 1.7229,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2858183584264206,
+      "grad_norm": 0.6060776114463806,
+      "learning_rate": 0.0001,
+      "loss": 1.8268,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2860611947547353,
+      "grad_norm": 0.5575780868530273,
+      "learning_rate": 0.0001,
+      "loss": 1.8252,
+      "step": 1178
+    },
+    {
+      "epoch": 0.28630403108305,
+      "grad_norm": 0.5623884201049805,
+      "learning_rate": 0.0001,
+      "loss": 1.8873,
+      "step": 1179
+    },
+    {
+      "epoch": 0.28654686741136476,
+      "grad_norm": 0.5704313516616821,
+      "learning_rate": 0.0001,
+      "loss": 1.8121,
+      "step": 1180
+    },
+    {
+      "epoch": 0.28678970373967944,
+      "grad_norm": 0.5479267239570618,
+      "learning_rate": 0.0001,
+      "loss": 1.804,
+      "step": 1181
+    },
+    {
+      "epoch": 0.28703254006799417,
+      "grad_norm": 0.5762827396392822,
+      "learning_rate": 0.0001,
+      "loss": 1.8986,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2872753763963089,
+      "grad_norm": 0.592075526714325,
+      "learning_rate": 0.0001,
+      "loss": 1.8693,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2875182127246236,
+      "grad_norm": 0.5763819813728333,
+      "learning_rate": 0.0001,
+      "loss": 1.7756,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2877610490529383,
+      "grad_norm": 0.5444159507751465,
+      "learning_rate": 0.0001,
+      "loss": 1.7511,
+      "step": 1185
+    },
+    {
+      "epoch": 0.28800388538125304,
+      "grad_norm": 0.6072043776512146,
+      "learning_rate": 0.0001,
+      "loss": 1.9858,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2882467217095678,
+      "grad_norm": 0.5676252841949463,
+      "learning_rate": 0.0001,
+      "loss": 1.8784,
+      "step": 1187
+    },
+    {
+      "epoch": 0.28848955803788245,
+      "grad_norm": 0.5364545583724976,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2887323943661972,
+      "grad_norm": 0.5934111475944519,
+      "learning_rate": 0.0001,
+      "loss": 1.7471,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2889752306945119,
+      "grad_norm": 0.5661386847496033,
+      "learning_rate": 0.0001,
+      "loss": 1.694,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2892180670228266,
+      "grad_norm": 0.5890941023826599,
+      "learning_rate": 0.0001,
+      "loss": 1.7736,
+      "step": 1191
+    },
+    {
+      "epoch": 0.28946090335114133,
+      "grad_norm": 0.572519063949585,
+      "learning_rate": 0.0001,
+      "loss": 1.8875,
+      "step": 1192
+    },
+    {
+      "epoch": 0.28970373967945606,
+      "grad_norm": 0.5531225204467773,
+      "learning_rate": 0.0001,
+      "loss": 1.771,
+      "step": 1193
+    },
+    {
+      "epoch": 0.28994657600777074,
+      "grad_norm": 0.5932171940803528,
+      "learning_rate": 0.0001,
+      "loss": 1.9663,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2901894123360855,
+      "grad_norm": 0.5722930431365967,
+      "learning_rate": 0.0001,
+      "loss": 1.803,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2904322486644002,
+      "grad_norm": 0.6092216372489929,
+      "learning_rate": 0.0001,
+      "loss": 1.6879,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2906750849927149,
+      "grad_norm": 0.5634588003158569,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2909179213210296,
+      "grad_norm": 0.5632761120796204,
+      "learning_rate": 0.0001,
+      "loss": 1.7852,
+      "step": 1198
+    },
+    {
+      "epoch": 0.29116075764934435,
+      "grad_norm": 0.5588541626930237,
+      "learning_rate": 0.0001,
+      "loss": 1.6781,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2914035939776591,
+      "grad_norm": 0.5605027675628662,
+      "learning_rate": 0.0001,
+      "loss": 1.8683,
+      "step": 1200
+    },
+    {
+      "epoch": 0.29164643030597376,
+      "grad_norm": 0.508496105670929,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2918892666342885,
+      "grad_norm": 0.5228567123413086,
+      "learning_rate": 0.0001,
+      "loss": 1.7026,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2921321029626032,
+      "grad_norm": 0.6024715304374695,
+      "learning_rate": 0.0001,
+      "loss": 1.9143,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2923749392909179,
+      "grad_norm": 0.5535398125648499,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 1204
+    },
+    {
+      "epoch": 0.29261777561923263,
+      "grad_norm": 0.5778299570083618,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 1205
+    },
+    {
+      "epoch": 0.29286061194754737,
+      "grad_norm": 0.5712555050849915,
+      "learning_rate": 0.0001,
+      "loss": 1.6741,
+      "step": 1206
+    },
+    {
+      "epoch": 0.29310344827586204,
+      "grad_norm": 0.5476194024085999,
+      "learning_rate": 0.0001,
+      "loss": 1.7123,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2933462846041768,
+      "grad_norm": 0.6042921543121338,
+      "learning_rate": 0.0001,
+      "loss": 1.8572,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2935891209324915,
+      "grad_norm": 0.5517978668212891,
+      "learning_rate": 0.0001,
+      "loss": 1.7138,
+      "step": 1209
+    },
+    {
+      "epoch": 0.29383195726080624,
+      "grad_norm": 0.57857346534729,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2940747935891209,
+      "grad_norm": 0.6032268404960632,
+      "learning_rate": 0.0001,
+      "loss": 1.9074,
+      "step": 1211
+    },
+    {
+      "epoch": 0.29431762991743565,
+      "grad_norm": 0.5374449491500854,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2945604662457504,
+      "grad_norm": 0.5887454152107239,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 1213
+    },
+    {
+      "epoch": 0.29480330257406506,
+      "grad_norm": 0.6122771501541138,
+      "learning_rate": 0.0001,
+      "loss": 1.8,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2950461389023798,
+      "grad_norm": 0.554463803768158,
+      "learning_rate": 0.0001,
+      "loss": 1.802,
+      "step": 1215
+    },
+    {
+      "epoch": 0.29528897523069453,
+      "grad_norm": 0.5563312768936157,
+      "learning_rate": 0.0001,
+      "loss": 1.727,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2955318115590092,
+      "grad_norm": 0.5625701546669006,
+      "learning_rate": 0.0001,
+      "loss": 1.6385,
+      "step": 1217
+    },
+    {
+      "epoch": 0.29577464788732394,
+      "grad_norm": 0.5303876996040344,
+      "learning_rate": 0.0001,
+      "loss": 1.6915,
+      "step": 1218
+    },
+    {
+      "epoch": 0.29601748421563867,
+      "grad_norm": 0.5515774488449097,
+      "learning_rate": 0.0001,
+      "loss": 1.4945,
+      "step": 1219
+    },
+    {
+      "epoch": 0.29626032054395335,
+      "grad_norm": 0.5511413216590881,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2965031568722681,
+      "grad_norm": 0.5535469055175781,
+      "learning_rate": 0.0001,
+      "loss": 1.7205,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2967459932005828,
+      "grad_norm": 0.5677886009216309,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 1222
+    },
+    {
+      "epoch": 0.29698882952889755,
+      "grad_norm": 0.5623748302459717,
+      "learning_rate": 0.0001,
+      "loss": 1.7513,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2972316658572122,
+      "grad_norm": 0.5762066841125488,
+      "learning_rate": 0.0001,
+      "loss": 1.9384,
+      "step": 1224
+    },
+    {
+      "epoch": 0.29747450218552696,
+      "grad_norm": 0.5834498405456543,
+      "learning_rate": 0.0001,
+      "loss": 1.807,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2977173385138417,
+      "grad_norm": 0.6799818277359009,
+      "learning_rate": 0.0001,
+      "loss": 2.0577,
+      "step": 1226
+    },
+    {
+      "epoch": 0.29796017484215637,
+      "grad_norm": 0.5903156995773315,
+      "learning_rate": 0.0001,
+      "loss": 1.9314,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2982030111704711,
+      "grad_norm": 0.626448392868042,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 1228
+    },
+    {
+      "epoch": 0.29844584749878583,
+      "grad_norm": 0.5882961750030518,
+      "learning_rate": 0.0001,
+      "loss": 1.8942,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2986886838271005,
+      "grad_norm": 0.5272478461265564,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 1230
+    },
+    {
+      "epoch": 0.29893152015541524,
+      "grad_norm": 0.5757167339324951,
+      "learning_rate": 0.0001,
+      "loss": 1.7958,
+      "step": 1231
+    },
+    {
+      "epoch": 0.29917435648373,
+      "grad_norm": 0.5639874935150146,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2994171928120447,
+      "grad_norm": 0.540255606174469,
+      "learning_rate": 0.0001,
+      "loss": 1.6874,
+      "step": 1233
+    },
+    {
+      "epoch": 0.2996600291403594,
+      "grad_norm": 0.6375817060470581,
+      "learning_rate": 0.0001,
+      "loss": 1.7894,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2999028654686741,
+      "grad_norm": 0.5961909890174866,
+      "learning_rate": 0.0001,
+      "loss": 1.8181,
+      "step": 1235
+    },
+    {
+      "epoch": 0.30014570179698885,
+      "grad_norm": 0.5758340358734131,
+      "learning_rate": 0.0001,
+      "loss": 1.8422,
+      "step": 1236
+    },
+    {
+      "epoch": 0.30038853812530353,
+      "grad_norm": 0.5907910466194153,
+      "learning_rate": 0.0001,
+      "loss": 1.9534,
+      "step": 1237
+    },
+    {
+      "epoch": 0.30063137445361826,
+      "grad_norm": 0.5534744262695312,
+      "learning_rate": 0.0001,
+      "loss": 1.6591,
+      "step": 1238
+    },
+    {
+      "epoch": 0.300874210781933,
+      "grad_norm": 0.5392131209373474,
+      "learning_rate": 0.0001,
+      "loss": 1.7531,
+      "step": 1239
+    },
+    {
+      "epoch": 0.30111704711024767,
+      "grad_norm": 0.5619401335716248,
+      "learning_rate": 0.0001,
+      "loss": 1.8698,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3013598834385624,
+      "grad_norm": 0.5474029183387756,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 1241
+    },
+    {
+      "epoch": 0.30160271976687714,
+      "grad_norm": 0.5693953037261963,
+      "learning_rate": 0.0001,
+      "loss": 1.7505,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3018455560951918,
+      "grad_norm": 0.5436322689056396,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 1243
+    },
+    {
+      "epoch": 0.30208839242350655,
+      "grad_norm": 0.5292843580245972,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3023312287518213,
+      "grad_norm": 0.579740583896637,
+      "learning_rate": 0.0001,
+      "loss": 1.8215,
+      "step": 1245
+    },
+    {
+      "epoch": 0.302574065080136,
+      "grad_norm": 0.5974770188331604,
+      "learning_rate": 0.0001,
+      "loss": 1.7842,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3028169014084507,
+      "grad_norm": 0.5881436467170715,
+      "learning_rate": 0.0001,
+      "loss": 1.7556,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3030597377367654,
+      "grad_norm": 0.5402075052261353,
+      "learning_rate": 0.0001,
+      "loss": 1.6768,
+      "step": 1248
+    },
+    {
+      "epoch": 0.30330257406508015,
+      "grad_norm": 0.5831952691078186,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 1249
+    },
+    {
+      "epoch": 0.30354541039339483,
+      "grad_norm": 0.5571645498275757,
+      "learning_rate": 0.0001,
+      "loss": 1.7934,
+      "step": 1250
+    },
+    {
+      "epoch": 0.30378824672170957,
+      "grad_norm": 0.5791111588478088,
+      "learning_rate": 0.0001,
+      "loss": 1.8274,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3040310830500243,
+      "grad_norm": 0.5482165217399597,
+      "learning_rate": 0.0001,
+      "loss": 1.7685,
+      "step": 1252
+    },
+    {
+      "epoch": 0.304273919378339,
+      "grad_norm": 0.5379577875137329,
+      "learning_rate": 0.0001,
+      "loss": 1.7371,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3045167557066537,
+      "grad_norm": 0.5821003913879395,
+      "learning_rate": 0.0001,
+      "loss": 1.9647,
+      "step": 1254
+    },
+    {
+      "epoch": 0.30475959203496844,
+      "grad_norm": 0.573422908782959,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3050024283632832,
+      "grad_norm": 0.595234215259552,
+      "learning_rate": 0.0001,
+      "loss": 1.7981,
+      "step": 1256
+    },
+    {
+      "epoch": 0.30524526469159785,
+      "grad_norm": 0.5558188557624817,
+      "learning_rate": 0.0001,
+      "loss": 1.7843,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3054881010199126,
+      "grad_norm": 0.5586723685264587,
+      "learning_rate": 0.0001,
+      "loss": 1.6505,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3057309373482273,
+      "grad_norm": 0.6428223848342896,
+      "learning_rate": 0.0001,
+      "loss": 1.8015,
+      "step": 1259
+    },
+    {
+      "epoch": 0.305973773676542,
+      "grad_norm": 0.5657460689544678,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3062166100048567,
+      "grad_norm": 0.5800066590309143,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 1261
+    },
+    {
+      "epoch": 0.30645944633317146,
+      "grad_norm": 0.5351372361183167,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 1262
+    },
+    {
+      "epoch": 0.30670228266148614,
+      "grad_norm": 0.5728726983070374,
+      "learning_rate": 0.0001,
+      "loss": 1.7704,
+      "step": 1263
+    },
+    {
+      "epoch": 0.30694511898980087,
+      "grad_norm": 0.7134020328521729,
+      "learning_rate": 0.0001,
+      "loss": 1.8761,
+      "step": 1264
+    },
+    {
+      "epoch": 0.3071879553181156,
+      "grad_norm": 0.6183131337165833,
+      "learning_rate": 0.0001,
+      "loss": 1.9261,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3074307916464303,
+      "grad_norm": 1.3757343292236328,
+      "learning_rate": 0.0001,
+      "loss": 1.6129,
+      "step": 1266
+    },
+    {
+      "epoch": 0.307673627974745,
+      "grad_norm": 0.5625571012496948,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 1267
+    },
+    {
+      "epoch": 0.30791646430305974,
+      "grad_norm": 0.5531574487686157,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3081593006313745,
+      "grad_norm": 0.8341613411903381,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 1269
+    },
+    {
+      "epoch": 0.30840213695968915,
+      "grad_norm": 0.5458047389984131,
+      "learning_rate": 0.0001,
+      "loss": 1.9236,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3086449732880039,
+      "grad_norm": 0.5536592602729797,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3088878096163186,
+      "grad_norm": 0.5676348805427551,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 1272
+    },
+    {
+      "epoch": 0.3091306459446333,
+      "grad_norm": 0.6149638891220093,
+      "learning_rate": 0.0001,
+      "loss": 1.7848,
+      "step": 1273
+    },
+    {
+      "epoch": 0.30937348227294803,
+      "grad_norm": 0.5674256682395935,
+      "learning_rate": 0.0001,
+      "loss": 1.751,
+      "step": 1274
+    },
+    {
+      "epoch": 0.30961631860126276,
+      "grad_norm": 0.5894138813018799,
+      "learning_rate": 0.0001,
+      "loss": 1.8041,
+      "step": 1275
+    },
+    {
+      "epoch": 0.30985915492957744,
+      "grad_norm": 0.531599223613739,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3101019912578922,
+      "grad_norm": 0.5443134903907776,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3103448275862069,
+      "grad_norm": 0.5844945907592773,
+      "learning_rate": 0.0001,
+      "loss": 1.7385,
+      "step": 1278
+    },
+    {
+      "epoch": 0.31058766391452164,
+      "grad_norm": 0.5538145899772644,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3108305002428363,
+      "grad_norm": 0.5729904174804688,
+      "learning_rate": 0.0001,
+      "loss": 1.6952,
+      "step": 1280
+    },
+    {
+      "epoch": 0.31107333657115105,
+      "grad_norm": 0.5446047186851501,
+      "learning_rate": 0.0001,
+      "loss": 1.7144,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3113161728994658,
+      "grad_norm": 0.5747256278991699,
+      "learning_rate": 0.0001,
+      "loss": 1.9335,
+      "step": 1282
+    },
+    {
+      "epoch": 0.31155900922778046,
+      "grad_norm": 0.5712779760360718,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3118018455560952,
+      "grad_norm": 0.5434122085571289,
+      "learning_rate": 0.0001,
+      "loss": 1.7904,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3120446818844099,
+      "grad_norm": 0.5477941036224365,
+      "learning_rate": 0.0001,
+      "loss": 1.6354,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3122875182127246,
+      "grad_norm": 0.5642449259757996,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 1286
+    },
+    {
+      "epoch": 0.31253035454103933,
+      "grad_norm": 0.5494605302810669,
+      "learning_rate": 0.0001,
+      "loss": 1.722,
+      "step": 1287
+    },
+    {
+      "epoch": 0.31277319086935407,
+      "grad_norm": 0.5640792846679688,
+      "learning_rate": 0.0001,
+      "loss": 1.8397,
+      "step": 1288
+    },
+    {
+      "epoch": 0.31301602719766874,
+      "grad_norm": 0.5498180985450745,
+      "learning_rate": 0.0001,
+      "loss": 1.7621,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3132588635259835,
+      "grad_norm": 0.5406601428985596,
+      "learning_rate": 0.0001,
+      "loss": 1.7247,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3135016998542982,
+      "grad_norm": 0.5511451363563538,
+      "learning_rate": 0.0001,
+      "loss": 1.8245,
+      "step": 1291
+    },
+    {
+      "epoch": 0.31374453618261294,
+      "grad_norm": 0.5950665473937988,
+      "learning_rate": 0.0001,
+      "loss": 1.7691,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3139873725109276,
+      "grad_norm": 0.5692906379699707,
+      "learning_rate": 0.0001,
+      "loss": 1.6992,
+      "step": 1293
+    },
+    {
+      "epoch": 0.31423020883924235,
+      "grad_norm": 0.5657053589820862,
+      "learning_rate": 0.0001,
+      "loss": 1.7324,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3144730451675571,
+      "grad_norm": 0.5812044143676758,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 1295
+    },
+    {
+      "epoch": 0.31471588149587176,
+      "grad_norm": 0.5734257102012634,
+      "learning_rate": 0.0001,
+      "loss": 1.808,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3149587178241865,
+      "grad_norm": 0.5991582274436951,
+      "learning_rate": 0.0001,
+      "loss": 1.8522,
+      "step": 1297
+    },
+    {
+      "epoch": 0.31520155415250123,
+      "grad_norm": 0.5761125683784485,
+      "learning_rate": 0.0001,
+      "loss": 1.867,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3154443904808159,
+      "grad_norm": 0.5132241249084473,
+      "learning_rate": 0.0001,
+      "loss": 1.3686,
+      "step": 1299
+    },
+    {
+      "epoch": 0.31568722680913064,
+      "grad_norm": 0.5863776206970215,
+      "learning_rate": 0.0001,
+      "loss": 1.7227,
+      "step": 1300
+    },
+    {
+      "epoch": 0.31593006313744537,
+      "grad_norm": 0.6152379512786865,
+      "learning_rate": 0.0001,
+      "loss": 1.987,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3161728994657601,
+      "grad_norm": 0.5339058041572571,
+      "learning_rate": 0.0001,
+      "loss": 1.6471,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3164157357940748,
+      "grad_norm": 0.5914755463600159,
+      "learning_rate": 0.0001,
+      "loss": 1.7393,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3166585721223895,
+      "grad_norm": 0.6291932463645935,
+      "learning_rate": 0.0001,
+      "loss": 1.8296,
+      "step": 1304
+    },
+    {
+      "epoch": 0.31690140845070425,
+      "grad_norm": 0.6238904595375061,
+      "learning_rate": 0.0001,
+      "loss": 1.9846,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3171442447790189,
+      "grad_norm": 0.5747682452201843,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 1306
+    },
+    {
+      "epoch": 0.31738708110733366,
+      "grad_norm": 0.5602648854255676,
+      "learning_rate": 0.0001,
+      "loss": 1.6642,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3176299174356484,
+      "grad_norm": 0.5592392683029175,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 1308
+    },
+    {
+      "epoch": 0.31787275376396307,
+      "grad_norm": 0.6572312116622925,
+      "learning_rate": 0.0001,
+      "loss": 2.1072,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3181155900922778,
+      "grad_norm": 0.5818151831626892,
+      "learning_rate": 0.0001,
+      "loss": 1.7513,
+      "step": 1310
+    },
+    {
+      "epoch": 0.31835842642059253,
+      "grad_norm": 0.5654920935630798,
+      "learning_rate": 0.0001,
+      "loss": 1.7266,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3186012627489072,
+      "grad_norm": 0.5874744653701782,
+      "learning_rate": 0.0001,
+      "loss": 1.8714,
+      "step": 1312
+    },
+    {
+      "epoch": 0.31884409907722194,
+      "grad_norm": 0.6022887229919434,
+      "learning_rate": 0.0001,
+      "loss": 1.7934,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3190869354055367,
+      "grad_norm": 0.5618531703948975,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3193297717338514,
+      "grad_norm": 0.564096987247467,
+      "learning_rate": 0.0001,
+      "loss": 1.7697,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3195726080621661,
+      "grad_norm": 0.5624117851257324,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3198154443904808,
+      "grad_norm": 0.6024875640869141,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 1317
+    },
+    {
+      "epoch": 0.32005828071879555,
+      "grad_norm": 0.5781903862953186,
+      "learning_rate": 0.0001,
+      "loss": 1.7032,
+      "step": 1318
+    },
+    {
+      "epoch": 0.32030111704711023,
+      "grad_norm": 0.5536518096923828,
+      "learning_rate": 0.0001,
+      "loss": 1.7281,
+      "step": 1319
+    },
+    {
+      "epoch": 0.32054395337542496,
+      "grad_norm": 0.577639102935791,
+      "learning_rate": 0.0001,
+      "loss": 1.7655,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3207867897037397,
+      "grad_norm": 0.5466935038566589,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 1321
+    },
+    {
+      "epoch": 0.32102962603205437,
+      "grad_norm": 0.6217895150184631,
+      "learning_rate": 0.0001,
+      "loss": 1.9077,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3212724623603691,
+      "grad_norm": 0.6197120547294617,
+      "learning_rate": 0.0001,
+      "loss": 1.9127,
+      "step": 1323
+    },
+    {
+      "epoch": 0.32151529868868384,
+      "grad_norm": 0.5837169289588928,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 1324
+    },
+    {
+      "epoch": 0.32175813501699857,
+      "grad_norm": 0.5738924741744995,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 1325
+    },
+    {
+      "epoch": 0.32200097134531325,
+      "grad_norm": 0.5743533968925476,
+      "learning_rate": 0.0001,
+      "loss": 1.6758,
+      "step": 1326
+    },
+    {
+      "epoch": 0.322243807673628,
+      "grad_norm": 0.5642489194869995,
+      "learning_rate": 0.0001,
+      "loss": 1.8368,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3224866440019427,
+      "grad_norm": 0.5734957456588745,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3227294803302574,
+      "grad_norm": 0.576200008392334,
+      "learning_rate": 0.0001,
+      "loss": 1.8198,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3229723166585721,
+      "grad_norm": 0.5713797807693481,
+      "learning_rate": 0.0001,
+      "loss": 1.8013,
+      "step": 1330
+    },
+    {
+      "epoch": 0.32321515298688686,
+      "grad_norm": 0.6150364279747009,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 1331
+    },
+    {
+      "epoch": 0.32345798931520153,
+      "grad_norm": 0.5615043640136719,
+      "learning_rate": 0.0001,
+      "loss": 1.6789,
+      "step": 1332
+    },
+    {
+      "epoch": 0.32370082564351627,
+      "grad_norm": 0.5571615695953369,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 1333
+    },
+    {
+      "epoch": 0.323943661971831,
+      "grad_norm": 0.5916815996170044,
+      "learning_rate": 0.0001,
+      "loss": 1.8651,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3241864983001457,
+      "grad_norm": 0.5488576889038086,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3244293346284604,
+      "grad_norm": 0.5989965796470642,
+      "learning_rate": 0.0001,
+      "loss": 1.83,
+      "step": 1336
+    },
+    {
+      "epoch": 0.32467217095677514,
+      "grad_norm": 0.5402348041534424,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3249150072850899,
+      "grad_norm": 0.5602414608001709,
+      "learning_rate": 0.0001,
+      "loss": 1.7319,
+      "step": 1338
+    },
+    {
+      "epoch": 0.32515784361340455,
+      "grad_norm": 0.5434054732322693,
+      "learning_rate": 0.0001,
+      "loss": 1.7123,
+      "step": 1339
+    },
+    {
+      "epoch": 0.3254006799417193,
+      "grad_norm": 0.5134353637695312,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 1340
+    },
+    {
+      "epoch": 0.325643516270034,
+      "grad_norm": 0.568906307220459,
+      "learning_rate": 0.0001,
+      "loss": 1.7518,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3258863525983487,
+      "grad_norm": 0.5728622078895569,
+      "learning_rate": 0.0001,
+      "loss": 1.6901,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3261291889266634,
+      "grad_norm": 0.5689023733139038,
+      "learning_rate": 0.0001,
+      "loss": 1.7392,
+      "step": 1343
+    },
+    {
+      "epoch": 0.32637202525497816,
+      "grad_norm": 0.5729114413261414,
+      "learning_rate": 0.0001,
+      "loss": 1.7333,
+      "step": 1344
+    },
+    {
+      "epoch": 0.32661486158329284,
+      "grad_norm": 0.5677634477615356,
+      "learning_rate": 0.0001,
+      "loss": 1.7947,
+      "step": 1345
+    },
+    {
+      "epoch": 0.32685769791160757,
+      "grad_norm": 0.5566307306289673,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3271005342399223,
+      "grad_norm": 0.566073477268219,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 1347
+    },
+    {
+      "epoch": 0.32734337056823704,
+      "grad_norm": 0.5132454037666321,
+      "learning_rate": 0.0001,
+      "loss": 1.6655,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3275862068965517,
+      "grad_norm": 0.5575150847434998,
+      "learning_rate": 0.0001,
+      "loss": 1.6852,
+      "step": 1349
+    },
+    {
+      "epoch": 0.32782904322486645,
+      "grad_norm": 0.5818497538566589,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3280718795531812,
+      "grad_norm": 0.5753150582313538,
+      "learning_rate": 0.0001,
+      "loss": 1.7942,
+      "step": 1351
+    },
+    {
+      "epoch": 0.32831471588149586,
+      "grad_norm": 0.5964739918708801,
+      "learning_rate": 0.0001,
+      "loss": 1.9348,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3285575522098106,
+      "grad_norm": 0.5620943307876587,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3288003885381253,
+      "grad_norm": 0.5440359711647034,
+      "learning_rate": 0.0001,
+      "loss": 1.7556,
+      "step": 1354
+    },
+    {
+      "epoch": 0.32904322486644,
+      "grad_norm": 0.5927224159240723,
+      "learning_rate": 0.0001,
+      "loss": 1.7276,
+      "step": 1355
+    },
+    {
+      "epoch": 0.32928606119475473,
+      "grad_norm": 0.5532374382019043,
+      "learning_rate": 0.0001,
+      "loss": 1.6874,
+      "step": 1356
+    },
+    {
+      "epoch": 0.32952889752306946,
+      "grad_norm": 0.5345386266708374,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 1357
+    },
+    {
+      "epoch": 0.32977173385138414,
+      "grad_norm": 0.5551013350486755,
+      "learning_rate": 0.0001,
+      "loss": 1.7162,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3300145701796989,
+      "grad_norm": 0.5722214579582214,
+      "learning_rate": 0.0001,
+      "loss": 1.8486,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3302574065080136,
+      "grad_norm": 0.5920416712760925,
+      "learning_rate": 0.0001,
+      "loss": 1.8309,
+      "step": 1360
+    },
+    {
+      "epoch": 0.33050024283632834,
+      "grad_norm": 0.5732881426811218,
+      "learning_rate": 0.0001,
+      "loss": 1.8577,
+      "step": 1361
+    },
+    {
+      "epoch": 0.330743079164643,
+      "grad_norm": 0.5619862675666809,
+      "learning_rate": 0.0001,
+      "loss": 1.8305,
+      "step": 1362
+    },
+    {
+      "epoch": 0.33098591549295775,
+      "grad_norm": 0.6099917888641357,
+      "learning_rate": 0.0001,
+      "loss": 1.8949,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3312287518212725,
+      "grad_norm": 0.5454998016357422,
+      "learning_rate": 0.0001,
+      "loss": 1.6676,
+      "step": 1364
+    },
+    {
+      "epoch": 0.33147158814958716,
+      "grad_norm": 0.5881146192550659,
+      "learning_rate": 0.0001,
+      "loss": 2.0026,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3317144244779019,
+      "grad_norm": 0.5716860294342041,
+      "learning_rate": 0.0001,
+      "loss": 1.7442,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3319572608062166,
+      "grad_norm": 0.5615752339363098,
+      "learning_rate": 0.0001,
+      "loss": 1.7291,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3322000971345313,
+      "grad_norm": 0.5943424105644226,
+      "learning_rate": 0.0001,
+      "loss": 1.9471,
+      "step": 1368
+    },
+    {
+      "epoch": 0.33244293346284604,
+      "grad_norm": 0.5808683633804321,
+      "learning_rate": 0.0001,
+      "loss": 1.7641,
+      "step": 1369
+    },
+    {
+      "epoch": 0.33268576979116077,
+      "grad_norm": 0.5758082270622253,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3329286061194755,
+      "grad_norm": 0.5595842599868774,
+      "learning_rate": 0.0001,
+      "loss": 1.7815,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3331714424477902,
+      "grad_norm": 0.5441301465034485,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3334142787761049,
+      "grad_norm": 0.6101175546646118,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 1373
+    },
+    {
+      "epoch": 0.33365711510441964,
+      "grad_norm": 0.5867419838905334,
+      "learning_rate": 0.0001,
+      "loss": 1.8394,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3338999514327343,
+      "grad_norm": 0.58198481798172,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 1375
+    },
+    {
+      "epoch": 0.33414278776104905,
+      "grad_norm": 0.5561076998710632,
+      "learning_rate": 0.0001,
+      "loss": 1.7582,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3343856240893638,
+      "grad_norm": 0.5937936902046204,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 1377
+    },
+    {
+      "epoch": 0.33462846041767846,
+      "grad_norm": 0.6016126275062561,
+      "learning_rate": 0.0001,
+      "loss": 1.7945,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3348712967459932,
+      "grad_norm": 0.6054165959358215,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 1379
+    },
+    {
+      "epoch": 0.33511413307430793,
+      "grad_norm": 0.6050477623939514,
+      "learning_rate": 0.0001,
+      "loss": 1.8576,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3353569694026226,
+      "grad_norm": 0.5650911331176758,
+      "learning_rate": 0.0001,
+      "loss": 1.7675,
+      "step": 1381
+    },
+    {
+      "epoch": 0.33559980573093734,
+      "grad_norm": 0.5832107663154602,
+      "learning_rate": 0.0001,
+      "loss": 1.8276,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3358426420592521,
+      "grad_norm": 0.5333999991416931,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3360854783875668,
+      "grad_norm": 0.5361559391021729,
+      "learning_rate": 0.0001,
+      "loss": 1.7057,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3363283147158815,
+      "grad_norm": 0.5865080952644348,
+      "learning_rate": 0.0001,
+      "loss": 1.739,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3365711510441962,
+      "grad_norm": 0.5516999959945679,
+      "learning_rate": 0.0001,
+      "loss": 1.7335,
+      "step": 1386
+    },
+    {
+      "epoch": 0.33681398737251095,
+      "grad_norm": 0.5868236422538757,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3370568237008256,
+      "grad_norm": 0.5723714828491211,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 1388
+    },
+    {
+      "epoch": 0.33729966002914036,
+      "grad_norm": 0.6132412552833557,
+      "learning_rate": 0.0001,
+      "loss": 1.8463,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3375424963574551,
+      "grad_norm": 0.5300971865653992,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 1390
+    },
+    {
+      "epoch": 0.33778533268576977,
+      "grad_norm": 0.5721192955970764,
+      "learning_rate": 0.0001,
+      "loss": 1.7578,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3380281690140845,
+      "grad_norm": 0.522925615310669,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 1392
+    },
+    {
+      "epoch": 0.33827100534239923,
+      "grad_norm": 0.5766683220863342,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 1393
+    },
+    {
+      "epoch": 0.33851384167071397,
+      "grad_norm": 0.5427753329277039,
+      "learning_rate": 0.0001,
+      "loss": 1.7914,
+      "step": 1394
+    },
+    {
+      "epoch": 0.33875667799902864,
+      "grad_norm": 0.5608222484588623,
+      "learning_rate": 0.0001,
+      "loss": 1.7307,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3389995143273434,
+      "grad_norm": 0.5690093040466309,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3392423506556581,
+      "grad_norm": 0.5487404465675354,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3394851869839728,
+      "grad_norm": 0.6467126607894897,
+      "learning_rate": 0.0001,
+      "loss": 1.8987,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3397280233122875,
+      "grad_norm": 0.5495607852935791,
+      "learning_rate": 0.0001,
+      "loss": 1.6867,
+      "step": 1399
+    },
+    {
+      "epoch": 0.33997085964060225,
+      "grad_norm": 0.5042486786842346,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 1400
+    },
+    {
+      "epoch": 0.34021369596891693,
+      "grad_norm": 0.5387740731239319,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 1401
+    },
+    {
+      "epoch": 0.34045653229723166,
+      "grad_norm": 0.5586868524551392,
+      "learning_rate": 0.0001,
+      "loss": 1.7908,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3406993686255464,
+      "grad_norm": 0.5448033809661865,
+      "learning_rate": 0.0001,
+      "loss": 1.8032,
+      "step": 1403
+    },
+    {
+      "epoch": 0.34094220495386107,
+      "grad_norm": 0.5725580453872681,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 1404
+    },
+    {
+      "epoch": 0.3411850412821758,
+      "grad_norm": 0.5539155006408691,
+      "learning_rate": 0.0001,
+      "loss": 1.829,
+      "step": 1405
+    },
+    {
+      "epoch": 0.34142787761049054,
+      "grad_norm": 0.5847489237785339,
+      "learning_rate": 0.0001,
+      "loss": 1.6599,
+      "step": 1406
+    },
+    {
+      "epoch": 0.34167071393880527,
+      "grad_norm": 0.6161686778068542,
+      "learning_rate": 0.0001,
+      "loss": 1.8224,
+      "step": 1407
+    },
+    {
+      "epoch": 0.34191355026711995,
+      "grad_norm": 0.5436552166938782,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 1408
+    },
+    {
+      "epoch": 0.3421563865954347,
+      "grad_norm": 0.5373216867446899,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3423992229237494,
+      "grad_norm": 0.5849260091781616,
+      "learning_rate": 0.0001,
+      "loss": 1.7975,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3426420592520641,
+      "grad_norm": 0.5963196754455566,
+      "learning_rate": 0.0001,
+      "loss": 1.8656,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3428848955803788,
+      "grad_norm": 0.5400733947753906,
+      "learning_rate": 0.0001,
+      "loss": 1.7551,
+      "step": 1412
+    },
+    {
+      "epoch": 0.34312773190869356,
+      "grad_norm": 0.5988975167274475,
+      "learning_rate": 0.0001,
+      "loss": 1.8325,
+      "step": 1413
+    },
+    {
+      "epoch": 0.34337056823700823,
+      "grad_norm": 0.5726484060287476,
+      "learning_rate": 0.0001,
+      "loss": 1.8216,
+      "step": 1414
+    },
+    {
+      "epoch": 0.34361340456532297,
+      "grad_norm": 0.5422387719154358,
+      "learning_rate": 0.0001,
+      "loss": 1.7552,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3438562408936377,
+      "grad_norm": 0.5612909197807312,
+      "learning_rate": 0.0001,
+      "loss": 1.7423,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3440990772219524,
+      "grad_norm": 0.5863240957260132,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3443419135502671,
+      "grad_norm": 0.5583364367485046,
+      "learning_rate": 0.0001,
+      "loss": 1.7598,
+      "step": 1418
+    },
+    {
+      "epoch": 0.34458474987858184,
+      "grad_norm": 0.5578181147575378,
+      "learning_rate": 0.0001,
+      "loss": 1.7413,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 0.5600872039794922,
+      "learning_rate": 0.0001,
+      "loss": 1.7814,
+      "step": 1420
+    },
+    {
+      "epoch": 0.34507042253521125,
+      "grad_norm": 0.5490904450416565,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 1421
+    },
+    {
+      "epoch": 0.345313258863526,
+      "grad_norm": 0.613351583480835,
+      "learning_rate": 0.0001,
+      "loss": 1.787,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3455560951918407,
+      "grad_norm": 0.5493449568748474,
+      "learning_rate": 0.0001,
+      "loss": 1.8616,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3457989315201554,
+      "grad_norm": 0.6109333038330078,
+      "learning_rate": 0.0001,
+      "loss": 1.7587,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3460417678484701,
+      "grad_norm": 0.5629788041114807,
+      "learning_rate": 0.0001,
+      "loss": 1.8761,
+      "step": 1425
+    },
+    {
+      "epoch": 0.34628460417678486,
+      "grad_norm": 0.5619919300079346,
+      "learning_rate": 0.0001,
+      "loss": 1.7521,
+      "step": 1426
+    },
+    {
+      "epoch": 0.34652744050509954,
+      "grad_norm": 0.5841615200042725,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 1427
+    },
+    {
+      "epoch": 0.34677027683341427,
+      "grad_norm": 0.5627974271774292,
+      "learning_rate": 0.0001,
+      "loss": 1.7341,
+      "step": 1428
+    },
+    {
+      "epoch": 0.347013113161729,
+      "grad_norm": 0.5274537801742554,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 1429
+    },
+    {
+      "epoch": 0.34725594949004374,
+      "grad_norm": 0.5390434861183167,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3474987858183584,
+      "grad_norm": 0.5722873210906982,
+      "learning_rate": 0.0001,
+      "loss": 1.7464,
+      "step": 1431
+    },
+    {
+      "epoch": 0.34774162214667315,
+      "grad_norm": 0.5416739583015442,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3479844584749879,
+      "grad_norm": 0.5875145196914673,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 1433
+    },
+    {
+      "epoch": 0.34822729480330256,
+      "grad_norm": 0.563105583190918,
+      "learning_rate": 0.0001,
+      "loss": 1.747,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3484701311316173,
+      "grad_norm": 0.5475865006446838,
+      "learning_rate": 0.0001,
+      "loss": 1.5961,
+      "step": 1435
+    },
+    {
+      "epoch": 0.348712967459932,
+      "grad_norm": 0.5591480135917664,
+      "learning_rate": 0.0001,
+      "loss": 1.7393,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3489558037882467,
+      "grad_norm": 0.6027525663375854,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 1437
+    },
+    {
+      "epoch": 0.34919864011656143,
+      "grad_norm": 0.5859847068786621,
+      "learning_rate": 0.0001,
+      "loss": 1.7547,
+      "step": 1438
+    },
+    {
+      "epoch": 0.34944147644487616,
+      "grad_norm": 0.6301659941673279,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 1439
+    },
+    {
+      "epoch": 0.34968431277319084,
+      "grad_norm": 0.5705239772796631,
+      "learning_rate": 0.0001,
+      "loss": 1.7873,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3499271491015056,
+      "grad_norm": 0.5834831595420837,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3501699854298203,
+      "grad_norm": 0.6054890751838684,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 1442
+    },
+    {
+      "epoch": 0.35041282175813504,
+      "grad_norm": 0.5621503591537476,
+      "learning_rate": 0.0001,
+      "loss": 1.747,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3506556580864497,
+      "grad_norm": 0.5485848784446716,
+      "learning_rate": 0.0001,
+      "loss": 1.7866,
+      "step": 1444
+    },
+    {
+      "epoch": 0.35089849441476445,
+      "grad_norm": 0.571904718875885,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3511413307430792,
+      "grad_norm": 0.5643053650856018,
+      "learning_rate": 0.0001,
+      "loss": 1.8232,
+      "step": 1446
+    },
+    {
+      "epoch": 0.35138416707139386,
+      "grad_norm": 0.5391576290130615,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3516270033997086,
+      "grad_norm": 0.5638372898101807,
+      "learning_rate": 0.0001,
+      "loss": 1.7338,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3518698397280233,
+      "grad_norm": 0.5746908187866211,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 1449
+    },
+    {
+      "epoch": 0.352112676056338,
+      "grad_norm": 0.5585685968399048,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 1450
+    },
+    {
+      "epoch": 0.35235551238465274,
+      "grad_norm": 0.5530781149864197,
+      "learning_rate": 0.0001,
+      "loss": 1.6715,
+      "step": 1451
+    },
+    {
+      "epoch": 0.35259834871296747,
+      "grad_norm": 0.5826124548912048,
+      "learning_rate": 0.0001,
+      "loss": 1.7969,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3528411850412822,
+      "grad_norm": 0.553278386592865,
+      "learning_rate": 0.0001,
+      "loss": 1.7855,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3530840213695969,
+      "grad_norm": 0.5775020718574524,
+      "learning_rate": 0.0001,
+      "loss": 1.7579,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3533268576979116,
+      "grad_norm": 0.581123948097229,
+      "learning_rate": 0.0001,
+      "loss": 1.8847,
+      "step": 1455
+    },
+    {
+      "epoch": 0.35356969402622634,
+      "grad_norm": 0.5525007247924805,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 1456
+    },
+    {
+      "epoch": 0.353812530354541,
+      "grad_norm": 0.5545541048049927,
+      "learning_rate": 0.0001,
+      "loss": 1.7563,
+      "step": 1457
+    },
+    {
+      "epoch": 0.35405536668285575,
+      "grad_norm": 0.5880481004714966,
+      "learning_rate": 0.0001,
+      "loss": 1.8606,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3542982030111705,
+      "grad_norm": 0.5368393063545227,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 1459
+    },
+    {
+      "epoch": 0.35454103933948516,
+      "grad_norm": 0.541566014289856,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3547838756677999,
+      "grad_norm": 0.5559922456741333,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 1461
+    },
+    {
+      "epoch": 0.35502671199611463,
+      "grad_norm": 0.5847586393356323,
+      "learning_rate": 0.0001,
+      "loss": 1.6755,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3552695483244293,
+      "grad_norm": 0.5786769390106201,
+      "learning_rate": 0.0001,
+      "loss": 1.8127,
+      "step": 1463
+    },
+    {
+      "epoch": 0.35551238465274404,
+      "grad_norm": 0.5552936792373657,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3557552209810588,
+      "grad_norm": 0.570335865020752,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3559980573093735,
+      "grad_norm": 0.5759338736534119,
+      "learning_rate": 0.0001,
+      "loss": 1.7253,
+      "step": 1466
+    },
+    {
+      "epoch": 0.3562408936376882,
+      "grad_norm": 0.5608747601509094,
+      "learning_rate": 0.0001,
+      "loss": 1.7129,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3564837299660029,
+      "grad_norm": 0.6293925046920776,
+      "learning_rate": 0.0001,
+      "loss": 1.8531,
+      "step": 1468
+    },
+    {
+      "epoch": 0.35672656629431765,
+      "grad_norm": 0.5296218991279602,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3569694026226323,
+      "grad_norm": 0.5386657118797302,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 1470
+    },
+    {
+      "epoch": 0.35721223895094706,
+      "grad_norm": 0.5375521779060364,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3574550752792618,
+      "grad_norm": 0.5659185647964478,
+      "learning_rate": 0.0001,
+      "loss": 1.8617,
+      "step": 1472
+    },
+    {
+      "epoch": 0.35769791160757647,
+      "grad_norm": 0.589073121547699,
+      "learning_rate": 0.0001,
+      "loss": 1.9014,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3579407479358912,
+      "grad_norm": 0.5113967061042786,
+      "learning_rate": 0.0001,
+      "loss": 1.6166,
+      "step": 1474
+    },
+    {
+      "epoch": 0.35818358426420593,
+      "grad_norm": 0.5525660514831543,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 1475
+    },
+    {
+      "epoch": 0.35842642059252067,
+      "grad_norm": 0.5673325657844543,
+      "learning_rate": 0.0001,
+      "loss": 1.733,
+      "step": 1476
+    },
+    {
+      "epoch": 0.35866925692083534,
+      "grad_norm": 0.5445988178253174,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3589120932491501,
+      "grad_norm": 0.5666174292564392,
+      "learning_rate": 0.0001,
+      "loss": 1.7183,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3591549295774648,
+      "grad_norm": 0.5427615642547607,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3593977659057795,
+      "grad_norm": 0.5578546524047852,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3596406022340942,
+      "grad_norm": 0.5783149003982544,
+      "learning_rate": 0.0001,
+      "loss": 1.803,
+      "step": 1481
+    },
+    {
+      "epoch": 0.35988343856240895,
+      "grad_norm": 0.5721765756607056,
+      "learning_rate": 0.0001,
+      "loss": 1.8746,
+      "step": 1482
+    },
+    {
+      "epoch": 0.36012627489072363,
+      "grad_norm": 0.5949157476425171,
+      "learning_rate": 0.0001,
+      "loss": 1.8768,
+      "step": 1483
+    },
+    {
+      "epoch": 0.36036911121903836,
+      "grad_norm": 0.5486117601394653,
+      "learning_rate": 0.0001,
+      "loss": 1.7638,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3606119475473531,
+      "grad_norm": 0.5355470180511475,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3608547838756678,
+      "grad_norm": 0.5477654337882996,
+      "learning_rate": 0.0001,
+      "loss": 1.879,
+      "step": 1486
+    },
+    {
+      "epoch": 0.3610976202039825,
+      "grad_norm": 0.5449265241622925,
+      "learning_rate": 0.0001,
+      "loss": 1.7839,
+      "step": 1487
+    },
+    {
+      "epoch": 0.36134045653229724,
+      "grad_norm": 0.5875566601753235,
+      "learning_rate": 0.0001,
+      "loss": 1.8003,
+      "step": 1488
+    },
+    {
+      "epoch": 0.36158329286061197,
+      "grad_norm": 0.596920371055603,
+      "learning_rate": 0.0001,
+      "loss": 1.8118,
+      "step": 1489
+    },
+    {
+      "epoch": 0.36182612918892665,
+      "grad_norm": 0.5583705306053162,
+      "learning_rate": 0.0001,
+      "loss": 1.7894,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3620689655172414,
+      "grad_norm": 0.5918691754341125,
+      "learning_rate": 0.0001,
+      "loss": 1.8412,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3623118018455561,
+      "grad_norm": 0.5425227284431458,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3625546381738708,
+      "grad_norm": 0.56281578540802,
+      "learning_rate": 0.0001,
+      "loss": 1.6218,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3627974745021855,
+      "grad_norm": 0.5747039318084717,
+      "learning_rate": 0.0001,
+      "loss": 1.7746,
+      "step": 1494
+    },
+    {
+      "epoch": 0.36304031083050026,
+      "grad_norm": 0.539703905582428,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 1495
+    },
+    {
+      "epoch": 0.36328314715881493,
+      "grad_norm": 0.5434957146644592,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 1496
+    },
+    {
+      "epoch": 0.36352598348712967,
+      "grad_norm": 0.5798366069793701,
+      "learning_rate": 0.0001,
+      "loss": 1.7758,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3637688198154444,
+      "grad_norm": 0.5288923382759094,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 1498
+    },
+    {
+      "epoch": 0.36401165614375913,
+      "grad_norm": 0.5663448572158813,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3642544924720738,
+      "grad_norm": 0.5524390339851379,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 1500
+    },
+    {
+      "epoch": 0.36449732880038854,
+      "grad_norm": 0.6314302086830139,
+      "learning_rate": 0.0001,
+      "loss": 1.8825,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3647401651287033,
+      "grad_norm": 0.5598084330558777,
+      "learning_rate": 0.0001,
+      "loss": 1.7589,
+      "step": 1502
+    },
+    {
+      "epoch": 0.36498300145701795,
+      "grad_norm": 0.5555556416511536,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3652258377853327,
+      "grad_norm": 0.552541971206665,
+      "learning_rate": 0.0001,
+      "loss": 1.6995,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3654686741136474,
+      "grad_norm": 0.5386813879013062,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3657115104419621,
+      "grad_norm": 0.5780212879180908,
+      "learning_rate": 0.0001,
+      "loss": 1.7706,
+      "step": 1506
+    },
+    {
+      "epoch": 0.36595434677027683,
+      "grad_norm": 0.5641130805015564,
+      "learning_rate": 0.0001,
+      "loss": 1.7096,
+      "step": 1507
+    },
+    {
+      "epoch": 0.36619718309859156,
+      "grad_norm": 0.5723060965538025,
+      "learning_rate": 0.0001,
+      "loss": 1.8688,
+      "step": 1508
+    },
+    {
+      "epoch": 0.36644001942690624,
+      "grad_norm": 0.5916156768798828,
+      "learning_rate": 0.0001,
+      "loss": 1.6937,
+      "step": 1509
+    },
+    {
+      "epoch": 0.36668285575522097,
+      "grad_norm": 0.5428485870361328,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3669256920835357,
+      "grad_norm": 0.5606679916381836,
+      "learning_rate": 0.0001,
+      "loss": 1.8716,
+      "step": 1511
+    },
+    {
+      "epoch": 0.36716852841185044,
+      "grad_norm": 0.5421441793441772,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3674113647401651,
+      "grad_norm": 0.5790001153945923,
+      "learning_rate": 0.0001,
+      "loss": 1.7519,
+      "step": 1513
+    },
+    {
+      "epoch": 0.36765420106847985,
+      "grad_norm": 0.5827850103378296,
+      "learning_rate": 0.0001,
+      "loss": 1.8111,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3678970373967946,
+      "grad_norm": 0.5159997940063477,
+      "learning_rate": 0.0001,
+      "loss": 1.5164,
+      "step": 1515
+    },
+    {
+      "epoch": 0.36813987372510926,
+      "grad_norm": 0.5459283590316772,
+      "learning_rate": 0.0001,
+      "loss": 1.77,
+      "step": 1516
+    },
+    {
+      "epoch": 0.368382710053424,
+      "grad_norm": 0.5678669810295105,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3686255463817387,
+      "grad_norm": 0.6005099415779114,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3688683827100534,
+      "grad_norm": 0.553765058517456,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 1519
+    },
+    {
+      "epoch": 0.36911121903836813,
+      "grad_norm": 0.5875484347343445,
+      "learning_rate": 0.0001,
+      "loss": 1.8644,
+      "step": 1520
+    },
+    {
+      "epoch": 0.36935405536668287,
+      "grad_norm": 0.5853337049484253,
+      "learning_rate": 0.0001,
+      "loss": 1.836,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3695968916949976,
+      "grad_norm": 0.5599927306175232,
+      "learning_rate": 0.0001,
+      "loss": 1.773,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3698397280233123,
+      "grad_norm": 0.5492868423461914,
+      "learning_rate": 0.0001,
+      "loss": 1.538,
+      "step": 1523
+    },
+    {
+      "epoch": 0.370082564351627,
+      "grad_norm": 0.5999908447265625,
+      "learning_rate": 0.0001,
+      "loss": 1.773,
+      "step": 1524
+    },
+    {
+      "epoch": 0.37032540067994174,
+      "grad_norm": 0.5405163168907166,
+      "learning_rate": 0.0001,
+      "loss": 1.7824,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3705682370082564,
+      "grad_norm": 0.5858007669448853,
+      "learning_rate": 0.0001,
+      "loss": 1.7849,
+      "step": 1526
+    },
+    {
+      "epoch": 0.37081107333657115,
+      "grad_norm": 0.566652238368988,
+      "learning_rate": 0.0001,
+      "loss": 1.8518,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3710539096648859,
+      "grad_norm": 0.5848403573036194,
+      "learning_rate": 0.0001,
+      "loss": 1.7684,
+      "step": 1528
+    },
+    {
+      "epoch": 0.37129674599320056,
+      "grad_norm": 0.533105194568634,
+      "learning_rate": 0.0001,
+      "loss": 1.6261,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3715395823215153,
+      "grad_norm": 0.5535341501235962,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 1530
+    },
+    {
+      "epoch": 0.37178241864983,
+      "grad_norm": 0.5954918265342712,
+      "learning_rate": 0.0001,
+      "loss": 1.8216,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3720252549781447,
+      "grad_norm": 0.5533576607704163,
+      "learning_rate": 0.0001,
+      "loss": 1.753,
+      "step": 1532
+    },
+    {
+      "epoch": 0.37226809130645944,
+      "grad_norm": 0.5351048707962036,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 1533
+    },
+    {
+      "epoch": 0.37251092763477417,
+      "grad_norm": 0.5829824209213257,
+      "learning_rate": 0.0001,
+      "loss": 1.7375,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3727537639630889,
+      "grad_norm": 0.6141155362129211,
+      "learning_rate": 0.0001,
+      "loss": 1.753,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3729966002914036,
+      "grad_norm": 0.5319721698760986,
+      "learning_rate": 0.0001,
+      "loss": 1.6866,
+      "step": 1536
+    },
+    {
+      "epoch": 0.3732394366197183,
+      "grad_norm": 0.5934935808181763,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 1537
+    },
+    {
+      "epoch": 0.37348227294803304,
+      "grad_norm": 0.5703699588775635,
+      "learning_rate": 0.0001,
+      "loss": 1.6895,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3737251092763477,
+      "grad_norm": 0.5700774192810059,
+      "learning_rate": 0.0001,
+      "loss": 1.6917,
+      "step": 1539
+    },
+    {
+      "epoch": 0.37396794560466246,
