diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12320 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 9.99786248664054,
+  "eval_steps": 50000,
+  "global_step": 17540,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0057000356252226575,
+      "grad_norm": 0.8329347851096954,
+      "learning_rate": 1.1402508551881415e-07,
+      "loss": 1.6812,
+      "step": 10
+    },
+    {
+      "epoch": 0.011400071250445315,
+      "grad_norm": 0.8173922269493806,
+      "learning_rate": 2.280501710376283e-07,
+      "loss": 1.6858,
+      "step": 20
+    },
+    {
+      "epoch": 0.017100106875667972,
+      "grad_norm": 0.809557505770668,
+      "learning_rate": 3.4207525655644247e-07,
+      "loss": 1.6882,
+      "step": 30
+    },
+    {
+      "epoch": 0.02280014250089063,
+      "grad_norm": 0.8748268159634193,
+      "learning_rate": 4.561003420752566e-07,
+      "loss": 1.684,
+      "step": 40
+    },
+    {
+      "epoch": 0.028500178126113287,
+      "grad_norm": 0.8347602557149199,
+      "learning_rate": 5.701254275940708e-07,
+      "loss": 1.6798,
+      "step": 50
+    },
+    {
+      "epoch": 0.034200213751335945,
+      "grad_norm": 0.8430580040769771,
+      "learning_rate": 6.841505131128849e-07,
+      "loss": 1.6845,
+      "step": 60
+    },
+    {
+      "epoch": 0.0399002493765586,
+      "grad_norm": 0.8808469992546447,
+      "learning_rate": 7.98175598631699e-07,
+      "loss": 1.6805,
+      "step": 70
+    },
+    {
+      "epoch": 0.04560028500178126,
+      "grad_norm": 0.9332195994860901,
+      "learning_rate": 9.122006841505132e-07,
+      "loss": 1.6737,
+      "step": 80
+    },
+    {
+      "epoch": 0.05130032062700392,
+      "grad_norm": 0.9292762787168932,
+      "learning_rate": 1.0262257696693273e-06,
+      "loss": 1.6761,
+      "step": 90
+    },
+    {
+      "epoch": 0.057000356252226575,
+      "grad_norm": 0.8959401505265313,
+      "learning_rate": 1.1402508551881415e-06,
+      "loss": 1.672,
+      "step": 100
+    },
+    {
+      "epoch": 0.06270039187744923,
+      "grad_norm": 0.9753219116814298,
+      "learning_rate": 1.2542759407069557e-06,
+      "loss": 1.6788,
+      "step": 110
+    },
+    {
+      "epoch": 0.06840042750267189,
+      "grad_norm": 1.0355003267680642,
+      "learning_rate": 1.3683010262257699e-06,
+      "loss": 1.6521,
+      "step": 120
+    },
+    {
+      "epoch": 0.07410046312789455,
+      "grad_norm": 0.9854284833432406,
+      "learning_rate": 1.4823261117445838e-06,
+      "loss": 1.6542,
+      "step": 130
+    },
+    {
+      "epoch": 0.0798004987531172,
+      "grad_norm": 0.8995458835476428,
+      "learning_rate": 1.596351197263398e-06,
+      "loss": 1.6266,
+      "step": 140
+    },
+    {
+      "epoch": 0.08550053437833986,
+      "grad_norm": 0.8937279797895432,
+      "learning_rate": 1.7103762827822124e-06,
+      "loss": 1.6018,
+      "step": 150
+    },
+    {
+      "epoch": 0.09120057000356252,
+      "grad_norm": 0.8508463445168406,
+      "learning_rate": 1.8244013683010263e-06,
+      "loss": 1.5702,
+      "step": 160
+    },
+    {
+      "epoch": 0.09690060562878518,
+      "grad_norm": 0.826282613868742,
+      "learning_rate": 1.9384264538198407e-06,
+      "loss": 1.568,
+      "step": 170
+    },
+    {
+      "epoch": 0.10260064125400783,
+      "grad_norm": 0.7919479763516626,
+      "learning_rate": 2.0524515393386547e-06,
+      "loss": 1.5417,
+      "step": 180
+    },
+    {
+      "epoch": 0.10830067687923049,
+      "grad_norm": 0.684561861348144,
+      "learning_rate": 2.166476624857469e-06,
+      "loss": 1.5114,
+      "step": 190
+    },
+    {
+      "epoch": 0.11400071250445315,
+      "grad_norm": 0.680205774626441,