+      "grad_norm": 0.6150509119033813,
+      "learning_rate": 0.0001,
+      "loss": 1.8295,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3742107819329772,
+      "grad_norm": 0.570936381816864,
+      "learning_rate": 0.0001,
+      "loss": 1.8789,
+      "step": 1541
+    },
+    {
+      "epoch": 0.37445361826129187,
+      "grad_norm": 0.5791286826133728,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3746964545896066,
+      "grad_norm": 0.5663446187973022,
+      "learning_rate": 0.0001,
+      "loss": 1.7994,
+      "step": 1543
+    },
+    {
+      "epoch": 0.37493929091792133,
+      "grad_norm": 0.5743141770362854,
+      "learning_rate": 0.0001,
+      "loss": 1.7445,
+      "step": 1544
+    },
+    {
+      "epoch": 0.37518212724623606,
+      "grad_norm": 0.5314855575561523,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 1545
+    },
+    {
+      "epoch": 0.37542496357455074,
+      "grad_norm": 0.6015771627426147,
+      "learning_rate": 0.0001,
+      "loss": 1.7693,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3756677999028655,
+      "grad_norm": 0.5668714642524719,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3759106362311802,
+      "grad_norm": 0.5673467516899109,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3761534725594949,
+      "grad_norm": 0.5929018259048462,
+      "learning_rate": 0.0001,
+      "loss": 1.8452,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3763963088878096,
+      "grad_norm": 0.5479089021682739,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 1550
+    },
+    {
+      "epoch": 0.37663914521612435,
+      "grad_norm": 0.5798628926277161,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 1551
+    },
+    {
+      "epoch": 0.376881981544439,
+      "grad_norm": 0.5535653829574585,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 1552
+    },
+    {
+      "epoch": 0.37712481787275376,
+      "grad_norm": 0.540377140045166,
+      "learning_rate": 0.0001,
+      "loss": 1.6866,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3773676542010685,
+      "grad_norm": 0.5681852698326111,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 1554
+    },
+    {
+      "epoch": 0.37761049052938317,
+      "grad_norm": 0.5816860795021057,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3778533268576979,
+      "grad_norm": 0.5638381838798523,
+      "learning_rate": 0.0001,
+      "loss": 1.8003,
+      "step": 1556
+    },
+    {
+      "epoch": 0.37809616318601263,
+      "grad_norm": 0.5493277907371521,
+      "learning_rate": 0.0001,
+      "loss": 1.7024,
+      "step": 1557
+    },
+    {
+      "epoch": 0.37833899951432737,
+      "grad_norm": 0.6079229712486267,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 1558
+    },
+    {
+      "epoch": 0.37858183584264204,
+      "grad_norm": 0.5536306500434875,
+      "learning_rate": 0.0001,
+      "loss": 1.7327,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3788246721709568,
+      "grad_norm": 0.6091067790985107,
+      "learning_rate": 0.0001,
+      "loss": 1.7779,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3790675084992715,
+      "grad_norm": 0.5523210167884827,
+      "learning_rate": 0.0001,
+      "loss": 1.7955,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3793103448275862,
+      "grad_norm": 0.5820581316947937,
+      "learning_rate": 0.0001,
+      "loss": 1.8316,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3795531811559009,
+      "grad_norm": 0.5861195921897888,
+      "learning_rate": 0.0001,
+      "loss": 1.7992,
+      "step": 1563
+    },
+    {
+      "epoch": 0.37979601748421565,
+      "grad_norm": 0.6205547451972961,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 1564
+    },
+    {
+      "epoch": 0.38003885381253033,
+      "grad_norm": 0.5986005663871765,
+      "learning_rate": 0.0001,
+      "loss": 1.722,
+      "step": 1565
+    },
+    {
+      "epoch": 0.38028169014084506,
+      "grad_norm": 0.5632820129394531,
+      "learning_rate": 0.0001,
+      "loss": 1.8611,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3805245264691598,
+      "grad_norm": 0.5678552985191345,
+      "learning_rate": 0.0001,
+      "loss": 1.7909,
+      "step": 1567
+    },
+    {
+      "epoch": 0.38076736279747453,
+      "grad_norm": 0.5708531141281128,
+      "learning_rate": 0.0001,
+      "loss": 1.8951,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3810101991257892,
+      "grad_norm": 0.5985879302024841,
+      "learning_rate": 0.0001,
+      "loss": 1.7499,
+      "step": 1569
+    },
+    {
+      "epoch": 0.38125303545410394,
+      "grad_norm": 0.546149492263794,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 1570
+    },
+    {
+      "epoch": 0.38149587178241867,
+      "grad_norm": 0.5527278780937195,
+      "learning_rate": 0.0001,
+      "loss": 1.8572,
+      "step": 1571
+    },
+    {
+      "epoch": 0.38173870811073335,
+      "grad_norm": 0.5659234523773193,
+      "learning_rate": 0.0001,
+      "loss": 1.752,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3819815444390481,
+      "grad_norm": 0.5430263876914978,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3822243807673628,
+      "grad_norm": 0.5225880146026611,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3824672170956775,
+      "grad_norm": 0.5456332564353943,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3827100534239922,
+      "grad_norm": 0.5807480216026306,
+      "learning_rate": 0.0001,
+      "loss": 1.8378,
+      "step": 1576
+    },
+    {
+      "epoch": 0.38295288975230696,
+      "grad_norm": 0.5756118297576904,
+      "learning_rate": 0.0001,
+      "loss": 1.8539,
+      "step": 1577
+    },
+    {
+      "epoch": 0.38319572608062163,
+      "grad_norm": 0.6021025776863098,
+      "learning_rate": 0.0001,
+      "loss": 1.728,
+      "step": 1578
+    },
+    {
+      "epoch": 0.38343856240893637,
+      "grad_norm": 0.5744970440864563,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3836813987372511,
+      "grad_norm": 0.5658952593803406,
+      "learning_rate": 0.0001,
+      "loss": 1.7816,
+      "step": 1580
+    },
+    {
+      "epoch": 0.38392423506556583,
+      "grad_norm": 0.5663431882858276,
+      "learning_rate": 0.0001,
+      "loss": 1.7463,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3841670713938805,
+      "grad_norm": 0.5760250091552734,
+      "learning_rate": 0.0001,
+      "loss": 1.879,
+      "step": 1582
+    },
+    {
+      "epoch": 0.38440990772219524,
+      "grad_norm": 0.5781668424606323,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 1583
+    },
+    {
+      "epoch": 0.38465274405051,
+      "grad_norm": 0.5917073488235474,
+      "learning_rate": 0.0001,
+      "loss": 1.758,
+      "step": 1584
+    },
+    {
+      "epoch": 0.38489558037882465,
+      "grad_norm": 0.6140726208686829,
+      "learning_rate": 0.0001,
+      "loss": 1.8021,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3851384167071394,
+      "grad_norm": 0.5443862080574036,
+      "learning_rate": 0.0001,
+      "loss": 1.7591,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3853812530354541,
+      "grad_norm": 0.5863355398178101,
+      "learning_rate": 0.0001,
+      "loss": 1.6797,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3856240893637688,
+      "grad_norm": 0.5166279673576355,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 1588
+    },
+    {
+      "epoch": 0.38586692569208353,
+      "grad_norm": 0.5556029081344604,
+      "learning_rate": 0.0001,
+      "loss": 1.8708,
+      "step": 1589
+    },
+    {
+      "epoch": 0.38610976202039826,
+      "grad_norm": 0.5802101492881775,
+      "learning_rate": 0.0001,
+      "loss": 1.8732,
+      "step": 1590
+    },
+    {
+      "epoch": 0.386352598348713,
+      "grad_norm": 0.514442503452301,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 1591
+    },
+    {
+      "epoch": 0.38659543467702767,
+      "grad_norm": 0.5633143782615662,
+      "learning_rate": 0.0001,
+      "loss": 1.7854,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3868382710053424,
+      "grad_norm": 0.5631378889083862,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 1593
+    },
+    {
+      "epoch": 0.38708110733365714,
+      "grad_norm": 0.5619434714317322,
+      "learning_rate": 0.0001,
+      "loss": 1.7912,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3873239436619718,
+      "grad_norm": 0.5826120972633362,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 1595
+    },
+    {
+      "epoch": 0.38756677999028655,
+      "grad_norm": 0.5944365859031677,
+      "learning_rate": 0.0001,
+      "loss": 1.7426,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3878096163186013,
+      "grad_norm": 0.5516484379768372,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 1597
+    },
+    {
+      "epoch": 0.38805245264691596,
+      "grad_norm": 0.6092392802238464,
+      "learning_rate": 0.0001,
+      "loss": 1.8984,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3882952889752307,
+      "grad_norm": 0.5654324889183044,
+      "learning_rate": 0.0001,
+      "loss": 1.7587,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3885381253035454,
+      "grad_norm": 0.5673196315765381,
+      "learning_rate": 0.0001,
+      "loss": 1.6392,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3887809616318601,
+      "grad_norm": 0.5889737606048584,
+      "learning_rate": 0.0001,
+      "loss": 1.8059,
+      "step": 1601
+    },
+    {
+      "epoch": 0.38902379796017483,
+      "grad_norm": 0.5544754862785339,
+      "learning_rate": 0.0001,
+      "loss": 1.7381,
+      "step": 1602
+    },
+    {
+      "epoch": 0.38926663428848957,
+      "grad_norm": 0.5846825838088989,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3895094706168043,
+      "grad_norm": 0.5588352680206299,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 1604
+    },
+    {
+      "epoch": 0.389752306945119,
+      "grad_norm": 0.5897813439369202,
+      "learning_rate": 0.0001,
+      "loss": 1.7817,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3899951432734337,
+      "grad_norm": 0.5556011199951172,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 1606
+    },
+    {
+      "epoch": 0.39023797960174844,
+      "grad_norm": 0.5782007575035095,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3904808159300631,
+      "grad_norm": 0.5817544460296631,
+      "learning_rate": 0.0001,
+      "loss": 1.7412,
+      "step": 1608
+    },
+    {
+      "epoch": 0.39072365225837785,
+      "grad_norm": 0.5802628397941589,
+      "learning_rate": 0.0001,
+      "loss": 1.898,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3909664885866926,
+      "grad_norm": 0.5531176924705505,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 1610
+    },
+    {
+      "epoch": 0.39120932491500726,
+      "grad_norm": 0.5454622507095337,
+      "learning_rate": 0.0001,
+      "loss": 1.7513,
+      "step": 1611
+    },
+    {
+      "epoch": 0.391452161243322,
+      "grad_norm": 0.5817705392837524,
+      "learning_rate": 0.0001,
+      "loss": 1.6988,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3916949975716367,
+      "grad_norm": 0.5824382901191711,
+      "learning_rate": 0.0001,
+      "loss": 1.7906,
+      "step": 1613
+    },
+    {
+      "epoch": 0.39193783389995146,
+      "grad_norm": 0.5714345574378967,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 1614
+    },
+    {
+      "epoch": 0.39218067022826614,
+      "grad_norm": 0.6089634895324707,
+      "learning_rate": 0.0001,
+      "loss": 1.8484,
+      "step": 1615
+    },
+    {
+      "epoch": 0.39242350655658087,
+      "grad_norm": 0.5468596816062927,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3926663428848956,
+      "grad_norm": 0.5312457084655762,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3929091792132103,
+      "grad_norm": 0.5698146224021912,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 1618
+    },
+    {
+      "epoch": 0.393152015541525,
+      "grad_norm": 0.5599279403686523,
+      "learning_rate": 0.0001,
+      "loss": 1.8704,
+      "step": 1619
+    },
+    {
+      "epoch": 0.39339485186983975,
+      "grad_norm": 0.5623302459716797,
+      "learning_rate": 0.0001,
+      "loss": 1.7474,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3936376881981544,
+      "grad_norm": 0.6181963682174683,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 1621
+    },
+    {
+      "epoch": 0.39388052452646916,
+      "grad_norm": 0.5787026286125183,
+      "learning_rate": 0.0001,
+      "loss": 1.7176,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3941233608547839,
+      "grad_norm": 0.6048277616500854,
+      "learning_rate": 0.0001,
+      "loss": 1.7601,
+      "step": 1623
+    },
+    {
+      "epoch": 0.39436619718309857,
+      "grad_norm": 0.576568067073822,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3946090335114133,
+      "grad_norm": 0.6017733812332153,
+      "learning_rate": 0.0001,
+      "loss": 1.9806,
+      "step": 1625
+    },
+    {
+      "epoch": 0.39485186983972803,
+      "grad_norm": 0.5605730414390564,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 1626
+    },
+    {
+      "epoch": 0.39509470616804276,
+      "grad_norm": 0.5670095086097717,
+      "learning_rate": 0.0001,
+      "loss": 1.7742,
+      "step": 1627
+    },
+    {
+      "epoch": 0.39533754249635744,
+      "grad_norm": 0.5428698658943176,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3955803788246722,
+      "grad_norm": 0.5755293369293213,
+      "learning_rate": 0.0001,
+      "loss": 1.6695,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3958232151529869,
+      "grad_norm": 0.562969982624054,
+      "learning_rate": 0.0001,
+      "loss": 1.7664,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3960660514813016,
+      "grad_norm": 0.5230188965797424,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3963088878096163,
+      "grad_norm": 0.553209125995636,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 1632
+    },
+    {
+      "epoch": 0.39655172413793105,
+      "grad_norm": 0.5449209213256836,
+      "learning_rate": 0.0001,
+      "loss": 1.744,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3967945604662457,
+      "grad_norm": 0.5767896175384521,
+      "learning_rate": 0.0001,
+      "loss": 1.7867,
+      "step": 1634
+    },
+    {
+      "epoch": 0.39703739679456046,
+      "grad_norm": 0.6298505663871765,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3972802331228752,
+      "grad_norm": 0.6241861581802368,
+      "learning_rate": 0.0001,
+      "loss": 1.8101,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3975230694511899,
+      "grad_norm": 0.577860951423645,
+      "learning_rate": 0.0001,
+      "loss": 1.9489,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3977659057795046,
+      "grad_norm": 0.5398596525192261,
+      "learning_rate": 0.0001,
+      "loss": 1.7749,
+      "step": 1638
+    },
+    {
+      "epoch": 0.39800874210781934,
+      "grad_norm": 0.5384426712989807,
+      "learning_rate": 0.0001,
+      "loss": 1.7272,
+      "step": 1639
+    },
+    {
+      "epoch": 0.39825157843613407,
+      "grad_norm": 0.5840216279029846,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 1640
+    },
+    {
+      "epoch": 0.39849441476444875,
+      "grad_norm": 0.5602217316627502,
+      "learning_rate": 0.0001,
+      "loss": 1.7066,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3987372510927635,
+      "grad_norm": 0.551922082901001,
+      "learning_rate": 0.0001,
+      "loss": 1.7214,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3989800874210782,
+      "grad_norm": 0.570489764213562,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3992229237493929,
+      "grad_norm": 0.5829172134399414,
+      "learning_rate": 0.0001,
+      "loss": 1.6867,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3994657600777076,
+      "grad_norm": 0.5585073828697205,
+      "learning_rate": 0.0001,
+      "loss": 1.8045,
+      "step": 1645
+    },
+    {
+      "epoch": 0.39970859640602235,
+      "grad_norm": 0.564818799495697,
+      "learning_rate": 0.0001,
+      "loss": 1.7552,
+      "step": 1646
+    },
+    {
+      "epoch": 0.39995143273433703,
+      "grad_norm": 0.575394868850708,
+      "learning_rate": 0.0001,
+      "loss": 1.7556,
+      "step": 1647
+    },
+    {
+      "epoch": 0.40019426906265176,
+      "grad_norm": 0.569500744342804,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 1648
+    },
+    {
+      "epoch": 0.4004371053909665,
+      "grad_norm": 0.5607755780220032,
+      "learning_rate": 0.0001,
+      "loss": 1.9207,
+      "step": 1649
+    },
+    {
+      "epoch": 0.40067994171928123,
+      "grad_norm": 0.6055318117141724,
+      "learning_rate": 0.0001,
+      "loss": 1.7892,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4009227780475959,
+      "grad_norm": 0.5560562014579773,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 1651
+    },
+    {
+      "epoch": 0.40116561437591064,
+      "grad_norm": 0.5481667518615723,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4014084507042254,
+      "grad_norm": 0.5663360953330994,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 1653
+    },
+    {
+      "epoch": 0.40165128703254005,
+      "grad_norm": 0.5803428292274475,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4018941233608548,
+      "grad_norm": 0.5931940078735352,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4021369596891695,
+      "grad_norm": 0.5731521844863892,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4023797960174842,
+      "grad_norm": 0.573701798915863,
+      "learning_rate": 0.0001,
+      "loss": 1.7434,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4026226323457989,
+      "grad_norm": 0.5959265232086182,
+      "learning_rate": 0.0001,
+      "loss": 1.9286,
+      "step": 1658
+    },
+    {
+      "epoch": 0.40286546867411366,
+      "grad_norm": 0.5944108963012695,
+      "learning_rate": 0.0001,
+      "loss": 1.8003,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4031083050024284,
+      "grad_norm": 0.5664130449295044,
+      "learning_rate": 0.0001,
+      "loss": 1.8043,
+      "step": 1660
+    },
+    {
+      "epoch": 0.40335114133074307,
+      "grad_norm": 0.6348901391029358,
+      "learning_rate": 0.0001,
+      "loss": 1.6961,
+      "step": 1661
+    },
+    {
+      "epoch": 0.4035939776590578,
+      "grad_norm": 0.5614269375801086,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 1662
+    },
+    {
+      "epoch": 0.40383681398737253,
+      "grad_norm": 0.5885881185531616,
+      "learning_rate": 0.0001,
+      "loss": 1.8601,
+      "step": 1663
+    },
+    {
+      "epoch": 0.4040796503156872,
+      "grad_norm": 0.5628127455711365,
+      "learning_rate": 0.0001,
+      "loss": 1.8385,
+      "step": 1664
+    },
+    {
+      "epoch": 0.40432248664400194,
+      "grad_norm": 0.5846302509307861,
+      "learning_rate": 0.0001,
+      "loss": 1.7258,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4045653229723167,
+      "grad_norm": 0.5827339291572571,
+      "learning_rate": 0.0001,
+      "loss": 1.8895,
+      "step": 1666
+    },
+    {
+      "epoch": 0.40480815930063135,
+      "grad_norm": 0.566876232624054,
+      "learning_rate": 0.0001,
+      "loss": 1.7546,
+      "step": 1667
+    },
+    {
+      "epoch": 0.4050509956289461,
+      "grad_norm": 0.6059743165969849,
+      "learning_rate": 0.0001,
+      "loss": 1.7,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4052938319572608,
+      "grad_norm": 0.5789825320243835,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4055366682855755,
+      "grad_norm": 0.5450209975242615,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 1670
+    },
+    {
+      "epoch": 0.40577950461389023,
+      "grad_norm": 0.5666068196296692,
+      "learning_rate": 0.0001,
+      "loss": 1.8344,
+      "step": 1671
+    },
+    {
+      "epoch": 0.40602234094220496,
+      "grad_norm": 0.5817674398422241,
+      "learning_rate": 0.0001,
+      "loss": 1.8103,
+      "step": 1672
+    },
+    {
+      "epoch": 0.4062651772705197,
+      "grad_norm": 0.5866352915763855,
+      "learning_rate": 0.0001,
+      "loss": 1.8503,
+      "step": 1673
+    },
+    {
+      "epoch": 0.4065080135988344,
+      "grad_norm": 0.5393112897872925,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4067508499271491,
+      "grad_norm": 0.5831518769264221,
+      "learning_rate": 0.0001,
+      "loss": 1.8358,
+      "step": 1675
+    },
+    {
+      "epoch": 0.40699368625546384,
+      "grad_norm": 0.5524880886077881,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 1676
+    },
+    {
+      "epoch": 0.4072365225837785,
+      "grad_norm": 0.5588178634643555,
+      "learning_rate": 0.0001,
+      "loss": 1.7048,
+      "step": 1677
+    },
+    {
+      "epoch": 0.40747935891209325,
+      "grad_norm": 0.5812271237373352,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 1678
+    },
+    {
+      "epoch": 0.407722195240408,
+      "grad_norm": 0.5578488707542419,
+      "learning_rate": 0.0001,
+      "loss": 1.7485,
+      "step": 1679
+    },
+    {
+      "epoch": 0.40796503156872266,
+      "grad_norm": 0.573398768901825,
+      "learning_rate": 0.0001,
+      "loss": 1.8274,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4082078678970374,
+      "grad_norm": 0.6085888147354126,
+      "learning_rate": 0.0001,
+      "loss": 1.939,
+      "step": 1681
+    },
+    {
+      "epoch": 0.4084507042253521,
+      "grad_norm": 0.5662161111831665,
+      "learning_rate": 0.0001,
+      "loss": 1.7705,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4086935405536668,
+      "grad_norm": 0.5128927230834961,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 1683
+    },
+    {
+      "epoch": 0.40893637688198153,
+      "grad_norm": 0.55360347032547,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 1684
+    },
+    {
+      "epoch": 0.40917921321029627,
+      "grad_norm": 0.544804036617279,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 1685
+    },
+    {
+      "epoch": 0.409422049538611,
+      "grad_norm": 0.5657835006713867,
+      "learning_rate": 0.0001,
+      "loss": 1.7755,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4096648858669257,
+      "grad_norm": 0.5455095767974854,
+      "learning_rate": 0.0001,
+      "loss": 1.6197,
+      "step": 1687
+    },
+    {
+      "epoch": 0.4099077221952404,
+      "grad_norm": 0.6149754524230957,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 1688
+    },
+    {
+      "epoch": 0.41015055852355514,
+      "grad_norm": 0.582294762134552,
+      "learning_rate": 0.0001,
+      "loss": 1.8127,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4103933948518698,
+      "grad_norm": 0.5286328792572021,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 1690
+    },
+    {
+      "epoch": 0.41063623118018455,
+      "grad_norm": 0.5528634786605835,
+      "learning_rate": 0.0001,
+      "loss": 1.7693,
+      "step": 1691
+    },
+    {
+      "epoch": 0.4108790675084993,
+      "grad_norm": 0.5821068286895752,
+      "learning_rate": 0.0001,
+      "loss": 1.8451,
+      "step": 1692
+    },
+    {
+      "epoch": 0.41112190383681396,
+      "grad_norm": 0.5589388012886047,
+      "learning_rate": 0.0001,
+      "loss": 1.8007,
+      "step": 1693
+    },
+    {
+      "epoch": 0.4113647401651287,
+      "grad_norm": 0.5556886792182922,
+      "learning_rate": 0.0001,
+      "loss": 1.7214,
+      "step": 1694
+    },
+    {
+      "epoch": 0.41160757649344343,
+      "grad_norm": 0.5554993748664856,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 1695
+    },
+    {
+      "epoch": 0.41185041282175816,
+      "grad_norm": 0.5385516881942749,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 1696
+    },
+    {
+      "epoch": 0.41209324915007284,
+      "grad_norm": 0.5801841020584106,
+      "learning_rate": 0.0001,
+      "loss": 1.9049,
+      "step": 1697
+    },
+    {
+      "epoch": 0.41233608547838757,
+      "grad_norm": 0.5936272144317627,
+      "learning_rate": 0.0001,
+      "loss": 1.7948,
+      "step": 1698
+    },
+    {
+      "epoch": 0.4125789218067023,
+      "grad_norm": 0.5968410968780518,
+      "learning_rate": 0.0001,
+      "loss": 1.932,
+      "step": 1699
+    },
+    {
+      "epoch": 0.412821758135017,
+      "grad_norm": 0.6293395161628723,
+      "learning_rate": 0.0001,
+      "loss": 1.8595,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4130645944633317,
+      "grad_norm": 0.5809662342071533,
+      "learning_rate": 0.0001,
+      "loss": 1.8307,
+      "step": 1701
+    },
+    {
+      "epoch": 0.41330743079164645,
+      "grad_norm": 0.5780881643295288,
+      "learning_rate": 0.0001,
+      "loss": 1.7059,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4135502671199611,
+      "grad_norm": 0.5809916853904724,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 1703
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 0.5669182538986206,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4140359397765906,
+      "grad_norm": 0.5528964996337891,
+      "learning_rate": 0.0001,
+      "loss": 1.6834,
+      "step": 1705
+    },
+    {
+      "epoch": 0.41427877610490527,
+      "grad_norm": 0.5690479278564453,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 1706
+    },
+    {
+      "epoch": 0.41452161243322,
+      "grad_norm": 0.5857531428337097,
+      "learning_rate": 0.0001,
+      "loss": 1.8466,
+      "step": 1707
+    },
+    {
+      "epoch": 0.41476444876153473,
+      "grad_norm": 0.5584151744842529,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 1708
+    },
+    {
+      "epoch": 0.41500728508984946,
+      "grad_norm": 0.5964641571044922,
+      "learning_rate": 0.0001,
+      "loss": 1.8333,
+      "step": 1709
+    },
+    {
+      "epoch": 0.41525012141816414,
+      "grad_norm": 0.6007984280586243,
+      "learning_rate": 0.0001,
+      "loss": 1.7439,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4154929577464789,
+      "grad_norm": 0.574708104133606,
+      "learning_rate": 0.0001,
+      "loss": 1.6967,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4157357940747936,
+      "grad_norm": 0.5792802572250366,
+      "learning_rate": 0.0001,
+      "loss": 1.7642,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4159786304031083,
+      "grad_norm": 0.610641360282898,
+      "learning_rate": 0.0001,
+      "loss": 1.9078,
+      "step": 1713
+    },
+    {
+      "epoch": 0.416221466731423,
+      "grad_norm": 0.5648356676101685,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 1714
+    },
+    {
+      "epoch": 0.41646430305973775,
+      "grad_norm": 0.5931963324546814,
+      "learning_rate": 0.0001,
+      "loss": 1.9544,
+      "step": 1715
+    },
+    {
+      "epoch": 0.4167071393880524,
+      "grad_norm": 0.5581526160240173,
+      "learning_rate": 0.0001,
+      "loss": 1.8014,
+      "step": 1716
+    },
+    {
+      "epoch": 0.41694997571636716,
+      "grad_norm": 0.6139750480651855,
+      "learning_rate": 0.0001,
+      "loss": 1.8504,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4171928120446819,
+      "grad_norm": 0.5767641663551331,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4174356483729966,
+      "grad_norm": 0.5608420372009277,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4176784847013113,
+      "grad_norm": 0.5488410592079163,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 1720
+    },
+    {
+      "epoch": 0.41792132102962604,
+      "grad_norm": 0.5951592922210693,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 1721
+    },
+    {
+      "epoch": 0.41816415735794077,
+      "grad_norm": 0.5684256553649902,
+      "learning_rate": 0.0001,
+      "loss": 1.6858,
+      "step": 1722
+    },
+    {
+      "epoch": 0.41840699368625545,
+      "grad_norm": 0.5633500814437866,
+      "learning_rate": 0.0001,
+      "loss": 1.7456,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4186498300145702,
+      "grad_norm": 0.5915912985801697,
+      "learning_rate": 0.0001,
+      "loss": 1.8124,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4188926663428849,
+      "grad_norm": 0.5751018524169922,
+      "learning_rate": 0.0001,
+      "loss": 1.7569,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4191355026711996,
+      "grad_norm": 0.539318859577179,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4193783389995143,
+      "grad_norm": 0.5930345058441162,
+      "learning_rate": 0.0001,
+      "loss": 1.8064,
+      "step": 1727
+    },
+    {
+      "epoch": 0.41962117532782905,
+      "grad_norm": 0.5691484808921814,
+      "learning_rate": 0.0001,
+      "loss": 1.8677,
+      "step": 1728
+    },
+    {
+      "epoch": 0.41986401165614373,
+      "grad_norm": 0.5631294250488281,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 1729
+    },
+    {
+      "epoch": 0.42010684798445846,
+      "grad_norm": 0.5319895148277283,
+      "learning_rate": 0.0001,
+      "loss": 1.6579,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4203496843127732,
+      "grad_norm": 0.5654844641685486,
+      "learning_rate": 0.0001,
+      "loss": 1.6974,
+      "step": 1731
+    },
+    {
+      "epoch": 0.42059252064108793,
+      "grad_norm": 0.5834603905677795,
+      "learning_rate": 0.0001,
+      "loss": 1.7598,
+      "step": 1732
+    },
+    {
+      "epoch": 0.4208353569694026,
+      "grad_norm": 0.6204494833946228,
+      "learning_rate": 0.0001,
+      "loss": 1.8967,
+      "step": 1733
+    },
+    {
+      "epoch": 0.42107819329771734,
+      "grad_norm": 0.6163272261619568,
+      "learning_rate": 0.0001,
+      "loss": 1.8135,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4213210296260321,
+      "grad_norm": 0.5401644110679626,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 1735
+    },
+    {
+      "epoch": 0.42156386595434675,
+      "grad_norm": 0.5756164193153381,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4218067022826615,
+      "grad_norm": 0.5235920548439026,
+      "learning_rate": 0.0001,
+      "loss": 1.5956,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4220495386109762,
+      "grad_norm": 0.5800116658210754,
+      "learning_rate": 0.0001,
+      "loss": 1.8049,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4222923749392909,
+      "grad_norm": 0.5359517335891724,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4225352112676056,
+      "grad_norm": 0.5693761110305786,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 1740
+    },
+    {
+      "epoch": 0.42277804759592036,
+      "grad_norm": 0.565039873123169,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 1741
+    },
+    {
+      "epoch": 0.4230208839242351,
+      "grad_norm": 0.5437293648719788,
+      "learning_rate": 0.0001,
+      "loss": 1.6931,
+      "step": 1742
+    },
+    {
+      "epoch": 0.42326372025254977,
+      "grad_norm": 0.5825754404067993,
+      "learning_rate": 0.0001,
+      "loss": 1.8266,
+      "step": 1743
+    },
+    {
+      "epoch": 0.4235065565808645,
+      "grad_norm": 0.5717706680297852,
+      "learning_rate": 0.0001,
+      "loss": 1.8816,
+      "step": 1744
+    },
+    {
+      "epoch": 0.42374939290917923,
+      "grad_norm": 0.5643588304519653,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4239922292374939,
+      "grad_norm": 0.5338857769966125,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 1746
+    },
+    {
+      "epoch": 0.42423506556580864,
+      "grad_norm": 0.5591118931770325,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4244779018941234,
+      "grad_norm": 0.6191600561141968,
+      "learning_rate": 0.0001,
+      "loss": 1.8901,
+      "step": 1748
+    },
+    {
+      "epoch": 0.42472073822243805,
+      "grad_norm": 0.5633436441421509,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 1749
+    },
+    {
+      "epoch": 0.4249635745507528,
+      "grad_norm": 0.564781904220581,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4252064108790675,
+      "grad_norm": 0.5725958943367004,
+      "learning_rate": 0.0001,
+      "loss": 1.7598,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4254492472073822,
+      "grad_norm": 0.5628660917282104,
+      "learning_rate": 0.0001,
+      "loss": 1.8647,
+      "step": 1752
+    },
+    {
+      "epoch": 0.42569208353569693,
+      "grad_norm": 0.5747968554496765,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 1753
+    },
+    {
+      "epoch": 0.42593491986401166,
+      "grad_norm": 0.5645371675491333,
+      "learning_rate": 0.0001,
+      "loss": 1.6864,
+      "step": 1754
+    },
+    {
+      "epoch": 0.4261777561923264,
+      "grad_norm": 0.5752917528152466,
+      "learning_rate": 0.0001,
+      "loss": 1.8961,
+      "step": 1755
+    },
+    {
+      "epoch": 0.4264205925206411,
+      "grad_norm": 0.5931205749511719,
+      "learning_rate": 0.0001,
+      "loss": 1.888,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4266634288489558,
+      "grad_norm": 0.581861138343811,
+      "learning_rate": 0.0001,
+      "loss": 1.8166,
+      "step": 1757
+    },
+    {
+      "epoch": 0.42690626517727054,
+      "grad_norm": 0.5772149562835693,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4271491015055852,
+      "grad_norm": 0.5850484371185303,
+      "learning_rate": 0.0001,
+      "loss": 1.7696,
+      "step": 1759
+    },
+    {
+      "epoch": 0.42739193783389995,
+      "grad_norm": 0.5533158779144287,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4276347741622147,
+      "grad_norm": 0.55548495054245,
+      "learning_rate": 0.0001,
+      "loss": 1.8209,
+      "step": 1761
+    },
+    {
+      "epoch": 0.42787761049052936,
+      "grad_norm": 0.5674953460693359,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4281204468188441,
+      "grad_norm": 0.5729069709777832,
+      "learning_rate": 0.0001,
+      "loss": 1.7524,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4283632831471588,
+      "grad_norm": 0.5497011542320251,
+      "learning_rate": 0.0001,
+      "loss": 1.5961,
+      "step": 1764
+    },
+    {
+      "epoch": 0.42860611947547356,
+      "grad_norm": 0.561162531375885,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 1765
+    },
+    {
+      "epoch": 0.42884895580378823,
+      "grad_norm": 0.5849171280860901,
+      "learning_rate": 0.0001,
+      "loss": 1.8342,
+      "step": 1766
+    },
+    {
+      "epoch": 0.42909179213210297,
+      "grad_norm": 0.56157386302948,
+      "learning_rate": 0.0001,
+      "loss": 1.7408,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4293346284604177,
+      "grad_norm": 0.6153436303138733,
+      "learning_rate": 0.0001,
+      "loss": 1.8358,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4295774647887324,
+      "grad_norm": 0.584125816822052,
+      "learning_rate": 0.0001,
+      "loss": 1.8792,
+      "step": 1769
+    },
+    {
+      "epoch": 0.4298203011170471,
+      "grad_norm": 0.6181815266609192,
+      "learning_rate": 0.0001,
+      "loss": 1.8735,
+      "step": 1770
+    },
+    {
+      "epoch": 0.43006313744536184,
+      "grad_norm": 0.579014241695404,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4303059737736765,
+      "grad_norm": 0.5735815763473511,
+      "learning_rate": 0.0001,
+      "loss": 1.779,
+      "step": 1772
+    },
+    {
+      "epoch": 0.43054881010199125,
+      "grad_norm": 0.5597633719444275,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 1773
+    },
+    {
+      "epoch": 0.430791646430306,
+      "grad_norm": 0.5619441270828247,
+      "learning_rate": 0.0001,
+      "loss": 1.9089,
+      "step": 1774
+    },
+    {
+      "epoch": 0.43103448275862066,
+      "grad_norm": 0.5433188676834106,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4312773190869354,
+      "grad_norm": 0.6146637201309204,
+      "learning_rate": 0.0001,
+      "loss": 1.7771,
+      "step": 1776
+    },
+    {
+      "epoch": 0.43152015541525013,
+      "grad_norm": 0.5701290965080261,
+      "learning_rate": 0.0001,
+      "loss": 1.7436,
+      "step": 1777
+    },
+    {
+      "epoch": 0.43176299174356486,
+      "grad_norm": 0.6096238493919373,
+      "learning_rate": 0.0001,
+      "loss": 1.8751,
+      "step": 1778
+    },
+    {
+      "epoch": 0.43200582807187954,
+      "grad_norm": 0.6130304336547852,
+      "learning_rate": 0.0001,
+      "loss": 1.8221,
+      "step": 1779
+    },
+    {
+      "epoch": 0.43224866440019427,
+      "grad_norm": 0.5769259333610535,
+      "learning_rate": 0.0001,
+      "loss": 1.8647,
+      "step": 1780
+    },
+    {
+      "epoch": 0.432491500728509,
+      "grad_norm": 0.5667563080787659,
+      "learning_rate": 0.0001,
+      "loss": 1.6987,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4327343370568237,
+      "grad_norm": 0.5886138081550598,
+      "learning_rate": 0.0001,
+      "loss": 1.7908,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4329771733851384,
+      "grad_norm": 0.6100742816925049,
+      "learning_rate": 0.0001,
+      "loss": 1.7411,
+      "step": 1783
+    },
+    {
+      "epoch": 0.43322000971345315,
+      "grad_norm": 0.5517228245735168,
+      "learning_rate": 0.0001,
+      "loss": 1.617,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4334628460417678,
+      "grad_norm": 0.5412666201591492,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 1785
+    },
+    {
+      "epoch": 0.43370568237008256,
+      "grad_norm": 0.5847234129905701,
+      "learning_rate": 0.0001,
+      "loss": 1.8383,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4339485186983973,
+      "grad_norm": 0.5502403974533081,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 1787
+    },
+    {
+      "epoch": 0.434191355026712,
+      "grad_norm": 0.5902757048606873,
+      "learning_rate": 0.0001,
+      "loss": 1.8038,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4344341913550267,
+      "grad_norm": 0.5674586296081543,
+      "learning_rate": 0.0001,
+      "loss": 1.7138,
+      "step": 1789
+    },
+    {
+      "epoch": 0.43467702768334143,
+      "grad_norm": 0.5742107629776001,
+      "learning_rate": 0.0001,
+      "loss": 1.8375,
+      "step": 1790
+    },
+    {
+      "epoch": 0.43491986401165617,
+      "grad_norm": 0.579260528087616,
+      "learning_rate": 0.0001,
+      "loss": 1.6283,
+      "step": 1791
+    },
+    {
+      "epoch": 0.43516270033997084,
+      "grad_norm": 0.5684863328933716,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4354055366682856,
+      "grad_norm": 0.5778998732566833,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4356483729966003,
+      "grad_norm": 0.5546566843986511,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 1794
+    },
+    {
+      "epoch": 0.435891209324915,
+      "grad_norm": 0.581365704536438,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4361340456532297,
+      "grad_norm": 0.5921564698219299,
+      "learning_rate": 0.0001,
+      "loss": 1.7575,
+      "step": 1796
+    },
+    {
+      "epoch": 0.43637688198154445,
+      "grad_norm": 0.5925065875053406,
+      "learning_rate": 0.0001,
+      "loss": 1.9726,
+      "step": 1797
+    },
+    {
+      "epoch": 0.43661971830985913,
+      "grad_norm": 0.5550657510757446,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 1798
+    },
+    {
+      "epoch": 0.43686255463817386,
+      "grad_norm": 0.5925093293190002,
+      "learning_rate": 0.0001,
+      "loss": 1.6507,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4371053909664886,
+      "grad_norm": 0.5593153238296509,
+      "learning_rate": 0.0001,
+      "loss": 1.6802,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4373482272948033,
+      "grad_norm": 0.5875351428985596,
+      "learning_rate": 0.0001,
+      "loss": 1.7225,
+      "step": 1801
+    },
+    {
+      "epoch": 0.437591063623118,
+      "grad_norm": 0.6166852116584778,
+      "learning_rate": 0.0001,
+      "loss": 1.8489,
+      "step": 1802
+    },
+    {
+      "epoch": 0.43783389995143274,
+      "grad_norm": 0.5772126913070679,
+      "learning_rate": 0.0001,
+      "loss": 1.9313,
+      "step": 1803
+    },
+    {
+      "epoch": 0.43807673627974747,
+      "grad_norm": 0.5607519745826721,
+      "learning_rate": 0.0001,
+      "loss": 1.7712,
+      "step": 1804
+    },
+    {
+      "epoch": 0.43831957260806215,
+      "grad_norm": 0.5917442440986633,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4385624089363769,
+      "grad_norm": 0.599422037601471,
+      "learning_rate": 0.0001,
+      "loss": 1.9742,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4388052452646916,
+      "grad_norm": 0.5593117475509644,
+      "learning_rate": 0.0001,
+      "loss": 1.8367,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4390480815930063,
+      "grad_norm": 0.5732071399688721,
+      "learning_rate": 0.0001,
+      "loss": 1.8717,
+      "step": 1808
+    },
+    {
+      "epoch": 0.439290917921321,
+      "grad_norm": 0.5580980777740479,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 1809
+    },
+    {
+      "epoch": 0.43953375424963576,
+      "grad_norm": 0.615973949432373,
+      "learning_rate": 0.0001,
+      "loss": 1.8811,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4397765905779505,
+      "grad_norm": 0.6223271489143372,
+      "learning_rate": 0.0001,
+      "loss": 1.9497,
+      "step": 1811
+    },
+    {
+      "epoch": 0.44001942690626517,
+      "grad_norm": 0.5596504211425781,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 1812
+    },
+    {
+      "epoch": 0.4402622632345799,
+      "grad_norm": 0.5751744508743286,
+      "learning_rate": 0.0001,
+      "loss": 1.8838,
+      "step": 1813
+    },
+    {
+      "epoch": 0.44050509956289463,
+      "grad_norm": 0.5983855724334717,
+      "learning_rate": 0.0001,
+      "loss": 1.8151,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4407479358912093,
+      "grad_norm": 0.5524014830589294,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 1815
+    },
+    {
+      "epoch": 0.44099077221952404,
+      "grad_norm": 0.5575211644172668,
+      "learning_rate": 0.0001,
+      "loss": 1.7025,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4412336085478388,
+      "grad_norm": 0.5798826217651367,
+      "learning_rate": 0.0001,
+      "loss": 1.788,
+      "step": 1817
+    },
+    {
+      "epoch": 0.44147644487615345,
+      "grad_norm": 0.5255656242370605,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4417192812044682,
+      "grad_norm": 0.560876727104187,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4419621175327829,
+      "grad_norm": 0.5471630692481995,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4422049538610976,
+      "grad_norm": 0.5710832476615906,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4424477901894123,
+      "grad_norm": 0.582895815372467,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 1822
+    },
+    {
+      "epoch": 0.44269062651772706,
+      "grad_norm": 0.5692479014396667,
+      "learning_rate": 0.0001,
+      "loss": 1.7125,
+      "step": 1823
+    },
+    {
+      "epoch": 0.4429334628460418,
+      "grad_norm": 0.5563973188400269,
+      "learning_rate": 0.0001,
+      "loss": 1.7806,
+      "step": 1824
+    },
+    {
+      "epoch": 0.44317629917435647,
+      "grad_norm": 0.6028217673301697,
+      "learning_rate": 0.0001,
+      "loss": 1.7916,
+      "step": 1825
+    },
+    {
+      "epoch": 0.4434191355026712,
+      "grad_norm": 0.6247011423110962,
+      "learning_rate": 0.0001,
+      "loss": 1.7032,
+      "step": 1826
+    },
+    {
+      "epoch": 0.44366197183098594,
+      "grad_norm": 0.5845817923545837,
+      "learning_rate": 0.0001,
+      "loss": 1.9361,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4439048081593006,
+      "grad_norm": 0.586754322052002,
+      "learning_rate": 0.0001,
+      "loss": 1.8163,
+      "step": 1828
+    },
+    {
+      "epoch": 0.44414764448761535,
+      "grad_norm": 0.6034722924232483,
+      "learning_rate": 0.0001,
+      "loss": 1.8628,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4443904808159301,
+      "grad_norm": 0.5657421946525574,
+      "learning_rate": 0.0001,
+      "loss": 1.7418,
+      "step": 1830
+    },
+    {
+      "epoch": 0.44463331714424476,
+      "grad_norm": 0.6093968152999878,
+      "learning_rate": 0.0001,
+      "loss": 1.7401,
+      "step": 1831
+    },
+    {
+      "epoch": 0.4448761534725595,
+      "grad_norm": 0.5818173885345459,
+      "learning_rate": 0.0001,
+      "loss": 1.6932,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4451189898008742,
+      "grad_norm": 0.578579306602478,
+      "learning_rate": 0.0001,
+      "loss": 1.8248,
+      "step": 1833
+    },
+    {
+      "epoch": 0.44536182612918895,
+      "grad_norm": 0.5778454542160034,
+      "learning_rate": 0.0001,
+      "loss": 1.7586,
+      "step": 1834
+    },
+    {
+      "epoch": 0.44560466245750363,
+      "grad_norm": 0.5978708267211914,
+      "learning_rate": 0.0001,
+      "loss": 1.7819,
+      "step": 1835
+    },
+    {
+      "epoch": 0.44584749878581836,
+      "grad_norm": 0.5598545670509338,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4460903351141331,
+      "grad_norm": 0.6154178977012634,
+      "learning_rate": 0.0001,
+      "loss": 1.7779,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4463331714424478,
+      "grad_norm": 0.5966653823852539,
+      "learning_rate": 0.0001,
+      "loss": 1.9132,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4465760077707625,
+      "grad_norm": 0.5512944459915161,
+      "learning_rate": 0.0001,
+      "loss": 1.6269,
+      "step": 1839
+    },
+    {
+      "epoch": 0.44681884409907724,
+      "grad_norm": 0.5656938552856445,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4470616804273919,
+      "grad_norm": 0.5822314023971558,
+      "learning_rate": 0.0001,
+      "loss": 1.7625,
+      "step": 1841
+    },
+    {
+      "epoch": 0.44730451675570665,
+      "grad_norm": 0.6166737675666809,
+      "learning_rate": 0.0001,
+      "loss": 1.8711,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4475473530840214,
+      "grad_norm": 0.573864221572876,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 1843
+    },
+    {
+      "epoch": 0.44779018941233606,
+      "grad_norm": 0.5893234014511108,
+      "learning_rate": 0.0001,
+      "loss": 1.7329,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4480330257406508,
+      "grad_norm": 0.6403665542602539,
+      "learning_rate": 0.0001,
+      "loss": 1.8165,
+      "step": 1845
+    },
+    {
+      "epoch": 0.4482758620689655,
+      "grad_norm": 0.5796035528182983,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 1846
+    },
+    {
+      "epoch": 0.44851869839728026,
+      "grad_norm": 0.5376893281936646,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 1847
+    },
+    {
+      "epoch": 0.44876153472559493,
+      "grad_norm": 0.5643702745437622,
+      "learning_rate": 0.0001,
+      "loss": 1.7544,
+      "step": 1848
+    },
+    {
+      "epoch": 0.44900437105390967,
+      "grad_norm": 0.5793332457542419,
+      "learning_rate": 0.0001,
+      "loss": 1.8197,
+      "step": 1849
+    },
+    {
+      "epoch": 0.4492472073822244,
+      "grad_norm": 0.5740710496902466,
+      "learning_rate": 0.0001,
+      "loss": 1.8129,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4494900437105391,
+      "grad_norm": 0.5284385681152344,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4497328800388538,
+      "grad_norm": 0.5854473114013672,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 1852
+    },
+    {
+      "epoch": 0.44997571636716854,
+      "grad_norm": 0.5556619763374329,
+      "learning_rate": 0.0001,
+      "loss": 1.7712,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4502185526954832,
+      "grad_norm": 0.596434473991394,
+      "learning_rate": 0.0001,
+      "loss": 1.8985,
+      "step": 1854
+    },
+    {
+      "epoch": 0.45046138902379795,
+      "grad_norm": 0.5931633710861206,
+      "learning_rate": 0.0001,
+      "loss": 1.8248,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4507042253521127,
+      "grad_norm": 0.5966574549674988,
+      "learning_rate": 0.0001,
+      "loss": 1.8493,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4509470616804274,
+      "grad_norm": 0.5613736510276794,
+      "learning_rate": 0.0001,
+      "loss": 1.7375,
+      "step": 1857
+    },
+    {
+      "epoch": 0.4511898980087421,
+      "grad_norm": 0.6010239720344543,
+      "learning_rate": 0.0001,
+      "loss": 1.8562,
+      "step": 1858
+    },
+    {
+      "epoch": 0.45143273433705683,
+      "grad_norm": 0.566714346408844,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 1859
+    },
+    {
+      "epoch": 0.45167557066537156,
+      "grad_norm": 0.5681232810020447,
+      "learning_rate": 0.0001,
+      "loss": 1.7613,
+      "step": 1860
+    },
+    {
+      "epoch": 0.45191840699368624,
+      "grad_norm": 0.5798001289367676,
+      "learning_rate": 0.0001,
+      "loss": 1.9256,
+      "step": 1861
+    },
+    {
+      "epoch": 0.45216124332200097,
+      "grad_norm": 0.553661048412323,
+      "learning_rate": 0.0001,
+      "loss": 1.6513,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4524040796503157,
+      "grad_norm": 0.5698717832565308,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 1863
+    },
+    {
+      "epoch": 0.4526469159786304,
+      "grad_norm": 0.5184240937232971,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 1864
+    },
+    {
+      "epoch": 0.4528897523069451,
+      "grad_norm": 0.5855940580368042,
+      "learning_rate": 0.0001,
+      "loss": 1.6505,
+      "step": 1865
+    },
+    {
+      "epoch": 0.45313258863525985,
+      "grad_norm": 0.5758311152458191,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4533754249635745,
+      "grad_norm": 0.615971028804779,
+      "learning_rate": 0.0001,
+      "loss": 1.735,
+      "step": 1867
+    },
+    {
+      "epoch": 0.45361826129188926,
+      "grad_norm": 0.5524644255638123,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 1868
+    },
+    {
+      "epoch": 0.453861097620204,
+      "grad_norm": 0.5362394452095032,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 1869
+    },
+    {
+      "epoch": 0.4541039339485187,
+      "grad_norm": 0.585970938205719,
+      "learning_rate": 0.0001,
+      "loss": 1.6943,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4543467702768334,
+      "grad_norm": 0.5870539546012878,
+      "learning_rate": 0.0001,
+      "loss": 1.741,
+      "step": 1871
+    },
+    {
+      "epoch": 0.45458960660514813,
+      "grad_norm": 0.5760737061500549,
+      "learning_rate": 0.0001,
+      "loss": 1.8845,
+      "step": 1872
+    },
+    {
+      "epoch": 0.45483244293346287,
+      "grad_norm": 0.5538679957389832,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 1873
+    },
+    {
+      "epoch": 0.45507527926177754,
+      "grad_norm": 0.5464616417884827,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4553181155900923,
+      "grad_norm": 0.5502505898475647,
+      "learning_rate": 0.0001,
+      "loss": 1.72,
+      "step": 1875
+    },
+    {
+      "epoch": 0.455560951918407,
+      "grad_norm": 0.5644949078559875,
+      "learning_rate": 0.0001,
+      "loss": 1.7139,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4558037882467217,
+      "grad_norm": 0.5552821159362793,
+      "learning_rate": 0.0001,
+      "loss": 1.7832,
+      "step": 1877
+    },
+    {
+      "epoch": 0.4560466245750364,
+      "grad_norm": 0.6095209717750549,
+      "learning_rate": 0.0001,
+      "loss": 1.8575,
+      "step": 1878
+    },
+    {
+      "epoch": 0.45628946090335115,
+      "grad_norm": 0.5591077208518982,
+      "learning_rate": 0.0001,
+      "loss": 1.5926,
+      "step": 1879
+    },
+    {
+      "epoch": 0.4565322972316659,
+      "grad_norm": 0.5598617196083069,
+      "learning_rate": 0.0001,
+      "loss": 1.7178,
+      "step": 1880
+    },
+    {
+      "epoch": 0.45677513355998056,
+      "grad_norm": 0.5825666189193726,
+      "learning_rate": 0.0001,
+      "loss": 1.6868,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4570179698882953,
+      "grad_norm": 0.614214301109314,
+      "learning_rate": 0.0001,
+      "loss": 1.8865,
+      "step": 1882
+    },
+    {
+      "epoch": 0.45726080621661,
+      "grad_norm": 0.5513318777084351,
+      "learning_rate": 0.0001,
+      "loss": 1.7676,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4575036425449247,
+      "grad_norm": 0.6541092991828918,
+      "learning_rate": 0.0001,
+      "loss": 1.9426,
+      "step": 1884
+    },
+    {
+      "epoch": 0.45774647887323944,
+      "grad_norm": 0.6022918224334717,
+      "learning_rate": 0.0001,
+      "loss": 1.8112,
+      "step": 1885
+    },
+    {
+      "epoch": 0.45798931520155417,
+      "grad_norm": 0.5467543601989746,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 1886
+    },
+    {
+      "epoch": 0.45823215152986885,
+      "grad_norm": 0.593727171421051,
+      "learning_rate": 0.0001,
+      "loss": 1.8278,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4584749878581836,
+      "grad_norm": 0.5079317688941956,
+      "learning_rate": 0.0001,
+      "loss": 1.6571,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4587178241864983,
+      "grad_norm": 0.5998428463935852,
+      "learning_rate": 0.0001,
+      "loss": 1.8235,
+      "step": 1889
+    },
+    {
+      "epoch": 0.458960660514813,
+      "grad_norm": 0.622458279132843,
+      "learning_rate": 0.0001,
+      "loss": 1.8577,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4592034968431277,
+      "grad_norm": 0.5983126759529114,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 1891
+    },
+    {
+      "epoch": 0.45944633317144246,
+      "grad_norm": 0.5393468737602234,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 1892
+    },
+    {
+      "epoch": 0.4596891694997572,
+      "grad_norm": 0.5763282775878906,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 1893
+    },
+    {
+      "epoch": 0.45993200582807187,
+      "grad_norm": 0.6060768961906433,
+      "learning_rate": 0.0001,
+      "loss": 1.8217,
+      "step": 1894
+    },
+    {
+      "epoch": 0.4601748421563866,
+      "grad_norm": 0.5790578722953796,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 1895
+    },
+    {
+      "epoch": 0.46041767848470133,
+      "grad_norm": 0.5628831386566162,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 1896
+    },
+    {
+      "epoch": 0.460660514813016,
+      "grad_norm": 0.583311140537262,
+      "learning_rate": 0.0001,
+      "loss": 1.7067,
+      "step": 1897
+    },
+    {
+      "epoch": 0.46090335114133074,
+      "grad_norm": 0.5306105613708496,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4611461874696455,
+      "grad_norm": 0.6257902383804321,
+      "learning_rate": 0.0001,
+      "loss": 1.749,
+      "step": 1899
+    },
+    {
+      "epoch": 0.46138902379796015,
+      "grad_norm": 0.5784273147583008,
+      "learning_rate": 0.0001,
+      "loss": 1.7965,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4616318601262749,
+      "grad_norm": 0.5503388047218323,
+      "learning_rate": 0.0001,
+      "loss": 1.6824,
+      "step": 1901
+    },
+    {
+      "epoch": 0.4618746964545896,
+      "grad_norm": 0.5989645719528198,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 1902
+    },
+    {
+      "epoch": 0.46211753278290435,
+      "grad_norm": 0.5790361166000366,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 1903
+    },
+    {
+      "epoch": 0.462360369111219,
+      "grad_norm": 0.5661267042160034,
+      "learning_rate": 0.0001,
+      "loss": 1.6861,
+      "step": 1904
+    },
+    {
+      "epoch": 0.46260320543953376,
+      "grad_norm": 0.5857040286064148,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 1905
+    },
+    {
+      "epoch": 0.4628460417678485,
+      "grad_norm": 0.6014676690101624,
+      "learning_rate": 0.0001,
+      "loss": 1.7985,
+      "step": 1906
+    },
+    {
+      "epoch": 0.46308887809616317,
+      "grad_norm": 0.5533243417739868,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 1907
+    },
+    {
+      "epoch": 0.4633317144244779,
+      "grad_norm": 0.5554340481758118,
+      "learning_rate": 0.0001,
+      "loss": 1.5683,
+      "step": 1908
+    },
+    {
+      "epoch": 0.46357455075279264,
+      "grad_norm": 0.5753559470176697,
+      "learning_rate": 0.0001,
+      "loss": 1.6389,
+      "step": 1909
+    },
+    {
+      "epoch": 0.4638173870811073,
+      "grad_norm": 0.6063196659088135,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 1910
+    },
+    {
+      "epoch": 0.46406022340942205,
+      "grad_norm": 0.6287242770195007,
+      "learning_rate": 0.0001,
+      "loss": 1.883,
+      "step": 1911
+    },
+    {
+      "epoch": 0.4643030597377368,
+      "grad_norm": 0.5789053440093994,
+      "learning_rate": 0.0001,
+      "loss": 1.8475,
+      "step": 1912
+    },
+    {
+      "epoch": 0.46454589606605146,
+      "grad_norm": 0.5780322551727295,
+      "learning_rate": 0.0001,
+      "loss": 1.8763,
+      "step": 1913
+    },
+    {
+      "epoch": 0.4647887323943662,
+      "grad_norm": 0.5103904008865356,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 1914
+    },
+    {
+      "epoch": 0.4650315687226809,
+      "grad_norm": 0.5886533856391907,
+      "learning_rate": 0.0001,
+      "loss": 1.7561,
+      "step": 1915
+    },
+    {
+      "epoch": 0.46527440505099565,
+      "grad_norm": 0.5434485077857971,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 1916
+    },
+    {
+      "epoch": 0.46551724137931033,
+      "grad_norm": 0.5687652230262756,
+      "learning_rate": 0.0001,
+      "loss": 1.7215,
+      "step": 1917
+    },
+    {
+      "epoch": 0.46576007770762506,
+      "grad_norm": 0.5779525637626648,
+      "learning_rate": 0.0001,
+      "loss": 1.7172,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4660029140359398,
+      "grad_norm": 0.551944375038147,
+      "learning_rate": 0.0001,
+      "loss": 1.6293,
+      "step": 1919
+    },
+    {
+      "epoch": 0.4662457503642545,
+      "grad_norm": 0.5759638547897339,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4664885866925692,
+      "grad_norm": 0.5612455010414124,
+      "learning_rate": 0.0001,
+      "loss": 1.803,
+      "step": 1921
+    },
+    {
+      "epoch": 0.46673142302088394,
+      "grad_norm": 0.5707778334617615,
+      "learning_rate": 0.0001,
+      "loss": 1.7959,
+      "step": 1922
+    },
+    {
+      "epoch": 0.4669742593491986,
+      "grad_norm": 0.6270820498466492,
+      "learning_rate": 0.0001,
+      "loss": 1.7296,
+      "step": 1923
+    },
+    {
+      "epoch": 0.46721709567751335,
+      "grad_norm": 0.5518383383750916,
+      "learning_rate": 0.0001,
+      "loss": 1.6417,
+      "step": 1924
+    },
+    {
+      "epoch": 0.4674599320058281,
+      "grad_norm": 0.5935297012329102,
+      "learning_rate": 0.0001,
+      "loss": 1.7797,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4677027683341428,
+      "grad_norm": 0.617348849773407,
+      "learning_rate": 0.0001,
+      "loss": 1.7296,
+      "step": 1926
+    },
+    {
+      "epoch": 0.4679456046624575,
+      "grad_norm": 0.5728745460510254,
+      "learning_rate": 0.0001,
+      "loss": 1.7844,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4681884409907722,
+      "grad_norm": 0.5723167657852173,
+      "learning_rate": 0.0001,
+      "loss": 1.7022,
+      "step": 1928
+    },
+    {
+      "epoch": 0.46843127731908696,
+      "grad_norm": 0.5914849638938904,
+      "learning_rate": 0.0001,
+      "loss": 1.8989,
+      "step": 1929
+    },
+    {
+      "epoch": 0.46867411364740164,
+      "grad_norm": 0.5882318019866943,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 1930
+    },
+    {
+      "epoch": 0.46891694997571637,
+      "grad_norm": 0.5570792555809021,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4691597863040311,
+      "grad_norm": 0.5721078515052795,
+      "learning_rate": 0.0001,
+      "loss": 1.8832,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4694026226323458,
+      "grad_norm": 0.6023865342140198,
+      "learning_rate": 0.0001,
+      "loss": 1.9534,
+      "step": 1933
+    },
+    {
+      "epoch": 0.4696454589606605,
+      "grad_norm": 0.6025052666664124,
+      "learning_rate": 0.0001,
+      "loss": 1.8732,
+      "step": 1934
+    },
+    {
+      "epoch": 0.46988829528897524,
+      "grad_norm": 0.5464321374893188,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4701311316172899,
+      "grad_norm": 0.6009752750396729,
+      "learning_rate": 0.0001,
+      "loss": 1.8822,
+      "step": 1936
+    },
+    {
+      "epoch": 0.47037396794560465,
+      "grad_norm": 0.6073281764984131,
+      "learning_rate": 0.0001,
+      "loss": 1.8904,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4706168042739194,
+      "grad_norm": 0.5457764863967896,
+      "learning_rate": 0.0001,
+      "loss": 1.7498,
+      "step": 1938
+    },
+    {
+      "epoch": 0.4708596406022341,
+      "grad_norm": 0.5712628364562988,
+      "learning_rate": 0.0001,
+      "loss": 1.8484,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4711024769305488,
+      "grad_norm": 0.5740649700164795,
+      "learning_rate": 0.0001,
+      "loss": 1.7276,
+      "step": 1940
+    },
+    {
+      "epoch": 0.47134531325886353,
+      "grad_norm": 0.5414160490036011,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 1941
+    },
+    {
+      "epoch": 0.47158814958717826,
+      "grad_norm": 0.5896227359771729,
+      "learning_rate": 0.0001,
+      "loss": 1.7745,
+      "step": 1942
+    },
+    {
+      "epoch": 0.47183098591549294,
+      "grad_norm": 0.5514649152755737,
+      "learning_rate": 0.0001,
+      "loss": 1.5694,
+      "step": 1943
+    },
+    {
+      "epoch": 0.4720738222438077,
+      "grad_norm": 0.5307028293609619,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 1944
+    },
+    {
+      "epoch": 0.4723166585721224,
+      "grad_norm": 0.5563642978668213,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 1945
+    },
+    {
+      "epoch": 0.4725594949004371,
+      "grad_norm": 0.5585283637046814,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4728023312287518,
+      "grad_norm": 0.5462757349014282,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 1947
+    },
+    {
+      "epoch": 0.47304516755706655,
+      "grad_norm": 0.5939182043075562,
+      "learning_rate": 0.0001,
+      "loss": 1.9834,
+      "step": 1948
+    },
+    {
+      "epoch": 0.4732880038853812,
+      "grad_norm": 0.5892949104309082,
+      "learning_rate": 0.0001,
+      "loss": 1.5758,
+      "step": 1949
+    },
+    {
+      "epoch": 0.47353084021369596,
+      "grad_norm": 0.5596404075622559,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4737736765420107,
+      "grad_norm": 0.551993727684021,
+      "learning_rate": 0.0001,
+      "loss": 1.7142,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4740165128703254,
+      "grad_norm": 0.5357285141944885,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4742593491986401,
+      "grad_norm": 0.6192750334739685,
+      "learning_rate": 0.0001,
+      "loss": 1.8688,
+      "step": 1953
+    },
+    {
+      "epoch": 0.47450218552695483,
+      "grad_norm": 0.5935304164886475,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 1954
+    },
+    {
+      "epoch": 0.47474502185526957,
+      "grad_norm": 0.5641899108886719,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 1955
+    },
+    {
+      "epoch": 0.47498785818358424,
+      "grad_norm": 0.5489999651908875,
+      "learning_rate": 0.0001,
+      "loss": 1.7741,
+      "step": 1956
+    },
+    {
+      "epoch": 0.475230694511899,
+      "grad_norm": 0.5819328427314758,
+      "learning_rate": 0.0001,
+      "loss": 1.7622,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4754735308402137,
+      "grad_norm": 0.5315662622451782,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 1958
+    },
+    {
+      "epoch": 0.4757163671685284,
+      "grad_norm": 0.5343338251113892,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 1959
+    },
+    {
+      "epoch": 0.4759592034968431,
+      "grad_norm": 0.5986287593841553,
+      "learning_rate": 0.0001,
+      "loss": 1.7476,
+      "step": 1960
+    },
+    {
+      "epoch": 0.47620203982515785,
+      "grad_norm": 0.5720230937004089,
+      "learning_rate": 0.0001,
+      "loss": 1.7423,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4764448761534726,
+      "grad_norm": 0.5954564809799194,
+      "learning_rate": 0.0001,
+      "loss": 1.8605,
+      "step": 1962
+    },
+    {
+      "epoch": 0.47668771248178726,
+      "grad_norm": 0.5579463243484497,
+      "learning_rate": 0.0001,
+      "loss": 1.7147,
+      "step": 1963
+    },
+    {
+      "epoch": 0.476930548810102,
+      "grad_norm": 0.5727841258049011,
+      "learning_rate": 0.0001,
+      "loss": 1.7308,
+      "step": 1964
+    },
+    {
+      "epoch": 0.47717338513841673,
+      "grad_norm": 0.6127378940582275,
+      "learning_rate": 0.0001,
+      "loss": 1.8088,
+      "step": 1965
+    },
+    {
+      "epoch": 0.4774162214667314,
+      "grad_norm": 0.5378556251525879,
+      "learning_rate": 0.0001,
+      "loss": 1.7187,
+      "step": 1966
+    },
+    {
+      "epoch": 0.47765905779504614,
+      "grad_norm": 0.5923590660095215,
+      "learning_rate": 0.0001,
+      "loss": 1.7997,
+      "step": 1967
+    },
+    {
+      "epoch": 0.47790189412336087,
+      "grad_norm": 0.6295888423919678,
+      "learning_rate": 0.0001,
+      "loss": 1.5496,
+      "step": 1968
+    },
+    {
+      "epoch": 0.47814473045167555,
+      "grad_norm": 0.5650759935379028,
+      "learning_rate": 0.0001,
+      "loss": 1.7427,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4783875667799903,
+      "grad_norm": 0.5341917872428894,
+      "learning_rate": 0.0001,
+      "loss": 1.4833,
+      "step": 1970
+    },
+    {
+      "epoch": 0.478630403108305,
+      "grad_norm": 0.5882410407066345,
+      "learning_rate": 0.0001,
+      "loss": 1.7442,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4788732394366197,
+      "grad_norm": 0.5801606774330139,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4791160757649344,
+      "grad_norm": 0.6176870465278625,
+      "learning_rate": 0.0001,
+      "loss": 1.7344,
+      "step": 1973
+    },
+    {
+      "epoch": 0.47935891209324916,
+      "grad_norm": 0.583313524723053,
+      "learning_rate": 0.0001,
+      "loss": 1.6592,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4796017484215639,
+      "grad_norm": 0.5627610087394714,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 1975
+    },
+    {
+      "epoch": 0.47984458474987857,
+      "grad_norm": 0.5924595594406128,
+      "learning_rate": 0.0001,
+      "loss": 1.8331,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4800874210781933,
+      "grad_norm": 0.6066086888313293,
+      "learning_rate": 0.0001,
+      "loss": 1.8329,
+      "step": 1977
+    },
+    {
+      "epoch": 0.48033025740650803,
+      "grad_norm": 0.5486518144607544,
+      "learning_rate": 0.0001,
+      "loss": 1.7316,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4805730937348227,
+      "grad_norm": 0.5555726885795593,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 1979
+    },
+    {
+      "epoch": 0.48081593006313744,
+      "grad_norm": 0.6334558725357056,
+      "learning_rate": 0.0001,
+      "loss": 1.8482,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4810587663914522,
+      "grad_norm": 0.5876926183700562,
+      "learning_rate": 0.0001,
+      "loss": 1.8004,
+      "step": 1981
+    },
+    {
+      "epoch": 0.48130160271976685,
+      "grad_norm": 0.5859856009483337,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4815444390480816,
+      "grad_norm": 0.6316081881523132,
+      "learning_rate": 0.0001,
+      "loss": 2.1124,
+      "step": 1983
+    },
+    {
+      "epoch": 0.4817872753763963,
+      "grad_norm": 0.5542670488357544,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 1984
+    },
+    {
+      "epoch": 0.48203011170471105,
+      "grad_norm": 0.6528927683830261,
+      "learning_rate": 0.0001,
+      "loss": 1.7741,
+      "step": 1985
+    },
+    {
+      "epoch": 0.48227294803302573,
+      "grad_norm": 0.6087407469749451,
+      "learning_rate": 0.0001,
+      "loss": 1.8804,
+      "step": 1986
+    },
+    {
+      "epoch": 0.48251578436134046,
+      "grad_norm": 0.642598032951355,
+      "learning_rate": 0.0001,
+      "loss": 1.8036,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 0.5882740020751953,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 1988
+    },
+    {
+      "epoch": 0.48300145701796987,
+      "grad_norm": 0.5687981843948364,
+      "learning_rate": 0.0001,
+      "loss": 1.8813,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4832442933462846,
+      "grad_norm": 0.5380786061286926,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 1990
+    },
+    {
+      "epoch": 0.48348712967459934,
+      "grad_norm": 0.5477362275123596,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 1991
+    },
+    {
+      "epoch": 0.483729966002914,
+      "grad_norm": 0.5760575532913208,
+      "learning_rate": 0.0001,
+      "loss": 1.8807,
+      "step": 1992
+    },
+    {
+      "epoch": 0.48397280233122875,
+      "grad_norm": 0.5555697083473206,
+      "learning_rate": 0.0001,
+      "loss": 1.701,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4842156386595435,
+      "grad_norm": 0.5567694306373596,
+      "learning_rate": 0.0001,
+      "loss": 1.8007,
+      "step": 1994
+    },
+    {
+      "epoch": 0.48445847498785816,
+      "grad_norm": 0.5891985297203064,
+      "learning_rate": 0.0001,
+      "loss": 1.7212,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4847013113161729,
+      "grad_norm": 0.62554532289505,
+      "learning_rate": 0.0001,
+      "loss": 1.8061,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4849441476444876,
+      "grad_norm": 0.6055716276168823,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 1997
+    },
+    {
+      "epoch": 0.48518698397280235,
+      "grad_norm": 0.5879396200180054,
+      "learning_rate": 0.0001,
+      "loss": 1.8191,
+      "step": 1998
+    },
+    {
+      "epoch": 0.48542982030111703,
+      "grad_norm": 0.5919262170791626,
+      "learning_rate": 0.0001,
+      "loss": 1.8164,
+      "step": 1999
+    },
+    {
+      "epoch": 0.48567265662943176,
+      "grad_norm": 0.5956177115440369,
+      "learning_rate": 0.0001,
+      "loss": 1.7904,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4859154929577465,
+      "grad_norm": 0.5396212935447693,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 2001
+    },
+    {
+      "epoch": 0.4861583292860612,
+      "grad_norm": 0.5972726941108704,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4864011656143759,
+      "grad_norm": 0.6242453455924988,
+      "learning_rate": 0.0001,
+      "loss": 1.7777,
+      "step": 2003
+    },
+    {
+      "epoch": 0.48664400194269064,
+      "grad_norm": 0.5980928540229797,
+      "learning_rate": 0.0001,
+      "loss": 1.8408,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4868868382710053,
+      "grad_norm": 0.5718754529953003,
+      "learning_rate": 0.0001,
+      "loss": 1.8969,
+      "step": 2005
+    },
+    {
+      "epoch": 0.48712967459932005,
+      "grad_norm": 0.5300527215003967,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4873725109276348,
+      "grad_norm": 0.5508219599723816,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4876153472559495,
+      "grad_norm": 0.5847325325012207,
+      "learning_rate": 0.0001,
+      "loss": 1.8637,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4878581835842642,
+      "grad_norm": 0.594556987285614,
+      "learning_rate": 0.0001,
+      "loss": 1.8994,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4881010199125789,
+      "grad_norm": 0.5872136950492859,
+      "learning_rate": 0.0001,
+      "loss": 1.8082,
+      "step": 2010
+    },
+    {
+      "epoch": 0.48834385624089366,
+      "grad_norm": 0.5755164623260498,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 2011
+    },
+    {
+      "epoch": 0.48858669256920834,
+      "grad_norm": 0.6021437048912048,
+      "learning_rate": 0.0001,
+      "loss": 1.817,
+      "step": 2012
+    },
+    {
+      "epoch": 0.48882952889752307,
+      "grad_norm": 0.5879169702529907,
+      "learning_rate": 0.0001,
+      "loss": 1.8699,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4890723652258378,
+      "grad_norm": 0.5833615660667419,
+      "learning_rate": 0.0001,
+      "loss": 1.7398,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4893152015541525,
+      "grad_norm": 0.5776499509811401,
+      "learning_rate": 0.0001,
+      "loss": 1.7552,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4895580378824672,
+      "grad_norm": 0.5966989398002625,
+      "learning_rate": 0.0001,
+      "loss": 1.9394,
+      "step": 2016
+    },
+    {
+      "epoch": 0.48980087421078194,
+      "grad_norm": 0.596585750579834,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4900437105390966,
+      "grad_norm": 0.6130557060241699,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 2018
+    },
+    {
+      "epoch": 0.49028654686741135,
+      "grad_norm": 0.5656613707542419,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4905293831957261,
+      "grad_norm": 0.5924555659294128,
+      "learning_rate": 0.0001,
+      "loss": 1.7926,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4907722195240408,
+      "grad_norm": 0.5761740207672119,
+      "learning_rate": 0.0001,
+      "loss": 1.8054,
+      "step": 2021
+    },
+    {
+      "epoch": 0.4910150558523555,
+      "grad_norm": 0.5992533564567566,
+      "learning_rate": 0.0001,
+      "loss": 1.8102,
+      "step": 2022
+    },
+    {
+      "epoch": 0.49125789218067023,
+      "grad_norm": 0.5910313129425049,
+      "learning_rate": 0.0001,
+      "loss": 1.8815,
+      "step": 2023
+    },
+    {
+      "epoch": 0.49150072850898496,
+      "grad_norm": 0.5652232766151428,
+      "learning_rate": 0.0001,
+      "loss": 1.729,
+      "step": 2024
+    },
+    {
+      "epoch": 0.49174356483729964,
+      "grad_norm": 0.5745852589607239,
+      "learning_rate": 0.0001,
+      "loss": 1.8081,
+      "step": 2025
+    },
+    {
+      "epoch": 0.4919864011656144,
+      "grad_norm": 0.5289776921272278,
+      "learning_rate": 0.0001,
+      "loss": 1.7566,
+      "step": 2026
+    },
+    {
+      "epoch": 0.4922292374939291,
+      "grad_norm": 0.5878112316131592,
+      "learning_rate": 0.0001,
+      "loss": 1.7305,
+      "step": 2027
+    },
+    {
+      "epoch": 0.4924720738222438,
+      "grad_norm": 0.5823032259941101,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4927149101505585,
+      "grad_norm": 0.5838625431060791,
+      "learning_rate": 0.0001,
+      "loss": 1.8703,
+      "step": 2029
+    },
+    {
+      "epoch": 0.49295774647887325,
+      "grad_norm": 0.578701913356781,
+      "learning_rate": 0.0001,
+      "loss": 1.7124,
+      "step": 2030
+    },
+    {
+      "epoch": 0.493200582807188,
+      "grad_norm": 0.5995233058929443,
+      "learning_rate": 0.0001,
+      "loss": 1.768,
+      "step": 2031
+    },
+    {
+      "epoch": 0.49344341913550266,
+      "grad_norm": 0.5891994833946228,
+      "learning_rate": 0.0001,
+      "loss": 1.8702,
+      "step": 2032
+    },
+    {
+      "epoch": 0.4936862554638174,
+      "grad_norm": 0.565297544002533,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4939290917921321,
+      "grad_norm": 0.563851535320282,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4941719281204468,
+      "grad_norm": 0.5984541177749634,
+      "learning_rate": 0.0001,
+      "loss": 1.9099,
+      "step": 2035
+    },
+    {
+      "epoch": 0.49441476444876153,
+      "grad_norm": 0.5796719789505005,
+      "learning_rate": 0.0001,
+      "loss": 1.7586,
+      "step": 2036
+    },
+    {
+      "epoch": 0.49465760077707627,
+      "grad_norm": 0.6082152128219604,
+      "learning_rate": 0.0001,
+      "loss": 1.8833,
+      "step": 2037
+    },
+    {
+      "epoch": 0.49490043710539094,
+      "grad_norm": 0.5553922057151794,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 2038
+    },
+    {
+      "epoch": 0.4951432734337057,
+      "grad_norm": 0.5671054720878601,
+      "learning_rate": 0.0001,
+      "loss": 1.7251,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4953861097620204,
+      "grad_norm": 0.6090965270996094,
+      "learning_rate": 0.0001,
+      "loss": 1.7635,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4956289460903351,
+      "grad_norm": 0.5869699120521545,
+      "learning_rate": 0.0001,
+      "loss": 1.7948,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4958717824186498,
+      "grad_norm": 0.5642756819725037,
+      "learning_rate": 0.0001,
+      "loss": 1.8111,
+      "step": 2042
+    },
+    {
+      "epoch": 0.49611461874696455,
+      "grad_norm": 0.6093323230743408,
+      "learning_rate": 0.0001,
+      "loss": 1.77,
+      "step": 2043
+    },
+    {
+      "epoch": 0.4963574550752793,
+      "grad_norm": 0.6150794625282288,
+      "learning_rate": 0.0001,
+      "loss": 1.7954,
+      "step": 2044
+    },
+    {
+      "epoch": 0.49660029140359396,
+      "grad_norm": 0.6348615884780884,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4968431277319087,
+      "grad_norm": 0.6043843626976013,
+      "learning_rate": 0.0001,
+      "loss": 1.7093,
+      "step": 2046
+    },
+    {
+      "epoch": 0.49708596406022343,
+      "grad_norm": 0.5615968108177185,
+      "learning_rate": 0.0001,
+      "loss": 1.7402,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4973288003885381,
+      "grad_norm": 0.552333414554596,
+      "learning_rate": 0.0001,
+      "loss": 1.6705,
+      "step": 2048
+    },
+    {
+      "epoch": 0.49757163671685284,
+      "grad_norm": 0.5569483637809753,
+      "learning_rate": 0.0001,
+      "loss": 1.7713,
+      "step": 2049
+    },
+    {
+      "epoch": 0.49781447304516757,
+      "grad_norm": 0.637229859828949,
+      "learning_rate": 0.0001,
+      "loss": 1.8701,
+      "step": 2050
+    },
+    {
+      "epoch": 0.49805730937348225,
+      "grad_norm": 0.5754112005233765,
+      "learning_rate": 0.0001,
+      "loss": 1.829,
+      "step": 2051
+    },
+    {
+      "epoch": 0.498300145701797,
+      "grad_norm": 0.5642938017845154,
+      "learning_rate": 0.0001,
+      "loss": 1.6451,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4985429820301117,
+      "grad_norm": 0.5582892894744873,
+      "learning_rate": 0.0001,
+      "loss": 1.5854,
+      "step": 2053
+    },
+    {
+      "epoch": 0.49878581835842645,
+      "grad_norm": 0.5798816084861755,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 2054
+    },
+    {
+      "epoch": 0.4990286546867411,
+      "grad_norm": 0.5920463800430298,
+      "learning_rate": 0.0001,
+      "loss": 1.6858,
+      "step": 2055
+    },
+    {
+      "epoch": 0.49927149101505586,
+      "grad_norm": 0.5674306154251099,
+      "learning_rate": 0.0001,
+      "loss": 1.6466,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4995143273433706,
+      "grad_norm": 0.5592432618141174,
+      "learning_rate": 0.0001,
+      "loss": 1.755,
+      "step": 2057
+    },
+    {
+      "epoch": 0.49975716367168527,
+      "grad_norm": 0.5635419487953186,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5989596843719482,
+      "learning_rate": 0.0001,
+      "loss": 1.7961,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5002428363283147,
+      "grad_norm": 0.5399906635284424,
+      "learning_rate": 0.0001,
+      "loss": 1.6511,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5004856726566295,
+      "grad_norm": 0.5654640793800354,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5007285089849441,
+      "grad_norm": 0.5929504036903381,
+      "learning_rate": 0.0001,
+      "loss": 1.783,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5009713453132588,
+      "grad_norm": 0.5613128542900085,
+      "learning_rate": 0.0001,
+      "loss": 1.7739,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5012141816415736,
+      "grad_norm": 0.5874308347702026,
+      "learning_rate": 0.0001,
+      "loss": 1.6707,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5014570179698883,
+      "grad_norm": 0.5515894889831543,
+      "learning_rate": 0.0001,
+      "loss": 1.7933,
+      "step": 2065
+    },
+    {
+      "epoch": 0.501699854298203,
+      "grad_norm": 0.5673650503158569,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5019426906265178,
+      "grad_norm": 0.5600554347038269,
+      "learning_rate": 0.0001,
+      "loss": 1.7876,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5021855269548324,
+      "grad_norm": 0.5556074380874634,
+      "learning_rate": 0.0001,
+      "loss": 1.7885,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5024283632831471,
+      "grad_norm": 0.5716009140014648,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5026711996114619,
+      "grad_norm": 0.5660889744758606,
+      "learning_rate": 0.0001,
+      "loss": 1.8036,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5029140359397766,
+      "grad_norm": 0.6146003007888794,
+      "learning_rate": 0.0001,
+      "loss": 1.746,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5031568722680914,
+      "grad_norm": 0.5939356088638306,
+      "learning_rate": 0.0001,
+      "loss": 1.7597,
+      "step": 2072
+    },
+    {
+      "epoch": 0.503399708596406,
+      "grad_norm": 0.5672742128372192,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5036425449247207,
+      "grad_norm": 0.5487650036811829,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5038853812530355,
+      "grad_norm": 0.6096526980400085,
+      "learning_rate": 0.0001,
+      "loss": 1.8213,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5041282175813502,
+      "grad_norm": 0.587897539138794,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5043710539096649,
+      "grad_norm": 0.5733286142349243,
+      "learning_rate": 0.0001,
+      "loss": 1.7982,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5046138902379796,
+      "grad_norm": 0.6002563834190369,
+      "learning_rate": 0.0001,
+      "loss": 1.7007,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5048567265662943,
+      "grad_norm": 0.5647587180137634,
+      "learning_rate": 0.0001,
+      "loss": 1.7711,
+      "step": 2079
+    },
+    {
+      "epoch": 0.505099562894609,
+      "grad_norm": 0.5878034830093384,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5053423992229238,
+      "grad_norm": 0.51972895860672,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5055852355512385,
+      "grad_norm": 0.5663842558860779,
+      "learning_rate": 0.0001,
+      "loss": 1.695,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5058280718795531,
+      "grad_norm": 0.6086951494216919,
+      "learning_rate": 0.0001,
+      "loss": 1.7639,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5060709082078679,
+      "grad_norm": 0.5852351784706116,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5063137445361826,
+      "grad_norm": 0.5884329676628113,
+      "learning_rate": 0.0001,
+      "loss": 1.7277,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5065565808644973,
+      "grad_norm": 0.5802691578865051,
+      "learning_rate": 0.0001,
+      "loss": 1.7907,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5067994171928121,
+      "grad_norm": 0.6157693266868591,
+      "learning_rate": 0.0001,
+      "loss": 1.7991,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5070422535211268,
+      "grad_norm": 0.5663011074066162,
+      "learning_rate": 0.0001,
+      "loss": 1.7187,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5072850898494414,
+      "grad_norm": 0.5462061762809753,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5075279261777562,
+      "grad_norm": 0.6452285051345825,
+      "learning_rate": 0.0001,
+      "loss": 1.7736,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5077707625060709,
+      "grad_norm": 0.6200202703475952,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5080135988343856,
+      "grad_norm": 0.5755290389060974,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5082564351627004,
+      "grad_norm": 0.5852632522583008,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 2093
+    },
+    {
+      "epoch": 0.508499271491015,
+      "grad_norm": 0.561608612537384,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5087421078193298,
+      "grad_norm": 0.5299578309059143,
+      "learning_rate": 0.0001,
+      "loss": 1.5823,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5089849441476445,
+      "grad_norm": 0.5728558897972107,
+      "learning_rate": 0.0001,
+      "loss": 1.7869,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5092277804759592,
+      "grad_norm": 0.5560303926467896,
+      "learning_rate": 0.0001,
+      "loss": 1.7693,
+      "step": 2097
+    },
+    {
+      "epoch": 0.509470616804274,
+      "grad_norm": 0.6147969365119934,
+      "learning_rate": 0.0001,
+      "loss": 1.6741,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5097134531325886,
+      "grad_norm": 0.5178264379501343,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5099562894609033,
+      "grad_norm": 0.550970196723938,
+      "learning_rate": 0.0001,
+      "loss": 1.499,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5101991257892181,
+      "grad_norm": 0.5818623900413513,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5104419621175328,
+      "grad_norm": 0.5710617899894714,
+      "learning_rate": 0.0001,
+      "loss": 1.7024,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5106847984458475,
+      "grad_norm": 0.629630982875824,
+      "learning_rate": 0.0001,
+      "loss": 1.9133,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5109276347741623,
+      "grad_norm": 0.558758556842804,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5111704711024769,
+      "grad_norm": 0.5871068239212036,
+      "learning_rate": 0.0001,
+      "loss": 1.7599,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5114133074307916,
+      "grad_norm": 0.5836141109466553,
+      "learning_rate": 0.0001,
+      "loss": 1.8942,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5116561437591064,
+      "grad_norm": 0.563770055770874,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5118989800874211,
+      "grad_norm": 0.5678231120109558,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5121418164157358,
+      "grad_norm": 0.5830565094947815,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5123846527440505,
+      "grad_norm": 0.5920232534408569,
+      "learning_rate": 0.0001,
+      "loss": 1.8429,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5126274890723652,
+      "grad_norm": 0.5994880199432373,
+      "learning_rate": 0.0001,
+      "loss": 1.7464,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5128703254006799,
+      "grad_norm": 0.5636540651321411,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5131131617289947,
+      "grad_norm": 0.5844842195510864,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5133559980573094,
+      "grad_norm": 0.5925378203392029,
+      "learning_rate": 0.0001,
+      "loss": 1.8161,
+      "step": 2114
+    },
+    {
+      "epoch": 0.513598834385624,
+      "grad_norm": 0.5761093497276306,
+      "learning_rate": 0.0001,
+      "loss": 1.5473,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5138416707139388,
+      "grad_norm": 0.5909484028816223,
+      "learning_rate": 0.0001,
+      "loss": 1.8494,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5140845070422535,
+      "grad_norm": 0.575232982635498,
+      "learning_rate": 0.0001,
+      "loss": 1.6683,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5143273433705683,
+      "grad_norm": 0.608296811580658,
+      "learning_rate": 0.0001,
+      "loss": 2.0095,
+      "step": 2118
+    },
+    {
+      "epoch": 0.514570179698883,
+      "grad_norm": 0.5402919054031372,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5148130160271976,
+      "grad_norm": 0.6317681074142456,
+      "learning_rate": 0.0001,
+      "loss": 1.8573,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5150558523555124,
+      "grad_norm": 0.545809268951416,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5152986886838271,
+      "grad_norm": 0.5815280675888062,
+      "learning_rate": 0.0001,
+      "loss": 1.7033,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5155415250121418,
+      "grad_norm": 0.6006367206573486,
+      "learning_rate": 0.0001,
+      "loss": 1.8273,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5157843613404566,
+      "grad_norm": 0.6038702130317688,
+      "learning_rate": 0.0001,
+      "loss": 1.7826,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5160271976687713,
+      "grad_norm": 0.5876898169517517,
+      "learning_rate": 0.0001,
+      "loss": 1.7943,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5162700339970859,
+      "grad_norm": 0.5861945152282715,
+      "learning_rate": 0.0001,
+      "loss": 1.8508,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5165128703254007,
+      "grad_norm": 0.5722491145133972,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5167557066537154,
+      "grad_norm": 0.6067075729370117,
+      "learning_rate": 0.0001,
+      "loss": 1.8338,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5169985429820301,
+      "grad_norm": 0.5551809668540955,
+      "learning_rate": 0.0001,
+      "loss": 1.7478,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 0.602811336517334,
+      "learning_rate": 0.0001,
+      "loss": 1.7956,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5174842156386595,
+      "grad_norm": 0.5505642890930176,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5177270519669742,
+      "grad_norm": 0.5739842057228088,
+      "learning_rate": 0.0001,
+      "loss": 1.7405,
+      "step": 2132
+    },
+    {
+      "epoch": 0.517969888295289,
+      "grad_norm": 0.6024134755134583,
+      "learning_rate": 0.0001,
+      "loss": 1.8489,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5182127246236037,
+      "grad_norm": 0.533274233341217,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5184555609519184,
+      "grad_norm": 0.5912434458732605,
+      "learning_rate": 0.0001,
+      "loss": 1.7793,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5186983972802331,
+      "grad_norm": 0.5838918089866638,
+      "learning_rate": 0.0001,
+      "loss": 1.824,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5189412336085478,
+      "grad_norm": 0.5628232359886169,
+      "learning_rate": 0.0001,
+      "loss": 1.7335,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5191840699368625,
+      "grad_norm": 0.571304202079773,
+      "learning_rate": 0.0001,
+      "loss": 1.8103,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5194269062651773,
+      "grad_norm": 0.5905882120132446,
+      "learning_rate": 0.0001,
+      "loss": 1.8236,
+      "step": 2139
+    },
+    {
+      "epoch": 0.519669742593492,
+      "grad_norm": 0.5618475079536438,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5199125789218068,
+      "grad_norm": 0.6003167629241943,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5201554152501214,
+      "grad_norm": 0.6050346493721008,
+      "learning_rate": 0.0001,
+      "loss": 1.7486,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5203982515784361,
+      "grad_norm": 0.5786466598510742,
+      "learning_rate": 0.0001,
+      "loss": 1.7552,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5206410879067509,
+      "grad_norm": 0.5782291889190674,
+      "learning_rate": 0.0001,
+      "loss": 1.7007,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5208839242350656,
+      "grad_norm": 0.6254963278770447,
+      "learning_rate": 0.0001,
+      "loss": 1.6921,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5211267605633803,
+      "grad_norm": 0.6038153767585754,
+      "learning_rate": 0.0001,
+      "loss": 1.7569,
+      "step": 2146
+    },
+    {
+      "epoch": 0.521369596891695,
+      "grad_norm": 0.6252439022064209,
+      "learning_rate": 0.0001,
+      "loss": 1.831,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5216124332200097,
+      "grad_norm": 0.5462331175804138,
+      "learning_rate": 0.0001,
+      "loss": 1.6698,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5218552695483244,
+      "grad_norm": 0.569652259349823,
+      "learning_rate": 0.0001,
+      "loss": 1.7232,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5220981058766392,
+      "grad_norm": 0.6135126352310181,
+      "learning_rate": 0.0001,
+      "loss": 1.8183,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5223409422049539,
+      "grad_norm": 0.6126598715782166,
+      "learning_rate": 0.0001,
+      "loss": 1.768,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5225837785332685,
+      "grad_norm": 0.5486062169075012,
+      "learning_rate": 0.0001,
+      "loss": 1.5589,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5228266148615833,
+      "grad_norm": 0.5582675933837891,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 2153
+    },
+    {
+      "epoch": 0.523069451189898,
+      "grad_norm": 0.6054219603538513,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5233122875182127,
+      "grad_norm": 0.5597957372665405,
+      "learning_rate": 0.0001,
+      "loss": 1.7998,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5235551238465275,
+      "grad_norm": 0.5427129864692688,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5237979601748421,
+      "grad_norm": 0.57682865858078,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5240407965031568,
+      "grad_norm": 0.5827271938323975,
+      "learning_rate": 0.0001,
+      "loss": 1.8182,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5242836328314716,
+      "grad_norm": 0.586132287979126,
+      "learning_rate": 0.0001,
+      "loss": 1.7707,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5245264691597863,
+      "grad_norm": 0.5706914663314819,
+      "learning_rate": 0.0001,
+      "loss": 1.6457,
+      "step": 2160
+    },
+    {
+      "epoch": 0.524769305488101,
+      "grad_norm": 0.5372036695480347,
+      "learning_rate": 0.0001,
+      "loss": 1.676,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5250121418164158,
+      "grad_norm": 0.5486818552017212,
+      "learning_rate": 0.0001,
+      "loss": 1.6701,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5252549781447304,
+      "grad_norm": 0.5342246294021606,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5254978144730452,
+      "grad_norm": 0.5594509243965149,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5257406508013599,
+      "grad_norm": 0.6321483850479126,
+      "learning_rate": 0.0001,
+      "loss": 1.8696,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5259834871296746,
+      "grad_norm": 0.5574361085891724,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5262263234579894,
+      "grad_norm": 0.6299461126327515,
+      "learning_rate": 0.0001,
+      "loss": 1.905,
+      "step": 2167
+    },
+    {
+      "epoch": 0.526469159786304,
+      "grad_norm": 0.5777122378349304,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5267119961146187,
+      "grad_norm": 0.5658559799194336,
+      "learning_rate": 0.0001,
+      "loss": 1.7185,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5269548324429335,
+      "grad_norm": 0.5544305443763733,
+      "learning_rate": 0.0001,
+      "loss": 1.7855,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5271976687712482,
+      "grad_norm": 0.5999164581298828,
+      "learning_rate": 0.0001,
+      "loss": 1.823,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5274405050995629,
+      "grad_norm": 0.5797856450080872,
+      "learning_rate": 0.0001,
+      "loss": 1.7401,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5276833414278777,
+      "grad_norm": 0.59946209192276,
+      "learning_rate": 0.0001,
+      "loss": 1.8566,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5279261777561923,
+      "grad_norm": 0.572341799736023,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 2174
+    },
+    {
+      "epoch": 0.528169014084507,
+      "grad_norm": 0.537335991859436,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5284118504128218,
+      "grad_norm": 0.5918770432472229,
+      "learning_rate": 0.0001,
+      "loss": 1.7216,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5286546867411365,
+      "grad_norm": 0.5716724991798401,
+      "learning_rate": 0.0001,
+      "loss": 1.7665,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5288975230694511,
+      "grad_norm": 0.5834047198295593,
+      "learning_rate": 0.0001,
+      "loss": 1.6417,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5291403593977659,
+      "grad_norm": 0.6267532110214233,
+      "learning_rate": 0.0001,
+      "loss": 1.99,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5293831957260806,
+      "grad_norm": 0.6181809306144714,
+      "learning_rate": 0.0001,
+      "loss": 1.9405,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5296260320543953,
+      "grad_norm": 0.5811281800270081,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5298688683827101,
+      "grad_norm": 0.5770442485809326,
+      "learning_rate": 0.0001,
+      "loss": 1.8281,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5301117047110248,
+      "grad_norm": 0.5526422262191772,
+      "learning_rate": 0.0001,
+      "loss": 1.7686,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5303545410393394,
+      "grad_norm": 0.579846978187561,
+      "learning_rate": 0.0001,
+      "loss": 1.8797,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5305973773676542,
+      "grad_norm": 0.5629196166992188,
+      "learning_rate": 0.0001,
+      "loss": 1.728,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5308402136959689,
+      "grad_norm": 0.554466962814331,
+      "learning_rate": 0.0001,
+      "loss": 1.5575,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5310830500242837,
+      "grad_norm": 0.6198550462722778,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5313258863525984,
+      "grad_norm": 0.5318722724914551,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 2188
+    },
+    {
+      "epoch": 0.531568722680913,
+      "grad_norm": 0.6216217875480652,
+      "learning_rate": 0.0001,
+      "loss": 1.8666,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5318115590092278,
+      "grad_norm": 0.5881581902503967,
+      "learning_rate": 0.0001,
+      "loss": 1.783,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5320543953375425,
+      "grad_norm": 0.5604850649833679,
+      "learning_rate": 0.0001,
+      "loss": 1.68,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5322972316658572,
+      "grad_norm": 0.6313543319702148,
+      "learning_rate": 0.0001,
+      "loss": 1.9791,
+      "step": 2192
+    },
+    {
+      "epoch": 0.532540067994172,
+      "grad_norm": 0.5774602890014648,
+      "learning_rate": 0.0001,
+      "loss": 1.8134,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5327829043224867,
+      "grad_norm": 0.6007572412490845,
+      "learning_rate": 0.0001,
+      "loss": 1.8626,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5330257406508013,
+      "grad_norm": 0.5974249839782715,
+      "learning_rate": 0.0001,
+      "loss": 1.8335,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5332685769791161,
+      "grad_norm": 0.5923995971679688,
+      "learning_rate": 0.0001,
+      "loss": 1.9244,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5335114133074308,
+      "grad_norm": 0.5441774725914001,
+      "learning_rate": 0.0001,
+      "loss": 1.7939,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5337542496357455,
+      "grad_norm": 0.580784022808075,
+      "learning_rate": 0.0001,
+      "loss": 1.7619,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5339970859640603,
+      "grad_norm": 0.5619747638702393,
+      "learning_rate": 0.0001,
+      "loss": 1.8213,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5342399222923749,
+      "grad_norm": 0.5800536274909973,
+      "learning_rate": 0.0001,
+      "loss": 1.9446,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5344827586206896,
+      "grad_norm": 0.6046760678291321,
+      "learning_rate": 0.0001,
+      "loss": 1.7507,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5347255949490044,
+      "grad_norm": 0.5831290483474731,
+      "learning_rate": 0.0001,
+      "loss": 1.8288,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5349684312773191,
+      "grad_norm": 0.5606281757354736,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5352112676056338,
+      "grad_norm": 0.5448996424674988,
+      "learning_rate": 0.0001,
+      "loss": 1.6451,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5354541039339485,
+      "grad_norm": 0.5596613883972168,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5356969402622632,
+      "grad_norm": 0.5989713072776794,
+      "learning_rate": 0.0001,
+      "loss": 1.8789,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5359397765905779,
+      "grad_norm": 0.6066573858261108,
+      "learning_rate": 0.0001,
+      "loss": 1.4983,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5361826129188927,
+      "grad_norm": 0.5666881799697876,
+      "learning_rate": 0.0001,
+      "loss": 1.6933,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5364254492472074,
+      "grad_norm": 0.5524195432662964,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 2209
+    },
+    {
+      "epoch": 0.5366682855755222,
+      "grad_norm": 0.6427804231643677,
+      "learning_rate": 0.0001,
+      "loss": 1.921,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5369111219038368,
+      "grad_norm": 0.6162734627723694,
+      "learning_rate": 0.0001,
+      "loss": 1.8159,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5371539582321515,
+      "grad_norm": 0.55301833152771,
+      "learning_rate": 0.0001,
+      "loss": 1.7398,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5373967945604663,
+      "grad_norm": 0.5881041288375854,
+      "learning_rate": 0.0001,
+      "loss": 1.6758,
+      "step": 2213
+    },
+    {
+      "epoch": 0.537639630888781,
+      "grad_norm": 0.6024550199508667,
+      "learning_rate": 0.0001,
+      "loss": 1.8827,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5378824672170957,
+      "grad_norm": 0.515755295753479,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5381253035454104,
+      "grad_norm": 0.6048948168754578,
+      "learning_rate": 0.0001,
+      "loss": 1.7744,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5383681398737251,
+      "grad_norm": 0.5852948427200317,
+      "learning_rate": 0.0001,
+      "loss": 1.7616,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5386109762020398,
+      "grad_norm": 0.5605524182319641,
+      "learning_rate": 0.0001,
+      "loss": 1.6569,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5388538125303546,
+      "grad_norm": 0.5623579621315002,
+      "learning_rate": 0.0001,
+      "loss": 1.6597,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5390966488586693,
+      "grad_norm": 0.5543200373649597,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5393394851869839,
+      "grad_norm": 0.5789135098457336,
+      "learning_rate": 0.0001,
+      "loss": 1.6436,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5395823215152987,
+      "grad_norm": 0.574054479598999,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5398251578436134,
+      "grad_norm": 0.615116536617279,
+      "learning_rate": 0.0001,
+      "loss": 1.8487,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5400679941719281,
+      "grad_norm": 0.5967799425125122,
+      "learning_rate": 0.0001,
+      "loss": 1.8072,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5403108305002429,
+      "grad_norm": 0.5706263780593872,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5405536668285575,
+      "grad_norm": 0.6475127339363098,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5407965031568722,
+      "grad_norm": 0.6100888848304749,
+      "learning_rate": 0.0001,
+      "loss": 1.8501,
+      "step": 2227
+    },
+    {
+      "epoch": 0.541039339485187,
+      "grad_norm": 0.57435542345047,
+      "learning_rate": 0.0001,
+      "loss": 1.7222,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5412821758135017,
+      "grad_norm": 0.6083694100379944,
+      "learning_rate": 0.0001,
+      "loss": 1.6986,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5415250121418164,
+      "grad_norm": 0.5984795093536377,
+      "learning_rate": 0.0001,
+      "loss": 1.8236,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5417678484701312,
+      "grad_norm": 0.5543849468231201,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5420106847984458,
+      "grad_norm": 0.5587611794471741,
+      "learning_rate": 0.0001,
+      "loss": 1.771,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5422535211267606,
+      "grad_norm": 0.5826887488365173,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5424963574550753,
+      "grad_norm": 0.5647484660148621,
+      "learning_rate": 0.0001,
+      "loss": 1.8202,
+      "step": 2234
+    },
+    {
+      "epoch": 0.54273919378339,
+      "grad_norm": 0.5916162133216858,
+      "learning_rate": 0.0001,
+      "loss": 1.7418,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5429820301117048,
+      "grad_norm": 0.5663244724273682,
+      "learning_rate": 0.0001,
+      "loss": 1.8817,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5432248664400194,
+      "grad_norm": 0.6044777631759644,
+      "learning_rate": 0.0001,
+      "loss": 1.8875,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5434677027683341,
+      "grad_norm": 0.5563994646072388,
+      "learning_rate": 0.0001,
+      "loss": 1.6417,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5437105390966489,
+      "grad_norm": 0.5902057886123657,
+      "learning_rate": 0.0001,
+      "loss": 1.8409,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5439533754249636,
+      "grad_norm": 0.5353649854660034,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5441962117532783,
+      "grad_norm": 0.574381947517395,
+      "learning_rate": 0.0001,
+      "loss": 1.7432,
+      "step": 2241
+    },
+    {
+      "epoch": 0.544439048081593,
+      "grad_norm": 0.5697255730628967,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5446818844099077,
+      "grad_norm": 0.5618753433227539,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5449247207382224,
+      "grad_norm": 0.6283233761787415,
+      "learning_rate": 0.0001,
+      "loss": 1.849,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5451675570665372,
+      "grad_norm": 0.572043240070343,
+      "learning_rate": 0.0001,
+      "loss": 1.6105,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5454103933948519,
+      "grad_norm": 0.5859431624412537,
+      "learning_rate": 0.0001,
+      "loss": 1.7307,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5456532297231665,
+      "grad_norm": 0.5741754770278931,
+      "learning_rate": 0.0001,
+      "loss": 1.6827,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5458960660514813,
+      "grad_norm": 0.5581367611885071,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 2248
+    },
+    {
+      "epoch": 0.546138902379796,
+      "grad_norm": 0.5916844606399536,
+      "learning_rate": 0.0001,
+      "loss": 1.8795,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5463817387081107,
+      "grad_norm": 0.5180323123931885,
+      "learning_rate": 0.0001,
+      "loss": 1.5186,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5466245750364255,
+      "grad_norm": 0.516656219959259,
+      "learning_rate": 0.0001,
+      "loss": 1.4862,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5468674113647402,
+      "grad_norm": 0.5897921919822693,
+      "learning_rate": 0.0001,
+      "loss": 1.6883,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5471102476930548,
+      "grad_norm": 0.5344215035438538,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5473530840213696,
+      "grad_norm": 0.5945500731468201,
+      "learning_rate": 0.0001,
+      "loss": 1.7989,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5475959203496843,
+      "grad_norm": 0.5791717767715454,
+      "learning_rate": 0.0001,
+      "loss": 1.7545,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5478387566779991,
+      "grad_norm": 0.591839075088501,
+      "learning_rate": 0.0001,
+      "loss": 1.8236,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5480815930063138,
+      "grad_norm": 0.5432397127151489,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5483244293346284,
+      "grad_norm": 0.5551950335502625,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5485672656629432,
+      "grad_norm": 0.5756182670593262,
+      "learning_rate": 0.0001,
+      "loss": 1.7265,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5488101019912579,
+      "grad_norm": 0.5714510679244995,
+      "learning_rate": 0.0001,
+      "loss": 1.7944,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5490529383195726,
+      "grad_norm": 0.5553797483444214,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 2261
+    },
+    {
+      "epoch": 0.5492957746478874,
+      "grad_norm": 0.561791718006134,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 2262
+    },
+    {
+      "epoch": 0.549538610976202,
+      "grad_norm": 0.548929750919342,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5497814473045167,
+      "grad_norm": 0.6032873392105103,
+      "learning_rate": 0.0001,
+      "loss": 1.8544,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5500242836328315,
+      "grad_norm": 0.6131521463394165,
+      "learning_rate": 0.0001,
+      "loss": 1.7348,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5502671199611462,
+      "grad_norm": 0.6067385673522949,
+      "learning_rate": 0.0001,
+      "loss": 1.7989,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5505099562894609,
+      "grad_norm": 0.5752245783805847,
+      "learning_rate": 0.0001,
+      "loss": 1.7589,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5507527926177757,
+      "grad_norm": 0.6058968305587769,
+      "learning_rate": 0.0001,
+      "loss": 1.7689,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5509956289460903,
+      "grad_norm": 0.522637665271759,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 2269
+    },
+    {
+      "epoch": 0.551238465274405,
+      "grad_norm": 0.5843346118927002,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5514813016027198,
+      "grad_norm": 0.5794123411178589,
+      "learning_rate": 0.0001,
+      "loss": 1.7654,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 0.5385357737541199,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5519669742593492,