+      "learning_rate": 2.280501710376283e-06,
+      "loss": 1.4744,
+      "step": 200
+    },
+    {
+      "epoch": 0.11970074812967581,
+      "grad_norm": 0.7006502228163867,
+      "learning_rate": 2.394526795895097e-06,
+      "loss": 1.4864,
+      "step": 210
+    },
+    {
+      "epoch": 0.12540078375489846,
+      "grad_norm": 0.6771625316054201,
+      "learning_rate": 2.5085518814139114e-06,
+      "loss": 1.4508,
+      "step": 220
+    },
+    {
+      "epoch": 0.13110081938012114,
+      "grad_norm": 0.6383811243685706,
+      "learning_rate": 2.6225769669327258e-06,
+      "loss": 1.4298,
+      "step": 230
+    },
+    {
+      "epoch": 0.13680085500534378,
+      "grad_norm": 0.6383478063230229,
+      "learning_rate": 2.7366020524515397e-06,
+      "loss": 1.4201,
+      "step": 240
+    },
+    {
+      "epoch": 0.14250089063056645,
+      "grad_norm": 0.664438350643699,
+      "learning_rate": 2.8506271379703537e-06,
+      "loss": 1.4037,
+      "step": 250
+    },
+    {
+      "epoch": 0.1482009262557891,
+      "grad_norm": 0.6120680534196937,
+      "learning_rate": 2.9646522234891676e-06,
+      "loss": 1.3977,
+      "step": 260
+    },
+    {
+      "epoch": 0.15390096188101177,
+      "grad_norm": 0.6152217588311774,
+      "learning_rate": 3.078677309007982e-06,
+      "loss": 1.3728,
+      "step": 270
+    },
+    {
+      "epoch": 0.1596009975062344,
+      "grad_norm": 0.6131574126061529,
+      "learning_rate": 3.192702394526796e-06,
+      "loss": 1.3524,
+      "step": 280
+    },
+    {
+      "epoch": 0.16530103313145708,
+      "grad_norm": 0.626799181295027,
+      "learning_rate": 3.30672748004561e-06,
+      "loss": 1.3344,
+      "step": 290
+    },
+    {
+      "epoch": 0.17100106875667972,
+      "grad_norm": 0.5974951223625137,
+      "learning_rate": 3.4207525655644248e-06,
+      "loss": 1.3409,
+      "step": 300
+    },
+    {
+      "epoch": 0.1767011043819024,
+      "grad_norm": 0.6112769459431191,
+      "learning_rate": 3.5347776510832387e-06,
+      "loss": 1.3241,
+      "step": 310
+    },
+    {
+      "epoch": 0.18240114000712504,
+      "grad_norm": 0.6229753422183877,
+      "learning_rate": 3.6488027366020527e-06,
+      "loss": 1.3017,
+      "step": 320
+    },
+    {
+      "epoch": 0.1881011756323477,
+      "grad_norm": 0.6105785881673802,
+      "learning_rate": 3.762827822120867e-06,
+      "loss": 1.3024,
+      "step": 330
+    },
+    {
+      "epoch": 0.19380121125757035,
+      "grad_norm": 0.6244573917539824,
+      "learning_rate": 3.8768529076396815e-06,
+      "loss": 1.2862,
+      "step": 340
+    },
+    {
+      "epoch": 0.19950124688279303,
+      "grad_norm": 0.6380126761223662,
+      "learning_rate": 3.990877993158495e-06,
+      "loss": 1.2763,
+      "step": 350
+    },
+    {
+      "epoch": 0.20520128250801567,
+      "grad_norm": 0.6234244975836795,
+      "learning_rate": 4.104903078677309e-06,
+      "loss": 1.2578,
+      "step": 360
+    },
+    {
+      "epoch": 0.21090131813323834,
+      "grad_norm": 0.640686344376765,
+      "learning_rate": 4.218928164196123e-06,
+      "loss": 1.264,
+      "step": 370
+    },
+    {
+      "epoch": 0.21660135375846098,
+      "grad_norm": 0.6664146960033351,
+      "learning_rate": 4.332953249714938e-06,
+      "loss": 1.2566,
+      "step": 380
+    },
+    {
+      "epoch": 0.22230138938368366,
+      "grad_norm": 0.6791004999182745,
+      "learning_rate": 4.446978335233752e-06,
+      "loss": 1.2506,
+      "step": 390
+    },
+    {
+      "epoch": 0.2280014250089063,
+      "grad_norm": 0.660782823423085,
+      "learning_rate": 4.561003420752566e-06,
+      "loss": 1.2341,
+      "step": 400
+    },
+    {