+      "grad_norm": 0.5631776452064514,
+      "learning_rate": 0.0001,
+      "loss": 1.7648,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5522098105876639,
+      "grad_norm": 0.5676831007003784,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5524526469159786,
+      "grad_norm": 0.5649113655090332,
+      "learning_rate": 0.0001,
+      "loss": 1.7205,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5526954832442933,
+      "grad_norm": 0.6241009831428528,
+      "learning_rate": 0.0001,
+      "loss": 1.6669,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5529383195726081,
+      "grad_norm": 0.5783674120903015,
+      "learning_rate": 0.0001,
+      "loss": 1.7265,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5531811559009228,
+      "grad_norm": 0.5854872465133667,
+      "learning_rate": 0.0001,
+      "loss": 1.8464,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5534239922292375,
+      "grad_norm": 0.5910096168518066,
+      "learning_rate": 0.0001,
+      "loss": 1.7909,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5536668285575522,
+      "grad_norm": 0.6107466220855713,
+      "learning_rate": 0.0001,
+      "loss": 1.8017,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5539096648858669,
+      "grad_norm": 0.5732429623603821,
+      "learning_rate": 0.0001,
+      "loss": 1.7704,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5541525012141817,
+      "grad_norm": 0.5652581453323364,
+      "learning_rate": 0.0001,
+      "loss": 1.6169,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5543953375424964,
+      "grad_norm": 0.5679978728294373,
+      "learning_rate": 0.0001,
+      "loss": 1.6261,
+      "step": 2283
+    },
+    {
+      "epoch": 0.554638173870811,
+      "grad_norm": 0.5831108689308167,
+      "learning_rate": 0.0001,
+      "loss": 1.7858,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5548810101991258,
+      "grad_norm": 0.6095585227012634,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5551238465274405,
+      "grad_norm": 0.5800433158874512,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5553666828557552,
+      "grad_norm": 0.5661283731460571,
+      "learning_rate": 0.0001,
+      "loss": 1.7831,
+      "step": 2287
+    },
+    {
+      "epoch": 0.55560951918407,
+      "grad_norm": 0.5875678658485413,
+      "learning_rate": 0.0001,
+      "loss": 1.8227,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5558523555123847,
+      "grad_norm": 0.5663167834281921,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5560951918406993,
+      "grad_norm": 0.5791200995445251,
+      "learning_rate": 0.0001,
+      "loss": 1.7324,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5563380281690141,
+      "grad_norm": 0.5939375162124634,
+      "learning_rate": 0.0001,
+      "loss": 1.8266,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5565808644973288,
+      "grad_norm": 0.5692607164382935,
+      "learning_rate": 0.0001,
+      "loss": 1.7979,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5568237008256435,
+      "grad_norm": 0.5719442963600159,
+      "learning_rate": 0.0001,
+      "loss": 1.7008,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5570665371539583,
+      "grad_norm": 0.6055617332458496,
+      "learning_rate": 0.0001,
+      "loss": 1.8149,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5573093734822729,
+      "grad_norm": 0.5880705118179321,
+      "learning_rate": 0.0001,
+      "loss": 1.729,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5575522098105876,
+      "grad_norm": 0.6210213303565979,
+      "learning_rate": 0.0001,
+      "loss": 1.6504,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5577950461389024,
+      "grad_norm": 0.6164421439170837,
+      "learning_rate": 0.0001,
+      "loss": 1.8222,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5580378824672171,
+      "grad_norm": 0.5638343095779419,
+      "learning_rate": 0.0001,
+      "loss": 1.6657,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5582807187955318,
+      "grad_norm": 0.5995288491249084,
+      "learning_rate": 0.0001,
+      "loss": 1.9264,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5585235551238465,
+      "grad_norm": 0.6023948788642883,
+      "learning_rate": 0.0001,
+      "loss": 1.8343,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5587663914521612,
+      "grad_norm": 0.6314024329185486,
+      "learning_rate": 0.0001,
+      "loss": 1.7337,
+      "step": 2301
+    },
+    {
+      "epoch": 0.559009227780476,
+      "grad_norm": 0.5966470241546631,
+      "learning_rate": 0.0001,
+      "loss": 1.743,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5592520641087907,
+      "grad_norm": 0.5859267115592957,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5594949004371054,
+      "grad_norm": 0.5759007334709167,
+      "learning_rate": 0.0001,
+      "loss": 1.7801,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5597377367654202,
+      "grad_norm": 0.604795515537262,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5599805730937348,
+      "grad_norm": 0.5576354265213013,
+      "learning_rate": 0.0001,
+      "loss": 1.6984,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5602234094220495,
+      "grad_norm": 0.5717795491218567,
+      "learning_rate": 0.0001,
+      "loss": 1.7677,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5604662457503643,
+      "grad_norm": 0.6093893051147461,
+      "learning_rate": 0.0001,
+      "loss": 1.6873,
+      "step": 2308
+    },
+    {
+      "epoch": 0.560709082078679,
+      "grad_norm": 0.5409795641899109,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5609519184069937,
+      "grad_norm": 0.5724250078201294,
+      "learning_rate": 0.0001,
+      "loss": 1.7439,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5611947547353084,
+      "grad_norm": 0.5454828143119812,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5614375910636231,
+      "grad_norm": 0.5908055305480957,
+      "learning_rate": 0.0001,
+      "loss": 1.7521,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5616804273919378,
+      "grad_norm": 0.5490610003471375,
+      "learning_rate": 0.0001,
+      "loss": 1.772,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5619232637202526,
+      "grad_norm": 0.5860586762428284,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5621661000485673,
+      "grad_norm": 0.5701284408569336,
+      "learning_rate": 0.0001,
+      "loss": 1.7656,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5624089363768819,
+      "grad_norm": 0.6234085559844971,
+      "learning_rate": 0.0001,
+      "loss": 1.9436,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5626517727051967,
+      "grad_norm": 0.6046967506408691,
+      "learning_rate": 0.0001,
+      "loss": 1.8797,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5628946090335114,
+      "grad_norm": 0.532106339931488,
+      "learning_rate": 0.0001,
+      "loss": 1.6235,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5631374453618261,
+      "grad_norm": 0.6197608113288879,
+      "learning_rate": 0.0001,
+      "loss": 1.782,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5633802816901409,
+      "grad_norm": 0.5382350087165833,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5636231180184555,
+      "grad_norm": 0.593874454498291,
+      "learning_rate": 0.0001,
+      "loss": 1.7379,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5638659543467702,
+      "grad_norm": 0.5685871243476868,
+      "learning_rate": 0.0001,
+      "loss": 1.5871,
+      "step": 2322
+    },
+    {
+      "epoch": 0.564108790675085,
+      "grad_norm": 0.5940852761268616,
+      "learning_rate": 0.0001,
+      "loss": 1.7865,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5643516270033997,
+      "grad_norm": 0.5474042296409607,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5645944633317144,
+      "grad_norm": 0.5839331150054932,
+      "learning_rate": 0.0001,
+      "loss": 1.7634,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5648372996600292,
+      "grad_norm": 0.6073717474937439,
+      "learning_rate": 0.0001,
+      "loss": 1.8117,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5650801359883438,
+      "grad_norm": 0.5777013897895813,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5653229723166586,
+      "grad_norm": 0.6037007570266724,
+      "learning_rate": 0.0001,
+      "loss": 1.8422,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5655658086449733,
+      "grad_norm": 0.5606606006622314,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 2329
+    },
+    {
+      "epoch": 0.565808644973288,
+      "grad_norm": 0.5418534874916077,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5660514813016028,
+      "grad_norm": 0.608181357383728,
+      "learning_rate": 0.0001,
+      "loss": 1.844,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5662943176299174,
+      "grad_norm": 0.59039306640625,
+      "learning_rate": 0.0001,
+      "loss": 1.6876,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5665371539582321,
+      "grad_norm": 0.5667287111282349,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5667799902865469,
+      "grad_norm": 0.5706629753112793,
+      "learning_rate": 0.0001,
+      "loss": 1.6236,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5670228266148616,
+      "grad_norm": 0.6076309084892273,
+      "learning_rate": 0.0001,
+      "loss": 1.746,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5672656629431763,
+      "grad_norm": 0.592994749546051,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 2336
+    },
+    {
+      "epoch": 0.567508499271491,
+      "grad_norm": 0.5950530767440796,
+      "learning_rate": 0.0001,
+      "loss": 1.7201,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5677513355998057,
+      "grad_norm": 0.5949794054031372,
+      "learning_rate": 0.0001,
+      "loss": 1.7501,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5679941719281204,
+      "grad_norm": 0.5638046264648438,
+      "learning_rate": 0.0001,
+      "loss": 1.7436,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5682370082564352,
+      "grad_norm": 0.5714852809906006,
+      "learning_rate": 0.0001,
+      "loss": 1.7623,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5684798445847499,
+      "grad_norm": 0.6423383355140686,
+      "learning_rate": 0.0001,
+      "loss": 1.7754,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5687226809130645,
+      "grad_norm": 0.6077386140823364,
+      "learning_rate": 0.0001,
+      "loss": 1.8693,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5689655172413793,
+      "grad_norm": 0.5809692740440369,
+      "learning_rate": 0.0001,
+      "loss": 1.728,
+      "step": 2343
+    },
+    {
+      "epoch": 0.569208353569694,
+      "grad_norm": 0.5595994591712952,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5694511898980087,
+      "grad_norm": 0.5897711515426636,
+      "learning_rate": 0.0001,
+      "loss": 1.935,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5696940262263235,
+      "grad_norm": 0.5622484683990479,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5699368625546382,
+      "grad_norm": 0.5812405347824097,
+      "learning_rate": 0.0001,
+      "loss": 1.7894,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5701796988829528,
+      "grad_norm": 0.5791590213775635,
+      "learning_rate": 0.0001,
+      "loss": 1.7274,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5704225352112676,
+      "grad_norm": 0.5815815329551697,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5706653715395823,
+      "grad_norm": 0.6123338937759399,
+      "learning_rate": 0.0001,
+      "loss": 1.7184,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5709082078678971,
+      "grad_norm": 0.5597078204154968,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5711510441962118,
+      "grad_norm": 0.5812671184539795,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5713938805245264,
+      "grad_norm": 0.587066650390625,
+      "learning_rate": 0.0001,
+      "loss": 1.7552,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5716367168528412,
+      "grad_norm": 0.5483470559120178,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5718795531811559,
+      "grad_norm": 0.5875955820083618,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5721223895094706,
+      "grad_norm": 0.5432369112968445,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5723652258377854,
+      "grad_norm": 0.5850716233253479,
+      "learning_rate": 0.0001,
+      "loss": 1.8068,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5726080621661,
+      "grad_norm": 0.5770143866539001,
+      "learning_rate": 0.0001,
+      "loss": 1.7533,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5728508984944147,
+      "grad_norm": 0.5965167880058289,
+      "learning_rate": 0.0001,
+      "loss": 1.7634,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5730937348227295,
+      "grad_norm": 0.5683436989784241,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5733365711510442,
+      "grad_norm": 0.5722373723983765,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5735794074793589,
+      "grad_norm": 0.5822277069091797,
+      "learning_rate": 0.0001,
+      "loss": 1.7077,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5738222438076737,
+      "grad_norm": 0.5640177130699158,
+      "learning_rate": 0.0001,
+      "loss": 1.7373,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5740650801359883,
+      "grad_norm": 0.5672557950019836,
+      "learning_rate": 0.0001,
+      "loss": 1.6969,
+      "step": 2364
+    },
+    {
+      "epoch": 0.574307916464303,
+      "grad_norm": 0.5683900117874146,
+      "learning_rate": 0.0001,
+      "loss": 1.7601,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5745507527926178,
+      "grad_norm": 0.5778457522392273,
+      "learning_rate": 0.0001,
+      "loss": 1.7324,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5747935891209325,
+      "grad_norm": 0.5638473629951477,
+      "learning_rate": 0.0001,
+      "loss": 1.7839,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5750364254492472,
+      "grad_norm": 0.5594433546066284,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 2368
+    },
+    {
+      "epoch": 0.575279261777562,
+      "grad_norm": 0.5993980169296265,
+      "learning_rate": 0.0001,
+      "loss": 1.7092,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5755220981058766,
+      "grad_norm": 0.6353656649589539,
+      "learning_rate": 0.0001,
+      "loss": 1.845,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5757649344341913,
+      "grad_norm": 0.5711445808410645,
+      "learning_rate": 0.0001,
+      "loss": 1.6876,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5760077707625061,
+      "grad_norm": 0.5684404969215393,
+      "learning_rate": 0.0001,
+      "loss": 1.7879,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5762506070908208,
+      "grad_norm": 0.5605738759040833,
+      "learning_rate": 0.0001,
+      "loss": 1.7602,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5764934434191356,
+      "grad_norm": 0.5888994932174683,
+      "learning_rate": 0.0001,
+      "loss": 1.8354,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5767362797474502,
+      "grad_norm": 0.5309900641441345,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5769791160757649,
+      "grad_norm": 0.630192756652832,
+      "learning_rate": 0.0001,
+      "loss": 1.8859,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5772219524040797,
+      "grad_norm": 0.5861800312995911,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5774647887323944,
+      "grad_norm": 0.578201949596405,
+      "learning_rate": 0.0001,
+      "loss": 1.7953,
+      "step": 2378
+    },
+    {
+      "epoch": 0.577707625060709,
+      "grad_norm": 0.5666414499282837,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5779504613890238,
+      "grad_norm": 0.5398064255714417,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5781932977173385,
+      "grad_norm": 0.5318560004234314,
+      "learning_rate": 0.0001,
+      "loss": 1.4524,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5784361340456532,
+      "grad_norm": 0.5985003113746643,
+      "learning_rate": 0.0001,
+      "loss": 1.8397,
+      "step": 2382
+    },
+    {
+      "epoch": 0.578678970373968,
+      "grad_norm": 0.5568950772285461,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5789218067022827,
+      "grad_norm": 0.5775126814842224,
+      "learning_rate": 0.0001,
+      "loss": 1.7422,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5791646430305973,
+      "grad_norm": 0.5712935924530029,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5794074793589121,
+      "grad_norm": 0.5984039902687073,
+      "learning_rate": 0.0001,
+      "loss": 1.727,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5796503156872268,
+      "grad_norm": 0.5886059999465942,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5798931520155415,
+      "grad_norm": 0.5869618654251099,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5801359883438563,
+      "grad_norm": 0.6158575415611267,
+      "learning_rate": 0.0001,
+      "loss": 1.7997,
+      "step": 2389
+    },
+    {
+      "epoch": 0.580378824672171,
+      "grad_norm": 0.5518700480461121,
+      "learning_rate": 0.0001,
+      "loss": 1.6309,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5806216610004856,
+      "grad_norm": 0.5720113515853882,
+      "learning_rate": 0.0001,
+      "loss": 1.7135,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5808644973288004,
+      "grad_norm": 0.5739331245422363,
+      "learning_rate": 0.0001,
+      "loss": 1.6715,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5811073336571151,
+      "grad_norm": 0.5607396364212036,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5813501699854298,
+      "grad_norm": 0.5959517359733582,
+      "learning_rate": 0.0001,
+      "loss": 1.8453,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5815930063137446,
+      "grad_norm": 0.5785279273986816,
+      "learning_rate": 0.0001,
+      "loss": 1.7878,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5818358426420592,
+      "grad_norm": 0.5693400502204895,
+      "learning_rate": 0.0001,
+      "loss": 1.6801,
+      "step": 2396
+    },
+    {
+      "epoch": 0.582078678970374,
+      "grad_norm": 0.5761942267417908,
+      "learning_rate": 0.0001,
+      "loss": 1.7075,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5823215152986887,
+      "grad_norm": 0.5831998586654663,
+      "learning_rate": 0.0001,
+      "loss": 1.701,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5825643516270034,
+      "grad_norm": 0.5638182759284973,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5828071879553182,
+      "grad_norm": 0.5755643248558044,
+      "learning_rate": 0.0001,
+      "loss": 1.72,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5830500242836328,
+      "grad_norm": 0.5585417747497559,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5832928606119475,
+      "grad_norm": 0.5857945084571838,
+      "learning_rate": 0.0001,
+      "loss": 1.7711,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5835356969402623,
+      "grad_norm": 0.5903569459915161,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 2403
+    },
+    {
+      "epoch": 0.583778533268577,
+      "grad_norm": 0.5820091962814331,
+      "learning_rate": 0.0001,
+      "loss": 1.6464,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5840213695968917,
+      "grad_norm": 0.6163525581359863,
+      "learning_rate": 0.0001,
+      "loss": 1.9778,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5842642059252064,
+      "grad_norm": 0.5677646398544312,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5845070422535211,
+      "grad_norm": 0.5462453365325928,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5847498785818358,
+      "grad_norm": 0.5642171502113342,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5849927149101506,
+      "grad_norm": 0.5937358140945435,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5852355512384653,
+      "grad_norm": 0.5900371074676514,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 2410
+    },
+    {
+      "epoch": 0.58547838756678,
+      "grad_norm": 0.5788990259170532,
+      "learning_rate": 0.0001,
+      "loss": 1.7227,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5857212238950947,
+      "grad_norm": 0.5753214955329895,
+      "learning_rate": 0.0001,
+      "loss": 1.6505,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5859640602234094,
+      "grad_norm": 0.5644556283950806,
+      "learning_rate": 0.0001,
+      "loss": 1.7429,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5862068965517241,
+      "grad_norm": 0.6227849721908569,
+      "learning_rate": 0.0001,
+      "loss": 1.718,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5864497328800389,
+      "grad_norm": 0.5930759906768799,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5866925692083536,
+      "grad_norm": 0.5450453162193298,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5869354055366682,
+      "grad_norm": 0.6010074019432068,
+      "learning_rate": 0.0001,
+      "loss": 1.8531,
+      "step": 2417
+    },
+    {
+      "epoch": 0.587178241864983,
+      "grad_norm": 0.5836543440818787,
+      "learning_rate": 0.0001,
+      "loss": 1.8302,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5874210781932977,
+      "grad_norm": 0.5937181711196899,
+      "learning_rate": 0.0001,
+      "loss": 1.7584,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5876639145216125,
+      "grad_norm": 0.5370633602142334,
+      "learning_rate": 0.0001,
+      "loss": 1.5638,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5879067508499272,
+      "grad_norm": 0.5745272636413574,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5881495871782418,
+      "grad_norm": 0.5497100353240967,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5883924235065566,
+      "grad_norm": 0.5470163822174072,
+      "learning_rate": 0.0001,
+      "loss": 1.6796,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5886352598348713,
+      "grad_norm": 0.5787187814712524,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 2424
+    },
+    {
+      "epoch": 0.588878096163186,
+      "grad_norm": 0.5292855501174927,
+      "learning_rate": 0.0001,
+      "loss": 1.6425,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5891209324915008,
+      "grad_norm": 0.5789694786071777,
+      "learning_rate": 0.0001,
+      "loss": 1.7506,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5893637688198154,
+      "grad_norm": 0.5826478600502014,
+      "learning_rate": 0.0001,
+      "loss": 1.7998,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5896066051481301,
+      "grad_norm": 0.5840790867805481,
+      "learning_rate": 0.0001,
+      "loss": 1.7444,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5898494414764449,
+      "grad_norm": 0.5900298953056335,
+      "learning_rate": 0.0001,
+      "loss": 1.7615,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5900922778047596,
+      "grad_norm": 0.5789086222648621,
+      "learning_rate": 0.0001,
+      "loss": 1.7267,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5903351141330743,
+      "grad_norm": 0.5412051677703857,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5905779504613891,
+      "grad_norm": 0.5917304754257202,
+      "learning_rate": 0.0001,
+      "loss": 1.8933,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5908207867897037,
+      "grad_norm": 0.583175778388977,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5910636231180184,
+      "grad_norm": 0.5745235681533813,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5913064594463332,
+      "grad_norm": 0.5677094459533691,
+      "learning_rate": 0.0001,
+      "loss": 1.6825,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5915492957746479,
+      "grad_norm": 0.5684336423873901,
+      "learning_rate": 0.0001,
+      "loss": 1.7585,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5917921321029626,
+      "grad_norm": 0.6031253337860107,
+      "learning_rate": 0.0001,
+      "loss": 1.8517,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5920349684312773,
+      "grad_norm": 0.5518312454223633,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 2438
+    },
+    {
+      "epoch": 0.592277804759592,
+      "grad_norm": 0.5937772393226624,
+      "learning_rate": 0.0001,
+      "loss": 1.7455,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5925206410879067,
+      "grad_norm": 0.5688191652297974,
+      "learning_rate": 0.0001,
+      "loss": 1.8297,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5927634774162215,
+      "grad_norm": 0.601302981376648,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5930063137445362,
+      "grad_norm": 0.572617769241333,
+      "learning_rate": 0.0001,
+      "loss": 1.6983,
+      "step": 2442
+    },
+    {
+      "epoch": 0.593249150072851,
+      "grad_norm": 0.566998302936554,
+      "learning_rate": 0.0001,
+      "loss": 1.7456,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5934919864011656,
+      "grad_norm": 0.5750413537025452,
+      "learning_rate": 0.0001,
+      "loss": 1.7105,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5937348227294803,
+      "grad_norm": 0.5915107131004333,
+      "learning_rate": 0.0001,
+      "loss": 1.6796,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5939776590577951,
+      "grad_norm": 0.5885311961174011,
+      "learning_rate": 0.0001,
+      "loss": 1.701,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5942204953861098,
+      "grad_norm": 0.614865779876709,
+      "learning_rate": 0.0001,
+      "loss": 1.7882,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5944633317144244,
+      "grad_norm": 0.5743029117584229,
+      "learning_rate": 0.0001,
+      "loss": 1.7656,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5947061680427392,
+      "grad_norm": 0.561428427696228,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5949490043710539,
+      "grad_norm": 0.5786658525466919,
+      "learning_rate": 0.0001,
+      "loss": 1.8408,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5951918406993686,
+      "grad_norm": 0.5619210600852966,
+      "learning_rate": 0.0001,
+      "loss": 1.7701,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5954346770276834,
+      "grad_norm": 0.5815402269363403,
+      "learning_rate": 0.0001,
+      "loss": 1.8588,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5956775133559981,
+      "grad_norm": 0.5524076223373413,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5959203496843127,
+      "grad_norm": 0.6057606339454651,
+      "learning_rate": 0.0001,
+      "loss": 1.7731,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5961631860126275,
+      "grad_norm": 0.5456556081771851,
+      "learning_rate": 0.0001,
+      "loss": 1.575,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5964060223409422,
+      "grad_norm": 0.5837120413780212,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5966488586692569,
+      "grad_norm": 0.5915725231170654,
+      "learning_rate": 0.0001,
+      "loss": 1.753,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5968916949975717,
+      "grad_norm": 1.2732526063919067,
+      "learning_rate": 0.0001,
+      "loss": 1.7561,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5971345313258863,
+      "grad_norm": 0.5702220797538757,
+      "learning_rate": 0.0001,
+      "loss": 1.7971,
+      "step": 2459
+    },
+    {
+      "epoch": 0.597377367654201,
+      "grad_norm": 0.5905393362045288,
+      "learning_rate": 0.0001,
+      "loss": 1.7459,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5976202039825158,
+      "grad_norm": 0.6075118780136108,
+      "learning_rate": 0.0001,
+      "loss": 1.6879,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5978630403108305,
+      "grad_norm": 0.5428155064582825,
+      "learning_rate": 0.0001,
+      "loss": 1.5011,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5981058766391452,
+      "grad_norm": 0.6202692985534668,
+      "learning_rate": 0.0001,
+      "loss": 1.8806,
+      "step": 2463
+    },
+    {
+      "epoch": 0.59834871296746,
+      "grad_norm": 0.6165500283241272,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5985915492957746,
+      "grad_norm": 0.6068257093429565,
+      "learning_rate": 0.0001,
+      "loss": 1.5594,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5988343856240894,
+      "grad_norm": 0.6311659216880798,
+      "learning_rate": 0.0001,
+      "loss": 1.8102,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5990772219524041,
+      "grad_norm": 0.5966998338699341,
+      "learning_rate": 0.0001,
+      "loss": 1.8661,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5993200582807188,
+      "grad_norm": 0.5616307258605957,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5995628946090336,
+      "grad_norm": 0.5898672342300415,
+      "learning_rate": 0.0001,
+      "loss": 1.7186,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5998057309373482,
+      "grad_norm": 0.5857921242713928,
+      "learning_rate": 0.0001,
+      "loss": 1.8196,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6000485672656629,
+      "grad_norm": 0.5745912790298462,
+      "learning_rate": 0.0001,
+      "loss": 1.7887,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6002914035939777,
+      "grad_norm": 0.5690114498138428,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6005342399222924,
+      "grad_norm": 0.5422451496124268,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6007770762506071,
+      "grad_norm": 0.5912676453590393,
+      "learning_rate": 0.0001,
+      "loss": 1.7619,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6010199125789218,
+      "grad_norm": 0.5485222935676575,
+      "learning_rate": 0.0001,
+      "loss": 1.6598,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6012627489072365,
+      "grad_norm": 0.5810349583625793,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6015055852355512,
+      "grad_norm": 0.6028423309326172,
+      "learning_rate": 0.0001,
+      "loss": 1.851,
+      "step": 2477
+    },
+    {
+      "epoch": 0.601748421563866,
+      "grad_norm": 0.5798795223236084,
+      "learning_rate": 0.0001,
+      "loss": 1.7714,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6019912578921807,
+      "grad_norm": 0.5962965488433838,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6022340942204953,
+      "grad_norm": 0.575608491897583,
+      "learning_rate": 0.0001,
+      "loss": 1.7201,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6024769305488101,
+      "grad_norm": 0.5525558590888977,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6027197668771248,
+      "grad_norm": 0.5977734923362732,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6029626032054395,
+      "grad_norm": 0.5953426361083984,
+      "learning_rate": 0.0001,
+      "loss": 1.7852,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6032054395337543,
+      "grad_norm": 0.58254474401474,
+      "learning_rate": 0.0001,
+      "loss": 1.6766,
+      "step": 2484
+    },
+    {
+      "epoch": 0.603448275862069,
+      "grad_norm": 0.5625193119049072,
+      "learning_rate": 0.0001,
+      "loss": 1.6301,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6036911121903836,
+      "grad_norm": 0.5599878430366516,
+      "learning_rate": 0.0001,
+      "loss": 1.752,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6039339485186984,
+      "grad_norm": 0.5602197051048279,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 2487
+    },
+    {
+      "epoch": 0.6041767848470131,
+      "grad_norm": 0.5864724516868591,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6044196211753279,
+      "grad_norm": 0.580513060092926,
+      "learning_rate": 0.0001,
+      "loss": 1.7135,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6046624575036426,
+      "grad_norm": 0.617057740688324,
+      "learning_rate": 0.0001,
+      "loss": 1.9021,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6049052938319572,
+      "grad_norm": 0.5644077658653259,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 2491
+    },
+    {
+      "epoch": 0.605148130160272,
+      "grad_norm": 0.5731435418128967,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6053909664885867,
+      "grad_norm": 0.6410675644874573,
+      "learning_rate": 0.0001,
+      "loss": 1.8625,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6056338028169014,
+      "grad_norm": 0.5538236498832703,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6058766391452162,
+      "grad_norm": 0.5494367480278015,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6061194754735308,
+      "grad_norm": 0.6226809024810791,
+      "learning_rate": 0.0001,
+      "loss": 1.7613,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6063623118018455,
+      "grad_norm": 0.6167261004447937,
+      "learning_rate": 0.0001,
+      "loss": 1.7512,
+      "step": 2497
+    },
+    {
+      "epoch": 0.6066051481301603,
+      "grad_norm": 0.5871309041976929,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 2498
+    },
+    {
+      "epoch": 0.606847984458475,
+      "grad_norm": 0.6313231587409973,
+      "learning_rate": 0.0001,
+      "loss": 1.7515,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6070908207867897,
+      "grad_norm": 0.5782124400138855,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6073336571151045,
+      "grad_norm": 0.6116856932640076,
+      "learning_rate": 0.0001,
+      "loss": 1.794,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6075764934434191,
+      "grad_norm": 0.5600969195365906,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6078193297717338,
+      "grad_norm": 0.5679277777671814,
+      "learning_rate": 0.0001,
+      "loss": 1.626,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6080621661000486,
+      "grad_norm": 0.5670769810676575,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6083050024283633,
+      "grad_norm": 0.5758386254310608,
+      "learning_rate": 0.0001,
+      "loss": 1.8142,
+      "step": 2505
+    },
+    {
+      "epoch": 0.608547838756678,
+      "grad_norm": 0.5373086929321289,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6087906750849927,
+      "grad_norm": 0.5614203214645386,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6090335114133074,
+      "grad_norm": 0.5441398024559021,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6092763477416221,
+      "grad_norm": 0.619110107421875,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6095191840699369,
+      "grad_norm": 0.603743851184845,
+      "learning_rate": 0.0001,
+      "loss": 1.8676,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6097620203982516,
+      "grad_norm": 0.5662248134613037,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6100048567265663,
+      "grad_norm": 0.5333821177482605,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 2512
+    },
+    {
+      "epoch": 0.610247693054881,
+      "grad_norm": 0.6046711206436157,
+      "learning_rate": 0.0001,
+      "loss": 1.7985,
+      "step": 2513
+    },
+    {
+      "epoch": 0.6104905293831957,
+      "grad_norm": 0.5740635395050049,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6107333657115105,
+      "grad_norm": 0.6029295325279236,
+      "learning_rate": 0.0001,
+      "loss": 1.8217,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6109762020398252,
+      "grad_norm": 0.5581886768341064,
+      "learning_rate": 0.0001,
+      "loss": 1.8342,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6112190383681398,
+      "grad_norm": 0.5729279518127441,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6114618746964546,
+      "grad_norm": 0.5676250457763672,
+      "learning_rate": 0.0001,
+      "loss": 1.6116,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6117047110247693,
+      "grad_norm": 0.5726991295814514,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 2519
+    },
+    {
+      "epoch": 0.611947547353084,
+      "grad_norm": 0.5791887640953064,
+      "learning_rate": 0.0001,
+      "loss": 1.6938,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6121903836813988,
+      "grad_norm": 0.5699273943901062,
+      "learning_rate": 0.0001,
+      "loss": 1.7318,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6124332200097135,
+      "grad_norm": 0.564121663570404,
+      "learning_rate": 0.0001,
+      "loss": 1.6655,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6126760563380281,
+      "grad_norm": 0.5962718725204468,
+      "learning_rate": 0.0001,
+      "loss": 1.7788,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6129188926663429,
+      "grad_norm": 0.5704603791236877,
+      "learning_rate": 0.0001,
+      "loss": 1.7749,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6131617289946576,
+      "grad_norm": 0.6029075384140015,
+      "learning_rate": 0.0001,
+      "loss": 1.8248,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6134045653229723,
+      "grad_norm": 0.5665852427482605,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6136474016512871,
+      "grad_norm": 0.5830678343772888,
+      "learning_rate": 0.0001,
+      "loss": 1.6869,
+      "step": 2527
+    },
+    {
+      "epoch": 0.6138902379796017,
+      "grad_norm": 0.5895706415176392,
+      "learning_rate": 0.0001,
+      "loss": 1.7805,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6141330743079164,
+      "grad_norm": 0.5577168464660645,
+      "learning_rate": 0.0001,
+      "loss": 1.7499,
+      "step": 2529
+    },
+    {
+      "epoch": 0.6143759106362312,
+      "grad_norm": 0.5509077906608582,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6146187469645459,
+      "grad_norm": 0.5555626749992371,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6148615832928606,
+      "grad_norm": 0.5771478414535522,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6151044196211753,
+      "grad_norm": 0.55790776014328,
+      "learning_rate": 0.0001,
+      "loss": 1.5604,
+      "step": 2533
+    },
+    {
+      "epoch": 0.61534725594949,
+      "grad_norm": 0.5720113515853882,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6155900922778048,
+      "grad_norm": 0.5889668464660645,
+      "learning_rate": 0.0001,
+      "loss": 1.688,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6158329286061195,
+      "grad_norm": 0.5640578866004944,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6160757649344342,
+      "grad_norm": 0.605247974395752,
+      "learning_rate": 0.0001,
+      "loss": 1.733,
+      "step": 2537
+    },
+    {
+      "epoch": 0.616318601262749,
+      "grad_norm": 0.5761142373085022,
+      "learning_rate": 0.0001,
+      "loss": 1.7619,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6165614375910636,
+      "grad_norm": 0.5925478935241699,
+      "learning_rate": 0.0001,
+      "loss": 1.784,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6168042739193783,
+      "grad_norm": 0.5719945430755615,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6170471102476931,
+      "grad_norm": 0.57254558801651,
+      "learning_rate": 0.0001,
+      "loss": 1.6022,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6172899465760078,
+      "grad_norm": 0.5895758867263794,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6175327829043225,
+      "grad_norm": 0.5915838479995728,
+      "learning_rate": 0.0001,
+      "loss": 1.8678,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6177756192326372,
+      "grad_norm": 0.5661805868148804,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6180184555609519,
+      "grad_norm": 0.6017798781394958,
+      "learning_rate": 0.0001,
+      "loss": 1.7785,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6182612918892666,
+      "grad_norm": 0.5868064761161804,
+      "learning_rate": 0.0001,
+      "loss": 1.8783,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6185041282175814,
+      "grad_norm": 0.5943471193313599,
+      "learning_rate": 0.0001,
+      "loss": 1.8255,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6187469645458961,
+      "grad_norm": 0.5721445083618164,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6189898008742107,
+      "grad_norm": 0.5788509845733643,
+      "learning_rate": 0.0001,
+      "loss": 1.7048,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6192326372025255,
+      "grad_norm": 0.5473672747612,
+      "learning_rate": 0.0001,
+      "loss": 1.5743,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6194754735308402,
+      "grad_norm": 0.5789993405342102,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 2551
+    },
+    {
+      "epoch": 0.6197183098591549,
+      "grad_norm": 0.5286515951156616,
+      "learning_rate": 0.0001,
+      "loss": 1.4971,
+      "step": 2552
+    },
+    {
+      "epoch": 0.6199611461874697,
+      "grad_norm": 0.5470941066741943,
+      "learning_rate": 0.0001,
+      "loss": 1.6914,
+      "step": 2553
+    },
+    {
+      "epoch": 0.6202039825157843,
+      "grad_norm": 0.6283156275749207,
+      "learning_rate": 0.0001,
+      "loss": 2.0322,
+      "step": 2554
+    },
+    {
+      "epoch": 0.620446818844099,
+      "grad_norm": 0.571928858757019,
+      "learning_rate": 0.0001,
+      "loss": 1.8533,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 0.5806577205657959,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6209324915007285,
+      "grad_norm": 0.6049844026565552,
+      "learning_rate": 0.0001,
+      "loss": 1.7491,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6211753278290433,
+      "grad_norm": 0.5663546323776245,
+      "learning_rate": 0.0001,
+      "loss": 1.7514,
+      "step": 2558
+    },
+    {
+      "epoch": 0.621418164157358,
+      "grad_norm": 0.5638301968574524,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6216610004856726,
+      "grad_norm": 0.5578255653381348,
+      "learning_rate": 0.0001,
+      "loss": 1.7402,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6219038368139874,
+      "grad_norm": 0.5995447039604187,
+      "learning_rate": 0.0001,
+      "loss": 1.8496,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6221466731423021,
+      "grad_norm": 0.5353391170501709,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6223895094706168,
+      "grad_norm": 0.5682584047317505,
+      "learning_rate": 0.0001,
+      "loss": 1.6659,
+      "step": 2563
+    },
+    {
+      "epoch": 0.6226323457989316,
+      "grad_norm": 0.5946477055549622,
+      "learning_rate": 0.0001,
+      "loss": 1.7719,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6228751821272462,
+      "grad_norm": 0.5968493819236755,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6231180184555609,
+      "grad_norm": 0.6130239367485046,
+      "learning_rate": 0.0001,
+      "loss": 1.7575,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6233608547838757,
+      "grad_norm": 0.6105282306671143,
+      "learning_rate": 0.0001,
+      "loss": 1.8922,
+      "step": 2567
+    },
+    {
+      "epoch": 0.6236036911121904,
+      "grad_norm": 0.5839565992355347,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6238465274405051,
+      "grad_norm": 0.5885269045829773,
+      "learning_rate": 0.0001,
+      "loss": 1.7915,
+      "step": 2569
+    },
+    {
+      "epoch": 0.6240893637688198,
+      "grad_norm": 0.5504341125488281,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6243322000971345,
+      "grad_norm": 0.5717918276786804,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6245750364254492,
+      "grad_norm": 0.5776162147521973,
+      "learning_rate": 0.0001,
+      "loss": 1.766,
+      "step": 2572
+    },
+    {
+      "epoch": 0.624817872753764,
+      "grad_norm": 0.5442912578582764,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6250607090820787,
+      "grad_norm": 0.5399508476257324,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6253035454103933,
+      "grad_norm": 0.601722776889801,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 2575
+    },
+    {
+      "epoch": 0.6255463817387081,
+      "grad_norm": 0.5953107476234436,
+      "learning_rate": 0.0001,
+      "loss": 1.7363,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6257892180670228,
+      "grad_norm": 0.5822069048881531,
+      "learning_rate": 0.0001,
+      "loss": 1.6664,
+      "step": 2577
+    },
+    {
+      "epoch": 0.6260320543953375,
+      "grad_norm": 0.6120938658714294,
+      "learning_rate": 0.0001,
+      "loss": 1.7251,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6262748907236523,
+      "grad_norm": 0.632987380027771,
+      "learning_rate": 0.0001,
+      "loss": 1.8687,
+      "step": 2579
+    },
+    {
+      "epoch": 0.626517727051967,
+      "grad_norm": 0.579434871673584,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6267605633802817,
+      "grad_norm": 0.6319934725761414,
+      "learning_rate": 0.0001,
+      "loss": 1.8018,
+      "step": 2581
+    },
+    {
+      "epoch": 0.6270033997085964,
+      "grad_norm": 0.5732057094573975,
+      "learning_rate": 0.0001,
+      "loss": 1.7314,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6272462360369111,
+      "grad_norm": 0.5757622122764587,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 2583
+    },
+    {
+      "epoch": 0.6274890723652259,
+      "grad_norm": 0.5308606624603271,
+      "learning_rate": 0.0001,
+      "loss": 1.6067,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6277319086935406,
+      "grad_norm": 0.5910230278968811,
+      "learning_rate": 0.0001,
+      "loss": 1.7544,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6279747450218552,
+      "grad_norm": 0.5853991508483887,
+      "learning_rate": 0.0001,
+      "loss": 1.7229,
+      "step": 2586
+    },
+    {
+      "epoch": 0.62821758135017,
+      "grad_norm": 0.5474392175674438,
+      "learning_rate": 0.0001,
+      "loss": 1.6707,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6284604176784847,
+      "grad_norm": 0.6053445339202881,
+      "learning_rate": 0.0001,
+      "loss": 1.8133,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6287032540067994,
+      "grad_norm": 0.577170729637146,
+      "learning_rate": 0.0001,
+      "loss": 1.781,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6289460903351142,
+      "grad_norm": 0.6177255511283875,
+      "learning_rate": 0.0001,
+      "loss": 1.8164,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6291889266634288,
+      "grad_norm": 0.5895721912384033,
+      "learning_rate": 0.0001,
+      "loss": 1.7574,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6294317629917435,
+      "grad_norm": 0.5663004517555237,
+      "learning_rate": 0.0001,
+      "loss": 1.7434,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6296745993200583,
+      "grad_norm": 0.6223627924919128,
+      "learning_rate": 0.0001,
+      "loss": 1.9011,
+      "step": 2593
+    },
+    {
+      "epoch": 0.629917435648373,
+      "grad_norm": 0.5563454627990723,
+      "learning_rate": 0.0001,
+      "loss": 1.6896,
+      "step": 2594
+    },
+    {
+      "epoch": 0.6301602719766877,
+      "grad_norm": 0.5253757238388062,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6304031083050025,
+      "grad_norm": 0.6091992855072021,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 2596
+    },
+    {
+      "epoch": 0.6306459446333171,
+      "grad_norm": 0.6034373641014099,
+      "learning_rate": 0.0001,
+      "loss": 1.8019,
+      "step": 2597
+    },
+    {
+      "epoch": 0.6308887809616318,
+      "grad_norm": 0.635011613368988,
+      "learning_rate": 0.0001,
+      "loss": 1.863,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6311316172899466,
+      "grad_norm": 0.5787094831466675,
+      "learning_rate": 0.0001,
+      "loss": 1.6395,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6313744536182613,
+      "grad_norm": 0.5938628911972046,
+      "learning_rate": 0.0001,
+      "loss": 1.8188,
+      "step": 2600
+    },
+    {
+      "epoch": 0.631617289946576,
+      "grad_norm": 0.6202360987663269,
+      "learning_rate": 0.0001,
+      "loss": 1.7701,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6318601262748907,
+      "grad_norm": 0.5813153386116028,
+      "learning_rate": 0.0001,
+      "loss": 1.7545,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6321029626032054,
+      "grad_norm": 0.643073320388794,
+      "learning_rate": 0.0001,
+      "loss": 1.8769,
+      "step": 2603
+    },
+    {
+      "epoch": 0.6323457989315202,
+      "grad_norm": 0.5803655385971069,
+      "learning_rate": 0.0001,
+      "loss": 1.7714,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6325886352598349,
+      "grad_norm": 0.5840656757354736,
+      "learning_rate": 0.0001,
+      "loss": 1.3808,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6328314715881496,
+      "grad_norm": 0.581245481967926,
+      "learning_rate": 0.0001,
+      "loss": 1.8951,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6330743079164644,
+      "grad_norm": 0.5805801153182983,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 2607
+    },
+    {
+      "epoch": 0.633317144244779,
+      "grad_norm": 0.6198165416717529,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6335599805730937,
+      "grad_norm": 0.6104231476783752,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6338028169014085,
+      "grad_norm": 0.5909473299980164,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6340456532297232,
+      "grad_norm": 0.5757015347480774,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6342884895580378,
+      "grad_norm": 0.5593287944793701,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6345313258863526,
+      "grad_norm": 0.6046830415725708,
+      "learning_rate": 0.0001,
+      "loss": 1.8173,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6347741622146673,
+      "grad_norm": 0.570202112197876,
+      "learning_rate": 0.0001,
+      "loss": 1.8575,
+      "step": 2614
+    },
+    {
+      "epoch": 0.635016998542982,
+      "grad_norm": 0.5984710454940796,
+      "learning_rate": 0.0001,
+      "loss": 1.8192,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6352598348712968,
+      "grad_norm": 0.6091638803482056,
+      "learning_rate": 0.0001,
+      "loss": 1.8031,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6355026711996115,
+      "grad_norm": 0.5655072927474976,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6357455075279261,
+      "grad_norm": 0.5606513023376465,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6359883438562409,
+      "grad_norm": 0.5860190987586975,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 2619
+    },
+    {
+      "epoch": 0.6362311801845556,
+      "grad_norm": 0.5981796979904175,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6364740165128703,
+      "grad_norm": 0.5950610041618347,
+      "learning_rate": 0.0001,
+      "loss": 1.7344,
+      "step": 2621
+    },
+    {
+      "epoch": 0.6367168528411851,
+      "grad_norm": 0.6207805275917053,
+      "learning_rate": 0.0001,
+      "loss": 1.9791,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6369596891694997,
+      "grad_norm": 0.5669253468513489,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 2623
+    },
+    {
+      "epoch": 0.6372025254978144,
+      "grad_norm": 0.5895827412605286,
+      "learning_rate": 0.0001,
+      "loss": 1.8426,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6374453618261292,
+      "grad_norm": 0.6235145926475525,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6376881981544439,
+      "grad_norm": 0.6031660437583923,
+      "learning_rate": 0.0001,
+      "loss": 1.8477,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6379310344827587,
+      "grad_norm": 0.6209763884544373,
+      "learning_rate": 0.0001,
+      "loss": 1.8003,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6381738708110734,
+      "grad_norm": 0.5706555247306824,
+      "learning_rate": 0.0001,
+      "loss": 1.8148,
+      "step": 2628
+    },
+    {
+      "epoch": 0.638416707139388,
+      "grad_norm": 0.6101211309432983,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6386595434677028,
+      "grad_norm": 0.6126018166542053,
+      "learning_rate": 0.0001,
+      "loss": 1.764,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6389023797960175,
+      "grad_norm": 0.6178079843521118,
+      "learning_rate": 0.0001,
+      "loss": 1.7531,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6391452161243322,
+      "grad_norm": 0.621501088142395,
+      "learning_rate": 0.0001,
+      "loss": 1.7783,
+      "step": 2632
+    },
+    {
+      "epoch": 0.639388052452647,
+      "grad_norm": 0.5543282628059387,
+      "learning_rate": 0.0001,
+      "loss": 1.7778,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6396308887809616,
+      "grad_norm": 0.5693379640579224,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6398737251092763,
+      "grad_norm": 0.6385512351989746,
+      "learning_rate": 0.0001,
+      "loss": 1.8175,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6401165614375911,
+      "grad_norm": 0.5857069492340088,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6403593977659058,
+      "grad_norm": 0.5730785131454468,
+      "learning_rate": 0.0001,
+      "loss": 1.6808,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6406022340942205,
+      "grad_norm": 0.5785220861434937,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6408450704225352,
+      "grad_norm": 0.5896817445755005,
+      "learning_rate": 0.0001,
+      "loss": 1.7995,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6410879067508499,
+      "grad_norm": 0.607143223285675,
+      "learning_rate": 0.0001,
+      "loss": 1.6896,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6413307430791646,
+      "grad_norm": 0.5732043385505676,
+      "learning_rate": 0.0001,
+      "loss": 1.7495,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6415735794074794,
+      "grad_norm": 0.592718780040741,
+      "learning_rate": 0.0001,
+      "loss": 1.7584,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6418164157357941,
+      "grad_norm": 0.5815568566322327,
+      "learning_rate": 0.0001,
+      "loss": 1.776,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6420592520641087,
+      "grad_norm": 0.573049008846283,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6423020883924235,
+      "grad_norm": 0.5804055333137512,
+      "learning_rate": 0.0001,
+      "loss": 1.7891,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6425449247207382,
+      "grad_norm": 0.5961693525314331,
+      "learning_rate": 0.0001,
+      "loss": 1.8251,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6427877610490529,
+      "grad_norm": 0.5870078206062317,
+      "learning_rate": 0.0001,
+      "loss": 1.6878,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6430305973773677,
+      "grad_norm": 0.6337385773658752,
+      "learning_rate": 0.0001,
+      "loss": 1.9097,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6432734337056824,
+      "grad_norm": 0.6195207834243774,
+      "learning_rate": 0.0001,
+      "loss": 1.7589,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6435162700339971,
+      "grad_norm": 0.5870786905288696,
+      "learning_rate": 0.0001,
+      "loss": 1.8101,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6437591063623118,
+      "grad_norm": 0.5888528227806091,
+      "learning_rate": 0.0001,
+      "loss": 1.7392,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6440019426906265,
+      "grad_norm": 0.5681914687156677,
+      "learning_rate": 0.0001,
+      "loss": 1.695,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6442447790189413,
+      "grad_norm": 0.5761615037918091,
+      "learning_rate": 0.0001,
+      "loss": 1.8565,
+      "step": 2653
+    },
+    {
+      "epoch": 0.644487615347256,
+      "grad_norm": 0.5865592360496521,
+      "learning_rate": 0.0001,
+      "loss": 1.8046,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6447304516755706,
+      "grad_norm": 0.5950198769569397,
+      "learning_rate": 0.0001,
+      "loss": 1.7282,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6449732880038854,
+      "grad_norm": 0.5902634859085083,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6452161243322001,
+      "grad_norm": 0.5786793828010559,
+      "learning_rate": 0.0001,
+      "loss": 1.7535,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6454589606605148,
+      "grad_norm": 0.5738018155097961,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6457017969888296,
+      "grad_norm": 0.5593220591545105,
+      "learning_rate": 0.0001,
+      "loss": 1.6467,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6459446333171442,
+      "grad_norm": 0.5977775454521179,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6461874696454589,
+      "grad_norm": 0.5915343761444092,
+      "learning_rate": 0.0001,
+      "loss": 1.7429,
+      "step": 2661
+    },
+    {
+      "epoch": 0.6464303059737737,
+      "grad_norm": 0.5825751423835754,
+      "learning_rate": 0.0001,
+      "loss": 1.7846,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6466731423020884,
+      "grad_norm": 0.601574718952179,
+      "learning_rate": 0.0001,
+      "loss": 1.7243,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6469159786304031,
+      "grad_norm": 0.5464960336685181,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6471588149587179,
+      "grad_norm": 0.5692195296287537,
+      "learning_rate": 0.0001,
+      "loss": 1.7974,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6474016512870325,
+      "grad_norm": 0.5517829060554504,
+      "learning_rate": 0.0001,
+      "loss": 1.7614,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6476444876153472,
+      "grad_norm": 0.5647340416908264,
+      "learning_rate": 0.0001,
+      "loss": 1.7064,
+      "step": 2667
+    },
+    {
+      "epoch": 0.647887323943662,
+      "grad_norm": 0.5393012166023254,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6481301602719767,
+      "grad_norm": 0.566716730594635,
+      "learning_rate": 0.0001,
+      "loss": 1.7787,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6483729966002914,
+      "grad_norm": 0.5588338971138,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6486158329286061,
+      "grad_norm": 0.5825018286705017,
+      "learning_rate": 0.0001,
+      "loss": 1.7021,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6488586692569208,
+      "grad_norm": 0.5807709693908691,
+      "learning_rate": 0.0001,
+      "loss": 1.833,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6491015055852356,
+      "grad_norm": 0.5714016556739807,
+      "learning_rate": 0.0001,
+      "loss": 1.7231,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6493443419135503,
+      "grad_norm": 0.5701686143875122,
+      "learning_rate": 0.0001,
+      "loss": 1.7271,
+      "step": 2674
+    },
+    {
+      "epoch": 0.649587178241865,
+      "grad_norm": 0.6211767196655273,
+      "learning_rate": 0.0001,
+      "loss": 1.8542,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6498300145701797,
+      "grad_norm": 0.5851500630378723,
+      "learning_rate": 0.0001,
+      "loss": 1.8296,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6500728508984944,
+      "grad_norm": 0.6012861728668213,
+      "learning_rate": 0.0001,
+      "loss": 1.8426,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6503156872268091,
+      "grad_norm": 0.5840545892715454,
+      "learning_rate": 0.0001,
+      "loss": 1.7581,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6505585235551239,
+      "grad_norm": 0.580803394317627,
+      "learning_rate": 0.0001,
+      "loss": 1.7851,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6508013598834386,
+      "grad_norm": 0.5769532322883606,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6510441962117532,
+      "grad_norm": 0.5599761605262756,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 2681
+    },
+    {
+      "epoch": 0.651287032540068,
+      "grad_norm": 0.6272950172424316,
+      "learning_rate": 0.0001,
+      "loss": 1.9708,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6515298688683827,
+      "grad_norm": 0.5727792978286743,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6517727051966974,
+      "grad_norm": 0.5690376162528992,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6520155415250122,
+      "grad_norm": 0.6194298267364502,
+      "learning_rate": 0.0001,
+      "loss": 1.8909,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6522583778533269,
+      "grad_norm": 0.585206151008606,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6525012141816415,
+      "grad_norm": 0.5940220355987549,
+      "learning_rate": 0.0001,
+      "loss": 1.8036,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6527440505099563,
+      "grad_norm": 0.5587031841278076,
+      "learning_rate": 0.0001,
+      "loss": 1.705,
+      "step": 2688
+    },
+    {
+      "epoch": 0.652986886838271,
+      "grad_norm": 0.5791237354278564,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6532297231665857,
+      "grad_norm": 0.6508381962776184,
+      "learning_rate": 0.0001,
+      "loss": 1.8489,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6534725594949005,
+      "grad_norm": 0.5821163058280945,
+      "learning_rate": 0.0001,
+      "loss": 1.7967,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6537153958232151,
+      "grad_norm": 0.5625880360603333,
+      "learning_rate": 0.0001,
+      "loss": 1.7925,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6539582321515298,
+      "grad_norm": 0.5700521469116211,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6542010684798446,
+      "grad_norm": 0.5617753863334656,
+      "learning_rate": 0.0001,
+      "loss": 1.6741,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6544439048081593,
+      "grad_norm": 0.5487257838249207,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6546867411364741,
+      "grad_norm": 0.6118389964103699,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6549295774647887,
+      "grad_norm": 0.6059249043464661,
+      "learning_rate": 0.0001,
+      "loss": 1.7782,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6551724137931034,
+      "grad_norm": 0.6007734537124634,
+      "learning_rate": 0.0001,
+      "loss": 1.7264,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6554152501214182,
+      "grad_norm": 0.5955468416213989,
+      "learning_rate": 0.0001,
+      "loss": 1.7981,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6556580864497329,
+      "grad_norm": 0.6089268326759338,
+      "learning_rate": 0.0001,
+      "loss": 1.7969,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6559009227780476,
+      "grad_norm": 0.5827802419662476,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6561437591063624,
+      "grad_norm": 0.5722429156303406,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 2702
+    },
+    {
+      "epoch": 0.656386595434677,
+      "grad_norm": 0.5856757164001465,
+      "learning_rate": 0.0001,
+      "loss": 1.9363,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6566294317629917,
+      "grad_norm": 0.5609936118125916,
+      "learning_rate": 0.0001,
+      "loss": 1.7985,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6568722680913065,
+      "grad_norm": 0.6096386909484863,
+      "learning_rate": 0.0001,
+      "loss": 1.8202,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6571151044196212,
+      "grad_norm": 0.5903800129890442,
+      "learning_rate": 0.0001,
+      "loss": 1.6566,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6573579407479359,
+      "grad_norm": 0.5459445714950562,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6576007770762506,
+      "grad_norm": 0.5889915823936462,
+      "learning_rate": 0.0001,
+      "loss": 1.7066,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6578436134045653,
+      "grad_norm": 0.5951976776123047,
+      "learning_rate": 0.0001,
+      "loss": 1.7497,
+      "step": 2709
+    },
+    {
+      "epoch": 0.65808644973288,
+      "grad_norm": 0.5950310826301575,
+      "learning_rate": 0.0001,
+      "loss": 1.743,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6583292860611948,
+      "grad_norm": 0.5863953232765198,
+      "learning_rate": 0.0001,
+      "loss": 1.9078,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6585721223895095,
+      "grad_norm": 0.5783514976501465,
+      "learning_rate": 0.0001,
+      "loss": 1.7259,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6588149587178241,
+      "grad_norm": 0.576743483543396,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6590577950461389,
+      "grad_norm": 0.6127098798751831,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6593006313744536,
+      "grad_norm": 0.5809155106544495,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6595434677027683,
+      "grad_norm": 0.5596655011177063,
+      "learning_rate": 0.0001,
+      "loss": 1.676,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6597863040310831,
+      "grad_norm": 0.560106635093689,
+      "learning_rate": 0.0001,
+      "loss": 1.7014,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6600291403593977,
+      "grad_norm": 0.6060172319412231,
+      "learning_rate": 0.0001,
+      "loss": 1.8205,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6602719766877125,
+      "grad_norm": 0.5648579001426697,
+      "learning_rate": 0.0001,
+      "loss": 1.7937,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6605148130160272,
+      "grad_norm": 0.5625224113464355,
+      "learning_rate": 0.0001,
+      "loss": 1.6573,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6607576493443419,
+      "grad_norm": 0.5313379168510437,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6610004856726567,
+      "grad_norm": 0.6125560998916626,
+      "learning_rate": 0.0001,
+      "loss": 1.9104,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6612433220009714,
+      "grad_norm": 0.5860222578048706,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 2723
+    },
+    {
+      "epoch": 0.661486158329286,
+      "grad_norm": 0.5965452790260315,
+      "learning_rate": 0.0001,
+      "loss": 1.7382,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6617289946576008,
+      "grad_norm": 0.5650197863578796,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6619718309859155,
+      "grad_norm": 0.5879565477371216,
+      "learning_rate": 0.0001,
+      "loss": 1.7021,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6622146673142302,
+      "grad_norm": 0.6002961993217468,
+      "learning_rate": 0.0001,
+      "loss": 1.8219,
+      "step": 2727
+    },
+    {
+      "epoch": 0.662457503642545,
+      "grad_norm": 0.5902764201164246,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6627003399708596,
+      "grad_norm": 0.6064320206642151,
+      "learning_rate": 0.0001,
+      "loss": 1.7249,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6629431762991743,
+      "grad_norm": 0.585027813911438,
+      "learning_rate": 0.0001,
+      "loss": 1.736,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6631860126274891,
+      "grad_norm": 0.6057623624801636,
+      "learning_rate": 0.0001,
+      "loss": 1.7291,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6634288489558038,
+      "grad_norm": 0.5800138711929321,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6636716852841185,
+      "grad_norm": 0.6094589233398438,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6639145216124333,
+      "grad_norm": 0.6267191767692566,
+      "learning_rate": 0.0001,
+      "loss": 1.8418,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6641573579407479,
+      "grad_norm": 0.6102715134620667,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6644001942690626,
+      "grad_norm": 0.569195032119751,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6646430305973774,
+      "grad_norm": 0.6051103472709656,
+      "learning_rate": 0.0001,
+      "loss": 1.8214,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6648858669256921,
+      "grad_norm": 0.6168873906135559,
+      "learning_rate": 0.0001,
+      "loss": 1.8553,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6651287032540067,
+      "grad_norm": 0.6061825752258301,
+      "learning_rate": 0.0001,
+      "loss": 1.7512,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6653715395823215,
+      "grad_norm": 0.5734617710113525,
+      "learning_rate": 0.0001,
+      "loss": 1.7128,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6656143759106362,
+      "grad_norm": 0.640291154384613,
+      "learning_rate": 0.0001,
+      "loss": 1.8902,
+      "step": 2741
+    },
+    {
+      "epoch": 0.665857212238951,
+      "grad_norm": 0.5751891732215881,
+      "learning_rate": 0.0001,
+      "loss": 1.7698,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6661000485672657,
+      "grad_norm": 0.6123544573783875,
+      "learning_rate": 0.0001,
+      "loss": 1.6901,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6663428848955804,
+      "grad_norm": 0.6082634329795837,
+      "learning_rate": 0.0001,
+      "loss": 1.7141,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6665857212238951,
+      "grad_norm": 0.5537638068199158,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6668285575522098,
+      "grad_norm": 0.5809677839279175,
+      "learning_rate": 0.0001,
+      "loss": 1.836,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6670713938805245,
+      "grad_norm": 0.5729596018791199,
+      "learning_rate": 0.0001,
+      "loss": 1.6577,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6673142302088393,
+      "grad_norm": 0.568006157875061,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 2748
+    },
+    {
+      "epoch": 0.667557066537154,
+      "grad_norm": 0.585652768611908,
+      "learning_rate": 0.0001,
+      "loss": 1.8467,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6677999028654686,
+      "grad_norm": 0.5728684663772583,
+      "learning_rate": 0.0001,
+      "loss": 1.7781,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6680427391937834,
+      "grad_norm": 0.563118040561676,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6682855755220981,
+      "grad_norm": 0.5746209621429443,
+      "learning_rate": 0.0001,
+      "loss": 1.6984,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6685284118504128,
+      "grad_norm": 0.5608879923820496,
+      "learning_rate": 0.0001,
+      "loss": 1.7711,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6687712481787276,
+      "grad_norm": 0.5681405067443848,
+      "learning_rate": 0.0001,
+      "loss": 1.69,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6690140845070423,
+      "grad_norm": 0.5153051018714905,
+      "learning_rate": 0.0001,
+      "loss": 1.3654,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6692569208353569,
+      "grad_norm": 0.5652470588684082,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6694997571636717,
+      "grad_norm": 0.5300597548484802,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6697425934919864,
+      "grad_norm": 0.5912163853645325,
+      "learning_rate": 0.0001,
+      "loss": 1.761,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6699854298203011,
+      "grad_norm": 0.6375007629394531,
+      "learning_rate": 0.0001,
+      "loss": 1.8046,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6702282661486159,
+      "grad_norm": 0.6044784784317017,
+      "learning_rate": 0.0001,
+      "loss": 1.9189,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6704711024769305,
+      "grad_norm": 0.603838324546814,
+      "learning_rate": 0.0001,
+      "loss": 1.9457,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6707139388052452,
+      "grad_norm": 0.5713114738464355,
+      "learning_rate": 0.0001,
+      "loss": 1.7561,
+      "step": 2762
+    },
+    {
+      "epoch": 0.67095677513356,
+      "grad_norm": 0.6627044677734375,
+      "learning_rate": 0.0001,
+      "loss": 1.8517,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6711996114618747,
+      "grad_norm": 0.5946755409240723,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6714424477901895,
+      "grad_norm": 0.575951337814331,
+      "learning_rate": 0.0001,
+      "loss": 1.8093,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6716852841185041,
+      "grad_norm": 0.6035037040710449,
+      "learning_rate": 0.0001,
+      "loss": 1.9085,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6719281204468188,
+      "grad_norm": 0.60302734375,
+      "learning_rate": 0.0001,
+      "loss": 1.6889,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6721709567751336,
+      "grad_norm": 0.5809454917907715,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6724137931034483,
+      "grad_norm": 0.5788617730140686,
+      "learning_rate": 0.0001,
+      "loss": 1.7135,
+      "step": 2769
+    },
+    {
+      "epoch": 0.672656629431763,
+      "grad_norm": 0.595129668712616,
+      "learning_rate": 0.0001,
+      "loss": 1.8324,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6728994657600778,
+      "grad_norm": 0.566836953163147,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6731423020883924,
+      "grad_norm": 0.5981171727180481,
+      "learning_rate": 0.0001,
+      "loss": 1.7693,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6733851384167071,
+      "grad_norm": 0.5812596678733826,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6736279747450219,
+      "grad_norm": 0.5905534029006958,
+      "learning_rate": 0.0001,
+      "loss": 1.8059,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6738708110733366,
+      "grad_norm": 0.6104099154472351,
+      "learning_rate": 0.0001,
+      "loss": 1.9028,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6741136474016513,
+      "grad_norm": 0.569607675075531,
+      "learning_rate": 0.0001,
+      "loss": 1.8641,
+      "step": 2776
+    },
+    {
+      "epoch": 0.674356483729966,
+      "grad_norm": 0.6469245553016663,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6745993200582807,
+      "grad_norm": 0.5621276497840881,
+      "learning_rate": 0.0001,
+      "loss": 1.7674,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6748421563865954,
+      "grad_norm": 0.5735675692558289,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6750849927149102,
+      "grad_norm": 0.5662938952445984,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6753278290432249,
+      "grad_norm": 0.5934909582138062,
+      "learning_rate": 0.0001,
+      "loss": 1.8791,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6755706653715395,
+      "grad_norm": 0.6003088355064392,
+      "learning_rate": 0.0001,
+      "loss": 1.6775,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6758135016998543,
+      "grad_norm": 0.5914050340652466,
+      "learning_rate": 0.0001,
+      "loss": 1.7438,
+      "step": 2783
+    },
+    {
+      "epoch": 0.676056338028169,
+      "grad_norm": 0.5427725911140442,
+      "learning_rate": 0.0001,
+      "loss": 1.4817,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6762991743564837,
+      "grad_norm": 0.5584941506385803,
+      "learning_rate": 0.0001,
+      "loss": 1.6642,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6765420106847985,
+      "grad_norm": 0.5654450058937073,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6767848470131131,
+      "grad_norm": 0.5703548192977905,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6770276833414279,
+      "grad_norm": 0.5988372564315796,
+      "learning_rate": 0.0001,
+      "loss": 1.8245,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6772705196697426,
+      "grad_norm": 0.5367932915687561,
+      "learning_rate": 0.0001,
+      "loss": 1.6302,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6775133559980573,
+      "grad_norm": 0.5318285226821899,
+      "learning_rate": 0.0001,
+      "loss": 1.5022,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6777561923263721,
+      "grad_norm": 0.6063349843025208,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6779990286546868,
+      "grad_norm": 0.6300348043441772,
+      "learning_rate": 0.0001,
+      "loss": 1.9982,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6782418649830014,
+      "grad_norm": 0.5618193745613098,
+      "learning_rate": 0.0001,
+      "loss": 1.5975,
+      "step": 2793
+    },
+    {
+      "epoch": 0.6784847013113162,
+      "grad_norm": 0.5824648141860962,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6787275376396309,
+      "grad_norm": 0.6013264656066895,
+      "learning_rate": 0.0001,
+      "loss": 1.7429,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6789703739679456,
+      "grad_norm": 0.5931627750396729,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6792132102962604,
+      "grad_norm": 0.5816960334777832,
+      "learning_rate": 0.0001,
+      "loss": 1.7652,
+      "step": 2797
+    },
+    {
+      "epoch": 0.679456046624575,
+      "grad_norm": 0.578484058380127,
+      "learning_rate": 0.0001,
+      "loss": 1.7473,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6796988829528897,
+      "grad_norm": 0.55427086353302,
+      "learning_rate": 0.0001,
+      "loss": 1.5989,
+      "step": 2799
+    },
+    {
+      "epoch": 0.6799417192812045,
+      "grad_norm": 0.6089683771133423,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6801845556095192,
+      "grad_norm": 0.5802733898162842,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6804273919378339,
+      "grad_norm": 0.5919795036315918,
+      "learning_rate": 0.0001,
+      "loss": 1.7557,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6806702282661486,
+      "grad_norm": 0.569189190864563,
+      "learning_rate": 0.0001,
+      "loss": 1.6756,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6809130645944633,
+      "grad_norm": 0.5664377808570862,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 2804
+    },
+    {
+      "epoch": 0.681155900922778,
+      "grad_norm": 0.6111416816711426,
+      "learning_rate": 0.0001,
+      "loss": 1.8656,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6813987372510928,
+      "grad_norm": 0.5777872204780579,
+      "learning_rate": 0.0001,
+      "loss": 1.6831,
+      "step": 2806
+    },
+    {
+      "epoch": 0.6816415735794075,
+      "grad_norm": 0.5720959305763245,
+      "learning_rate": 0.0001,
+      "loss": 1.8043,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6818844099077221,
+      "grad_norm": 0.5836105942726135,
+      "learning_rate": 0.0001,
+      "loss": 1.7525,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6821272462360369,
+      "grad_norm": 0.5708861351013184,
+      "learning_rate": 0.0001,
+      "loss": 1.7397,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6823700825643516,
+      "grad_norm": 0.6118195652961731,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6826129188926664,
+      "grad_norm": 0.6879733204841614,
+      "learning_rate": 0.0001,
+      "loss": 1.979,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6828557552209811,
+      "grad_norm": 0.5931374430656433,
+      "learning_rate": 0.0001,
+      "loss": 1.8943,
+      "step": 2812
+    },
+    {
+      "epoch": 0.6830985915492958,
+      "grad_norm": 0.5758197903633118,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 2813
+    },
+    {
+      "epoch": 0.6833414278776105,
+      "grad_norm": 0.6338828206062317,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6835842642059252,
+      "grad_norm": 0.6063835024833679,
+      "learning_rate": 0.0001,
+      "loss": 1.846,
+      "step": 2815
+    },
+    {
+      "epoch": 0.6838271005342399,
+      "grad_norm": 0.5691032409667969,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6840699368625547,
+      "grad_norm": 0.6327906250953674,
+      "learning_rate": 0.0001,
+      "loss": 1.8607,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6843127731908694,
+      "grad_norm": 0.5907915830612183,
+      "learning_rate": 0.0001,
+      "loss": 1.8197,
+      "step": 2818
+    },
+    {
+      "epoch": 0.684555609519184,
+      "grad_norm": 0.5953392386436462,
+      "learning_rate": 0.0001,
+      "loss": 1.804,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6847984458474988,
+      "grad_norm": 0.5982223153114319,
+      "learning_rate": 0.0001,
+      "loss": 1.6639,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6850412821758135,
+      "grad_norm": 0.5657848715782166,
+      "learning_rate": 0.0001,
+      "loss": 1.728,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6852841185041282,
+      "grad_norm": 0.603330671787262,
+      "learning_rate": 0.0001,
+      "loss": 1.7572,
+      "step": 2822
+    },
+    {
+      "epoch": 0.685526954832443,
+      "grad_norm": 0.5856655836105347,
+      "learning_rate": 0.0001,
+      "loss": 1.7514,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6857697911607576,
+      "grad_norm": 0.6200686693191528,
+      "learning_rate": 0.0001,
+      "loss": 1.7767,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6860126274890723,
+      "grad_norm": 0.5934330224990845,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 2825
+    },
+    {
+      "epoch": 0.6862554638173871,
+      "grad_norm": 0.5886359810829163,
+      "learning_rate": 0.0001,
+      "loss": 1.8898,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6864983001457018,
+      "grad_norm": 0.592971920967102,
+      "learning_rate": 0.0001,
+      "loss": 1.8712,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6867411364740165,
+      "grad_norm": 0.5992989540100098,
+      "learning_rate": 0.0001,
+      "loss": 1.7167,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6869839728023313,
+      "grad_norm": 0.5146058797836304,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 2829
+    },
+    {
+      "epoch": 0.6872268091306459,
+      "grad_norm": 0.576221764087677,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6874696454589606,
+      "grad_norm": 0.5706197619438171,
+      "learning_rate": 0.0001,
+      "loss": 1.7267,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6877124817872754,
+      "grad_norm": 0.5885552167892456,
+      "learning_rate": 0.0001,
+      "loss": 1.7293,
+      "step": 2832
+    },
+    {
+      "epoch": 0.6879553181155901,
+      "grad_norm": 0.5783635377883911,
+      "learning_rate": 0.0001,
+      "loss": 1.7528,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6881981544439048,
+      "grad_norm": 0.5897202491760254,
+      "learning_rate": 0.0001,
+      "loss": 1.7233,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6884409907722195,
+      "grad_norm": 0.5649927258491516,
+      "learning_rate": 0.0001,
+      "loss": 1.7574,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6886838271005342,
+      "grad_norm": 0.5524166822433472,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 2836
+    },
+    {
+      "epoch": 0.688926663428849,
+      "grad_norm": 0.5628504753112793,
+      "learning_rate": 0.0001,
+      "loss": 1.6541,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6891694997571637,
+      "grad_norm": 0.5964009165763855,
+      "learning_rate": 0.0001,
+      "loss": 1.802,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6894123360854784,
+      "grad_norm": 0.5477882623672485,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 0.6216742992401123,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6898980087421078,
+      "grad_norm": 0.5934717059135437,
+      "learning_rate": 0.0001,
+      "loss": 1.7309,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6901408450704225,
+      "grad_norm": 0.593638002872467,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6903836813987373,
+      "grad_norm": 0.5847916007041931,
+      "learning_rate": 0.0001,
+      "loss": 1.8504,
+      "step": 2843
+    },
+    {
+      "epoch": 0.690626517727052,
+      "grad_norm": 0.5497144460678101,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6908693540553666,
+      "grad_norm": 0.5658949613571167,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6911121903836814,
+      "grad_norm": 0.6022396087646484,
+      "learning_rate": 0.0001,
+      "loss": 1.8237,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6913550267119961,
+      "grad_norm": 0.6041214466094971,
+      "learning_rate": 0.0001,
+      "loss": 1.7601,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6915978630403108,
+      "grad_norm": 0.6111637949943542,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6918406993686256,
+      "grad_norm": 0.5883265733718872,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6920835356969403,
+      "grad_norm": 0.5720015168190002,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6923263720252549,
+      "grad_norm": 0.5846550464630127,
+      "learning_rate": 0.0001,
+      "loss": 1.7651,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6925692083535697,
+      "grad_norm": 0.5911680459976196,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6928120446818844,
+      "grad_norm": 0.5872257351875305,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6930548810101991,
+      "grad_norm": 0.5911620259284973,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6932977173385139,
+      "grad_norm": 0.6094566583633423,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6935405536668285,
+      "grad_norm": 0.5861377120018005,
+      "learning_rate": 0.0001,
+      "loss": 1.6755,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6937833899951432,
+      "grad_norm": 0.7181078791618347,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 2857
+    },
+    {
+      "epoch": 0.694026226323458,
+      "grad_norm": 0.583500862121582,
+      "learning_rate": 0.0001,
+      "loss": 1.6292,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6942690626517727,
+      "grad_norm": 0.5601634383201599,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 2859
+    },
+    {
+      "epoch": 0.6945118989800875,
+      "grad_norm": 0.5888351798057556,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6947547353084021,
+      "grad_norm": 0.6004803776741028,
+      "learning_rate": 0.0001,
+      "loss": 1.7305,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6949975716367168,
+      "grad_norm": 0.6880599856376648,
+      "learning_rate": 0.0001,
+      "loss": 1.8187,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6952404079650316,
+      "grad_norm": 0.5957933664321899,
+      "learning_rate": 0.0001,
+      "loss": 1.6591,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6954832442933463,
+      "grad_norm": 0.5726467370986938,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 2864
+    },
+    {
+      "epoch": 0.695726080621661,
+      "grad_norm": 0.5528472661972046,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6959689169499758,
+      "grad_norm": 0.6097925901412964,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6962117532782904,
+      "grad_norm": 0.5837544202804565,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6964545896066051,
+      "grad_norm": 0.532633900642395,
+      "learning_rate": 0.0001,
+      "loss": 1.6354,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6966974259349199,
+      "grad_norm": 0.5818161368370056,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6969402622632346,
+      "grad_norm": 0.5662007331848145,
+      "learning_rate": 0.0001,
+      "loss": 1.6389,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6971830985915493,
+      "grad_norm": 0.5478448271751404,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 2871
+    },
+    {
+      "epoch": 0.697425934919864,
+      "grad_norm": 0.5718416571617126,
+      "learning_rate": 0.0001,
+      "loss": 1.6713,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6976687712481787,
+      "grad_norm": 0.5725582242012024,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6979116075764934,
+      "grad_norm": 0.6120561957359314,
+      "learning_rate": 0.0001,
+      "loss": 1.8605,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6981544439048082,
+      "grad_norm": 0.5600008964538574,
+      "learning_rate": 0.0001,
+      "loss": 1.8146,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6983972802331229,
+      "grad_norm": 0.5600332021713257,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6986401165614375,
+      "grad_norm": 0.5910006165504456,
+      "learning_rate": 0.0001,
+      "loss": 1.7757,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6988829528897523,
+      "grad_norm": 0.5550327897071838,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 2878
+    },
+    {
+      "epoch": 0.699125789218067,
+      "grad_norm": 0.5717950463294983,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6993686255463817,
+      "grad_norm": 0.5722221732139587,
+      "learning_rate": 0.0001,
+      "loss": 1.7925,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6996114618746965,
+      "grad_norm": 0.6006417870521545,
+      "learning_rate": 0.0001,
+      "loss": 1.6632,
+      "step": 2881
+    },
+    {
+      "epoch": 0.6998542982030111,
+      "grad_norm": 0.5942846536636353,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7000971345313259,
+      "grad_norm": 0.5749700665473938,
+      "learning_rate": 0.0001,
+      "loss": 1.6598,
+      "step": 2883
+    },
+    {
+      "epoch": 0.7003399708596406,
+      "grad_norm": 0.5908930897712708,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7005828071879553,
+      "grad_norm": 0.5929084420204163,
+      "learning_rate": 0.0001,
+      "loss": 1.8005,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7008256435162701,
+      "grad_norm": 0.7345240116119385,
+      "learning_rate": 0.0001,
+      "loss": 1.7273,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7010684798445848,
+      "grad_norm": 0.5420176386833191,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 2887
+    },
+    {
+      "epoch": 0.7013113161728994,
+      "grad_norm": 0.5552915334701538,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7015541525012142,
+      "grad_norm": 0.5782786011695862,
+      "learning_rate": 0.0001,
+      "loss": 1.8232,
+      "step": 2889
+    },
+    {
+      "epoch": 0.7017969888295289,
+      "grad_norm": 0.836561381816864,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7020398251578436,
+      "grad_norm": 0.569405734539032,
+      "learning_rate": 0.0001,
+      "loss": 1.7464,
+      "step": 2891
+    },
+    {
+      "epoch": 0.7022826614861584,
+      "grad_norm": 0.5833035111427307,
+      "learning_rate": 0.0001,
+      "loss": 1.7176,
+      "step": 2892
+    },
+    {
+      "epoch": 0.702525497814473,
+      "grad_norm": 0.5674729943275452,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7027683341427877,
+      "grad_norm": 0.5252358317375183,
+      "learning_rate": 0.0001,
+      "loss": 1.3547,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7030111704711025,
+      "grad_norm": 0.5658174157142639,
+      "learning_rate": 0.0001,
+      "loss": 1.7606,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7032540067994172,
+      "grad_norm": 0.5492091774940491,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7034968431277319,
+      "grad_norm": 0.58930504322052,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7037396794560467,
+      "grad_norm": 0.6033841371536255,
+      "learning_rate": 0.0001,
+      "loss": 1.748,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7039825157843613,
+      "grad_norm": 0.5531272888183594,
+      "learning_rate": 0.0001,
+      "loss": 1.7049,
+      "step": 2899
+    },
+    {
+      "epoch": 0.704225352112676,
+      "grad_norm": 0.5709564685821533,
+      "learning_rate": 0.0001,
+      "loss": 1.6982,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7044681884409908,
+      "grad_norm": 0.5532764196395874,
+      "learning_rate": 0.0001,
+      "loss": 1.6809,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7047110247693055,
+      "grad_norm": 0.6018196940422058,
+      "learning_rate": 0.0001,
+      "loss": 1.764,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7049538610976201,
+      "grad_norm": 0.6237990856170654,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7051966974259349,
+      "grad_norm": 0.5766505599021912,
+      "learning_rate": 0.0001,
+      "loss": 1.7848,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7054395337542496,
+      "grad_norm": 0.5628949999809265,
+      "learning_rate": 0.0001,
+      "loss": 1.8078,
+      "step": 2905
+    },
+    {
+      "epoch": 0.7056823700825644,
+      "grad_norm": 0.5653403401374817,
+      "learning_rate": 0.0001,
+      "loss": 1.7858,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7059252064108791,
+      "grad_norm": 0.5594356060028076,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7061680427391938,
+      "grad_norm": 0.590790867805481,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7064108790675085,
+      "grad_norm": 0.6296771764755249,
+      "learning_rate": 0.0001,
+      "loss": 1.8381,
+      "step": 2909
+    },
+    {
+      "epoch": 0.7066537153958232,
+      "grad_norm": 0.6157625913619995,
+      "learning_rate": 0.0001,
+      "loss": 1.8493,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7068965517241379,
+      "grad_norm": 0.6185581684112549,
+      "learning_rate": 0.0001,
+      "loss": 1.8729,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7071393880524527,
+      "grad_norm": 0.6104115843772888,
+      "learning_rate": 0.0001,
+      "loss": 1.8072,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7073822243807674,
+      "grad_norm": 0.5635713934898376,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 2913
+    },
+    {
+      "epoch": 0.707625060709082,
+      "grad_norm": 0.5789936780929565,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7078678970373968,
+      "grad_norm": 0.5909901857376099,
+      "learning_rate": 0.0001,
+      "loss": 1.8809,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7081107333657115,
+      "grad_norm": 0.6040669083595276,
+      "learning_rate": 0.0001,
+      "loss": 1.7442,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7083535696940262,
+      "grad_norm": 0.5579518675804138,
+      "learning_rate": 0.0001,
+      "loss": 1.7682,
+      "step": 2917
+    },
+    {
+      "epoch": 0.708596406022341,
+      "grad_norm": 0.6221187114715576,
+      "learning_rate": 0.0001,
+      "loss": 1.7396,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7088392423506557,
+      "grad_norm": 0.5807497501373291,
+      "learning_rate": 0.0001,
+      "loss": 1.918,
+      "step": 2919
+    },
+    {
+      "epoch": 0.7090820786789703,
+      "grad_norm": 0.5804089307785034,
+      "learning_rate": 0.0001,
+      "loss": 1.7405,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7093249150072851,
+      "grad_norm": 0.6316050887107849,
+      "learning_rate": 0.0001,
+      "loss": 1.8899,
+      "step": 2921
+    },
+    {
+      "epoch": 0.7095677513355998,
+      "grad_norm": 0.598678708076477,
+      "learning_rate": 0.0001,
+      "loss": 1.8111,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7098105876639145,
+      "grad_norm": 0.5926578640937805,
+      "learning_rate": 0.0001,
+      "loss": 1.809,
+      "step": 2923
+    },
+    {
+      "epoch": 0.7100534239922293,
+      "grad_norm": 0.5795942544937134,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7102962603205439,
+      "grad_norm": 0.6003085970878601,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 2925
+    },
+    {
+      "epoch": 0.7105390966488586,
+      "grad_norm": 0.6050456166267395,
+      "learning_rate": 0.0001,
+      "loss": 1.7144,
+      "step": 2926
+    },
+    {
+      "epoch": 0.7107819329771734,
+      "grad_norm": 0.5947707295417786,
+      "learning_rate": 0.0001,
+      "loss": 1.7912,
+      "step": 2927
+    },
+    {
+      "epoch": 0.7110247693054881,
+      "grad_norm": 0.5814579129219055,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 2928
+    },
+    {
+      "epoch": 0.7112676056338029,
+      "grad_norm": 0.5886326432228088,
+      "learning_rate": 0.0001,
+      "loss": 1.7891,
+      "step": 2929
+    },
+    {
+      "epoch": 0.7115104419621175,
+      "grad_norm": 0.6063361167907715,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7117532782904322,
+      "grad_norm": 0.6063582897186279,
+      "learning_rate": 0.0001,
+      "loss": 1.83,
+      "step": 2931
+    },
+    {
+      "epoch": 0.711996114618747,
+      "grad_norm": 0.5683070421218872,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 2932
+    },
+    {
+      "epoch": 0.7122389509470617,
+      "grad_norm": 0.5507863759994507,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 2933
+    },
+    {
+      "epoch": 0.7124817872753764,
+      "grad_norm": 0.5840550065040588,
+      "learning_rate": 0.0001,
+      "loss": 1.848,
+      "step": 2934
+    },
+    {
+      "epoch": 0.7127246236036912,
+      "grad_norm": 0.5971823930740356,
+      "learning_rate": 0.0001,
+      "loss": 1.8814,
+      "step": 2935
+    },
+    {
+      "epoch": 0.7129674599320058,
+      "grad_norm": 0.5722901821136475,
+      "learning_rate": 0.0001,
+      "loss": 1.7001,
+      "step": 2936
+    },
+    {
+      "epoch": 0.7132102962603205,
+      "grad_norm": 0.6489852070808411,
+      "learning_rate": 0.0001,
+      "loss": 1.8752,
+      "step": 2937
+    },
+    {
+      "epoch": 0.7134531325886353,
+      "grad_norm": 0.5677469968795776,
+      "learning_rate": 0.0001,
+      "loss": 1.7549,
+      "step": 2938
+    },
+    {
+      "epoch": 0.71369596891695,
+      "grad_norm": 0.5717501044273376,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 2939
+    },
+    {
+      "epoch": 0.7139388052452647,
+      "grad_norm": 0.5825884342193604,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7141816415735794,
+      "grad_norm": 0.6087138056755066,
+      "learning_rate": 0.0001,
+      "loss": 1.8682,
+      "step": 2941
+    },
+    {
+      "epoch": 0.7144244779018941,
+      "grad_norm": 0.5498895645141602,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7146673142302088,
+      "grad_norm": 0.5925459861755371,
+      "learning_rate": 0.0001,
+      "loss": 1.7232,
+      "step": 2943
+    },
+    {
+      "epoch": 0.7149101505585236,
+      "grad_norm": 0.5935659408569336,
+      "learning_rate": 0.0001,
+      "loss": 1.8265,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7151529868868383,
+      "grad_norm": 0.5577383637428284,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 2945
+    },
+    {
+      "epoch": 0.7153958232151529,
+      "grad_norm": 0.5835193395614624,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7156386595434677,
+      "grad_norm": 0.5707915425300598,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 2947
+    },
+    {
+      "epoch": 0.7158814958717824,
+      "grad_norm": 0.6077779531478882,
+      "learning_rate": 0.0001,
+      "loss": 1.9231,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7161243322000971,
+      "grad_norm": 0.5671358108520508,
+      "learning_rate": 0.0001,
+      "loss": 1.7163,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7163671685284119,
+      "grad_norm": 0.8459880948066711,
+      "learning_rate": 0.0001,
+      "loss": 1.7304,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7166100048567265,
+      "grad_norm": 0.5568290948867798,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 2951
+    },
+    {
+      "epoch": 0.7168528411850413,
+      "grad_norm": 0.5902974009513855,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 2952
+    },
+    {
+      "epoch": 0.717095677513356,
+      "grad_norm": 0.6150861382484436,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 2953
+    },
+    {
+      "epoch": 0.7173385138416707,
+      "grad_norm": 0.5791119337081909,
+      "learning_rate": 0.0001,
+      "loss": 1.6013,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7175813501699855,
+      "grad_norm": 0.5551214814186096,
+      "learning_rate": 0.0001,
+      "loss": 1.5818,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7178241864983002,
+      "grad_norm": 0.5971116423606873,
+      "learning_rate": 0.0001,
+      "loss": 1.7887,
+      "step": 2956
+    },
+    {
+      "epoch": 0.7180670228266148,
+      "grad_norm": 0.5904876589775085,
+      "learning_rate": 0.0001,
+      "loss": 1.7936,
+      "step": 2957
+    },
+    {
+      "epoch": 0.7183098591549296,
+      "grad_norm": 0.5888270735740662,
+      "learning_rate": 0.0001,
+      "loss": 1.7204,
+      "step": 2958
+    },
+    {
+      "epoch": 0.7185526954832443,
+      "grad_norm": 0.5791041254997253,
+      "learning_rate": 0.0001,
+      "loss": 1.8105,
+      "step": 2959
+    },
+    {
+      "epoch": 0.718795531811559,
+      "grad_norm": 0.5695481300354004,
+      "learning_rate": 0.0001,
+      "loss": 1.684,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7190383681398738,
+      "grad_norm": 0.5556251406669617,
+      "learning_rate": 0.0001,
+      "loss": 1.6875,
+      "step": 2961
+    },
+    {
+      "epoch": 0.7192812044681884,
+      "grad_norm": 0.5751869082450867,
+      "learning_rate": 0.0001,
+      "loss": 1.6599,
+      "step": 2962
+    },
+    {
+      "epoch": 0.7195240407965031,
+      "grad_norm": 0.6280362010002136,
+      "learning_rate": 0.0001,
+      "loss": 2.0623,
+      "step": 2963
+    },
+    {
+      "epoch": 0.7197668771248179,
+      "grad_norm": 0.5435699820518494,
+      "learning_rate": 0.0001,
+      "loss": 1.8137,
+      "step": 2964
+    },
+    {
+      "epoch": 0.7200097134531326,
+      "grad_norm": 0.5561855435371399,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 2965
+    },
+    {
+      "epoch": 0.7202525497814473,
+      "grad_norm": 0.5420215725898743,
+      "learning_rate": 0.0001,
+      "loss": 1.6225,
+      "step": 2966
+    },
+    {
+      "epoch": 0.720495386109762,
+      "grad_norm": 0.5700600743293762,
+      "learning_rate": 0.0001,
+      "loss": 1.8171,
+      "step": 2967
+    },
+    {
+      "epoch": 0.7207382224380767,
+      "grad_norm": 0.6144974827766418,
+      "learning_rate": 0.0001,
+      "loss": 1.8919,
+      "step": 2968
+    },
+    {
+      "epoch": 0.7209810587663914,
+      "grad_norm": 0.6396198272705078,
+      "learning_rate": 0.0001,
+      "loss": 1.7337,
+      "step": 2969
+    },
+    {
+      "epoch": 0.7212238950947062,
+      "grad_norm": 0.584147572517395,
+      "learning_rate": 0.0001,
+      "loss": 1.8727,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7214667314230209,
+      "grad_norm": 0.552257239818573,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 2971
+    },
+    {
+      "epoch": 0.7217095677513355,
+      "grad_norm": 0.5884346961975098,
+      "learning_rate": 0.0001,
+      "loss": 1.7232,
+      "step": 2972
+    },
+    {
+      "epoch": 0.7219524040796503,
+      "grad_norm": 0.5640490651130676,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 2973
+    },
+    {
+      "epoch": 0.722195240407965,
+      "grad_norm": 0.5864781737327576,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7224380767362798,
+      "grad_norm": 0.5875230431556702,
+      "learning_rate": 0.0001,
+      "loss": 1.8265,
+      "step": 2975
+    },
+    {
+      "epoch": 0.7226809130645945,
+      "grad_norm": 0.6013203263282776,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 2976
+    },
+    {
+      "epoch": 0.7229237493929092,
+      "grad_norm": 0.5607584118843079,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 2977
+    },
+    {
+      "epoch": 0.7231665857212239,
+      "grad_norm": 0.6358622908592224,
+      "learning_rate": 0.0001,
+      "loss": 1.7346,
+      "step": 2978
+    },
+    {
+      "epoch": 0.7234094220495386,
+      "grad_norm": 0.5818342566490173,
+      "learning_rate": 0.0001,
+      "loss": 1.7539,
+      "step": 2979
+    },
+    {
+      "epoch": 0.7236522583778533,
+      "grad_norm": 0.5819612741470337,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7238950947061681,
+      "grad_norm": 0.5618541836738586,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 2981
+    },
+    {
+      "epoch": 0.7241379310344828,
+      "grad_norm": 0.5999396443367004,
+      "learning_rate": 0.0001,
+      "loss": 1.8407,
+      "step": 2982
+    },
+    {
+      "epoch": 0.7243807673627974,
+      "grad_norm": 0.5402510762214661,
+      "learning_rate": 0.0001,
+      "loss": 1.688,
+      "step": 2983
+    },
+    {
+      "epoch": 0.7246236036911122,
+      "grad_norm": 0.6189697980880737,
+      "learning_rate": 0.0001,
+      "loss": 1.7554,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7248664400194269,
+      "grad_norm": 0.6344261169433594,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7251092763477416,
+      "grad_norm": 0.6194679737091064,
+      "learning_rate": 0.0001,
+      "loss": 1.887,
+      "step": 2986
+    },
+    {
+      "epoch": 0.7253521126760564,
+      "grad_norm": 0.6036260724067688,
+      "learning_rate": 0.0001,
+      "loss": 1.7307,
+      "step": 2987
+    },
+    {
+      "epoch": 0.725594949004371,
+      "grad_norm": 0.5975467562675476,
+      "learning_rate": 0.0001,
+      "loss": 1.6809,
+      "step": 2988
+    },
+    {
+      "epoch": 0.7258377853326857,
+      "grad_norm": 0.5954137444496155,
+      "learning_rate": 0.0001,
+      "loss": 1.724,
+      "step": 2989
+    },
+    {
+      "epoch": 0.7260806216610005,
+      "grad_norm": 0.6179779171943665,
+      "learning_rate": 0.0001,
+      "loss": 1.8012,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7263234579893152,
+      "grad_norm": 0.5949034690856934,
+      "learning_rate": 0.0001,
+      "loss": 1.7704,
+      "step": 2991
+    },
+    {
+      "epoch": 0.7265662943176299,
+      "grad_norm": 0.5713704228401184,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 2992
+    },
+    {
+      "epoch": 0.7268091306459447,
+      "grad_norm": 0.5797938108444214,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 2993
+    },
+    {
+      "epoch": 0.7270519669742593,
+      "grad_norm": 0.6108956933021545,
+      "learning_rate": 0.0001,
+      "loss": 1.7338,
+      "step": 2994
+    },
+    {
+      "epoch": 0.727294803302574,
+      "grad_norm": 0.6109453439712524,
+      "learning_rate": 0.0001,
+      "loss": 1.6392,
+      "step": 2995
+    },
+    {
+      "epoch": 0.7275376396308888,
+      "grad_norm": 0.5588051676750183,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 2996
+    },
+    {
+      "epoch": 0.7277804759592035,
+      "grad_norm": 0.5878009796142578,
+      "learning_rate": 0.0001,
+      "loss": 1.7112,
+      "step": 2997
+    },
+    {
+      "epoch": 0.7280233122875183,
+      "grad_norm": 0.605579674243927,
+      "learning_rate": 0.0001,
+      "loss": 1.7491,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7282661486158329,
+      "grad_norm": 0.5869989395141602,
+      "learning_rate": 0.0001,
+      "loss": 1.7755,
+      "step": 2999
+    },
+    {
+      "epoch": 0.7285089849441476,
+      "grad_norm": 0.5857266783714294,
+      "learning_rate": 0.0001,
+      "loss": 1.8855,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7287518212724624,
+      "grad_norm": 0.5307328104972839,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 3001
+    },
+    {
+      "epoch": 0.7289946576007771,
+      "grad_norm": 0.5867645740509033,
+      "learning_rate": 0.0001,
+      "loss": 1.7953,
+      "step": 3002
+    },
+    {
+      "epoch": 0.7292374939290918,
+      "grad_norm": 0.5966058373451233,
+      "learning_rate": 0.0001,
+      "loss": 1.8166,
+      "step": 3003
+    },
+    {
+      "epoch": 0.7294803302574066,
+      "grad_norm": 0.5763681530952454,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 3004
+    },
+    {
+      "epoch": 0.7297231665857212,
+      "grad_norm": 0.5982712507247925,
+      "learning_rate": 0.0001,
+      "loss": 1.7192,
+      "step": 3005
+    },
+    {
+      "epoch": 0.7299660029140359,
+      "grad_norm": 0.5639309883117676,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7302088392423507,
+      "grad_norm": 0.6086863279342651,
+      "learning_rate": 0.0001,
+      "loss": 1.8071,
+      "step": 3007
+    },
+    {
+      "epoch": 0.7304516755706654,
+      "grad_norm": 0.6218518614768982,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 3008
+    },
+    {
+      "epoch": 0.73069451189898,
+      "grad_norm": 0.6090829372406006,
+      "learning_rate": 0.0001,
+      "loss": 1.915,
+      "step": 3009
+    },
+    {
+      "epoch": 0.7309373482272948,
+      "grad_norm": 0.6307433247566223,
+      "learning_rate": 0.0001,
+      "loss": 1.8065,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7311801845556095,
+      "grad_norm": 0.5762611627578735,
+      "learning_rate": 0.0001,
+      "loss": 1.8111,
+      "step": 3011
+    },
+    {
+      "epoch": 0.7314230208839242,
+      "grad_norm": 0.5527024865150452,
+      "learning_rate": 0.0001,
+      "loss": 1.513,
+      "step": 3012
+    },
+    {
+      "epoch": 0.731665857212239,
+      "grad_norm": 0.5604642033576965,
+      "learning_rate": 0.0001,
+      "loss": 1.6819,
+      "step": 3013
+    },
+    {
+      "epoch": 0.7319086935405537,
+      "grad_norm": 0.5796170830726624,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 3014
+    },
+    {
+      "epoch": 0.7321515298688683,
+      "grad_norm": 0.5859696269035339,
+      "learning_rate": 0.0001,
+      "loss": 1.7098,
+      "step": 3015
+    },
+    {
+      "epoch": 0.7323943661971831,
+      "grad_norm": 0.5520855188369751,
+      "learning_rate": 0.0001,
+      "loss": 1.6901,
+      "step": 3016
+    },
+    {
+      "epoch": 0.7326372025254978,
+      "grad_norm": 0.5656354427337646,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 3017
+    },
+    {
+      "epoch": 0.7328800388538125,
+      "grad_norm": 0.5668532848358154,
+      "learning_rate": 0.0001,
+      "loss": 1.7174,
+      "step": 3018
+    },
+    {
+      "epoch": 0.7331228751821273,
+      "grad_norm": 0.606231689453125,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 3019
+    },
+    {
+      "epoch": 0.7333657115104419,
+      "grad_norm": 0.5504002571105957,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7336085478387567,
+      "grad_norm": 0.6125749349594116,
+      "learning_rate": 0.0001,
+      "loss": 1.7903,
+      "step": 3021
+    },
+    {
+      "epoch": 0.7338513841670714,
+      "grad_norm": 0.6384183168411255,
+      "learning_rate": 0.0001,
+      "loss": 1.8925,
+      "step": 3022
+    },
+    {
+      "epoch": 0.7340942204953861,
+      "grad_norm": 0.5936228036880493,
+      "learning_rate": 0.0001,
+      "loss": 1.666,
+      "step": 3023
+    },
+    {
+      "epoch": 0.7343370568237009,
+      "grad_norm": 0.5869163870811462,
+      "learning_rate": 0.0001,
+      "loss": 1.7754,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7345798931520156,
+      "grad_norm": 0.6066444516181946,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 3025
+    },
+    {
+      "epoch": 0.7348227294803302,
+      "grad_norm": 0.5754687190055847,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 3026
+    },
+    {
+      "epoch": 0.735065565808645,
+      "grad_norm": 0.5544506311416626,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 3027
+    },
+    {
+      "epoch": 0.7353084021369597,
+      "grad_norm": 0.5787808895111084,
+      "learning_rate": 0.0001,
+      "loss": 1.6657,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7355512384652744,
+      "grad_norm": 0.5914311408996582,
+      "learning_rate": 0.0001,
+      "loss": 1.8207,
+      "step": 3029
+    },
+    {
+      "epoch": 0.7357940747935892,
+      "grad_norm": 0.5448651313781738,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7360369111219038,
+      "grad_norm": 0.572437584400177,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 3031
+    },
+    {
+      "epoch": 0.7362797474502185,
+      "grad_norm": 0.6246130466461182,
+      "learning_rate": 0.0001,
+      "loss": 1.7405,
+      "step": 3032
+    },
+    {
+      "epoch": 0.7365225837785333,
+      "grad_norm": 0.5604763627052307,
+      "learning_rate": 0.0001,
+      "loss": 1.6263,
+      "step": 3033
+    },
+    {
+      "epoch": 0.736765420106848,
+      "grad_norm": 0.5959675312042236,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 3034
+    },
+    {
+      "epoch": 0.7370082564351627,
+      "grad_norm": 0.5705462694168091,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 3035
+    },
+    {
+      "epoch": 0.7372510927634774,
+      "grad_norm": 0.6315643787384033,
+      "learning_rate": 0.0001,
+      "loss": 1.7861,
+      "step": 3036
+    },
+    {
+      "epoch": 0.7374939290917921,
+      "grad_norm": 0.5820364356040955,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 3037
+    },
+    {
+      "epoch": 0.7377367654201068,
+      "grad_norm": 0.5807769894599915,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7379796017484216,
+      "grad_norm": 0.5841823220252991,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 3039
+    },
+    {
+      "epoch": 0.7382224380767363,
+      "grad_norm": 0.5556533932685852,
+      "learning_rate": 0.0001,
+      "loss": 1.6971,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7384652744050509,
+      "grad_norm": 0.5819579362869263,
+      "learning_rate": 0.0001,
+      "loss": 1.7295,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7387081107333657,
+      "grad_norm": 0.6066816449165344,
+      "learning_rate": 0.0001,
+      "loss": 1.8619,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7389509470616804,
+      "grad_norm": 0.5626749396324158,
+      "learning_rate": 0.0001,
+      "loss": 1.696,
+      "step": 3043
+    },
+    {
+      "epoch": 0.7391937833899952,
+      "grad_norm": 0.5701555609703064,
+      "learning_rate": 0.0001,
+      "loss": 1.7388,
+      "step": 3044
+    },
+    {
+      "epoch": 0.7394366197183099,
+      "grad_norm": 0.6135877966880798,
+      "learning_rate": 0.0001,
+      "loss": 1.7617,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7396794560466246,
+      "grad_norm": 0.6123836040496826,
+      "learning_rate": 0.0001,
+      "loss": 1.7067,
+      "step": 3046
+    },
+    {
+      "epoch": 0.7399222923749393,
+      "grad_norm": 0.586012065410614,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 3047
+    },
+    {
+      "epoch": 0.740165128703254,
+      "grad_norm": 0.6175816655158997,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 3048
+    },
+    {
+      "epoch": 0.7404079650315687,
+      "grad_norm": 0.6169615387916565,
+      "learning_rate": 0.0001,
+      "loss": 1.7067,
+      "step": 3049
+    },
+    {
+      "epoch": 0.7406508013598835,
+      "grad_norm": 0.5519340634346008,
+      "learning_rate": 0.0001,
+      "loss": 1.7388,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7408936376881982,
+      "grad_norm": 0.5570865869522095,
+      "learning_rate": 0.0001,
+      "loss": 1.7541,
+      "step": 3051
+    },
+    {
+      "epoch": 0.7411364740165128,
+      "grad_norm": 0.56632399559021,
+      "learning_rate": 0.0001,
+      "loss": 1.7458,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7413793103448276,
+      "grad_norm": 0.6042296290397644,
+      "learning_rate": 0.0001,
+      "loss": 1.7227,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7416221466731423,
+      "grad_norm": 0.633659303188324,
+      "learning_rate": 0.0001,
+      "loss": 1.8576,
+      "step": 3054
+    },
+    {
+      "epoch": 0.741864983001457,
+      "grad_norm": 0.6423869729042053,
+      "learning_rate": 0.0001,
+      "loss": 1.7807,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7421078193297718,
+      "grad_norm": 0.5772718787193298,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 3056
+    },
+    {
+      "epoch": 0.7423506556580864,
+      "grad_norm": 0.5670206546783447,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7425934919864011,
+      "grad_norm": 0.5802311301231384,
+      "learning_rate": 0.0001,
+      "loss": 1.6043,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7428363283147159,
+      "grad_norm": 0.6183196306228638,
+      "learning_rate": 0.0001,
+      "loss": 1.6892,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7430791646430306,
+      "grad_norm": 0.5805888772010803,
+      "learning_rate": 0.0001,
+      "loss": 1.91,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7433220009713453,
+      "grad_norm": 0.5857672095298767,
+      "learning_rate": 0.0001,
+      "loss": 1.7233,
+      "step": 3061
+    },
+    {
+      "epoch": 0.74356483729966,
+      "grad_norm": 0.6274641156196594,
+      "learning_rate": 0.0001,
+      "loss": 1.7267,
+      "step": 3062
+    },
+    {
+      "epoch": 0.7438076736279747,
+      "grad_norm": 0.5870610475540161,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7440505099562894,
+      "grad_norm": 0.5928947925567627,
+      "learning_rate": 0.0001,
+      "loss": 1.7029,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7442933462846042,
+      "grad_norm": 0.6157515645027161,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7445361826129189,
+      "grad_norm": 0.5869420170783997,
+      "learning_rate": 0.0001,
+      "loss": 1.7315,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7447790189412337,
+      "grad_norm": 0.6070082783699036,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7450218552695483,
+      "grad_norm": 0.5500762462615967,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 3068
+    },
+    {
+      "epoch": 0.745264691597863,
+      "grad_norm": 0.5982913374900818,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7455075279261778,
+      "grad_norm": 0.5838366150856018,
+      "learning_rate": 0.0001,
+      "loss": 1.7077,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7457503642544925,
+      "grad_norm": 0.5614213943481445,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7459932005828072,
+      "grad_norm": 0.6322265863418579,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 3072
+    },
+    {
+      "epoch": 0.746236036911122,
+      "grad_norm": 0.594893753528595,
+      "learning_rate": 0.0001,
+      "loss": 1.8296,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7464788732394366,
+      "grad_norm": 0.6140832901000977,
+      "learning_rate": 0.0001,
+      "loss": 1.6699,
+      "step": 3074
+    },
+    {
+      "epoch": 0.7467217095677513,
+      "grad_norm": 0.6009077429771423,
+      "learning_rate": 0.0001,
+      "loss": 1.6844,
+      "step": 3075
+    },
+    {
+      "epoch": 0.7469645458960661,
+      "grad_norm": 0.608458399772644,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7472073822243808,
+      "grad_norm": 0.5779408812522888,
+      "learning_rate": 0.0001,
+      "loss": 1.7067,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7474502185526954,
+      "grad_norm": 0.5849694609642029,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7476930548810102,
+      "grad_norm": 0.6309943199157715,
+      "learning_rate": 0.0001,
+      "loss": 1.7543,
+      "step": 3079
+    },
+    {
+      "epoch": 0.7479358912093249,
+      "grad_norm": 0.6028780937194824,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7481787275376396,
+      "grad_norm": 0.6136448979377747,
+      "learning_rate": 0.0001,
+      "loss": 1.7322,
+      "step": 3081
+    },
+    {
+      "epoch": 0.7484215638659544,
+      "grad_norm": 0.6148863434791565,
+      "learning_rate": 0.0001,
+      "loss": 1.733,
+      "step": 3082
+    },
+    {
+      "epoch": 0.748664400194269,
+      "grad_norm": 0.5814813375473022,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 3083
+    },
+    {
+      "epoch": 0.7489072365225837,
+      "grad_norm": 0.6196994185447693,
+      "learning_rate": 0.0001,
+      "loss": 1.7411,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7491500728508985,
+      "grad_norm": 0.5938546657562256,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7493929091792132,
+      "grad_norm": 0.6048294305801392,
+      "learning_rate": 0.0001,
+      "loss": 1.7233,
+      "step": 3086
+    },
+    {
+      "epoch": 0.7496357455075279,
+      "grad_norm": 0.6158680319786072,
+      "learning_rate": 0.0001,
+      "loss": 1.765,
+      "step": 3087
+    },
+    {
+      "epoch": 0.7498785818358427,
+      "grad_norm": 0.6024149060249329,
+      "learning_rate": 0.0001,
+      "loss": 1.7101,
+      "step": 3088
+    },
+    {
+      "epoch": 0.7501214181641573,
+      "grad_norm": 0.6003633141517639,
+      "learning_rate": 0.0001,
+      "loss": 1.8116,
+      "step": 3089
+    },
+    {
+      "epoch": 0.7503642544924721,
+      "grad_norm": 0.5719137191772461,
+      "learning_rate": 0.0001,
+      "loss": 1.8171,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7506070908207868,
+      "grad_norm": 0.5652117729187012,
+      "learning_rate": 0.0001,
+      "loss": 1.7708,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7508499271491015,
+      "grad_norm": 0.5504488348960876,
+      "learning_rate": 0.0001,
+      "loss": 1.6541,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7510927634774163,
+      "grad_norm": 0.6070501804351807,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 3093
+    },
+    {
+      "epoch": 0.751335599805731,
+      "grad_norm": 0.6052151918411255,
+      "learning_rate": 0.0001,
+      "loss": 1.7172,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7515784361340456,
+      "grad_norm": 0.6147742867469788,
+      "learning_rate": 0.0001,
+      "loss": 1.7188,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7518212724623604,
+      "grad_norm": 0.5747918486595154,
+      "learning_rate": 0.0001,
+      "loss": 1.5887,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7520641087906751,
+      "grad_norm": 0.5973472595214844,
+      "learning_rate": 0.0001,
+      "loss": 1.7014,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7523069451189898,
+      "grad_norm": 0.6190682053565979,
+      "learning_rate": 0.0001,
+      "loss": 1.7363,
+      "step": 3098
+    },
+    {
+      "epoch": 0.7525497814473046,
+      "grad_norm": 0.5895213484764099,
+      "learning_rate": 0.0001,
+      "loss": 1.8408,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7527926177756192,
+      "grad_norm": 0.6008792519569397,
+      "learning_rate": 0.0001,
+      "loss": 1.8563,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7530354541039339,
+      "grad_norm": 5.0266032218933105,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 3101
+    },
+    {
+      "epoch": 0.7532782904322487,
+      "grad_norm": 0.5932336449623108,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7535211267605634,
+      "grad_norm": 0.5837019085884094,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 3103
+    },
+    {
+      "epoch": 0.753763963088878,
+      "grad_norm": 0.5850164890289307,
+      "learning_rate": 0.0001,
+      "loss": 1.7852,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7540067994171928,
+      "grad_norm": 0.5760244131088257,
+      "learning_rate": 0.0001,
+      "loss": 1.7942,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7542496357455075,
+      "grad_norm": 0.6135173439979553,
+      "learning_rate": 0.0001,
+      "loss": 1.9088,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7544924720738222,
+      "grad_norm": 0.597471296787262,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 3107
+    },
+    {
+      "epoch": 0.754735308402137,
+      "grad_norm": 0.606307864189148,
+      "learning_rate": 0.0001,
+      "loss": 1.8266,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7549781447304517,
+      "grad_norm": 0.5847671031951904,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7552209810587663,
+      "grad_norm": 0.5444035530090332,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7554638173870811,
+      "grad_norm": 0.5614792704582214,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7557066537153958,
+      "grad_norm": 0.5745176672935486,
+      "learning_rate": 0.0001,
+      "loss": 1.6814,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7559494900437106,
+      "grad_norm": 0.5440202355384827,
+      "learning_rate": 0.0001,
+      "loss": 1.7116,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7561923263720253,
+      "grad_norm": 0.5792099237442017,
+      "learning_rate": 0.0001,
+      "loss": 1.8194,
+      "step": 3114
+    },
+    {
+      "epoch": 0.75643516270034,
+      "grad_norm": 0.5704084634780884,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7566779990286547,
+      "grad_norm": 0.6286868453025818,
+      "learning_rate": 0.0001,
+      "loss": 1.8912,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7569208353569694,
+      "grad_norm": 0.620141863822937,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7571636716852841,
+      "grad_norm": 0.5992688536643982,
+      "learning_rate": 0.0001,
+      "loss": 1.7075,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7574065080135989,
+      "grad_norm": 0.5898826122283936,
+      "learning_rate": 0.0001,
+      "loss": 1.7085,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7576493443419136,
+      "grad_norm": 0.5879241228103638,
+      "learning_rate": 0.0001,
+      "loss": 1.6773,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7578921806702282,
+      "grad_norm": 0.5776136517524719,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 3121
+    },
+    {
+      "epoch": 0.758135016998543,
+      "grad_norm": 0.5695837736129761,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7583778533268577,
+      "grad_norm": 0.5705963969230652,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 0.601582407951355,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7588635259834872,
+      "grad_norm": 0.6535794138908386,
+      "learning_rate": 0.0001,
+      "loss": 1.7291,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7591063623118018,
+      "grad_norm": 0.5979652404785156,
+      "learning_rate": 0.0001,
+      "loss": 1.758,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7593491986401165,
+      "grad_norm": 0.5911663174629211,
+      "learning_rate": 0.0001,
+      "loss": 1.8147,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7595920349684313,
+      "grad_norm": 0.5932319760322571,
+      "learning_rate": 0.0001,
+      "loss": 1.8987,
+      "step": 3128
+    },
+    {
+      "epoch": 0.759834871296746,
+      "grad_norm": 0.6082956194877625,
+      "learning_rate": 0.0001,
+      "loss": 1.758,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7600777076250607,
+      "grad_norm": 0.5703206658363342,
+      "learning_rate": 0.0001,
+      "loss": 1.6945,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7603205439533754,
+      "grad_norm": 0.5616307854652405,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7605633802816901,
+      "grad_norm": 0.5814014673233032,
+      "learning_rate": 0.0001,
+      "loss": 1.7443,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7608062166100048,
+      "grad_norm": 0.5997304916381836,
+      "learning_rate": 0.0001,
+      "loss": 1.7166,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7610490529383196,
+      "grad_norm": 0.6064666509628296,
+      "learning_rate": 0.0001,
+      "loss": 1.9331,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7612918892666343,
+      "grad_norm": 0.5871558785438538,
+      "learning_rate": 0.0001,
+      "loss": 1.6392,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7615347255949491,
+      "grad_norm": 0.5561057925224304,
+      "learning_rate": 0.0001,
+      "loss": 1.6142,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7617775619232637,
+      "grad_norm": 0.6017075181007385,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7620203982515784,
+      "grad_norm": 0.5785053372383118,
+      "learning_rate": 0.0001,
+      "loss": 1.7708,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7622632345798932,
+      "grad_norm": 0.5785541534423828,
+      "learning_rate": 0.0001,
+      "loss": 1.7217,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7625060709082079,
+      "grad_norm": 0.6177810430526733,
+      "learning_rate": 0.0001,
+      "loss": 1.8104,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7627489072365226,
+      "grad_norm": 0.6104243993759155,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7629917435648373,
+      "grad_norm": 0.5768064260482788,
+      "learning_rate": 0.0001,
+      "loss": 1.761,
+      "step": 3142
+    },
+    {
+      "epoch": 0.763234579893152,
+      "grad_norm": 0.6506632566452026,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7634774162214667,
+      "grad_norm": 0.6132780313491821,
+      "learning_rate": 0.0001,
+      "loss": 1.7527,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7637202525497815,
+      "grad_norm": 0.5931320190429688,
+      "learning_rate": 0.0001,
+      "loss": 1.83,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7639630888780962,
+      "grad_norm": 0.5766931176185608,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7642059252064108,
+      "grad_norm": 0.5978641510009766,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7644487615347256,
+      "grad_norm": 0.5475472807884216,
+      "learning_rate": 0.0001,
+      "loss": 1.7076,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7646915978630403,
+      "grad_norm": 0.592376172542572,
+      "learning_rate": 0.0001,
+      "loss": 1.7808,
+      "step": 3149
+    },
+    {
+      "epoch": 0.764934434191355,
+      "grad_norm": 0.6536989808082581,
+      "learning_rate": 0.0001,
+      "loss": 1.753,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7651772705196698,
+      "grad_norm": 0.5901771187782288,
+      "learning_rate": 0.0001,
+      "loss": 1.7637,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7654201068479844,
+      "grad_norm": 0.599584698677063,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7656629431762991,
+      "grad_norm": 0.571835994720459,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7659057795046139,
+      "grad_norm": 0.5579686760902405,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7661486158329286,
+      "grad_norm": 0.59943026304245,
+      "learning_rate": 0.0001,
+      "loss": 1.7429,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7663914521612433,
+      "grad_norm": 0.6282011866569519,
+      "learning_rate": 0.0001,
+      "loss": 1.7826,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7666342884895581,
+      "grad_norm": 0.5895368456840515,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7668771248178727,
+      "grad_norm": 0.5622402429580688,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7671199611461875,
+      "grad_norm": 0.6002911329269409,
+      "learning_rate": 0.0001,
+      "loss": 1.778,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7673627974745022,
+      "grad_norm": 0.6253058910369873,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7676056338028169,
+      "grad_norm": 0.6376633644104004,
+      "learning_rate": 0.0001,
+      "loss": 1.6844,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7678484701311317,
+      "grad_norm": 0.6060060262680054,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7680913064594463,
+      "grad_norm": 0.604957103729248,
+      "learning_rate": 0.0001,
+      "loss": 1.6954,
+      "step": 3163
+    },
+    {
+      "epoch": 0.768334142787761,
+      "grad_norm": 0.6303216814994812,
+      "learning_rate": 0.0001,
+      "loss": 1.8798,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7685769791160758,
+      "grad_norm": 0.5767795443534851,
+      "learning_rate": 0.0001,
+      "loss": 1.8241,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7688198154443905,
+      "grad_norm": 0.5388797521591187,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7690626517727052,
+      "grad_norm": 0.5943977236747742,
+      "learning_rate": 0.0001,
+      "loss": 1.7706,
+      "step": 3167
+    },
+    {
+      "epoch": 0.76930548810102,
+      "grad_norm": 0.5794751644134521,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7695483244293346,
+      "grad_norm": 0.5764420628547668,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7697911607576493,
+      "grad_norm": 0.5977728962898254,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7700339970859641,
+      "grad_norm": 0.6072218418121338,
+      "learning_rate": 0.0001,
+      "loss": 1.7673,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7702768334142788,
+      "grad_norm": 0.6105883121490479,
+      "learning_rate": 0.0001,
+      "loss": 1.7475,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7705196697425934,
+      "grad_norm": 0.5928831696510315,
+      "learning_rate": 0.0001,
+      "loss": 1.6449,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7707625060709082,
+      "grad_norm": 0.5637102723121643,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7710053423992229,
+      "grad_norm": 0.5758238434791565,
+      "learning_rate": 0.0001,
+      "loss": 1.7284,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7712481787275376,
+      "grad_norm": 0.589185893535614,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7714910150558524,
+      "grad_norm": 0.6024256348609924,
+      "learning_rate": 0.0001,
+      "loss": 1.7998,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7717338513841671,
+      "grad_norm": 0.6078524589538574,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7719766877124817,
+      "grad_norm": 0.5936061143875122,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7722195240407965,
+      "grad_norm": 0.6246563792228699,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7724623603691112,
+      "grad_norm": 0.565122663974762,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 3181
+    },
+    {
+      "epoch": 0.772705196697426,
+      "grad_norm": 0.6005831956863403,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7729480330257407,
+      "grad_norm": 0.5594892501831055,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7731908693540553,
+      "grad_norm": 0.6021163463592529,
+      "learning_rate": 0.0001,
+      "loss": 1.8812,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7734337056823701,
+      "grad_norm": 0.5833166837692261,
+      "learning_rate": 0.0001,
+      "loss": 1.6699,
+      "step": 3185
+    },
+    {
+      "epoch": 0.7736765420106848,
+      "grad_norm": 0.5965452790260315,
+      "learning_rate": 0.0001,
+      "loss": 1.7835,
+      "step": 3186
+    },
+    {
+      "epoch": 0.7739193783389995,
+      "grad_norm": 0.581089973449707,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7741622146673143,
+      "grad_norm": 0.5958520770072937,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 3188
+    },
+    {
+      "epoch": 0.774405050995629,
+      "grad_norm": 0.6011066436767578,
+      "learning_rate": 0.0001,
+      "loss": 1.7536,
+      "step": 3189
+    },
+    {
+      "epoch": 0.7746478873239436,
+      "grad_norm": 0.5939875245094299,
+      "learning_rate": 0.0001,
+      "loss": 1.7563,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7748907236522584,
+      "grad_norm": 0.5792884230613708,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7751335599805731,
+      "grad_norm": 0.582841694355011,
+      "learning_rate": 0.0001,
+      "loss": 1.7502,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7753763963088878,
+      "grad_norm": 0.6032854914665222,
+      "learning_rate": 0.0001,
+      "loss": 1.9189,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7756192326372026,
+      "grad_norm": 0.5822592377662659,
+      "learning_rate": 0.0001,
+      "loss": 1.5918,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7758620689655172,
+      "grad_norm": 0.5904735922813416,
+      "learning_rate": 0.0001,
+      "loss": 1.7974,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7761049052938319,
+      "grad_norm": 0.5985095500946045,
+      "learning_rate": 0.0001,
+      "loss": 1.8372,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7763477416221467,
+      "grad_norm": 0.5882258415222168,
+      "learning_rate": 0.0001,
+      "loss": 1.8258,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7765905779504614,
+      "grad_norm": 0.6293593049049377,
+      "learning_rate": 0.0001,
+      "loss": 1.7996,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7768334142787761,
+      "grad_norm": 0.5894924402236938,
+      "learning_rate": 0.0001,
+      "loss": 1.7175,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7770762506070908,
+      "grad_norm": 0.5985170006752014,
+      "learning_rate": 0.0001,
+      "loss": 1.7009,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7773190869354055,
+      "grad_norm": 0.609933614730835,
+      "learning_rate": 0.0001,
+      "loss": 1.7568,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7775619232637202,
+      "grad_norm": 0.6414185166358948,
+      "learning_rate": 0.0001,
+      "loss": 1.8863,
+      "step": 3202
+    },
+    {
+      "epoch": 0.777804759592035,
+      "grad_norm": 0.6187397837638855,
+      "learning_rate": 0.0001,
+      "loss": 1.8666,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7780475959203497,
+      "grad_norm": 0.5837206840515137,
+      "learning_rate": 0.0001,
+      "loss": 1.8421,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7782904322486645,
+      "grad_norm": 0.5552575588226318,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7785332685769791,
+      "grad_norm": 0.5835022926330566,
+      "learning_rate": 0.0001,
+      "loss": 1.7375,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7787761049052938,
+      "grad_norm": 0.6458723545074463,
+      "learning_rate": 0.0001,
+      "loss": 1.8088,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7790189412336086,
+      "grad_norm": 0.6067644953727722,
+      "learning_rate": 0.0001,
+      "loss": 1.7628,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7792617775619233,
+      "grad_norm": 0.6323834657669067,
+      "learning_rate": 0.0001,
+      "loss": 1.8119,
+      "step": 3209
+    },
+    {
+      "epoch": 0.779504613890238,
+      "grad_norm": 0.648609459400177,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7797474502185527,
+      "grad_norm": 0.60872882604599,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7799902865468674,
+      "grad_norm": 0.6171422600746155,
+      "learning_rate": 0.0001,
+      "loss": 1.7076,
+      "step": 3212
+    },
+    {
+      "epoch": 0.7802331228751821,
+      "grad_norm": 0.5864462852478027,
+      "learning_rate": 0.0001,
+      "loss": 1.6012,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7804759592034969,
+      "grad_norm": 0.6133025288581848,
+      "learning_rate": 0.0001,
+      "loss": 1.7297,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7807187955318116,
+      "grad_norm": 0.617095410823822,
+      "learning_rate": 0.0001,
+      "loss": 1.7052,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7809616318601262,
+      "grad_norm": 0.6132461428642273,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 3216
+    },
+    {
+      "epoch": 0.781204468188441,
+      "grad_norm": 0.6561414003372192,
+      "learning_rate": 0.0001,
+      "loss": 1.9216,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7814473045167557,
+      "grad_norm": 0.6122778058052063,
+      "learning_rate": 0.0001,
+      "loss": 1.7763,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7816901408450704,
+      "grad_norm": 0.6221407055854797,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7819329771733852,
+      "grad_norm": 0.6318554282188416,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7821758135016998,
+      "grad_norm": 0.5872081518173218,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7824186498300145,
+      "grad_norm": 0.6384358406066895,
+      "learning_rate": 0.0001,
+      "loss": 1.7984,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7826614861583293,
+      "grad_norm": 0.6401801109313965,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 3223
+    },
+    {
+      "epoch": 0.782904322486644,
+      "grad_norm": 0.5717194676399231,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7831471588149587,
+      "grad_norm": 0.6136575937271118,
+      "learning_rate": 0.0001,
+      "loss": 1.7576,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7833899951432735,
+      "grad_norm": 0.5917444229125977,
+      "learning_rate": 0.0001,
+      "loss": 1.6342,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7836328314715881,
+      "grad_norm": 0.5684612989425659,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 3227
+    },
+    {
+      "epoch": 0.7838756677999029,
+      "grad_norm": 0.6119584441184998,
+      "learning_rate": 0.0001,
+      "loss": 1.793,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7841185041282176,
+      "grad_norm": 0.5921847224235535,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 3229
+    },
+    {
+      "epoch": 0.7843613404565323,
+      "grad_norm": 0.6124568581581116,
+      "learning_rate": 0.0001,
+      "loss": 1.7255,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7846041767848471,
+      "grad_norm": 0.6209405064582825,
+      "learning_rate": 0.0001,
+      "loss": 1.7774,
+      "step": 3231
+    },
+    {
+      "epoch": 0.7848470131131617,
+      "grad_norm": 0.5682047009468079,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7850898494414764,
+      "grad_norm": 0.5621352195739746,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 3233
+    },
+    {
+      "epoch": 0.7853326857697912,
+      "grad_norm": 0.5859969854354858,
+      "learning_rate": 0.0001,
+      "loss": 1.7187,
+      "step": 3234
+    },
+    {
+      "epoch": 0.7855755220981059,
+      "grad_norm": 0.5869945287704468,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7858183584264206,
+      "grad_norm": 0.5958589911460876,
+      "learning_rate": 0.0001,
+      "loss": 1.742,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7860611947547353,
+      "grad_norm": 0.608558714389801,
+      "learning_rate": 0.0001,
+      "loss": 1.7526,
+      "step": 3237
+    },
+    {
+      "epoch": 0.78630403108305,
+      "grad_norm": 0.5519112944602966,
+      "learning_rate": 0.0001,
+      "loss": 1.7003,
+      "step": 3238
+    },
+    {
+      "epoch": 0.7865468674113647,
+      "grad_norm": 0.5748507380485535,
+      "learning_rate": 0.0001,
+      "loss": 1.7509,
+      "step": 3239
+    },
+    {
+      "epoch": 0.7867897037396795,
+      "grad_norm": 0.6219350099563599,
+      "learning_rate": 0.0001,
+      "loss": 1.8431,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7870325400679942,
+      "grad_norm": 0.5954362154006958,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 3241
+    },
+    {
+      "epoch": 0.7872753763963088,
+      "grad_norm": 0.6077887415885925,
+      "learning_rate": 0.0001,
+      "loss": 1.7988,
+      "step": 3242
+    },
+    {
+      "epoch": 0.7875182127246236,
+      "grad_norm": 0.5785715579986572,
+      "learning_rate": 0.0001,
+      "loss": 1.6917,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7877610490529383,
+      "grad_norm": 0.6050088405609131,
+      "learning_rate": 0.0001,
+      "loss": 1.71,
+      "step": 3244
+    },
+    {
+      "epoch": 0.788003885381253,
+      "grad_norm": 0.5776847004890442,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7882467217095678,
+      "grad_norm": 0.5937278866767883,
+      "learning_rate": 0.0001,
+      "loss": 1.7619,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7884895580378825,
+      "grad_norm": 0.5709724426269531,
+      "learning_rate": 0.0001,
+      "loss": 1.7535,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7887323943661971,
+      "grad_norm": 0.5952240228652954,
+      "learning_rate": 0.0001,
+      "loss": 1.6862,
+      "step": 3248
+    },
+    {
+      "epoch": 0.7889752306945119,
+      "grad_norm": 0.5571580529212952,
+      "learning_rate": 0.0001,
+      "loss": 1.6971,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7892180670228266,
+      "grad_norm": 0.5808640122413635,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7894609033511414,
+      "grad_norm": 0.5837990045547485,
+      "learning_rate": 0.0001,
+      "loss": 1.7686,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7897037396794561,
+      "grad_norm": 0.5844842195510864,
+      "learning_rate": 0.0001,
+      "loss": 1.7822,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7899465760077707,
+      "grad_norm": 0.6063905954360962,
+      "learning_rate": 0.0001,
+      "loss": 1.7837,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7901894123360855,
+      "grad_norm": 0.6192396879196167,
+      "learning_rate": 0.0001,
+      "loss": 1.7271,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7904322486644002,
+      "grad_norm": 0.5675437450408936,
+      "learning_rate": 0.0001,
+      "loss": 1.7922,
+      "step": 3255
+    },
+    {
+      "epoch": 0.7906750849927149,
+      "grad_norm": 0.6188688278198242,
+      "learning_rate": 0.0001,
+      "loss": 1.9356,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7909179213210297,
+      "grad_norm": 0.5751339197158813,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 3257
+    },
+    {
+      "epoch": 0.7911607576493443,
+      "grad_norm": 0.5956962704658508,
+      "learning_rate": 0.0001,
+      "loss": 1.7142,
+      "step": 3258
+    },
+    {
+      "epoch": 0.791403593977659,
+      "grad_norm": 0.6013764142990112,
+      "learning_rate": 0.0001,
+      "loss": 1.8275,
+      "step": 3259
+    },
+    {
+      "epoch": 0.7916464303059738,
+      "grad_norm": 0.5936847925186157,
+      "learning_rate": 0.0001,
+      "loss": 1.8003,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7918892666342885,
+      "grad_norm": 0.5982014536857605,
+      "learning_rate": 0.0001,
+      "loss": 1.8204,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7921321029626032,
+      "grad_norm": 0.6026539206504822,
+      "learning_rate": 0.0001,
+      "loss": 1.7183,
+      "step": 3262
+    },
+    {
+      "epoch": 0.792374939290918,
+      "grad_norm": 0.6419681310653687,
+      "learning_rate": 0.0001,
+      "loss": 2.0105,
+      "step": 3263
+    },
+    {
+      "epoch": 0.7926177756192326,
+      "grad_norm": 0.6058057546615601,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7928606119475473,
+      "grad_norm": 0.6088738441467285,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7931034482758621,
+      "grad_norm": 0.5703161358833313,
+      "learning_rate": 0.0001,
+      "loss": 1.679,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7933462846041768,
+      "grad_norm": 0.6744810342788696,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7935891209324915,
+      "grad_norm": 0.6127514839172363,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7938319572608062,
+      "grad_norm": 0.5914934873580933,
+      "learning_rate": 0.0001,
+      "loss": 1.8143,
+      "step": 3269
+    },
+    {
+      "epoch": 0.7940747935891209,
+      "grad_norm": 0.5978199243545532,
+      "learning_rate": 0.0001,
+      "loss": 1.8426,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7943176299174356,
+      "grad_norm": 0.5905243158340454,
+      "learning_rate": 0.0001,
+      "loss": 1.7238,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7945604662457504,
+      "grad_norm": 0.5864617824554443,
+      "learning_rate": 0.0001,
+      "loss": 1.6579,
+      "step": 3272
+    },
+    {
+      "epoch": 0.7948033025740651,
+      "grad_norm": 0.5839855074882507,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 3273
+    },
+    {
+      "epoch": 0.7950461389023799,
+      "grad_norm": 0.5632340312004089,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7952889752306945,
+      "grad_norm": 0.5917242169380188,
+      "learning_rate": 0.0001,
+      "loss": 1.9173,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7955318115590092,
+      "grad_norm": 0.5705491304397583,
+      "learning_rate": 0.0001,
+      "loss": 1.7487,
+      "step": 3276
+    },
+    {
+      "epoch": 0.795774647887324,
+      "grad_norm": 0.5937005281448364,
+      "learning_rate": 0.0001,
+      "loss": 1.7422,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7960174842156387,
+      "grad_norm": 0.5753030180931091,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 3278
+    },
+    {
+      "epoch": 0.7962603205439533,
+      "grad_norm": 0.6627799868583679,
+      "learning_rate": 0.0001,
+      "loss": 1.7026,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7965031568722681,
+      "grad_norm": 0.6088401675224304,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7967459932005828,
+      "grad_norm": 0.6398965716362,
+      "learning_rate": 0.0001,
+      "loss": 1.7722,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7969888295288975,
+      "grad_norm": 0.6321691274642944,
+      "learning_rate": 0.0001,
+      "loss": 1.8379,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7972316658572123,
+      "grad_norm": 0.5635577440261841,
+      "learning_rate": 0.0001,
+      "loss": 1.6831,
+      "step": 3283
+    },
+    {
+      "epoch": 0.797474502185527,
+      "grad_norm": 0.6203828454017639,
+      "learning_rate": 0.0001,
+      "loss": 1.8552,
+      "step": 3284
+    },
+    {
+      "epoch": 0.7977173385138416,
+      "grad_norm": 0.6240285038948059,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 3285
+    },
+    {
+      "epoch": 0.7979601748421564,
+      "grad_norm": 0.5690562129020691,
+      "learning_rate": 0.0001,
+      "loss": 1.7832,
+      "step": 3286
+    },
+    {
+      "epoch": 0.7982030111704711,
+      "grad_norm": 0.610381543636322,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 3287
+    },
+    {
+      "epoch": 0.7984458474987858,
+      "grad_norm": 0.609774649143219,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 3288
+    },
+    {
+      "epoch": 0.7986886838271006,
+      "grad_norm": 0.563381016254425,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7989315201554152,
+      "grad_norm": 0.6335068345069885,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7991743564837299,
+      "grad_norm": 0.6035629510879517,
+      "learning_rate": 0.0001,
+      "loss": 1.7156,
+      "step": 3291
+    },
+    {
+      "epoch": 0.7994171928120447,
+      "grad_norm": 0.586322009563446,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 3292
+    },
+    {
+      "epoch": 0.7996600291403594,
+      "grad_norm": 0.6346733570098877,
+      "learning_rate": 0.0001,
+      "loss": 1.8079,
+      "step": 3293
+    },
+    {
+      "epoch": 0.7999028654686741,
+      "grad_norm": 0.5991312265396118,
+      "learning_rate": 0.0001,
+      "loss": 1.7526,
+      "step": 3294
+    },
+    {
+      "epoch": 0.8001457017969889,
+      "grad_norm": 0.6022869348526001,
+      "learning_rate": 0.0001,
+      "loss": 1.8159,
+      "step": 3295
+    },
+    {
+      "epoch": 0.8003885381253035,
+      "grad_norm": 0.5966434478759766,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 3296
+    },
+    {
+      "epoch": 0.8006313744536183,
+      "grad_norm": 0.578927755355835,
+      "learning_rate": 0.0001,
+      "loss": 1.7053,
+      "step": 3297
+    },
+    {
+      "epoch": 0.800874210781933,
+      "grad_norm": 0.6092239022254944,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 3298
+    },
+    {
+      "epoch": 0.8011170471102477,
+      "grad_norm": 0.6259116530418396,
+      "learning_rate": 0.0001,
+      "loss": 1.7511,
+      "step": 3299
+    },
+    {
+      "epoch": 0.8013598834385625,
+      "grad_norm": 0.6093431711196899,
+      "learning_rate": 0.0001,
+      "loss": 1.8281,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8016027197668771,
+      "grad_norm": 0.5632680058479309,
+      "learning_rate": 0.0001,
+      "loss": 1.6561,
+      "step": 3301
+    },
+    {
+      "epoch": 0.8018455560951918,
+      "grad_norm": 0.5798899531364441,
+      "learning_rate": 0.0001,
+      "loss": 1.859,
+      "step": 3302
+    },
+    {
+      "epoch": 0.8020883924235066,
+      "grad_norm": 0.5695531964302063,
+      "learning_rate": 0.0001,
+      "loss": 1.7348,
+      "step": 3303
+    },
+    {
+      "epoch": 0.8023312287518213,
+      "grad_norm": 0.6000730991363525,
+      "learning_rate": 0.0001,
+      "loss": 1.7621,
+      "step": 3304
+    },
+    {
+      "epoch": 0.802574065080136,
+      "grad_norm": 0.5967425107955933,
+      "learning_rate": 0.0001,
+      "loss": 1.8134,
+      "step": 3305
+    },
+    {
+      "epoch": 0.8028169014084507,
+      "grad_norm": 0.571355402469635,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 3306
+    },
+    {
+      "epoch": 0.8030597377367654,
+      "grad_norm": 0.5934798121452332,
+      "learning_rate": 0.0001,
+      "loss": 1.7581,
+      "step": 3307
+    },
+    {
+      "epoch": 0.8033025740650801,
+      "grad_norm": 0.6144223213195801,
+      "learning_rate": 0.0001,
+      "loss": 1.7315,
+      "step": 3308
+    },
+    {
+      "epoch": 0.8035454103933949,
+      "grad_norm": 0.5985466837882996,
+      "learning_rate": 0.0001,
+      "loss": 1.68,
+      "step": 3309
+    },
+    {
+      "epoch": 0.8037882467217096,
+      "grad_norm": 0.630021333694458,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8040310830500242,
+      "grad_norm": 0.5904750823974609,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 3311
+    },
+    {
+      "epoch": 0.804273919378339,
+      "grad_norm": 0.5716640949249268,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 3312
+    },
+    {
+      "epoch": 0.8045167557066537,
+      "grad_norm": 0.588955819606781,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 3313
+    },
+    {
+      "epoch": 0.8047595920349684,
+      "grad_norm": 0.596904993057251,
+      "learning_rate": 0.0001,
+      "loss": 1.7657,
+      "step": 3314
+    },
+    {
+      "epoch": 0.8050024283632832,
+      "grad_norm": 0.6122095584869385,
+      "learning_rate": 0.0001,
+      "loss": 1.8043,
+      "step": 3315
+    },
+    {
+      "epoch": 0.8052452646915979,
+      "grad_norm": 0.567223072052002,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 3316
+    },
+    {
+      "epoch": 0.8054881010199125,
+      "grad_norm": 0.6023831963539124,
+      "learning_rate": 0.0001,
+      "loss": 1.7581,
+      "step": 3317
+    },
+    {
+      "epoch": 0.8057309373482273,
+      "grad_norm": 0.5783039927482605,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 3318
+    },
+    {
+      "epoch": 0.805973773676542,
+      "grad_norm": 0.61045902967453,
+      "learning_rate": 0.0001,
+      "loss": 1.8467,
+      "step": 3319
+    },
+    {
+      "epoch": 0.8062166100048568,
+      "grad_norm": 0.6231280565261841,
+      "learning_rate": 0.0001,
+      "loss": 1.7842,
+      "step": 3320
+    },
+    {
+      "epoch": 0.8064594463331715,
+      "grad_norm": 0.5988601446151733,
+      "learning_rate": 0.0001,
+      "loss": 1.754,
+      "step": 3321
+    },
+    {
+      "epoch": 0.8067022826614861,
+      "grad_norm": 0.6404216885566711,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 3322
+    },
+    {
+      "epoch": 0.8069451189898009,
+      "grad_norm": 0.5820360779762268,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 3323
+    },
+    {
+      "epoch": 0.8071879553181156,
+      "grad_norm": 0.6122472882270813,
+      "learning_rate": 0.0001,
+      "loss": 1.8324,
+      "step": 3324
+    },
+    {
+      "epoch": 0.8074307916464303,
+      "grad_norm": 0.6082800030708313,
+      "learning_rate": 0.0001,
+      "loss": 1.9817,
+      "step": 3325
+    },
+    {
+      "epoch": 0.8076736279747451,
+      "grad_norm": 0.6259562373161316,
+      "learning_rate": 0.0001,
+      "loss": 1.7077,
+      "step": 3326
+    },
+    {
+      "epoch": 0.8079164643030597,
+      "grad_norm": 0.5752188563346863,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 3327
+    },
+    {
+      "epoch": 0.8081593006313744,
+      "grad_norm": 0.575911819934845,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 3328
+    },
+    {
+      "epoch": 0.8084021369596892,
+      "grad_norm": 0.5568429231643677,
+      "learning_rate": 0.0001,
+      "loss": 1.7373,
+      "step": 3329
+    },
+    {
+      "epoch": 0.8086449732880039,
+      "grad_norm": 0.6027485132217407,
+      "learning_rate": 0.0001,
+      "loss": 1.7404,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8088878096163186,
+      "grad_norm": 0.7463359236717224,
+      "learning_rate": 0.0001,
+      "loss": 1.6852,
+      "step": 3331
+    },
+    {
+      "epoch": 0.8091306459446334,
+      "grad_norm": 0.6194759011268616,
+      "learning_rate": 0.0001,
+      "loss": 1.8937,
+      "step": 3332
+    },
+    {
+      "epoch": 0.809373482272948,
+      "grad_norm": 0.5699089169502258,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 3333
+    },
+    {
+      "epoch": 0.8096163186012627,
+      "grad_norm": 0.6080254316329956,
+      "learning_rate": 0.0001,
+      "loss": 1.88,
+      "step": 3334
+    },
+    {
+      "epoch": 0.8098591549295775,
+      "grad_norm": 0.5942913889884949,
+      "learning_rate": 0.0001,
+      "loss": 1.6951,
+      "step": 3335
+    },
+    {
+      "epoch": 0.8101019912578922,
+      "grad_norm": 0.5429050326347351,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 3336
+    },
+    {
+      "epoch": 0.8103448275862069,
+      "grad_norm": 0.5945674180984497,
+      "learning_rate": 0.0001,
+      "loss": 1.7982,
+      "step": 3337
+    },
+    {
+      "epoch": 0.8105876639145216,
+      "grad_norm": 0.6173865795135498,
+      "learning_rate": 0.0001,
+      "loss": 1.9694,
+      "step": 3338
+    },
+    {
+      "epoch": 0.8108305002428363,
+      "grad_norm": 0.600135862827301,
+      "learning_rate": 0.0001,
+      "loss": 1.7166,
+      "step": 3339
+    },
+    {
+      "epoch": 0.811073336571151,
+      "grad_norm": 0.6154102087020874,
+      "learning_rate": 0.0001,
+      "loss": 1.9117,
+      "step": 3340
+    },
+    {
+      "epoch": 0.8113161728994658,
+      "grad_norm": 0.6211844682693481,
+      "learning_rate": 0.0001,
+      "loss": 1.9107,
+      "step": 3341
+    },
+    {
+      "epoch": 0.8115590092277805,
+      "grad_norm": 0.5779953002929688,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 3342
+    },
+    {
+      "epoch": 0.8118018455560952,
+      "grad_norm": 0.5749261975288391,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 3343
+    },
+    {
+      "epoch": 0.8120446818844099,
+      "grad_norm": 0.5764738321304321,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 3344
+    },
+    {
+      "epoch": 0.8122875182127246,
+      "grad_norm": 0.6124898791313171,
+      "learning_rate": 0.0001,
+      "loss": 1.7762,
+      "step": 3345
+    },
+    {
+      "epoch": 0.8125303545410394,
+      "grad_norm": 0.5406044125556946,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 3346
+    },
+    {
+      "epoch": 0.8127731908693541,
+      "grad_norm": 0.6130670309066772,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 3347
+    },
+    {
+      "epoch": 0.8130160271976687,
+      "grad_norm": 0.6192781329154968,
+      "learning_rate": 0.0001,
+      "loss": 1.779,
+      "step": 3348
+    },
+    {
+      "epoch": 0.8132588635259835,
+      "grad_norm": 0.6246933937072754,
+      "learning_rate": 0.0001,
+      "loss": 1.9329,
+      "step": 3349
+    },
+    {
+      "epoch": 0.8135016998542982,
+      "grad_norm": 0.633960485458374,
+      "learning_rate": 0.0001,
+      "loss": 1.8494,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8137445361826129,
+      "grad_norm": 0.5493264198303223,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 3351
+    },
+    {
+      "epoch": 0.8139873725109277,
+      "grad_norm": 0.5964555740356445,
+      "learning_rate": 0.0001,
+      "loss": 1.6861,
+      "step": 3352
+    },
+    {
+      "epoch": 0.8142302088392424,
+      "grad_norm": 0.5876238346099854,
+      "learning_rate": 0.0001,
+      "loss": 1.8228,
+      "step": 3353
+    },
+    {
+      "epoch": 0.814473045167557,
+      "grad_norm": 0.5785903334617615,
+      "learning_rate": 0.0001,
+      "loss": 1.6771,
+      "step": 3354
+    },
+    {
+      "epoch": 0.8147158814958718,
+      "grad_norm": 0.5826572775840759,
+      "learning_rate": 0.0001,
+      "loss": 1.7107,
+      "step": 3355
+    },
+    {
+      "epoch": 0.8149587178241865,
+      "grad_norm": 0.5708701014518738,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 3356
+    },
+    {
+      "epoch": 0.8152015541525012,
+      "grad_norm": 0.5835521221160889,
+      "learning_rate": 0.0001,
+      "loss": 1.7775,
+      "step": 3357
+    },
+    {
+      "epoch": 0.815444390480816,
+      "grad_norm": 0.6153636574745178,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 3358
+    },
+    {
+      "epoch": 0.8156872268091306,
+      "grad_norm": 0.5732671618461609,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 3359
+    },
+    {
+      "epoch": 0.8159300631374453,
+      "grad_norm": 0.5535603761672974,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8161728994657601,
+      "grad_norm": 0.6083071827888489,
+      "learning_rate": 0.0001,
+      "loss": 1.7317,
+      "step": 3361
+    },
+    {
+      "epoch": 0.8164157357940748,
+      "grad_norm": 0.6135209798812866,
+      "learning_rate": 0.0001,
+      "loss": 1.8898,
+      "step": 3362
+    },
+    {
+      "epoch": 0.8166585721223895,
+      "grad_norm": 0.6117151379585266,
+      "learning_rate": 0.0001,
+      "loss": 1.7949,
+      "step": 3363
+    },
+    {
+      "epoch": 0.8169014084507042,
+      "grad_norm": 0.6317639946937561,
+      "learning_rate": 0.0001,
+      "loss": 1.8458,
+      "step": 3364
+    },
+    {
+      "epoch": 0.8171442447790189,
+      "grad_norm": 0.6188638806343079,
+      "learning_rate": 0.0001,
+      "loss": 1.8251,
+      "step": 3365
+    },
+    {
+      "epoch": 0.8173870811073336,
+      "grad_norm": 0.620080828666687,
+      "learning_rate": 0.0001,
+      "loss": 1.7581,
+      "step": 3366
+    },
+    {
+      "epoch": 0.8176299174356484,
+      "grad_norm": 0.5871297121047974,
+      "learning_rate": 0.0001,
+      "loss": 1.7579,
+      "step": 3367
+    },
+    {
+      "epoch": 0.8178727537639631,
+      "grad_norm": 0.5991242527961731,
+      "learning_rate": 0.0001,
+      "loss": 1.8426,
+      "step": 3368
+    },
+    {
+      "epoch": 0.8181155900922779,
+      "grad_norm": 0.608951985836029,
+      "learning_rate": 0.0001,
+      "loss": 1.8797,
+      "step": 3369
+    },
+    {
+      "epoch": 0.8183584264205925,
+      "grad_norm": 0.614581823348999,
+      "learning_rate": 0.0001,
+      "loss": 1.9202,
+      "step": 3370
+    },
+    {
+      "epoch": 0.8186012627489072,
+      "grad_norm": 0.5930675864219666,
+      "learning_rate": 0.0001,
+      "loss": 1.739,
+      "step": 3371
+    },
+    {
+      "epoch": 0.818844099077222,
+      "grad_norm": 0.5880351066589355,
+      "learning_rate": 0.0001,
+      "loss": 1.7359,
+      "step": 3372
+    },
+    {
+      "epoch": 0.8190869354055367,
+      "grad_norm": 0.5454912185668945,
+      "learning_rate": 0.0001,
+      "loss": 1.4532,
+      "step": 3373
+    },
+    {
+      "epoch": 0.8193297717338514,
+      "grad_norm": 0.561832070350647,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 3374
+    },
+    {
+      "epoch": 0.8195726080621661,
+      "grad_norm": 0.5611982941627502,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 3375
+    },
+    {
+      "epoch": 0.8198154443904808,
+      "grad_norm": 0.5709531307220459,
+      "learning_rate": 0.0001,
+      "loss": 1.7507,
+      "step": 3376
+    },
+    {
+      "epoch": 0.8200582807187955,
+      "grad_norm": 0.5754792094230652,
+      "learning_rate": 0.0001,
+      "loss": 1.7595,
+      "step": 3377
+    },
+    {
+      "epoch": 0.8203011170471103,
+      "grad_norm": 0.5866820812225342,
+      "learning_rate": 0.0001,
+      "loss": 1.7232,
+      "step": 3378
+    },
+    {
+      "epoch": 0.820543953375425,
+      "grad_norm": 0.5608850717544556,
+      "learning_rate": 0.0001,
+      "loss": 1.7659,
+      "step": 3379
+    },
+    {
+      "epoch": 0.8207867897037396,
+      "grad_norm": 0.5590660572052002,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8210296260320544,
+      "grad_norm": 0.5848101377487183,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 3381
+    },
+    {
+      "epoch": 0.8212724623603691,
+      "grad_norm": 0.5834150314331055,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 3382
+    },
+    {
+      "epoch": 0.8215152986886838,
+      "grad_norm": 0.6223803162574768,
+      "learning_rate": 0.0001,
+      "loss": 1.8801,
+      "step": 3383
+    },
+    {
+      "epoch": 0.8217581350169986,
+      "grad_norm": 0.5955408811569214,
+      "learning_rate": 0.0001,
+      "loss": 1.8415,
+      "step": 3384
+    },
+    {
+      "epoch": 0.8220009713453132,
+      "grad_norm": 0.5798283815383911,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 3385
+    },
+    {
+      "epoch": 0.8222438076736279,
+      "grad_norm": 0.5866160988807678,
+      "learning_rate": 0.0001,
+      "loss": 1.8269,
+      "step": 3386
+    },
+    {
+      "epoch": 0.8224866440019427,
+      "grad_norm": 0.5723693370819092,
+      "learning_rate": 0.0001,
+      "loss": 1.6874,
+      "step": 3387
+    },
+    {
+      "epoch": 0.8227294803302574,
+      "grad_norm": 0.6119223237037659,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 3388
+    },
+    {
+      "epoch": 0.8229723166585721,
+      "grad_norm": 0.6263529062271118,
+      "learning_rate": 0.0001,
+      "loss": 1.8294,
+      "step": 3389
+    },
+    {
+      "epoch": 0.8232151529868869,
+      "grad_norm": 0.6026468873023987,
+      "learning_rate": 0.0001,
+      "loss": 1.7883,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8234579893152015,
+      "grad_norm": 0.6187738180160522,
+      "learning_rate": 0.0001,
+      "loss": 1.8427,
+      "step": 3391
+    },
+    {
+      "epoch": 0.8237008256435163,
+      "grad_norm": 0.5795246362686157,
+      "learning_rate": 0.0001,
+      "loss": 1.7458,
+      "step": 3392
+    },
+    {
+      "epoch": 0.823943661971831,
+      "grad_norm": 0.5572130680084229,
+      "learning_rate": 0.0001,
+      "loss": 1.7382,
+      "step": 3393
+    },
+    {
+      "epoch": 0.8241864983001457,
+      "grad_norm": 0.5740678906440735,
+      "learning_rate": 0.0001,
+      "loss": 1.8143,
+      "step": 3394
+    },
+    {
+      "epoch": 0.8244293346284605,
+      "grad_norm": 0.5776839852333069,
+      "learning_rate": 0.0001,
+      "loss": 1.8622,
+      "step": 3395
+    },
+    {
+      "epoch": 0.8246721709567751,
+      "grad_norm": 0.5869353413581848,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 3396
+    },
+    {
+      "epoch": 0.8249150072850898,
+      "grad_norm": 0.5583310127258301,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 3397
+    },
+    {
+      "epoch": 0.8251578436134046,
+      "grad_norm": 0.5862078666687012,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 3398
+    },
+    {
+      "epoch": 0.8254006799417193,
+      "grad_norm": 0.5608081817626953,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 3399
+    },
+    {
+      "epoch": 0.825643516270034,
+      "grad_norm": 0.6446062326431274,
+      "learning_rate": 0.0001,
+      "loss": 1.853,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8258863525983487,
+      "grad_norm": 0.5685895085334778,
+      "learning_rate": 0.0001,
+      "loss": 1.6336,
+      "step": 3401
+    },
+    {
+      "epoch": 0.8261291889266634,
+      "grad_norm": 0.5998561978340149,
+      "learning_rate": 0.0001,
+      "loss": 1.8534,
+      "step": 3402
+    },
+    {
+      "epoch": 0.8263720252549781,
+      "grad_norm": 0.5718868970870972,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 3403
+    },
+    {
+      "epoch": 0.8266148615832929,
+      "grad_norm": 0.6252078413963318,
+      "learning_rate": 0.0001,
+      "loss": 1.6878,
+      "step": 3404
+    },
+    {
+      "epoch": 0.8268576979116076,
+      "grad_norm": 0.6125273108482361,
+      "learning_rate": 0.0001,
+      "loss": 1.8284,
+      "step": 3405
+    },
+    {
+      "epoch": 0.8271005342399222,
+      "grad_norm": 0.608466625213623,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 3406
+    },
+    {
+      "epoch": 0.827343370568237,
+      "grad_norm": 0.599600613117218,
+      "learning_rate": 0.0001,
+      "loss": 1.7604,
+      "step": 3407
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 0.6215369701385498,
+      "learning_rate": 0.0001,
+      "loss": 1.7342,
+      "step": 3408
+    },
+    {
+      "epoch": 0.8278290432248664,
+      "grad_norm": 0.5949468016624451,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 3409
+    },
+    {
+      "epoch": 0.8280718795531812,
+      "grad_norm": 0.5503864288330078,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8283147158814959,
+      "grad_norm": 0.6335252523422241,
+      "learning_rate": 0.0001,
+      "loss": 1.9147,
+      "step": 3411
+    },
+    {
+      "epoch": 0.8285575522098105,
+      "grad_norm": 0.56465744972229,
+      "learning_rate": 0.0001,
+      "loss": 1.5786,
+      "step": 3412
+    },
+    {
+      "epoch": 0.8288003885381253,
+      "grad_norm": 0.592567503452301,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 3413
+    },
+    {
+      "epoch": 0.82904322486644,
+      "grad_norm": 0.5808621048927307,
+      "learning_rate": 0.0001,
+      "loss": 1.7253,
+      "step": 3414
+    },
+    {
+      "epoch": 0.8292860611947548,
+      "grad_norm": 0.5746657252311707,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 3415
+    },
+    {
+      "epoch": 0.8295288975230695,
+      "grad_norm": 0.5854600667953491,
+      "learning_rate": 0.0001,
+      "loss": 1.8233,
+      "step": 3416
+    },
+    {
+      "epoch": 0.8297717338513841,
+      "grad_norm": 0.5802686810493469,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 3417
+    },
+    {
+      "epoch": 0.8300145701796989,
+      "grad_norm": 0.6190335154533386,
+      "learning_rate": 0.0001,
+      "loss": 1.8894,
+      "step": 3418
+    },
+    {
+      "epoch": 0.8302574065080136,
+      "grad_norm": 0.5804746150970459,
+      "learning_rate": 0.0001,
+      "loss": 1.648,
+      "step": 3419
+    },
+    {
+      "epoch": 0.8305002428363283,
+      "grad_norm": 0.5803455114364624,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8307430791646431,
+      "grad_norm": 0.5573564767837524,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 3421
+    },
+    {
+      "epoch": 0.8309859154929577,
+      "grad_norm": 0.6028070449829102,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 3422
+    },
+    {
+      "epoch": 0.8312287518212724,
+      "grad_norm": 0.5921975374221802,
+      "learning_rate": 0.0001,
+      "loss": 1.5926,
+      "step": 3423
+    },
+    {
+      "epoch": 0.8314715881495872,
+      "grad_norm": 0.6960628032684326,
+      "learning_rate": 0.0001,
+      "loss": 1.7498,
+      "step": 3424
+    },
+    {
+      "epoch": 0.8317144244779019,
+      "grad_norm": 0.5475769639015198,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 3425
+    },
+    {
+      "epoch": 0.8319572608062166,
+      "grad_norm": 0.6360837817192078,
+      "learning_rate": 0.0001,
+      "loss": 1.8568,
+      "step": 3426
+    },
+    {
+      "epoch": 0.8322000971345314,
+      "grad_norm": 0.5973774790763855,
+      "learning_rate": 0.0001,
+      "loss": 1.8229,
+      "step": 3427
+    },
+    {
+      "epoch": 0.832442933462846,
+      "grad_norm": 0.663210391998291,
+      "learning_rate": 0.0001,
+      "loss": 1.7362,
+      "step": 3428
+    },
+    {
+      "epoch": 0.8326857697911607,
+      "grad_norm": 0.5464731454849243,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 3429
+    },
+    {
+      "epoch": 0.8329286061194755,
+      "grad_norm": 0.619607150554657,
+      "learning_rate": 0.0001,
+      "loss": 1.7336,
+      "step": 3430
+    },
+    {
+      "epoch": 0.8331714424477902,
+      "grad_norm": 0.5969376564025879,
+      "learning_rate": 0.0001,
+      "loss": 1.7484,
+      "step": 3431
+    },
+    {
+      "epoch": 0.8334142787761049,
+      "grad_norm": 0.5881327390670776,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 3432
+    },
+    {
+      "epoch": 0.8336571151044196,
+      "grad_norm": 0.5548758506774902,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 3433
+    },
+    {
+      "epoch": 0.8338999514327343,
+      "grad_norm": 0.6055938601493835,
+      "learning_rate": 0.0001,
+      "loss": 1.8663,
+      "step": 3434
+    },
+    {
+      "epoch": 0.834142787761049,
+      "grad_norm": 0.6541330814361572,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 3435
+    },
+    {
+      "epoch": 0.8343856240893638,
+      "grad_norm": 0.6338015198707581,
+      "learning_rate": 0.0001,
+      "loss": 1.7116,
+      "step": 3436
+    },
+    {
+      "epoch": 0.8346284604176785,
+      "grad_norm": 0.5681234002113342,
+      "learning_rate": 0.0001,
+      "loss": 1.689,
+      "step": 3437
+    },
+    {
+      "epoch": 0.8348712967459933,
+      "grad_norm": 0.6056068539619446,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 3438
+    },
+    {
+      "epoch": 0.8351141330743079,
+      "grad_norm": 0.5616917014122009,
+      "learning_rate": 0.0001,
+      "loss": 1.5896,
+      "step": 3439
+    },
+    {
+      "epoch": 0.8353569694026226,
+      "grad_norm": 0.6003168821334839,
+      "learning_rate": 0.0001,
+      "loss": 1.8037,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8355998057309374,
+      "grad_norm": 0.575546145439148,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 3441
+    },
+    {
+      "epoch": 0.8358426420592521,
+      "grad_norm": 0.5927971601486206,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 3442
+    },
+    {
+      "epoch": 0.8360854783875667,
+      "grad_norm": 0.5666440725326538,
+      "learning_rate": 0.0001,
+      "loss": 1.8136,
+      "step": 3443
+    },
+    {
+      "epoch": 0.8363283147158815,
+      "grad_norm": 0.621577262878418,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 3444
+    },
+    {
+      "epoch": 0.8365711510441962,
+      "grad_norm": 0.5634074211120605,
+      "learning_rate": 0.0001,
+      "loss": 1.7177,
+      "step": 3445
+    },
+    {
+      "epoch": 0.8368139873725109,
+      "grad_norm": 0.6412850022315979,
+      "learning_rate": 0.0001,
+      "loss": 1.7874,
+      "step": 3446
+    },
+    {
+      "epoch": 0.8370568237008257,
+      "grad_norm": 0.5438231825828552,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 3447
+    },
+    {
+      "epoch": 0.8372996600291404,
+      "grad_norm": 0.6114881038665771,
+      "learning_rate": 0.0001,
+      "loss": 1.8538,
+      "step": 3448
+    },
+    {
+      "epoch": 0.837542496357455,
+      "grad_norm": 0.5776000618934631,
+      "learning_rate": 0.0001,
+      "loss": 1.7675,
+      "step": 3449
+    },
+    {
+      "epoch": 0.8377853326857698,
+      "grad_norm": 0.5970380306243896,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8380281690140845,
+      "grad_norm": 0.6232032179832458,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 3451
+    },
+    {
+      "epoch": 0.8382710053423992,
+      "grad_norm": 0.5810390710830688,
+      "learning_rate": 0.0001,
+      "loss": 1.8324,
+      "step": 3452
+    },
+    {
+      "epoch": 0.838513841670714,
+      "grad_norm": 0.5402159690856934,
+      "learning_rate": 0.0001,
+      "loss": 1.5199,
+      "step": 3453
+    },
+    {
+      "epoch": 0.8387566779990286,
+      "grad_norm": 0.5929580330848694,
+      "learning_rate": 0.0001,
+      "loss": 1.7311,
+      "step": 3454
+    },
+    {
+      "epoch": 0.8389995143273433,
+      "grad_norm": 0.5634052753448486,
+      "learning_rate": 0.0001,
+      "loss": 1.6945,
+      "step": 3455
+    },
+    {
+      "epoch": 0.8392423506556581,
+      "grad_norm": 0.6263976693153381,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 3456
+    },
+    {
+      "epoch": 0.8394851869839728,
+      "grad_norm": 0.6204410195350647,
+      "learning_rate": 0.0001,
+      "loss": 1.9143,
+      "step": 3457
+    },
+    {
+      "epoch": 0.8397280233122875,
+      "grad_norm": 0.5862576961517334,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 3458
+    },
+    {
+      "epoch": 0.8399708596406023,
+      "grad_norm": 0.5897918343544006,
+      "learning_rate": 0.0001,
+      "loss": 1.6781,
+      "step": 3459
+    },
+    {
+      "epoch": 0.8402136959689169,
+      "grad_norm": 0.5764839053153992,
+      "learning_rate": 0.0001,
+      "loss": 1.7503,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8404565322972317,
+      "grad_norm": 0.593485951423645,
+      "learning_rate": 0.0001,
+      "loss": 1.8107,
+      "step": 3461
+    },
+    {
+      "epoch": 0.8406993686255464,
+      "grad_norm": 0.6075116991996765,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 3462
+    },
+    {
+      "epoch": 0.8409422049538611,
+      "grad_norm": 0.6165798902511597,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 3463
+    },
+    {
+      "epoch": 0.8411850412821759,
+      "grad_norm": 0.5829122066497803,
+      "learning_rate": 0.0001,
+      "loss": 1.7196,
+      "step": 3464
+    },
+    {
+      "epoch": 0.8414278776104905,
+      "grad_norm": 0.6220898032188416,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8416707139388052,
+      "grad_norm": 0.635849118232727,
+      "learning_rate": 0.0001,
+      "loss": 1.8856,
+      "step": 3466
+    },
+    {
+      "epoch": 0.84191355026712,
+      "grad_norm": 0.6103588938713074,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 3467
+    },
+    {
+      "epoch": 0.8421563865954347,
+      "grad_norm": 0.5577727556228638,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 3468
+    },
+    {
+      "epoch": 0.8423992229237494,
+      "grad_norm": 0.5995204448699951,
+      "learning_rate": 0.0001,
+      "loss": 1.7391,
+      "step": 3469
+    },
+    {
+      "epoch": 0.8426420592520641,
+      "grad_norm": 0.6427181363105774,
+      "learning_rate": 0.0001,
+      "loss": 1.8653,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8428848955803788,
+      "grad_norm": 0.5356771945953369,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 3471
+    },
+    {
+      "epoch": 0.8431277319086935,
+      "grad_norm": 0.5844012498855591,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 3472
+    },
+    {
+      "epoch": 0.8433705682370083,
+      "grad_norm": 0.60114985704422,
+      "learning_rate": 0.0001,
+      "loss": 1.7445,
+      "step": 3473
+    },
+    {
+      "epoch": 0.843613404565323,
+      "grad_norm": 0.6197024583816528,
+      "learning_rate": 0.0001,
+      "loss": 1.7657,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8438562408936376,
+      "grad_norm": 0.5999199748039246,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8440990772219524,
+      "grad_norm": 0.5859060883522034,
+      "learning_rate": 0.0001,
+      "loss": 1.8221,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8443419135502671,
+      "grad_norm": 0.5816472768783569,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 3477
+    },
+    {
+      "epoch": 0.8445847498785818,
+      "grad_norm": 0.5703580379486084,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 3478
+    },
+    {
+      "epoch": 0.8448275862068966,
+      "grad_norm": 0.6125529408454895,
+      "learning_rate": 0.0001,
+      "loss": 1.8035,
+      "step": 3479
+    },
+    {
+      "epoch": 0.8450704225352113,
+      "grad_norm": 0.5765839219093323,
+      "learning_rate": 0.0001,
+      "loss": 1.7085,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8453132588635259,
+      "grad_norm": 0.573478102684021,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 3481
+    },
+    {
+      "epoch": 0.8455560951918407,
+      "grad_norm": 0.6237002611160278,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 3482
+    },
+    {
+      "epoch": 0.8457989315201554,
+      "grad_norm": 0.5803018808364868,
+      "learning_rate": 0.0001,
+      "loss": 1.7738,
+      "step": 3483
+    },
+    {
+      "epoch": 0.8460417678484702,
+      "grad_norm": 0.576075553894043,
+      "learning_rate": 0.0001,
+      "loss": 1.6795,
+      "step": 3484
+    },
+    {
+      "epoch": 0.8462846041767849,
+      "grad_norm": 0.5590042471885681,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 3485
+    },
+    {
+      "epoch": 0.8465274405050995,
+      "grad_norm": 0.5912151336669922,
+      "learning_rate": 0.0001,
+      "loss": 1.7616,
+      "step": 3486
+    },
+    {
+      "epoch": 0.8467702768334143,
+      "grad_norm": 0.634704053401947,
+      "learning_rate": 0.0001,
+      "loss": 1.5935,
+      "step": 3487
+    },
+    {
+      "epoch": 0.847013113161729,
+      "grad_norm": 0.5654214024543762,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8472559494900437,
+      "grad_norm": 0.5535054802894592,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8474987858183585,
+      "grad_norm": 0.5521407127380371,
+      "learning_rate": 0.0001,
+      "loss": 1.5164,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8477416221466731,
+      "grad_norm": 0.5911891460418701,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 3491
+    },
+    {
+      "epoch": 0.8479844584749878,
+      "grad_norm": 0.5401911735534668,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 3492
+    },
+    {
+      "epoch": 0.8482272948033026,
+      "grad_norm": 0.5964999794960022,
+      "learning_rate": 0.0001,
+      "loss": 1.6824,
+      "step": 3493
+    },
+    {
+      "epoch": 0.8484701311316173,
+      "grad_norm": 0.5567051768302917,
+      "learning_rate": 0.0001,
+      "loss": 1.7838,
+      "step": 3494
+    },
+    {
+      "epoch": 0.848712967459932,
+      "grad_norm": 0.5711089372634888,
+      "learning_rate": 0.0001,
+      "loss": 1.632,
+      "step": 3495
+    },
+    {
+      "epoch": 0.8489558037882468,
+      "grad_norm": 0.6089751124382019,
+      "learning_rate": 0.0001,
+      "loss": 1.6986,
+      "step": 3496
+    },
+    {
+      "epoch": 0.8491986401165614,
+      "grad_norm": 0.6084120869636536,
+      "learning_rate": 0.0001,
+      "loss": 1.6776,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8494414764448761,
+      "grad_norm": 0.6146928668022156,
+      "learning_rate": 0.0001,
+      "loss": 1.8569,
+      "step": 3498
+    },
+    {
+      "epoch": 0.8496843127731909,
+      "grad_norm": 0.5824174880981445,
+      "learning_rate": 0.0001,
+      "loss": 1.7408,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8499271491015056,
+      "grad_norm": 0.5815324783325195,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8501699854298203,
+      "grad_norm": 0.6049876809120178,
+      "learning_rate": 0.0001,
+      "loss": 1.7525,
+      "step": 3501
+    },
+    {
+      "epoch": 0.850412821758135,
+      "grad_norm": 0.6228092908859253,
+      "learning_rate": 0.0001,
+      "loss": 1.7271,
+      "step": 3502
+    },
+    {
+      "epoch": 0.8506556580864497,
+      "grad_norm": 0.561995267868042,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8508984944147644,
+      "grad_norm": 0.6361596584320068,
+      "learning_rate": 0.0001,
+      "loss": 1.7825,
+      "step": 3504
+    },
+    {
+      "epoch": 0.8511413307430792,
+      "grad_norm": 0.5702955722808838,
+      "learning_rate": 0.0001,
+      "loss": 1.6949,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8513841670713939,
+      "grad_norm": 0.5604357123374939,
+      "learning_rate": 0.0001,
+      "loss": 1.5887,
+      "step": 3506
+    },
+    {
+      "epoch": 0.8516270033997086,
+      "grad_norm": 0.6207056045532227,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8518698397280233,
+      "grad_norm": 0.6111729145050049,
+      "learning_rate": 0.0001,
+      "loss": 1.8756,
+      "step": 3508
+    },
+    {
+      "epoch": 0.852112676056338,
+      "grad_norm": 0.615805983543396,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8523555123846528,
+      "grad_norm": 0.6363980770111084,
+      "learning_rate": 0.0001,
+      "loss": 1.7821,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8525983487129675,
+      "grad_norm": 0.5989717841148376,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 3511
+    },
+    {
+      "epoch": 0.8528411850412821,
+      "grad_norm": 0.5952607989311218,
+      "learning_rate": 0.0001,
+      "loss": 1.69,
+      "step": 3512
+    },
+    {
+      "epoch": 0.8530840213695969,
+      "grad_norm": 0.5801405310630798,
+      "learning_rate": 0.0001,
+      "loss": 1.7882,
+      "step": 3513
+    },
+    {
+      "epoch": 0.8533268576979116,
+      "grad_norm": 0.6062362194061279,
+      "learning_rate": 0.0001,
+      "loss": 1.6513,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8535696940262263,
+      "grad_norm": 0.5364546179771423,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8538125303545411,
+      "grad_norm": 0.5886363983154297,
+      "learning_rate": 0.0001,
+      "loss": 1.5536,
+      "step": 3516
+    },
+    {
+      "epoch": 0.8540553666828558,
+      "grad_norm": 0.6251400113105774,
+      "learning_rate": 0.0001,
+      "loss": 1.847,
+      "step": 3517
+    },
+    {
+      "epoch": 0.8542982030111704,
+      "grad_norm": 0.6041175723075867,
+      "learning_rate": 0.0001,
+      "loss": 1.8304,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8545410393394852,
+      "grad_norm": 0.6060884594917297,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8547838756677999,
+      "grad_norm": 0.6119658946990967,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8550267119961146,
+      "grad_norm": 0.6325762271881104,
+      "learning_rate": 0.0001,
+      "loss": 1.7807,
+      "step": 3521
+    },
+    {
+      "epoch": 0.8552695483244294,
+      "grad_norm": 0.5806691646575928,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 3522
+    },
+    {
+      "epoch": 0.855512384652744,
+      "grad_norm": 0.5474377870559692,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 3523
+    },
+    {
+      "epoch": 0.8557552209810587,
+      "grad_norm": 0.571455717086792,
+      "learning_rate": 0.0001,
+      "loss": 1.7621,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8559980573093735,
+      "grad_norm": 0.6741440892219543,
+      "learning_rate": 0.0001,
+      "loss": 1.7855,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8562408936376882,
+      "grad_norm": 0.6805163025856018,
+      "learning_rate": 0.0001,
+      "loss": 1.9816,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8564837299660029,
+      "grad_norm": 0.5746136903762817,
+      "learning_rate": 0.0001,
+      "loss": 1.6439,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8567265662943176,
+      "grad_norm": 0.652385413646698,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8569694026226323,
+      "grad_norm": 0.6076606512069702,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8572122389509471,
+      "grad_norm": 0.6327298879623413,
+      "learning_rate": 0.0001,
+      "loss": 1.8304,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8574550752792618,
+      "grad_norm": 0.5788837671279907,
+      "learning_rate": 0.0001,
+      "loss": 1.7163,
+      "step": 3531
+    },
+    {
+      "epoch": 0.8576979116075765,
+      "grad_norm": 0.6050636172294617,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 3532
+    },
+    {
+      "epoch": 0.8579407479358913,
+      "grad_norm": 0.6376311182975769,
+      "learning_rate": 0.0001,
+      "loss": 1.7723,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8581835842642059,
+      "grad_norm": 0.5577212572097778,
+      "learning_rate": 0.0001,
+      "loss": 1.6499,
+      "step": 3534
+    },
+    {
+      "epoch": 0.8584264205925206,
+      "grad_norm": 0.5583710074424744,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8586692569208354,
+      "grad_norm": 0.6112894415855408,
+      "learning_rate": 0.0001,
+      "loss": 1.7546,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8589120932491501,
+      "grad_norm": 0.5702512264251709,
+      "learning_rate": 0.0001,
+      "loss": 1.7296,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8591549295774648,
+      "grad_norm": 0.634347140789032,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8593977659057795,
+      "grad_norm": 0.5780573487281799,
+      "learning_rate": 0.0001,
+      "loss": 1.8952,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8596406022340942,
+      "grad_norm": 0.590006947517395,
+      "learning_rate": 0.0001,
+      "loss": 1.7142,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8598834385624089,
+      "grad_norm": 0.5899177193641663,
+      "learning_rate": 0.0001,
+      "loss": 1.8622,
+      "step": 3541
+    },
+    {
+      "epoch": 0.8601262748907237,
+      "grad_norm": 0.5887792706489563,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8603691112190384,
+      "grad_norm": 0.5811349749565125,
+      "learning_rate": 0.0001,
+      "loss": 1.703,
+      "step": 3543
+    },
+    {
+      "epoch": 0.860611947547353,
+      "grad_norm": 0.5674530863761902,
+      "learning_rate": 0.0001,
+      "loss": 1.7557,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8608547838756678,
+      "grad_norm": 0.6252373456954956,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8610976202039825,
+      "grad_norm": 0.5738163590431213,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 3546
+    },
+    {
+      "epoch": 0.8613404565322972,
+      "grad_norm": 0.6242806911468506,
+      "learning_rate": 0.0001,
+      "loss": 1.7783,
+      "step": 3547
+    },
+    {
+      "epoch": 0.861583292860612,
+      "grad_norm": 0.6254070401191711,
+      "learning_rate": 0.0001,
+      "loss": 1.7868,
+      "step": 3548
+    },
+    {
+      "epoch": 0.8618261291889266,
+      "grad_norm": 0.5892719030380249,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8620689655172413,
+      "grad_norm": 0.5632604360580444,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8623118018455561,
+      "grad_norm": 0.5731838941574097,
+      "learning_rate": 0.0001,
+      "loss": 1.6758,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8625546381738708,
+      "grad_norm": 0.5648952722549438,
+      "learning_rate": 0.0001,
+      "loss": 1.6706,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8627974745021856,
+      "grad_norm": 0.5980069637298584,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8630403108305003,
+      "grad_norm": 0.6209183931350708,
+      "learning_rate": 0.0001,
+      "loss": 1.6986,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8632831471588149,
+      "grad_norm": 0.6167494058609009,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8635259834871297,
+      "grad_norm": 0.6167067289352417,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8637688198154444,
+      "grad_norm": 0.5660021305084229,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8640116561437591,
+      "grad_norm": 0.6034141182899475,
+      "learning_rate": 0.0001,
+      "loss": 1.7734,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8642544924720739,
+      "grad_norm": 0.6104663014411926,
+      "learning_rate": 0.0001,
+      "loss": 1.6955,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8644973288003885,
+      "grad_norm": 0.6068470478057861,
+      "learning_rate": 0.0001,
+      "loss": 1.83,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8647401651287032,
+      "grad_norm": 0.6156007647514343,
+      "learning_rate": 0.0001,
+      "loss": 1.6015,
+      "step": 3561
+    },
+    {
+      "epoch": 0.864983001457018,
+      "grad_norm": 0.5979913473129272,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8652258377853327,
+      "grad_norm": 0.5854653716087341,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8654686741136474,
+      "grad_norm": 0.582108736038208,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8657115104419622,
+      "grad_norm": 0.6201448440551758,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8659543467702768,
+      "grad_norm": 0.5824017524719238,
+      "learning_rate": 0.0001,
+      "loss": 1.8216,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8661971830985915,
+      "grad_norm": 0.5889251232147217,
+      "learning_rate": 0.0001,
+      "loss": 1.7786,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8664400194269063,
+      "grad_norm": 0.598010778427124,
+      "learning_rate": 0.0001,
+      "loss": 1.7527,
+      "step": 3568
+    },
+    {
+      "epoch": 0.866682855755221,
+      "grad_norm": 0.6177082061767578,
+      "learning_rate": 0.0001,
+      "loss": 1.8827,
+      "step": 3569
+    },
+    {
+      "epoch": 0.8669256920835356,
+      "grad_norm": 0.624889612197876,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8671685284118504,
+      "grad_norm": 0.6158528327941895,
+      "learning_rate": 0.0001,
+      "loss": 1.8653,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8674113647401651,
+      "grad_norm": 0.5674200654029846,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8676542010684798,
+      "grad_norm": 0.5981341600418091,
+      "learning_rate": 0.0001,
+      "loss": 1.9083,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8678970373967946,
+      "grad_norm": 0.5735521912574768,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8681398737251093,
+      "grad_norm": 0.5744978189468384,
+      "learning_rate": 0.0001,
+      "loss": 1.7118,
+      "step": 3575
+    },
+    {
+      "epoch": 0.868382710053424,
+      "grad_norm": 0.566753625869751,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 3576
+    },
+    {
+      "epoch": 0.8686255463817387,
+      "grad_norm": 0.5849733352661133,
+      "learning_rate": 0.0001,
+      "loss": 1.8447,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8688683827100534,
+      "grad_norm": 0.5533862709999084,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8691112190383682,
+      "grad_norm": 0.5679661631584167,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 3579
+    },
+    {
+      "epoch": 0.8693540553666829,
+      "grad_norm": 0.6186648011207581,
+      "learning_rate": 0.0001,
+      "loss": 1.7538,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8695968916949975,
+      "grad_norm": 0.5672974586486816,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 3581
+    },
+    {
+      "epoch": 0.8698397280233123,
+      "grad_norm": 0.5962534546852112,
+      "learning_rate": 0.0001,
+      "loss": 1.8265,
+      "step": 3582
+    },
+    {
+      "epoch": 0.870082564351627,
+      "grad_norm": 0.5624101161956787,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 3583
+    },
+    {
+      "epoch": 0.8703254006799417,
+      "grad_norm": 0.5874091982841492,
+      "learning_rate": 0.0001,
+      "loss": 1.7295,
+      "step": 3584
+    },
+    {
+      "epoch": 0.8705682370082565,
+      "grad_norm": 0.5865095853805542,
+      "learning_rate": 0.0001,
+      "loss": 1.6617,
+      "step": 3585
+    },
+    {
+      "epoch": 0.8708110733365712,
+      "grad_norm": 0.5641353130340576,
+      "learning_rate": 0.0001,
+      "loss": 1.7834,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8710539096648858,
+      "grad_norm": 0.5696240663528442,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 3587
+    },
+    {
+      "epoch": 0.8712967459932006,
+      "grad_norm": 0.5919355750083923,
+      "learning_rate": 0.0001,
+      "loss": 1.7896,
+      "step": 3588
+    },
+    {
+      "epoch": 0.8715395823215153,
+      "grad_norm": 0.5535728931427002,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 3589
+    },
+    {
+      "epoch": 0.87178241864983,
+      "grad_norm": 0.6022486686706543,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8720252549781448,
+      "grad_norm": 0.7028996348381042,
+      "learning_rate": 0.0001,
+      "loss": 1.592,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8722680913064594,
+      "grad_norm": 0.6094292998313904,
+      "learning_rate": 0.0001,
+      "loss": 1.7961,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8725109276347741,
+      "grad_norm": 0.5781928300857544,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8727537639630889,
+      "grad_norm": 0.6018481850624084,
+      "learning_rate": 0.0001,
+      "loss": 1.7303,
+      "step": 3594
+    },
+    {
+      "epoch": 0.8729966002914036,
+      "grad_norm": 0.5868637561798096,
+      "learning_rate": 0.0001,
+      "loss": 1.7745,
+      "step": 3595
+    },
+    {
+      "epoch": 0.8732394366197183,
+      "grad_norm": 0.5851695537567139,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 3596
+    },
+    {
+      "epoch": 0.873482272948033,
+      "grad_norm": 0.617761492729187,
+      "learning_rate": 0.0001,
+      "loss": 1.8024,
+      "step": 3597
+    },
+    {
+      "epoch": 0.8737251092763477,
+      "grad_norm": 0.5883352756500244,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 3598
+    },
+    {
+      "epoch": 0.8739679456046625,
+      "grad_norm": 0.5812327861785889,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 3599
+    },
+    {
+      "epoch": 0.8742107819329772,
+      "grad_norm": 0.5833345055580139,
+      "learning_rate": 0.0001,
+      "loss": 1.7707,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8744536182612919,
+      "grad_norm": 0.5722734928131104,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 3601
+    },
+    {
+      "epoch": 0.8746964545896067,
+      "grad_norm": 0.5645515322685242,
+      "learning_rate": 0.0001,
+      "loss": 1.5211,
+      "step": 3602
+    },
+    {
+      "epoch": 0.8749392909179213,
+      "grad_norm": 0.5958465337753296,
+      "learning_rate": 0.0001,
+      "loss": 1.702,
+      "step": 3603
+    },
+    {
+      "epoch": 0.875182127246236,
+      "grad_norm": 0.5952925682067871,
+      "learning_rate": 0.0001,
+      "loss": 1.6857,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8754249635745508,
+      "grad_norm": 0.5824811458587646,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8756677999028655,
+      "grad_norm": 0.574558675289154,
+      "learning_rate": 0.0001,
+      "loss": 1.748,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8759106362311802,
+      "grad_norm": 0.5838613510131836,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 3607
+    },
+    {
+      "epoch": 0.8761534725594949,
+      "grad_norm": 0.5939115881919861,
+      "learning_rate": 0.0001,
+      "loss": 1.8715,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8763963088878096,
+      "grad_norm": 0.5994473099708557,
+      "learning_rate": 0.0001,
+      "loss": 1.7028,
+      "step": 3609
+    },
+    {
+      "epoch": 0.8766391452161243,
+      "grad_norm": 0.63933265209198,
+      "learning_rate": 0.0001,
+      "loss": 1.5971,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8768819815444391,
+      "grad_norm": 0.5982732772827148,
+      "learning_rate": 0.0001,
+      "loss": 1.8988,
+      "step": 3611
+    },
+    {
+      "epoch": 0.8771248178727538,
+      "grad_norm": 0.6075206995010376,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 3612
+    },
+    {
+      "epoch": 0.8773676542010684,
+      "grad_norm": 0.6145090460777283,
+      "learning_rate": 0.0001,
+      "loss": 1.8204,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8776104905293832,
+      "grad_norm": 0.5468227863311768,
+      "learning_rate": 0.0001,
+      "loss": 1.6884,
+      "step": 3614
+    },
+    {
+      "epoch": 0.8778533268576979,
+      "grad_norm": 0.5902441143989563,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 3615
+    },
+    {
+      "epoch": 0.8780961631860126,
+      "grad_norm": 0.6053389310836792,
+      "learning_rate": 0.0001,
+      "loss": 1.793,
+      "step": 3616
+    },
+    {
+      "epoch": 0.8783389995143274,
+      "grad_norm": 0.5707299113273621,
+      "learning_rate": 0.0001,
+      "loss": 1.7813,
+      "step": 3617
+    },
+    {
+      "epoch": 0.878581835842642,
+      "grad_norm": 0.5711703896522522,
+      "learning_rate": 0.0001,
+      "loss": 1.4658,
+      "step": 3618
+    },
+    {
+      "epoch": 0.8788246721709567,
+      "grad_norm": 0.5625070333480835,
+      "learning_rate": 0.0001,
+      "loss": 1.6934,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8790675084992715,
+      "grad_norm": 0.5734425783157349,
+      "learning_rate": 0.0001,
+      "loss": 1.7695,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8793103448275862,
+      "grad_norm": 0.6269510984420776,
+      "learning_rate": 0.0001,
+      "loss": 1.7358,
+      "step": 3621
+    },
+    {
+      "epoch": 0.879553181155901,
+      "grad_norm": 0.6240434646606445,
+      "learning_rate": 0.0001,
+      "loss": 1.9493,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8797960174842157,
+      "grad_norm": 0.6287503838539124,
+      "learning_rate": 0.0001,
+      "loss": 1.6903,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8800388538125303,
+      "grad_norm": 0.6086653470993042,
+      "learning_rate": 0.0001,
+      "loss": 1.6884,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8802816901408451,
+      "grad_norm": 0.5977360606193542,
+      "learning_rate": 0.0001,
+      "loss": 1.7675,
+      "step": 3625
+    },
+    {
+      "epoch": 0.8805245264691598,
+      "grad_norm": 0.5500513911247253,
+      "learning_rate": 0.0001,
+      "loss": 1.5043,
+      "step": 3626
+    },
+    {
+      "epoch": 0.8807673627974745,
+      "grad_norm": 0.5777398943901062,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8810101991257893,
+      "grad_norm": 0.5849345326423645,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 3628
+    },
+    {
+      "epoch": 0.8812530354541039,
+      "grad_norm": 0.5858832597732544,
+      "learning_rate": 0.0001,
+      "loss": 1.7539,
+      "step": 3629
+    },
+    {
+      "epoch": 0.8814958717824186,
+      "grad_norm": 0.5634268522262573,
+      "learning_rate": 0.0001,
+      "loss": 1.497,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8817387081107334,
+      "grad_norm": 0.5676572918891907,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8819815444390481,
+      "grad_norm": 0.6174838542938232,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8822243807673628,
+      "grad_norm": 0.5701363682746887,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8824672170956775,
+      "grad_norm": 0.6180095076560974,
+      "learning_rate": 0.0001,
+      "loss": 1.7446,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8827100534239922,
+      "grad_norm": 0.5902948379516602,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8829528897523069,
+      "grad_norm": 0.6113959550857544,
+      "learning_rate": 0.0001,
+      "loss": 1.8169,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8831957260806217,
+      "grad_norm": 0.5810447335243225,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8834385624089364,
+      "grad_norm": 0.646421492099762,
+      "learning_rate": 0.0001,
+      "loss": 1.8118,
+      "step": 3638
+    },
+    {
+      "epoch": 0.883681398737251,
+      "grad_norm": 0.566224217414856,
+      "learning_rate": 0.0001,
+      "loss": 1.8104,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8839242350655658,
+      "grad_norm": 0.5958870053291321,
+      "learning_rate": 0.0001,
+      "loss": 1.8216,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8841670713938805,
+      "grad_norm": 0.6059859395027161,
+      "learning_rate": 0.0001,
+      "loss": 1.7857,
+      "step": 3641
+    },
+    {
+      "epoch": 0.8844099077221952,
+      "grad_norm": 0.5715227127075195,
+      "learning_rate": 0.0001,
+      "loss": 1.7616,
+      "step": 3642
+    },
+    {
+      "epoch": 0.88465274405051,
+      "grad_norm": 0.6450254321098328,
+      "learning_rate": 0.0001,
+      "loss": 1.8297,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8848955803788247,
+      "grad_norm": 0.6196300983428955,
+      "learning_rate": 0.0001,
+      "loss": 1.7653,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8851384167071394,
+      "grad_norm": 0.5951306819915771,
+      "learning_rate": 0.0001,
+      "loss": 1.6767,
+      "step": 3645
+    },
+    {
+      "epoch": 0.8853812530354541,
+      "grad_norm": 0.6688085198402405,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 3646
+    },
+    {
+      "epoch": 0.8856240893637688,
+      "grad_norm": 0.5445377826690674,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 3647
+    },
+    {
+      "epoch": 0.8858669256920836,
+      "grad_norm": 0.6012933254241943,
+      "learning_rate": 0.0001,
+      "loss": 1.7783,
+      "step": 3648
+    },
+    {
+      "epoch": 0.8861097620203983,
+      "grad_norm": 0.6122232675552368,
+      "learning_rate": 0.0001,
+      "loss": 1.8086,
+      "step": 3649
+    },
+    {
+      "epoch": 0.8863525983487129,
+      "grad_norm": 0.5938394665718079,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8865954346770277,
+      "grad_norm": 0.610999584197998,
+      "learning_rate": 0.0001,
+      "loss": 1.8749,
+      "step": 3651
+    },
+    {
+      "epoch": 0.8868382710053424,
+      "grad_norm": 0.5855596661567688,
+      "learning_rate": 0.0001,
+      "loss": 1.725,
+      "step": 3652
+    },
+    {
+      "epoch": 0.8870811073336571,
+      "grad_norm": 0.5986215472221375,
+      "learning_rate": 0.0001,
+      "loss": 1.7427,
+      "step": 3653
+    },
+    {
+      "epoch": 0.8873239436619719,
+      "grad_norm": 0.6195816993713379,
+      "learning_rate": 0.0001,
+      "loss": 1.6951,
+      "step": 3654
+    },
+    {
+      "epoch": 0.8875667799902865,
+      "grad_norm": 0.5849098563194275,
+      "learning_rate": 0.0001,
+      "loss": 1.8607,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8878096163186012,
+      "grad_norm": 0.6396300792694092,
+      "learning_rate": 0.0001,
+      "loss": 1.8764,
+      "step": 3656
+    },
+    {
+      "epoch": 0.888052452646916,
+      "grad_norm": 0.5767935514450073,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8882952889752307,
+      "grad_norm": 0.5865722894668579,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8885381253035454,
+      "grad_norm": 0.5967360734939575,
+      "learning_rate": 0.0001,
+      "loss": 1.7648,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8887809616318602,
+      "grad_norm": 0.6081897616386414,
+      "learning_rate": 0.0001,
+      "loss": 1.7452,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8890237979601748,
+      "grad_norm": 0.6059684157371521,
+      "learning_rate": 0.0001,
+      "loss": 1.7785,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8892666342884895,
+      "grad_norm": 0.5943751931190491,
+      "learning_rate": 0.0001,
+      "loss": 1.7477,
+      "step": 3662
+    },
+    {
+      "epoch": 0.8895094706168043,
+      "grad_norm": 0.6292147636413574,
+      "learning_rate": 0.0001,
+      "loss": 1.8816,
+      "step": 3663
+    },
+    {
+      "epoch": 0.889752306945119,
+      "grad_norm": 0.6073547601699829,
+      "learning_rate": 0.0001,
+      "loss": 1.7417,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8899951432734337,
+      "grad_norm": 0.5393562912940979,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8902379796017484,
+      "grad_norm": 0.5665006637573242,
+      "learning_rate": 0.0001,
+      "loss": 1.5303,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8904808159300631,
+      "grad_norm": 0.5666619539260864,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 3667
+    },
+    {
+      "epoch": 0.8907236522583779,
+      "grad_norm": 0.5895118117332458,
+      "learning_rate": 0.0001,
+      "loss": 1.6932,
+      "step": 3668
+    },
+    {
+      "epoch": 0.8909664885866926,
+      "grad_norm": 0.5936942100524902,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8912093249150073,
+      "grad_norm": 0.6128193736076355,
+      "learning_rate": 0.0001,
+      "loss": 1.69,
+      "step": 3670
+    },
+    {
+      "epoch": 0.891452161243322,
+      "grad_norm": 0.627239465713501,
+      "learning_rate": 0.0001,
+      "loss": 1.823,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8916949975716367,
+      "grad_norm": 0.5960967540740967,
+      "learning_rate": 0.0001,
+      "loss": 1.6852,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8919378338999514,
+      "grad_norm": 0.6583183407783508,
+      "learning_rate": 0.0001,
+      "loss": 1.8119,
+      "step": 3673
+    },
+    {
+      "epoch": 0.8921806702282662,
+      "grad_norm": 0.5878483653068542,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8924235065565809,
+      "grad_norm": 0.5881945490837097,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 3675
+    },
+    {
+      "epoch": 0.8926663428848955,
+      "grad_norm": 0.5920316576957703,
+      "learning_rate": 0.0001,
+      "loss": 1.7318,
+      "step": 3676
+    },
+    {
+      "epoch": 0.8929091792132103,
+      "grad_norm": 0.6558576226234436,
+      "learning_rate": 0.0001,
+      "loss": 1.7684,
+      "step": 3677
+    },
+    {
+      "epoch": 0.893152015541525,
+      "grad_norm": 0.6191709637641907,
+      "learning_rate": 0.0001,
+      "loss": 1.823,
+      "step": 3678
+    },
+    {
+      "epoch": 0.8933948518698397,
+      "grad_norm": 0.6138198971748352,
+      "learning_rate": 0.0001,
+      "loss": 1.8532,
+      "step": 3679
+    },
+    {
+      "epoch": 0.8936376881981545,
+      "grad_norm": 0.6346641182899475,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 3680
+    },
+    {
+      "epoch": 0.8938805245264692,
+      "grad_norm": 0.6237822771072388,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 3681
+    },
+    {
+      "epoch": 0.8941233608547838,
+      "grad_norm": 0.6110655069351196,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8943661971830986,
+      "grad_norm": 0.5926641821861267,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 3683
+    },
+    {
+      "epoch": 0.8946090335114133,
+      "grad_norm": 0.6240578889846802,
+      "learning_rate": 0.0001,
+      "loss": 1.6809,
+      "step": 3684
+    },
+    {
+      "epoch": 0.894851869839728,
+      "grad_norm": 0.6072526574134827,
+      "learning_rate": 0.0001,
+      "loss": 1.6632,
+      "step": 3685
+    },
+    {
+      "epoch": 0.8950947061680428,
+      "grad_norm": 0.6060703992843628,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8953375424963574,
+      "grad_norm": 0.5971805453300476,
+      "learning_rate": 0.0001,
+      "loss": 1.676,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8955803788246721,
+      "grad_norm": 0.6302820444107056,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 3688
+    },
+    {
+      "epoch": 0.8958232151529869,
+      "grad_norm": 0.5734922289848328,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8960660514813016,
+      "grad_norm": 0.608064591884613,
+      "learning_rate": 0.0001,
+      "loss": 1.7129,
+      "step": 3690
+    },
+    {
+      "epoch": 0.8963088878096164,
+      "grad_norm": 0.605770468711853,
+      "learning_rate": 0.0001,
+      "loss": 1.7061,
+      "step": 3691
+    },
+    {
+      "epoch": 0.896551724137931,
+      "grad_norm": 0.5936496257781982,
+      "learning_rate": 0.0001,
+      "loss": 1.8021,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8967945604662457,
+      "grad_norm": 0.619182288646698,
+      "learning_rate": 0.0001,
+      "loss": 1.6411,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8970373967945605,
+      "grad_norm": 0.5974726676940918,
+      "learning_rate": 0.0001,
+      "loss": 1.8246,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8972802331228752,
+      "grad_norm": 0.6047124266624451,
+      "learning_rate": 0.0001,
+      "loss": 1.715,
+      "step": 3695
+    },
+    {
+      "epoch": 0.8975230694511899,
+      "grad_norm": 0.606947660446167,
+      "learning_rate": 0.0001,
+      "loss": 1.7117,
+      "step": 3696
+    },
+    {
+      "epoch": 0.8977659057795047,
+      "grad_norm": 0.5917686223983765,
+      "learning_rate": 0.0001,
+      "loss": 1.7008,
+      "step": 3697
+    },
+    {
+      "epoch": 0.8980087421078193,
+      "grad_norm": 0.6276222467422485,
+      "learning_rate": 0.0001,
+      "loss": 1.8742,
+      "step": 3698
+    },
+    {
+      "epoch": 0.898251578436134,
+      "grad_norm": 0.5839412212371826,
+      "learning_rate": 0.0001,
+      "loss": 1.7297,
+      "step": 3699
+    },
+    {
+      "epoch": 0.8984944147644488,
+      "grad_norm": 0.5966883301734924,
+      "learning_rate": 0.0001,
+      "loss": 1.6937,
+      "step": 3700
+    },
+    {
+      "epoch": 0.8987372510927635,
+      "grad_norm": 0.5819137692451477,
+      "learning_rate": 0.0001,
+      "loss": 1.6871,
+      "step": 3701
+    },
+    {
+      "epoch": 0.8989800874210782,
+      "grad_norm": 0.6237320899963379,
+      "learning_rate": 0.0001,
+      "loss": 1.7537,
+      "step": 3702
+    },
+    {
+      "epoch": 0.8992229237493929,
+      "grad_norm": 0.587004542350769,
+      "learning_rate": 0.0001,
+      "loss": 1.771,
+      "step": 3703
+    },
+    {
+      "epoch": 0.8994657600777076,
+      "grad_norm": 0.5957283973693848,
+      "learning_rate": 0.0001,
+      "loss": 1.8436,
+      "step": 3704
+    },
+    {
+      "epoch": 0.8997085964060223,
+      "grad_norm": 0.6247084736824036,
+      "learning_rate": 0.0001,
+      "loss": 1.7767,
+      "step": 3705
+    },
+    {
+      "epoch": 0.8999514327343371,
+      "grad_norm": 0.5712265968322754,
+      "learning_rate": 0.0001,
+      "loss": 1.7591,
+      "step": 3706
+    },
+    {
+      "epoch": 0.9001942690626518,
+      "grad_norm": 0.6066750884056091,
+      "learning_rate": 0.0001,
+      "loss": 1.6773,
+      "step": 3707
+    },
+    {
+      "epoch": 0.9004371053909664,
+      "grad_norm": 0.5812699198722839,
+      "learning_rate": 0.0001,
+      "loss": 1.6081,
+      "step": 3708
+    },
+    {
+      "epoch": 0.9006799417192812,
+      "grad_norm": 0.5693857669830322,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 3709
+    },
+    {
+      "epoch": 0.9009227780475959,
+      "grad_norm": 0.6075789332389832,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9011656143759106,
+      "grad_norm": 0.588006854057312,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 3711
+    },
+    {
+      "epoch": 0.9014084507042254,
+      "grad_norm": 0.5788747072219849,
+      "learning_rate": 0.0001,
+      "loss": 1.7602,
+      "step": 3712
+    },
+    {
+      "epoch": 0.90165128703254,
+      "grad_norm": 0.6109026074409485,
+      "learning_rate": 0.0001,
+      "loss": 1.7753,
+      "step": 3713
+    },
+    {
+      "epoch": 0.9018941233608548,
+      "grad_norm": 0.6000962853431702,
+      "learning_rate": 0.0001,
+      "loss": 1.7303,
+      "step": 3714
+    },
+    {
+      "epoch": 0.9021369596891695,
+      "grad_norm": 0.5677233934402466,
+      "learning_rate": 0.0001,
+      "loss": 1.7165,
+      "step": 3715
+    },
+    {
+      "epoch": 0.9023797960174842,
+      "grad_norm": 0.6030009984970093,
+      "learning_rate": 0.0001,
+      "loss": 1.8595,
+      "step": 3716
+    },
+    {
+      "epoch": 0.902622632345799,
+      "grad_norm": 0.5491967797279358,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 3717
+    },
+    {
+      "epoch": 0.9028654686741137,
+      "grad_norm": 0.6326006054878235,
+      "learning_rate": 0.0001,
+      "loss": 1.6389,
+      "step": 3718
+    },
+    {
+      "epoch": 0.9031083050024283,
+      "grad_norm": 0.6032041907310486,
+      "learning_rate": 0.0001,
+      "loss": 1.7333,
+      "step": 3719
+    },
+    {
+      "epoch": 0.9033511413307431,
+      "grad_norm": 0.6120818257331848,
+      "learning_rate": 0.0001,
+      "loss": 1.7694,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9035939776590578,
+      "grad_norm": 0.6020964980125427,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 3721
+    },
+    {
+      "epoch": 0.9038368139873725,
+      "grad_norm": 0.6394541263580322,
+      "learning_rate": 0.0001,
+      "loss": 1.8643,
+      "step": 3722
+    },
+    {
+      "epoch": 0.9040796503156873,
+      "grad_norm": 0.5862223505973816,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 3723
+    },
+    {
+      "epoch": 0.9043224866440019,
+      "grad_norm": 0.6442126035690308,
+      "learning_rate": 0.0001,
+      "loss": 1.9307,
+      "step": 3724
+    },
+    {
+      "epoch": 0.9045653229723166,
+      "grad_norm": 0.5906519293785095,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 3725
+    },
+    {
+      "epoch": 0.9048081593006314,
+      "grad_norm": 0.5946016311645508,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 3726
+    },
+    {
+      "epoch": 0.9050509956289461,
+      "grad_norm": 0.5882769227027893,
+      "learning_rate": 0.0001,
+      "loss": 1.7124,
+      "step": 3727
+    },
+    {
+      "epoch": 0.9052938319572608,
+      "grad_norm": 0.6075938940048218,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 3728
+    },
+    {
+      "epoch": 0.9055366682855756,
+      "grad_norm": 0.5872309803962708,
+      "learning_rate": 0.0001,
+      "loss": 1.7817,
+      "step": 3729
+    },
+    {
+      "epoch": 0.9057795046138902,
+      "grad_norm": 0.5707372426986694,
+      "learning_rate": 0.0001,
+      "loss": 1.7341,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9060223409422049,
+      "grad_norm": 0.5877151489257812,
+      "learning_rate": 0.0001,
+      "loss": 1.7046,
+      "step": 3731
+    },
+    {
+      "epoch": 0.9062651772705197,
+      "grad_norm": 0.5864531993865967,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 3732
+    },
+    {
+      "epoch": 0.9065080135988344,
+      "grad_norm": 0.6085807085037231,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 3733
+    },
+    {
+      "epoch": 0.906750849927149,
+      "grad_norm": 0.6236429214477539,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 3734
+    },
+    {
+      "epoch": 0.9069936862554638,
+      "grad_norm": 0.5786526799201965,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 3735
+    },
+    {
+      "epoch": 0.9072365225837785,
+      "grad_norm": 0.6803106665611267,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 3736
+    },
+    {
+      "epoch": 0.9074793589120933,
+      "grad_norm": 0.6150429844856262,
+      "learning_rate": 0.0001,
+      "loss": 1.7344,
+      "step": 3737
+    },
+    {
+      "epoch": 0.907722195240408,
+      "grad_norm": 0.5588183403015137,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 3738
+    },
+    {
+      "epoch": 0.9079650315687227,
+      "grad_norm": 0.6157742142677307,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 3739
+    },
+    {
+      "epoch": 0.9082078678970374,
+      "grad_norm": 0.6282200813293457,
+      "learning_rate": 0.0001,
+      "loss": 1.7407,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9084507042253521,
+      "grad_norm": 0.6336793303489685,
+      "learning_rate": 0.0001,
+      "loss": 1.7984,
+      "step": 3741
+    },
+    {
+      "epoch": 0.9086935405536668,
+      "grad_norm": 0.6165140271186829,
+      "learning_rate": 0.0001,
+      "loss": 1.8259,
+      "step": 3742
+    },
+    {
+      "epoch": 0.9089363768819816,
+      "grad_norm": 0.5867929458618164,
+      "learning_rate": 0.0001,
+      "loss": 1.7165,
+      "step": 3743
+    },
+    {
+      "epoch": 0.9091792132102963,
+      "grad_norm": 0.5789873600006104,
+      "learning_rate": 0.0001,
+      "loss": 1.4274,
+      "step": 3744
+    },
+    {
+      "epoch": 0.9094220495386109,
+      "grad_norm": 0.6115802526473999,
+      "learning_rate": 0.0001,
+      "loss": 1.8345,
+      "step": 3745
+    },
+    {
+      "epoch": 0.9096648858669257,
+      "grad_norm": 0.5854842066764832,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 3746
+    },
+    {
+      "epoch": 0.9099077221952404,
+      "grad_norm": 0.5675647854804993,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 3747
+    },
+    {
+      "epoch": 0.9101505585235551,
+      "grad_norm": 0.6123351454734802,
+      "learning_rate": 0.0001,
+      "loss": 1.691,
+      "step": 3748
+    },
+    {
+      "epoch": 0.9103933948518699,
+      "grad_norm": 0.6066859364509583,
+      "learning_rate": 0.0001,
+      "loss": 1.7858,
+      "step": 3749
+    },
+    {
+      "epoch": 0.9106362311801846,
+      "grad_norm": 0.6148110032081604,
+      "learning_rate": 0.0001,
+      "loss": 1.8063,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9108790675084992,
+      "grad_norm": 0.58165043592453,
+      "learning_rate": 0.0001,
+      "loss": 1.7182,
+      "step": 3751
+    },
+    {
+      "epoch": 0.911121903836814,
+      "grad_norm": 0.5953090190887451,
+      "learning_rate": 0.0001,
+      "loss": 1.7449,
+      "step": 3752
+    },
+    {
+      "epoch": 0.9113647401651287,
+      "grad_norm": 0.5710526704788208,
+      "learning_rate": 0.0001,
+      "loss": 1.7502,
+      "step": 3753
+    },
+    {
+      "epoch": 0.9116075764934434,
+      "grad_norm": 0.5716732144355774,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 3754
+    },
+    {
+      "epoch": 0.9118504128217582,
+      "grad_norm": 0.576026439666748,
+      "learning_rate": 0.0001,
+      "loss": 1.7577,
+      "step": 3755
+    },
+    {
+      "epoch": 0.9120932491500728,
+      "grad_norm": 0.5796546339988708,
+      "learning_rate": 0.0001,
+      "loss": 1.7471,
+      "step": 3756
+    },
+    {
+      "epoch": 0.9123360854783875,
+      "grad_norm": 0.5911909341812134,
+      "learning_rate": 0.0001,
+      "loss": 1.6457,
+      "step": 3757
+    },
+    {
+      "epoch": 0.9125789218067023,
+      "grad_norm": 0.5691782236099243,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 3758
+    },
+    {
+      "epoch": 0.912821758135017,
+      "grad_norm": 0.6342900395393372,
+      "learning_rate": 0.0001,
+      "loss": 1.7737,
+      "step": 3759
+    },
+    {
+      "epoch": 0.9130645944633318,
+      "grad_norm": 0.5813829302787781,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 3760
+    },
+    {
+      "epoch": 0.9133074307916464,
+      "grad_norm": 0.6647397875785828,
+      "learning_rate": 0.0001,
+      "loss": 1.7957,
+      "step": 3761
+    },
+    {
+      "epoch": 0.9135502671199611,
+      "grad_norm": 0.5722330212593079,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 3762
+    },
+    {
+      "epoch": 0.9137931034482759,
+      "grad_norm": 0.5854828357696533,
+      "learning_rate": 0.0001,
+      "loss": 1.6495,
+      "step": 3763
+    },
+    {
+      "epoch": 0.9140359397765906,
+      "grad_norm": 0.6135445237159729,
+      "learning_rate": 0.0001,
+      "loss": 1.7936,
+      "step": 3764
+    },
+    {
+      "epoch": 0.9142787761049053,
+      "grad_norm": 0.6264251470565796,
+      "learning_rate": 0.0001,
+      "loss": 1.869,
+      "step": 3765
+    },
+    {
+      "epoch": 0.91452161243322,
+      "grad_norm": 0.5948366522789001,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 3766
+    },
+    {
+      "epoch": 0.9147644487615347,
+      "grad_norm": 0.5939854383468628,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 3767
+    },
+    {
+      "epoch": 0.9150072850898494,
+      "grad_norm": 0.6219908595085144,
+      "learning_rate": 0.0001,
+      "loss": 1.7704,
+      "step": 3768
+    },
+    {
+      "epoch": 0.9152501214181642,
+      "grad_norm": 0.5622676610946655,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 3769
+    },
+    {
+      "epoch": 0.9154929577464789,
+      "grad_norm": 0.5718271732330322,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 3770
+    },
+    {
+      "epoch": 0.9157357940747936,
+      "grad_norm": 0.5890514254570007,
+      "learning_rate": 0.0001,
+      "loss": 1.6326,
+      "step": 3771
+    },
+    {
+      "epoch": 0.9159786304031083,
+      "grad_norm": 0.5604651570320129,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 3772
+    },
+    {
+      "epoch": 0.916221466731423,
+      "grad_norm": 0.6045302748680115,
+      "learning_rate": 0.0001,
+      "loss": 1.8132,
+      "step": 3773
+    },
+    {
+      "epoch": 0.9164643030597377,
+      "grad_norm": 0.5818734765052795,
+      "learning_rate": 0.0001,
+      "loss": 1.7577,
+      "step": 3774
+    },
+    {
+      "epoch": 0.9167071393880525,
+      "grad_norm": 0.5890612006187439,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 3775
+    },
+    {
+      "epoch": 0.9169499757163672,
+      "grad_norm": 0.5723394751548767,
+      "learning_rate": 0.0001,
+      "loss": 1.6875,
+      "step": 3776
+    },
+    {
+      "epoch": 0.9171928120446818,
+      "grad_norm": 0.6129171848297119,
+      "learning_rate": 0.0001,
+      "loss": 1.8051,
+      "step": 3777
+    },
+    {
+      "epoch": 0.9174356483729966,
+      "grad_norm": 0.5996651649475098,
+      "learning_rate": 0.0001,
+      "loss": 1.8676,
+      "step": 3778
+    },
+    {
+      "epoch": 0.9176784847013113,
+      "grad_norm": 0.6562318205833435,
+      "learning_rate": 0.0001,
+      "loss": 1.744,
+      "step": 3779
+    },
+    {
+      "epoch": 0.917921321029626,
+      "grad_norm": 0.5926910638809204,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9181641573579408,
+      "grad_norm": 0.6157199144363403,
+      "learning_rate": 0.0001,
+      "loss": 1.8089,
+      "step": 3781
+    },
+    {
+      "epoch": 0.9184069936862554,
+      "grad_norm": 0.6023468971252441,
+      "learning_rate": 0.0001,
+      "loss": 1.7532,
+      "step": 3782
+    },
+    {
+      "epoch": 0.9186498300145702,
+      "grad_norm": 0.6074219942092896,
+      "learning_rate": 0.0001,
+      "loss": 1.7703,
+      "step": 3783
+    },
+    {
+      "epoch": 0.9188926663428849,
+      "grad_norm": 0.9243726134300232,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 3784
+    },
+    {
+      "epoch": 0.9191355026711996,
+      "grad_norm": 0.5886449813842773,
+      "learning_rate": 0.0001,
+      "loss": 1.8595,
+      "step": 3785
+    },
+    {
+      "epoch": 0.9193783389995144,
+      "grad_norm": 0.6371552348136902,
+      "learning_rate": 0.0001,
+      "loss": 1.9148,
+      "step": 3786
+    },
+    {
+      "epoch": 0.919621175327829,
+      "grad_norm": 0.6345494389533997,
+      "learning_rate": 0.0001,
+      "loss": 1.7892,
+      "step": 3787
+    },
+    {
+      "epoch": 0.9198640116561437,
+      "grad_norm": 0.5974324345588684,
+      "learning_rate": 0.0001,
+      "loss": 1.8767,
+      "step": 3788
+    },
+    {
+      "epoch": 0.9201068479844585,
+      "grad_norm": 0.5730004906654358,
+      "learning_rate": 0.0001,
+      "loss": 1.6995,
+      "step": 3789
+    },
+    {
+      "epoch": 0.9203496843127732,
+      "grad_norm": 0.5770270824432373,
+      "learning_rate": 0.0001,
+      "loss": 1.8269,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9205925206410879,
+      "grad_norm": 0.5567600131034851,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 3791
+    },
+    {
+      "epoch": 0.9208353569694027,
+      "grad_norm": 0.5911000967025757,
+      "learning_rate": 0.0001,
+      "loss": 1.7848,
+      "step": 3792
+    },
+    {
+      "epoch": 0.9210781932977173,
+      "grad_norm": 0.5451039671897888,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 3793
+    },
+    {
+      "epoch": 0.921321029626032,
+      "grad_norm": 0.6299775242805481,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 3794
+    },
+    {
+      "epoch": 0.9215638659543468,
+      "grad_norm": 0.5983388423919678,
+      "learning_rate": 0.0001,
+      "loss": 1.692,
+      "step": 3795
+    },
+    {
+      "epoch": 0.9218067022826615,
+      "grad_norm": 0.5751836895942688,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 3796
+    },
+    {
+      "epoch": 0.9220495386109762,
+      "grad_norm": 0.5966198444366455,
+      "learning_rate": 0.0001,
+      "loss": 1.8375,
+      "step": 3797
+    },
+    {
+      "epoch": 0.922292374939291,
+      "grad_norm": 0.5846254825592041,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 3798
+    },
+    {
+      "epoch": 0.9225352112676056,
+      "grad_norm": 0.6296234726905823,
+      "learning_rate": 0.0001,
+      "loss": 1.8008,
+      "step": 3799
+    },
+    {
+      "epoch": 0.9227780475959203,
+      "grad_norm": 0.5613099336624146,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9230208839242351,
+      "grad_norm": 0.6608403325080872,
+      "learning_rate": 0.0001,
+      "loss": 1.7196,
+      "step": 3801
+    },
+    {
+      "epoch": 0.9232637202525498,
+      "grad_norm": 0.6562487483024597,
+      "learning_rate": 0.0001,
+      "loss": 1.7521,
+      "step": 3802
+    },
+    {
+      "epoch": 0.9235065565808644,
+      "grad_norm": 0.5922927260398865,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 3803
+    },
+    {
+      "epoch": 0.9237493929091792,
+      "grad_norm": 0.601926326751709,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 3804
+    },
+    {
+      "epoch": 0.9239922292374939,
+      "grad_norm": 0.6223918199539185,
+      "learning_rate": 0.0001,
+      "loss": 1.8285,
+      "step": 3805
+    },
+    {
+      "epoch": 0.9242350655658087,
+      "grad_norm": 0.6224992275238037,
+      "learning_rate": 0.0001,
+      "loss": 1.871,
+      "step": 3806
+    },
+    {
+      "epoch": 0.9244779018941234,
+      "grad_norm": 0.6175224184989929,
+      "learning_rate": 0.0001,
+      "loss": 1.7312,
+      "step": 3807
+    },
+    {
+      "epoch": 0.924720738222438,
+      "grad_norm": 0.5862079858779907,
+      "learning_rate": 0.0001,
+      "loss": 1.6354,
+      "step": 3808
+    },
+    {
+      "epoch": 0.9249635745507528,
+      "grad_norm": 0.6115992069244385,
+      "learning_rate": 0.0001,
+      "loss": 1.8493,
+      "step": 3809
+    },
+    {
+      "epoch": 0.9252064108790675,
+      "grad_norm": 0.620858371257782,
+      "learning_rate": 0.0001,
+      "loss": 1.7691,
+      "step": 3810
+    },
+    {
+      "epoch": 0.9254492472073822,
+      "grad_norm": 0.584667980670929,
+      "learning_rate": 0.0001,
+      "loss": 1.6955,
+      "step": 3811
+    },
+    {
+      "epoch": 0.925692083535697,
+      "grad_norm": 0.5902495384216309,
+      "learning_rate": 0.0001,
+      "loss": 1.6646,
+      "step": 3812
+    },
+    {
+      "epoch": 0.9259349198640117,
+      "grad_norm": 0.611920177936554,
+      "learning_rate": 0.0001,
+      "loss": 1.757,
+      "step": 3813
+    },
+    {
+      "epoch": 0.9261777561923263,
+      "grad_norm": 0.5571722388267517,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 3814
+    },
+    {
+      "epoch": 0.9264205925206411,
+      "grad_norm": 0.5735197067260742,
+      "learning_rate": 0.0001,
+      "loss": 1.7786,
+      "step": 3815
+    },
+    {
+      "epoch": 0.9266634288489558,
+      "grad_norm": 0.5902324318885803,
+      "learning_rate": 0.0001,
+      "loss": 1.6966,
+      "step": 3816
+    },
+    {
+      "epoch": 0.9269062651772705,
+      "grad_norm": 0.5405271649360657,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 3817
+    },
+    {
+      "epoch": 0.9271491015055853,
+      "grad_norm": 0.5784160494804382,
+      "learning_rate": 0.0001,
+      "loss": 1.7828,
+      "step": 3818
+    },
+    {
+      "epoch": 0.9273919378339,
+      "grad_norm": 0.6429668664932251,
+      "learning_rate": 0.0001,
+      "loss": 1.7877,
+      "step": 3819
+    },
+    {
+      "epoch": 0.9276347741622146,
+      "grad_norm": 0.6807743906974792,
+      "learning_rate": 0.0001,
+      "loss": 1.8208,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9278776104905294,
+      "grad_norm": 0.6062427163124084,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 3821
+    },
+    {
+      "epoch": 0.9281204468188441,
+      "grad_norm": 0.5669822096824646,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 3822
+    },
+    {
+      "epoch": 0.9283632831471588,
+      "grad_norm": 0.6149395704269409,
+      "learning_rate": 0.0001,
+      "loss": 1.788,
+      "step": 3823
+    },
+    {
+      "epoch": 0.9286061194754736,
+      "grad_norm": 0.6279157996177673,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 3824
+    },
+    {
+      "epoch": 0.9288489558037882,
+      "grad_norm": 0.5734469294548035,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 3825
+    },
+    {
+      "epoch": 0.9290917921321029,
+      "grad_norm": 0.5863770246505737,
+      "learning_rate": 0.0001,
+      "loss": 1.751,
+      "step": 3826
+    },
+    {
+      "epoch": 0.9293346284604177,
+      "grad_norm": 0.5849167704582214,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 3827
+    },
+    {
+      "epoch": 0.9295774647887324,
+      "grad_norm": 0.5744926333427429,
+      "learning_rate": 0.0001,
+      "loss": 1.8028,
+      "step": 3828
+    },
+    {
+      "epoch": 0.9298203011170472,
+      "grad_norm": 0.5908921360969543,
+      "learning_rate": 0.0001,
+      "loss": 1.8405,
+      "step": 3829
+    },
+    {
+      "epoch": 0.9300631374453618,
+      "grad_norm": 0.6141979098320007,
+      "learning_rate": 0.0001,
+      "loss": 1.7229,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9303059737736765,
+      "grad_norm": 0.6094731688499451,
+      "learning_rate": 0.0001,
+      "loss": 1.832,
+      "step": 3831
+    },
+    {
+      "epoch": 0.9305488101019913,
+      "grad_norm": 0.615892231464386,
+      "learning_rate": 0.0001,
+      "loss": 1.9006,
+      "step": 3832
+    },
+    {
+      "epoch": 0.930791646430306,
+      "grad_norm": 0.5927395820617676,
+      "learning_rate": 0.0001,
+      "loss": 1.7176,
+      "step": 3833
+    },
+    {
+      "epoch": 0.9310344827586207,
+      "grad_norm": 0.5887064337730408,
+      "learning_rate": 0.0001,
+      "loss": 1.6997,
+      "step": 3834
+    },
+    {
+      "epoch": 0.9312773190869355,
+      "grad_norm": 0.6049761772155762,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 3835
+    },
+    {
+      "epoch": 0.9315201554152501,
+      "grad_norm": 0.5994516015052795,
+      "learning_rate": 0.0001,
+      "loss": 1.7624,
+      "step": 3836
+    },
+    {
+      "epoch": 0.9317629917435648,
+      "grad_norm": 0.6419912576675415,
+      "learning_rate": 0.0001,
+      "loss": 1.8233,
+      "step": 3837
+    },
+    {
+      "epoch": 0.9320058280718796,
+      "grad_norm": 0.5899677276611328,
+      "learning_rate": 0.0001,
+      "loss": 1.7693,
+      "step": 3838
+    },
+    {
+      "epoch": 0.9322486644001943,
+      "grad_norm": 0.6384326219558716,
+      "learning_rate": 0.0001,
+      "loss": 1.8361,
+      "step": 3839
+    },
+    {
+      "epoch": 0.932491500728509,
+      "grad_norm": 0.5595104694366455,
+      "learning_rate": 0.0001,
+      "loss": 1.7417,
+      "step": 3840
+    },
+    {
+      "epoch": 0.9327343370568237,
+      "grad_norm": 0.5852187275886536,
+      "learning_rate": 0.0001,
+      "loss": 1.6915,
+      "step": 3841
+    },
+    {
+      "epoch": 0.9329771733851384,
+      "grad_norm": 0.6391562223434448,
+      "learning_rate": 0.0001,
+      "loss": 1.8735,
+      "step": 3842
+    },
+    {
+      "epoch": 0.9332200097134531,
+      "grad_norm": 0.5452216267585754,
+      "learning_rate": 0.0001,
+      "loss": 1.6466,
+      "step": 3843
+    },
+    {
+      "epoch": 0.9334628460417679,
+      "grad_norm": 0.5904660224914551,
+      "learning_rate": 0.0001,
+      "loss": 1.7699,
+      "step": 3844
+    },
+    {
+      "epoch": 0.9337056823700826,
+      "grad_norm": 0.6349517107009888,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 3845
+    },
+    {
+      "epoch": 0.9339485186983972,
+      "grad_norm": 0.5781106948852539,
+      "learning_rate": 0.0001,
+      "loss": 1.6949,
+      "step": 3846
+    },
+    {
+      "epoch": 0.934191355026712,
+      "grad_norm": 0.5469078421592712,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 3847
+    },
+    {
+      "epoch": 0.9344341913550267,
+      "grad_norm": 0.5812481641769409,
+      "learning_rate": 0.0001,
+      "loss": 1.477,
+      "step": 3848
+    },
+    {
+      "epoch": 0.9346770276833414,
+      "grad_norm": 0.5483930110931396,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 3849
+    },
+    {
+      "epoch": 0.9349198640116562,
+      "grad_norm": 0.6049747467041016,
+      "learning_rate": 0.0001,
+      "loss": 1.7515,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9351627003399708,
+      "grad_norm": 0.6384432315826416,
+      "learning_rate": 0.0001,
+      "loss": 1.9378,
+      "step": 3851
+    },
+    {
+      "epoch": 0.9354055366682856,
+      "grad_norm": 0.5857444405555725,
+      "learning_rate": 0.0001,
+      "loss": 1.7695,
+      "step": 3852
+    },
+    {
+      "epoch": 0.9356483729966003,
+      "grad_norm": 0.5994032025337219,
+      "learning_rate": 0.0001,
+      "loss": 1.831,
+      "step": 3853
+    },
+    {
+      "epoch": 0.935891209324915,
+      "grad_norm": 0.6487367749214172,
+      "learning_rate": 0.0001,
+      "loss": 1.8053,
+      "step": 3854
+    },
+    {
+      "epoch": 0.9361340456532298,
+      "grad_norm": 0.6017345786094666,
+      "learning_rate": 0.0001,
+      "loss": 1.7729,
+      "step": 3855
+    },
+    {
+      "epoch": 0.9363768819815445,
+      "grad_norm": 0.6166176199913025,
+      "learning_rate": 0.0001,
+      "loss": 1.8245,
+      "step": 3856
+    },
+    {
+      "epoch": 0.9366197183098591,
+      "grad_norm": 0.5821506977081299,
+      "learning_rate": 0.0001,
+      "loss": 1.6508,
+      "step": 3857
+    },
+    {
+      "epoch": 0.9368625546381739,
+      "grad_norm": 0.6487970948219299,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 3858
+    },
+    {
+      "epoch": 0.9371053909664886,
+      "grad_norm": 0.5837119221687317,
+      "learning_rate": 0.0001,
+      "loss": 1.7526,
+      "step": 3859
+    },
+    {
+      "epoch": 0.9373482272948033,
+      "grad_norm": 0.6360679268836975,
+      "learning_rate": 0.0001,
+      "loss": 1.8961,
+      "step": 3860
+    },
+    {
+      "epoch": 0.9375910636231181,
+      "grad_norm": 0.6194024085998535,
+      "learning_rate": 0.0001,
+      "loss": 1.8107,
+      "step": 3861
+    },
+    {
+      "epoch": 0.9378338999514327,
+      "grad_norm": 0.5554211139678955,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 3862
+    },
+    {
+      "epoch": 0.9380767362797474,
+      "grad_norm": 0.6048312783241272,
+      "learning_rate": 0.0001,
+      "loss": 1.8423,
+      "step": 3863
+    },
+    {
+      "epoch": 0.9383195726080622,
+      "grad_norm": 0.5703085064888,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 3864
+    },
+    {
+      "epoch": 0.9385624089363769,
+      "grad_norm": 0.5596864819526672,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 3865
+    },
+    {
+      "epoch": 0.9388052452646916,
+      "grad_norm": 0.6383177042007446,
+      "learning_rate": 0.0001,
+      "loss": 1.8643,
+      "step": 3866
+    },
+    {
+      "epoch": 0.9390480815930063,
+      "grad_norm": 0.5903500914573669,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 3867
+    },
+    {
+      "epoch": 0.939290917921321,
+      "grad_norm": 0.5811389684677124,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 3868
+    },
+    {
+      "epoch": 0.9395337542496357,
+      "grad_norm": 0.5726868510246277,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 3869
+    },
+    {
+      "epoch": 0.9397765905779505,
+      "grad_norm": 0.5806669592857361,
+      "learning_rate": 0.0001,
+      "loss": 1.5,
+      "step": 3870
+    },
+    {
+      "epoch": 0.9400194269062652,
+      "grad_norm": 0.606339156627655,
+      "learning_rate": 0.0001,
+      "loss": 1.7749,
+      "step": 3871
+    },
+    {
+      "epoch": 0.9402622632345798,
+      "grad_norm": 0.5703273415565491,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 3872
+    },
+    {
+      "epoch": 0.9405050995628946,
+      "grad_norm": 0.5913130044937134,
+      "learning_rate": 0.0001,
+      "loss": 1.763,
+      "step": 3873
+    },
+    {
+      "epoch": 0.9407479358912093,
+      "grad_norm": 0.5414733290672302,
+      "learning_rate": 0.0001,
+      "loss": 1.522,
+      "step": 3874
+    },
+    {
+      "epoch": 0.940990772219524,
+      "grad_norm": 0.570084273815155,
+      "learning_rate": 0.0001,
+      "loss": 1.7923,
+      "step": 3875
+    },
+    {
+      "epoch": 0.9412336085478388,
+      "grad_norm": 0.591760516166687,
+      "learning_rate": 0.0001,
+      "loss": 1.6011,
+      "step": 3876
+    },
+    {
+      "epoch": 0.9414764448761535,
+      "grad_norm": 0.6212313771247864,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 3877
+    },
+    {
+      "epoch": 0.9417192812044682,
+      "grad_norm": 0.5751554369926453,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 3878
+    },
+    {
+      "epoch": 0.9419621175327829,
+      "grad_norm": 0.6063977479934692,
+      "learning_rate": 0.0001,
+      "loss": 1.7936,
+      "step": 3879
+    },
+    {
+      "epoch": 0.9422049538610976,
+      "grad_norm": 0.5817692279815674,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9424477901894124,
+      "grad_norm": 0.580960214138031,
+      "learning_rate": 0.0001,
+      "loss": 1.7584,
+      "step": 3881
+    },
+    {
+      "epoch": 0.9426906265177271,
+      "grad_norm": 0.5761325359344482,
+      "learning_rate": 0.0001,
+      "loss": 1.6608,
+      "step": 3882
+    },
+    {
+      "epoch": 0.9429334628460417,
+      "grad_norm": 0.6162409782409668,
+      "learning_rate": 0.0001,
+      "loss": 1.8851,
+      "step": 3883
+    },
+    {
+      "epoch": 0.9431762991743565,
+      "grad_norm": 0.5872670412063599,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 3884
+    },
+    {
+      "epoch": 0.9434191355026712,
+      "grad_norm": 0.5654221773147583,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 3885
+    },
+    {
+      "epoch": 0.9436619718309859,
+      "grad_norm": 0.6001012921333313,
+      "learning_rate": 0.0001,
+      "loss": 1.775,
+      "step": 3886
+    },
+    {
+      "epoch": 0.9439048081593007,
+      "grad_norm": 0.6080231070518494,
+      "learning_rate": 0.0001,
+      "loss": 1.8407,
+      "step": 3887
+    },
+    {
+      "epoch": 0.9441476444876153,
+      "grad_norm": 0.6112394332885742,
+      "learning_rate": 0.0001,
+      "loss": 1.8503,
+      "step": 3888
+    },
+    {
+      "epoch": 0.94439048081593,
+      "grad_norm": 0.5936259627342224,
+      "learning_rate": 0.0001,
+      "loss": 1.7469,
+      "step": 3889
+    },
+    {
+      "epoch": 0.9446333171442448,
+      "grad_norm": 0.6158880591392517,
+      "learning_rate": 0.0001,
+      "loss": 1.8293,
+      "step": 3890
+    },
+    {
+      "epoch": 0.9448761534725595,
+      "grad_norm": 0.5968067646026611,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 3891
+    },
+    {
+      "epoch": 0.9451189898008742,
+      "grad_norm": 0.62282395362854,
+      "learning_rate": 0.0001,
+      "loss": 1.8189,
+      "step": 3892
+    },
+    {
+      "epoch": 0.945361826129189,
+      "grad_norm": 0.6170820593833923,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 3893
+    },
+    {
+      "epoch": 0.9456046624575036,
+      "grad_norm": 0.6034669876098633,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 3894
+    },
+    {
+      "epoch": 0.9458474987858183,
+      "grad_norm": 0.6236358880996704,
+      "learning_rate": 0.0001,
+      "loss": 1.729,
+      "step": 3895
+    },
+    {
+      "epoch": 0.9460903351141331,
+      "grad_norm": 0.5863955616950989,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 3896
+    },
+    {
+      "epoch": 0.9463331714424478,
+      "grad_norm": 0.6099282503128052,
+      "learning_rate": 0.0001,
+      "loss": 1.8143,
+      "step": 3897
+    },
+    {
+      "epoch": 0.9465760077707625,
+      "grad_norm": 0.5919102430343628,
+      "learning_rate": 0.0001,
+      "loss": 1.778,
+      "step": 3898
+    },
+    {
+      "epoch": 0.9468188440990772,
+      "grad_norm": 0.5809880495071411,
+      "learning_rate": 0.0001,
+      "loss": 1.8069,
+      "step": 3899
+    },
+    {
+      "epoch": 0.9470616804273919,
+      "grad_norm": 0.6201245188713074,
+      "learning_rate": 0.0001,
+      "loss": 1.7953,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9473045167557067,
+      "grad_norm": 0.5886535048484802,
+      "learning_rate": 0.0001,
+      "loss": 1.7421,
+      "step": 3901
+    },
+    {
+      "epoch": 0.9475473530840214,
+      "grad_norm": 0.5675178170204163,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 3902
+    },
+    {
+      "epoch": 0.9477901894123361,
+      "grad_norm": 0.5949831008911133,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 3903
+    },
+    {
+      "epoch": 0.9480330257406508,
+      "grad_norm": 0.6057093739509583,
+      "learning_rate": 0.0001,
+      "loss": 1.7499,
+      "step": 3904
+    },
+    {
+      "epoch": 0.9482758620689655,
+      "grad_norm": 0.6029499769210815,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 3905
+    },
+    {
+      "epoch": 0.9485186983972802,
+      "grad_norm": 0.5812211632728577,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 3906
+    },
+    {
+      "epoch": 0.948761534725595,
+      "grad_norm": 0.6465932726860046,
+      "learning_rate": 0.0001,
+      "loss": 1.7672,
+      "step": 3907
+    },
+    {
+      "epoch": 0.9490043710539097,
+      "grad_norm": 0.5721454620361328,
+      "learning_rate": 0.0001,
+      "loss": 1.6448,
+      "step": 3908
+    },
+    {
+      "epoch": 0.9492472073822243,
+      "grad_norm": 0.6058249473571777,
+      "learning_rate": 0.0001,
+      "loss": 1.8698,
+      "step": 3909
+    },
+    {
+      "epoch": 0.9494900437105391,
+      "grad_norm": 0.6070011854171753,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9497328800388538,
+      "grad_norm": 0.610149621963501,
+      "learning_rate": 0.0001,
+      "loss": 1.7704,
+      "step": 3911
+    },
+    {
+      "epoch": 0.9499757163671685,
+      "grad_norm": 0.6089284420013428,
+      "learning_rate": 0.0001,
+      "loss": 1.7174,
+      "step": 3912
+    },
+    {
+      "epoch": 0.9502185526954833,
+      "grad_norm": 0.5909119248390198,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 3913
+    },
+    {
+      "epoch": 0.950461389023798,
+      "grad_norm": 0.62584388256073,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 3914
+    },
+    {
+      "epoch": 0.9507042253521126,
+      "grad_norm": 0.5750866532325745,
+      "learning_rate": 0.0001,
+      "loss": 1.5785,
+      "step": 3915
+    },
+    {
+      "epoch": 0.9509470616804274,
+      "grad_norm": 0.6291926503181458,
+      "learning_rate": 0.0001,
+      "loss": 1.7933,
+      "step": 3916
+    },
+    {
+      "epoch": 0.9511898980087421,
+      "grad_norm": 0.5660816431045532,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 3917
+    },
+    {
+      "epoch": 0.9514327343370568,
+      "grad_norm": 0.5772616267204285,
+      "learning_rate": 0.0001,
+      "loss": 1.7692,
+      "step": 3918
+    },
+    {
+      "epoch": 0.9516755706653716,
+      "grad_norm": 0.6061490774154663,
+      "learning_rate": 0.0001,
+      "loss": 1.8757,
+      "step": 3919
+    },
+    {
+      "epoch": 0.9519184069936862,
+      "grad_norm": 0.5998229384422302,
+      "learning_rate": 0.0001,
+      "loss": 1.7443,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9521612433220009,
+      "grad_norm": 0.5965450406074524,
+      "learning_rate": 0.0001,
+      "loss": 1.6834,
+      "step": 3921
+    },
+    {
+      "epoch": 0.9524040796503157,
+      "grad_norm": 0.5807543396949768,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 3922
+    },
+    {
+      "epoch": 0.9526469159786304,
+      "grad_norm": 0.5883108973503113,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 3923
+    },
+    {
+      "epoch": 0.9528897523069452,
+      "grad_norm": 0.6153419017791748,
+      "learning_rate": 0.0001,
+      "loss": 1.7335,
+      "step": 3924
+    },
+    {
+      "epoch": 0.9531325886352598,
+      "grad_norm": 0.6349037885665894,
+      "learning_rate": 0.0001,
+      "loss": 1.6976,
+      "step": 3925
+    },
+    {
+      "epoch": 0.9533754249635745,
+      "grad_norm": 0.6264722347259521,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 3926
+    },
+    {
+      "epoch": 0.9536182612918893,
+      "grad_norm": 0.5886244177818298,
+      "learning_rate": 0.0001,
+      "loss": 1.689,
+      "step": 3927
+    },
+    {
+      "epoch": 0.953861097620204,
+      "grad_norm": 0.589606761932373,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 3928
+    },
+    {
+      "epoch": 0.9541039339485187,
+      "grad_norm": 0.6218578219413757,
+      "learning_rate": 0.0001,
+      "loss": 1.7403,
+      "step": 3929
+    },
+    {
+      "epoch": 0.9543467702768335,
+      "grad_norm": 0.5539392828941345,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9545896066051481,
+      "grad_norm": 0.6072409749031067,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 3931
+    },
+    {
+      "epoch": 0.9548324429334628,
+      "grad_norm": 0.6281037330627441,
+      "learning_rate": 0.0001,
+      "loss": 1.7151,
+      "step": 3932
+    },
+    {
+      "epoch": 0.9550752792617776,
+      "grad_norm": 0.7160969376564026,
+      "learning_rate": 0.0001,
+      "loss": 1.7379,
+      "step": 3933
+    },
+    {
+      "epoch": 0.9553181155900923,
+      "grad_norm": 0.616183340549469,
+      "learning_rate": 0.0001,
+      "loss": 1.7787,
+      "step": 3934
+    },
+    {
+      "epoch": 0.955560951918407,
+      "grad_norm": 0.594199538230896,
+      "learning_rate": 0.0001,
+      "loss": 1.8091,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9558037882467217,
+      "grad_norm": 0.5698363184928894,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 3936
+    },
+    {
+      "epoch": 0.9560466245750364,
+      "grad_norm": 0.6054688096046448,
+      "learning_rate": 0.0001,
+      "loss": 1.5187,
+      "step": 3937
+    },
+    {
+      "epoch": 0.9562894609033511,
+      "grad_norm": 0.6152597069740295,
+      "learning_rate": 0.0001,
+      "loss": 1.6591,
+      "step": 3938
+    },
+    {
+      "epoch": 0.9565322972316659,
+      "grad_norm": 0.6362252235412598,
+      "learning_rate": 0.0001,
+      "loss": 1.808,
+      "step": 3939
+    },
+    {
+      "epoch": 0.9567751335599806,
+      "grad_norm": 0.6119574308395386,
+      "learning_rate": 0.0001,
+      "loss": 1.8636,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9570179698882952,
+      "grad_norm": 0.612122118473053,
+      "learning_rate": 0.0001,
+      "loss": 1.7372,
+      "step": 3941
+    },
+    {
+      "epoch": 0.95726080621661,
+      "grad_norm": 0.5694190263748169,
+      "learning_rate": 0.0001,
+      "loss": 1.5142,
+      "step": 3942
+    },
+    {
+      "epoch": 0.9575036425449247,
+      "grad_norm": 0.5553438663482666,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 3943
+    },
+    {
+      "epoch": 0.9577464788732394,
+      "grad_norm": 0.5808476805686951,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 3944
+    },
+    {
+      "epoch": 0.9579893152015542,
+      "grad_norm": 0.580889105796814,
+      "learning_rate": 0.0001,
+      "loss": 1.7251,
+      "step": 3945
+    },
+    {
+      "epoch": 0.9582321515298688,
+      "grad_norm": 0.6135175824165344,
+      "learning_rate": 0.0001,
+      "loss": 1.8378,
+      "step": 3946
+    },
+    {
+      "epoch": 0.9584749878581836,
+      "grad_norm": 0.5823888778686523,
+      "learning_rate": 0.0001,
+      "loss": 1.7151,
+      "step": 3947
+    },
+    {
+      "epoch": 0.9587178241864983,
+      "grad_norm": 0.6322687864303589,
+      "learning_rate": 0.0001,
+      "loss": 1.8039,
+      "step": 3948
+    },
+    {
+      "epoch": 0.958960660514813,
+      "grad_norm": 0.6384401321411133,
+      "learning_rate": 0.0001,
+      "loss": 1.7832,
+      "step": 3949
+    },
+    {
+      "epoch": 0.9592034968431278,
+      "grad_norm": 0.6263468861579895,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9594463331714425,
+      "grad_norm": 0.6187636852264404,
+      "learning_rate": 0.0001,
+      "loss": 1.9863,
+      "step": 3951
+    },
+    {
+      "epoch": 0.9596891694997571,
+      "grad_norm": 0.5676218271255493,
+      "learning_rate": 0.0001,
+      "loss": 1.6597,
+      "step": 3952
+    },
+    {
+      "epoch": 0.9599320058280719,
+      "grad_norm": 0.6038702130317688,
+      "learning_rate": 0.0001,
+      "loss": 1.6809,
+      "step": 3953
+    },
+    {
+      "epoch": 0.9601748421563866,
+      "grad_norm": 0.5852795839309692,
+      "learning_rate": 0.0001,
+      "loss": 1.8271,
+      "step": 3954
+    },
+    {
+      "epoch": 0.9604176784847013,
+      "grad_norm": 0.5519470572471619,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 3955
+    },
+    {
+      "epoch": 0.9606605148130161,
+      "grad_norm": 0.6034982800483704,
+      "learning_rate": 0.0001,
+      "loss": 1.6728,
+      "step": 3956
+    },
+    {
+      "epoch": 0.9609033511413307,
+      "grad_norm": 0.6074084043502808,
+      "learning_rate": 0.0001,
+      "loss": 1.8403,
+      "step": 3957
+    },
+    {
+      "epoch": 0.9611461874696454,
+      "grad_norm": 0.6192346811294556,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 3958
+    },
+    {
+      "epoch": 0.9613890237979602,
+      "grad_norm": 0.5913572907447815,
+      "learning_rate": 0.0001,
+      "loss": 1.6842,
+      "step": 3959
+    },
+    {
+      "epoch": 0.9616318601262749,
+      "grad_norm": 0.5921082496643066,
+      "learning_rate": 0.0001,
+      "loss": 1.7138,
+      "step": 3960
+    },
+    {
+      "epoch": 0.9618746964545896,
+      "grad_norm": 0.5722637176513672,
+      "learning_rate": 0.0001,
+      "loss": 1.6918,
+      "step": 3961
+    },
+    {
+      "epoch": 0.9621175327829043,
+      "grad_norm": 0.6428258419036865,
+      "learning_rate": 0.0001,
+      "loss": 1.8678,
+      "step": 3962
+    },
+    {
+      "epoch": 0.962360369111219,
+      "grad_norm": 0.5630905628204346,
+      "learning_rate": 0.0001,
+      "loss": 1.7039,
+      "step": 3963
+    },
+    {
+      "epoch": 0.9626032054395337,
+      "grad_norm": 0.6388512253761292,
+      "learning_rate": 0.0001,
+      "loss": 1.9527,
+      "step": 3964
+    },
+    {
+      "epoch": 0.9628460417678485,
+      "grad_norm": 0.6157130002975464,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9630888780961632,
+      "grad_norm": 0.5768729448318481,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 3966
+    },
+    {
+      "epoch": 0.9633317144244778,
+      "grad_norm": 0.5799568891525269,
+      "learning_rate": 0.0001,
+      "loss": 1.7477,
+      "step": 3967
+    },
+    {
+      "epoch": 0.9635745507527926,
+      "grad_norm": 0.5764046311378479,
+      "learning_rate": 0.0001,
+      "loss": 1.782,
+      "step": 3968
+    },
+    {
+      "epoch": 0.9638173870811073,
+      "grad_norm": 0.590692937374115,
+      "learning_rate": 0.0001,
+      "loss": 1.8864,
+      "step": 3969
+    },
+    {
+      "epoch": 0.9640602234094221,
+      "grad_norm": 0.6519653797149658,
+      "learning_rate": 0.0001,
+      "loss": 1.9161,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9643030597377368,
+      "grad_norm": 0.6048383712768555,
+      "learning_rate": 0.0001,
+      "loss": 1.7699,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9645458960660515,
+      "grad_norm": 0.6238450407981873,
+      "learning_rate": 0.0001,
+      "loss": 1.8133,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9647887323943662,
+      "grad_norm": 0.5832772850990295,
+      "learning_rate": 0.0001,
+      "loss": 1.7083,
+      "step": 3973
+    },
+    {
+      "epoch": 0.9650315687226809,
+      "grad_norm": 0.5764178037643433,
+      "learning_rate": 0.0001,
+      "loss": 1.7431,
+      "step": 3974
+    },
+    {
+      "epoch": 0.9652744050509956,
+      "grad_norm": 0.6305683255195618,
+      "learning_rate": 0.0001,
+      "loss": 1.7541,
+      "step": 3975
+    },
+    {
+      "epoch": 0.9655172413793104,
+      "grad_norm": 0.6307585835456848,
+      "learning_rate": 0.0001,
+      "loss": 2.0108,
+      "step": 3976
+    },
+    {
+      "epoch": 0.9657600777076251,
+      "grad_norm": 0.6229358315467834,
+      "learning_rate": 0.0001,
+      "loss": 1.8295,
+      "step": 3977
+    },
+    {
+      "epoch": 0.9660029140359397,
+      "grad_norm": 0.597227156162262,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 3978
+    },
+    {
+      "epoch": 0.9662457503642545,
+      "grad_norm": 0.5758910179138184,
+      "learning_rate": 0.0001,
+      "loss": 1.7248,
+      "step": 3979
+    },
+    {
+      "epoch": 0.9664885866925692,
+      "grad_norm": 0.5859194397926331,
+      "learning_rate": 0.0001,
+      "loss": 1.7211,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9667314230208839,
+      "grad_norm": 0.595125675201416,
+      "learning_rate": 0.0001,
+      "loss": 1.7135,
+      "step": 3981
+    },
+    {
+      "epoch": 0.9669742593491987,
+      "grad_norm": 0.5936609506607056,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 3982
+    },
+    {
+      "epoch": 0.9672170956775133,
+      "grad_norm": 0.6093140244483948,
+      "learning_rate": 0.0001,
+      "loss": 1.8882,
+      "step": 3983
+    },
+    {
+      "epoch": 0.967459932005828,
+      "grad_norm": 0.6792926788330078,
+      "learning_rate": 0.0001,
+      "loss": 1.8406,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9677027683341428,
+      "grad_norm": 0.604537308216095,
+      "learning_rate": 0.0001,
+      "loss": 1.7376,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9679456046624575,
+      "grad_norm": 0.6674985289573669,
+      "learning_rate": 0.0001,
+      "loss": 1.8353,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9681884409907722,
+      "grad_norm": 0.6092299222946167,
+      "learning_rate": 0.0001,
+      "loss": 1.7397,
+      "step": 3987
+    },
+    {
+      "epoch": 0.968431277319087,
+      "grad_norm": 0.6411168575286865,
+      "learning_rate": 0.0001,
+      "loss": 1.788,
+      "step": 3988
+    },
+    {
+      "epoch": 0.9686741136474016,
+      "grad_norm": 0.5817126035690308,
+      "learning_rate": 0.0001,
+      "loss": 1.7381,
+      "step": 3989
+    },
+    {
+      "epoch": 0.9689169499757163,
+      "grad_norm": 0.6701199412345886,
+      "learning_rate": 0.0001,
+      "loss": 1.8624,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9691597863040311,
+      "grad_norm": 0.5746298432350159,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 3991
+    },
+    {
+      "epoch": 0.9694026226323458,
+      "grad_norm": 0.6374258995056152,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 3992
+    },
+    {
+      "epoch": 0.9696454589606606,
+      "grad_norm": 0.5993824005126953,
+      "learning_rate": 0.0001,
+      "loss": 1.6389,
+      "step": 3993
+    },
+    {
+      "epoch": 0.9698882952889752,
+      "grad_norm": 0.5874238610267639,
+      "learning_rate": 0.0001,
+      "loss": 1.5926,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9701311316172899,
+      "grad_norm": 0.550464391708374,
+      "learning_rate": 0.0001,
+      "loss": 1.5598,
+      "step": 3995
+    },
+    {
+      "epoch": 0.9703739679456047,
+      "grad_norm": 0.6151759028434753,
+      "learning_rate": 0.0001,
+      "loss": 1.6834,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9706168042739194,
+      "grad_norm": 0.6637762188911438,
+      "learning_rate": 0.0001,
+      "loss": 1.8155,
+      "step": 3997
+    },
+    {
+      "epoch": 0.9708596406022341,
+      "grad_norm": 0.5628734230995178,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 3998
+    },
+    {
+      "epoch": 0.9711024769305489,
+      "grad_norm": 0.579086184501648,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9713453132588635,
+      "grad_norm": 0.6538147330284119,
+      "learning_rate": 0.0001,
+      "loss": 1.9069,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9715881495871782,
+      "grad_norm": 0.5787370800971985,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 4001
+    },
+    {
+      "epoch": 0.971830985915493,
+      "grad_norm": 0.6167012453079224,
+      "learning_rate": 0.0001,
+      "loss": 1.8567,
+      "step": 4002
+    },
+    {
+      "epoch": 0.9720738222438077,
+      "grad_norm": 0.6138943433761597,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9723166585721223,
+      "grad_norm": 0.577503502368927,
+      "learning_rate": 0.0001,
+      "loss": 1.7194,
+      "step": 4004
+    },
+    {
+      "epoch": 0.9725594949004371,
+      "grad_norm": 0.5734987854957581,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 4005
+    },
+    {
+      "epoch": 0.9728023312287518,
+      "grad_norm": 0.5983854532241821,
+      "learning_rate": 0.0001,
+      "loss": 1.768,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9730451675570665,
+      "grad_norm": 0.6402791738510132,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 4007
+    },
+    {
+      "epoch": 0.9732880038853813,
+      "grad_norm": 0.5972356200218201,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 4008
+    },
+    {
+      "epoch": 0.973530840213696,
+      "grad_norm": 0.6086263060569763,
+      "learning_rate": 0.0001,
+      "loss": 1.7982,
+      "step": 4009
+    },
+    {
+      "epoch": 0.9737736765420106,
+      "grad_norm": 0.5835751295089722,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 4010
+    },
+    {
+      "epoch": 0.9740165128703254,
+      "grad_norm": 0.596291184425354,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 4011
+    },
+    {
+      "epoch": 0.9742593491986401,
+      "grad_norm": 0.5902321934700012,
+      "learning_rate": 0.0001,
+      "loss": 1.8671,
+      "step": 4012
+    },
+    {
+      "epoch": 0.9745021855269548,
+      "grad_norm": 0.6232100129127502,
+      "learning_rate": 0.0001,
+      "loss": 1.7251,
+      "step": 4013
+    },
+    {
+      "epoch": 0.9747450218552696,
+      "grad_norm": 0.5748656988143921,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 4014
+    },
+    {
+      "epoch": 0.9749878581835842,
+      "grad_norm": 0.5785793662071228,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 4015
+    },
+    {
+      "epoch": 0.975230694511899,
+      "grad_norm": 0.6358336806297302,
+      "learning_rate": 0.0001,
+      "loss": 1.8765,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9754735308402137,
+      "grad_norm": 0.5772257447242737,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 4017
+    },
+    {
+      "epoch": 0.9757163671685284,
+      "grad_norm": 0.6292567253112793,
+      "learning_rate": 0.0001,
+      "loss": 1.7887,
+      "step": 4018
+    },
+    {
+      "epoch": 0.9759592034968432,
+      "grad_norm": 0.6002832055091858,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 4019
+    },
+    {
+      "epoch": 0.9762020398251579,
+      "grad_norm": 0.6181592345237732,
+      "learning_rate": 0.0001,
+      "loss": 1.6487,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9764448761534725,
+      "grad_norm": 0.581417441368103,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 4021
+    },
+    {
+      "epoch": 0.9766877124817873,
+      "grad_norm": 0.5849958062171936,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 4022
+    },
+    {
+      "epoch": 0.976930548810102,
+      "grad_norm": 0.6539707779884338,
+      "learning_rate": 0.0001,
+      "loss": 1.8053,
+      "step": 4023
+    },
+    {
+      "epoch": 0.9771733851384167,
+      "grad_norm": 0.6116283535957336,
+      "learning_rate": 0.0001,
+      "loss": 1.7819,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9774162214667315,
+      "grad_norm": 0.5561971664428711,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 4025
+    },
+    {
+      "epoch": 0.9776590577950461,
+      "grad_norm": 0.6015164256095886,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 4026
+    },
+    {
+      "epoch": 0.9779018941233608,
+      "grad_norm": 0.5934497117996216,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 4027
+    },
+    {
+      "epoch": 0.9781447304516756,
+      "grad_norm": 0.5897216796875,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 4028
+    },
+    {
+      "epoch": 0.9783875667799903,
+      "grad_norm": 0.5853744745254517,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 4029
+    },
+    {
+      "epoch": 0.978630403108305,
+      "grad_norm": 0.5475685000419617,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9788732394366197,
+      "grad_norm": 0.5999184250831604,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 4031
+    },
+    {
+      "epoch": 0.9791160757649344,
+      "grad_norm": 0.6003595590591431,
+      "learning_rate": 0.0001,
+      "loss": 1.7322,
+      "step": 4032
+    },
+    {
+      "epoch": 0.9793589120932491,
+      "grad_norm": 0.5681313276290894,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 4033
+    },
+    {
+      "epoch": 0.9796017484215639,
+      "grad_norm": 0.603021502494812,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9798445847498786,
+      "grad_norm": 0.5827537178993225,
+      "learning_rate": 0.0001,
+      "loss": 1.6952,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9800874210781932,
+      "grad_norm": 0.6328116655349731,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 4036
+    },
+    {
+      "epoch": 0.980330257406508,
+      "grad_norm": 0.5913118720054626,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 4037
+    },
+    {
+      "epoch": 0.9805730937348227,
+      "grad_norm": 0.6244193911552429,
+      "learning_rate": 0.0001,
+      "loss": 1.8205,
+      "step": 4038
+    },
+    {
+      "epoch": 0.9808159300631375,
+      "grad_norm": 0.6432669162750244,
+      "learning_rate": 0.0001,
+      "loss": 1.9614,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9810587663914522,
+      "grad_norm": 0.6130082607269287,
+      "learning_rate": 0.0001,
+      "loss": 1.7674,
+      "step": 4040
+    },
+    {
+      "epoch": 0.9813016027197669,
+      "grad_norm": 0.6276662349700928,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9815444390480816,
+      "grad_norm": 0.5937540531158447,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9817872753763963,
+      "grad_norm": 0.5934786796569824,
+      "learning_rate": 0.0001,
+      "loss": 1.8394,
+      "step": 4043
+    },
+    {
+      "epoch": 0.982030111704711,
+      "grad_norm": 0.6066287159919739,
+      "learning_rate": 0.0001,
+      "loss": 1.7155,
+      "step": 4044
+    },
+    {
+      "epoch": 0.9822729480330258,
+      "grad_norm": 0.6309837698936462,
+      "learning_rate": 0.0001,
+      "loss": 1.8514,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9825157843613405,
+      "grad_norm": 0.5881063938140869,
+      "learning_rate": 0.0001,
+      "loss": 1.7806,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9827586206896551,
+      "grad_norm": 0.6197108626365662,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 4047
+    },
+    {
+      "epoch": 0.9830014570179699,
+      "grad_norm": 0.580711305141449,
+      "learning_rate": 0.0001,
+      "loss": 1.7516,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9832442933462846,
+      "grad_norm": 0.5490559935569763,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 4049
+    },
+    {
+      "epoch": 0.9834871296745993,
+      "grad_norm": 0.5962522625923157,
+      "learning_rate": 0.0001,
+      "loss": 1.7585,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9837299660029141,
+      "grad_norm": 0.6085889339447021,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 4051
+    },
+    {
+      "epoch": 0.9839728023312287,
+      "grad_norm": 0.589601993560791,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9842156386595434,
+      "grad_norm": 0.5880647897720337,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9844584749878582,
+      "grad_norm": 0.6051363945007324,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 4054
+    },
+    {
+      "epoch": 0.9847013113161729,
+      "grad_norm": 0.6068199276924133,
+      "learning_rate": 0.0001,
+      "loss": 1.8676,
+      "step": 4055
+    },
+    {
+      "epoch": 0.9849441476444876,
+      "grad_norm": 0.5522249937057495,
+      "learning_rate": 0.0001,
+      "loss": 1.6681,
+      "step": 4056
+    },
+    {
+      "epoch": 0.9851869839728024,
+      "grad_norm": 0.5948834419250488,
+      "learning_rate": 0.0001,
+      "loss": 1.8587,
+      "step": 4057
+    },
+    {
+      "epoch": 0.985429820301117,
+      "grad_norm": 0.5941211581230164,
+      "learning_rate": 0.0001,
+      "loss": 1.7026,
+      "step": 4058
+    },
+    {
+      "epoch": 0.9856726566294317,
+      "grad_norm": 0.6230162382125854,
+      "learning_rate": 0.0001,
+      "loss": 1.7857,
+      "step": 4059
+    },
+    {
+      "epoch": 0.9859154929577465,
+      "grad_norm": 0.6194135546684265,
+      "learning_rate": 0.0001,
+      "loss": 1.691,
+      "step": 4060
+    },
+    {
+      "epoch": 0.9861583292860612,
+      "grad_norm": 0.6586176156997681,
+      "learning_rate": 0.0001,
+      "loss": 1.8014,
+      "step": 4061
+    },
+    {
+      "epoch": 0.986401165614376,
+      "grad_norm": 0.5540204644203186,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 4062
+    },
+    {
+      "epoch": 0.9866440019426906,
+      "grad_norm": 0.5742043256759644,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 4063
+    },
+    {
+      "epoch": 0.9868868382710053,
+      "grad_norm": 0.5965940356254578,
+      "learning_rate": 0.0001,
+      "loss": 1.7155,
+      "step": 4064
+    },
+    {
+      "epoch": 0.9871296745993201,
+      "grad_norm": 0.62481290102005,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9873725109276348,
+      "grad_norm": 0.5996125340461731,
+      "learning_rate": 0.0001,
+      "loss": 1.82,
+      "step": 4066
+    },
+    {
+      "epoch": 0.9876153472559495,
+      "grad_norm": 0.5856594443321228,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 4067
+    },
+    {
+      "epoch": 0.9878581835842642,
+      "grad_norm": 0.6302304863929749,
+      "learning_rate": 0.0001,
+      "loss": 1.7396,
+      "step": 4068
+    },
+    {
+      "epoch": 0.9881010199125789,
+      "grad_norm": 0.5669844150543213,
+      "learning_rate": 0.0001,
+      "loss": 1.462,
+      "step": 4069
+    },
+    {
+      "epoch": 0.9883438562408936,
+      "grad_norm": 0.6028769612312317,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9885866925692084,
+      "grad_norm": 0.6249973177909851,
+      "learning_rate": 0.0001,
+      "loss": 1.7147,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9888295288975231,
+      "grad_norm": 0.5884829163551331,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 4072
+    },
+    {
+      "epoch": 0.9890723652258377,
+      "grad_norm": 0.5409967303276062,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 4073
+    },
+    {
+      "epoch": 0.9893152015541525,
+      "grad_norm": 0.7405964136123657,
+      "learning_rate": 0.0001,
+      "loss": 1.9007,
+      "step": 4074
+    },
+    {
+      "epoch": 0.9895580378824672,
+      "grad_norm": 0.581794798374176,
+      "learning_rate": 0.0001,
+      "loss": 1.711,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9898008742107819,
+      "grad_norm": 0.5944003462791443,
+      "learning_rate": 0.0001,
+      "loss": 1.6429,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9900437105390967,
+      "grad_norm": 0.6132760047912598,
+      "learning_rate": 0.0001,
+      "loss": 1.7843,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9902865468674114,
+      "grad_norm": 0.5919248461723328,
+      "learning_rate": 0.0001,
+      "loss": 1.7671,
+      "step": 4078
+    },
+    {
+      "epoch": 0.990529383195726,
+      "grad_norm": 0.5920957922935486,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 4079
+    },
+    {
+      "epoch": 0.9907722195240408,
+      "grad_norm": 0.6020213961601257,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9910150558523555,
+      "grad_norm": 0.6029257774353027,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 4081
+    },
+    {
+      "epoch": 0.9912578921806702,
+      "grad_norm": 0.5782517194747925,
+      "learning_rate": 0.0001,
+      "loss": 1.6464,
+      "step": 4082
+    },
+    {
+      "epoch": 0.991500728508985,
+      "grad_norm": 0.5874188542366028,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9917435648372996,
+      "grad_norm": 0.6039769649505615,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9919864011656144,
+      "grad_norm": 0.6164844036102295,
+      "learning_rate": 0.0001,
+      "loss": 1.8007,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9922292374939291,
+      "grad_norm": 0.6580904126167297,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9924720738222438,
+      "grad_norm": 0.6556647419929504,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9927149101505586,
+      "grad_norm": 0.5966843962669373,
+      "learning_rate": 0.0001,
+      "loss": 1.817,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9929577464788732,
+      "grad_norm": 0.6216123104095459,
+      "learning_rate": 0.0001,
+      "loss": 1.8487,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9932005828071879,
+      "grad_norm": 0.5934640169143677,
+      "learning_rate": 0.0001,
+      "loss": 1.7539,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9934434191355027,
+      "grad_norm": 0.6474573612213135,
+      "learning_rate": 0.0001,
+      "loss": 1.7928,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9936862554638174,
+      "grad_norm": 0.605293333530426,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9939290917921321,
+      "grad_norm": 0.6208190321922302,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 4093
+    },
+    {
+      "epoch": 0.9941719281204469,
+      "grad_norm": 0.6126958131790161,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9944147644487615,
+      "grad_norm": 0.6111500859260559,
+      "learning_rate": 0.0001,
+      "loss": 1.802,
+      "step": 4095
+    },
+    {
+      "epoch": 0.9946576007770762,
+      "grad_norm": 0.6183574795722961,
+      "learning_rate": 0.0001,
+      "loss": 1.7097,
+      "step": 4096
+    },
+    {
+      "epoch": 0.994900437105391,
+      "grad_norm": 0.5912253260612488,
+      "learning_rate": 0.0001,
+      "loss": 1.7948,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9951432734337057,
+      "grad_norm": 0.5707657933235168,
+      "learning_rate": 0.0001,
+      "loss": 1.8227,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9953861097620204,
+      "grad_norm": 0.6282753944396973,
+      "learning_rate": 0.0001,
+      "loss": 1.6504,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9956289460903351,
+      "grad_norm": 0.5808796286582947,
+      "learning_rate": 0.0001,
+      "loss": 1.8258,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9958717824186498,
+      "grad_norm": 0.6238116025924683,
+      "learning_rate": 0.0001,
+      "loss": 1.8399,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9961146187469645,
+      "grad_norm": 0.6109184622764587,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 4102
+    },
+    {
+      "epoch": 0.9963574550752793,
+      "grad_norm": 0.5754851698875427,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 4103
+    },
+    {
+      "epoch": 0.996600291403594,
+      "grad_norm": 0.600270688533783,
+      "learning_rate": 0.0001,
+      "loss": 1.7249,
+      "step": 4104
+    },
+    {
+      "epoch": 0.9968431277319086,
+      "grad_norm": 0.5850363969802856,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9970859640602234,
+      "grad_norm": 0.5854186415672302,
+      "learning_rate": 0.0001,
+      "loss": 1.6658,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9973288003885381,
+      "grad_norm": 0.5924867391586304,
+      "learning_rate": 0.0001,
+      "loss": 1.678,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9975716367168529,
+      "grad_norm": 0.6060158610343933,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9978144730451676,
+      "grad_norm": 0.5951091051101685,
+      "learning_rate": 0.0001,
+      "loss": 1.8025,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9980573093734822,
+      "grad_norm": 0.6506190896034241,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 4110
+    },
+    {
+      "epoch": 0.998300145701797,
+      "grad_norm": 0.6137253046035767,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 4111
+    },
+    {
+      "epoch": 0.9985429820301117,
+      "grad_norm": 0.5973034501075745,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 4112
+    },
+    {
+      "epoch": 0.9987858183584264,
+      "grad_norm": 0.6223469972610474,
+      "learning_rate": 0.0001,
+      "loss": 1.8826,
+      "step": 4113
+    },
+    {
+      "epoch": 0.9990286546867412,
+      "grad_norm": 0.6080133318901062,
+      "learning_rate": 0.0001,
+      "loss": 1.7747,
+      "step": 4114
+    },
+    {
+      "epoch": 0.9992714910150559,
+      "grad_norm": 0.7053512930870056,
+      "learning_rate": 0.0001,
+      "loss": 1.9432,
+      "step": 4115
+    },
+    {
+      "epoch": 0.9995143273433705,
+      "grad_norm": 0.6507899761199951,
+      "learning_rate": 0.0001,
+      "loss": 1.7915,
+      "step": 4116
+    },
+    {
+      "epoch": 0.9997571636716853,
+      "grad_norm": 0.6092012524604797,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 4117
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.592926561832428,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 4118
+    },
+    {
+      "epoch": 1.0,
+      "step": 4118,
+      "total_flos": 6.212058917275435e+18,
+      "train_loss": 1.7427738688708856,
+      "train_runtime": 80585.133,
+      "train_samples_per_second": 0.204,
+      "train_steps_per_second": 0.051
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4118,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.212058917275435e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}