+      "epoch": 0.23370146063412897,
+      "grad_norm": 0.6496810448682551,
+      "learning_rate": 4.67502850627138e-06,
+      "loss": 1.2303,
+      "step": 410
+    },
+    {
+      "epoch": 0.23940149625935161,
+      "grad_norm": 0.6760464986579647,
+      "learning_rate": 4.789053591790194e-06,
+      "loss": 1.2157,
+      "step": 420
+    },
+    {
+      "epoch": 0.2451015318845743,
+      "grad_norm": 0.7099546157920937,
+      "learning_rate": 4.903078677309008e-06,
+      "loss": 1.2119,
+      "step": 430
+    },
+    {
+      "epoch": 0.25080156750979693,
+      "grad_norm": 0.6985263486782688,
+      "learning_rate": 5.017103762827823e-06,
+      "loss": 1.1998,
+      "step": 440
+    },
+    {
+      "epoch": 0.2565016031350196,
+      "grad_norm": 0.7086874492220844,
+      "learning_rate": 5.131128848346637e-06,
+      "loss": 1.1845,
+      "step": 450
+    },
+    {
+      "epoch": 0.2622016387602423,
+      "grad_norm": 0.6952132915582485,
+      "learning_rate": 5.2451539338654515e-06,
+      "loss": 1.2045,
+      "step": 460
+    },
+    {
+      "epoch": 0.2679016743854649,
+      "grad_norm": 0.7563820343660739,
+      "learning_rate": 5.3591790193842655e-06,
+      "loss": 1.1816,
+      "step": 470
+    },
+    {
+      "epoch": 0.27360171001068756,
+      "grad_norm": 0.7383768041736016,
+      "learning_rate": 5.4732041049030794e-06,
+      "loss": 1.1867,
+      "step": 480
+    },
+    {
+      "epoch": 0.2793017456359102,
+      "grad_norm": 0.7377951397065144,
+      "learning_rate": 5.587229190421893e-06,
+      "loss": 1.1693,
+      "step": 490
+    },
+    {
+      "epoch": 0.2850017812611329,
+      "grad_norm": 0.7930492156928618,
+      "learning_rate": 5.701254275940707e-06,
+      "loss": 1.1769,
+      "step": 500
+    },
+    {
+      "epoch": 0.29070181688635555,
+      "grad_norm": 0.7804629910719232,
+      "learning_rate": 5.815279361459521e-06,
+      "loss": 1.1572,
+      "step": 510
+    },
+    {
+      "epoch": 0.2964018525115782,
+      "grad_norm": 0.7476754916217279,
+      "learning_rate": 5.929304446978335e-06,
+      "loss": 1.1559,
+      "step": 520
+    },
+    {
+      "epoch": 0.30210188813680083,
+      "grad_norm": 0.7928284277955501,
+      "learning_rate": 6.04332953249715e-06,
+      "loss": 1.1402,
+      "step": 530
+    },
+    {
+      "epoch": 0.30780192376202353,
+      "grad_norm": 0.8039246648014535,
+      "learning_rate": 6.157354618015964e-06,
+      "loss": 1.1456,
+      "step": 540
+    },
+    {
+      "epoch": 0.3135019593872462,
+      "grad_norm": 0.7914550658951943,
+      "learning_rate": 6.271379703534778e-06,
+      "loss": 1.138,
+      "step": 550
+    },
+    {
+      "epoch": 0.3192019950124688,
+      "grad_norm": 0.8477034908005336,
+      "learning_rate": 6.385404789053592e-06,
+      "loss": 1.1398,
+      "step": 560
+    },
+    {
+      "epoch": 0.32490203063769146,
+      "grad_norm": 0.8870257129031497,
+      "learning_rate": 6.499429874572406e-06,
+      "loss": 1.121,
+      "step": 570
+    },
+    {
+      "epoch": 0.33060206626291416,
+      "grad_norm": 0.8865720994018285,
+      "learning_rate": 6.61345496009122e-06,
+      "loss": 1.1116,
+      "step": 580
+    },
+    {
+      "epoch": 0.3363021018881368,
+      "grad_norm": 0.894221994899268,
+      "learning_rate": 6.727480045610034e-06,
+      "loss": 1.1134,
+      "step": 590
+    },
+    {
+      "epoch": 0.34200213751335945,
+      "grad_norm": 0.9002142031410411,
+      "learning_rate": 6.8415051311288495e-06,
+      "loss": 1.0982,
+      "step": 600
+    },
+    {
+      "epoch": 0.3477021731385821,
+      "grad_norm": 0.9405603084160153,
+      "learning_rate": 6.9555302166476635e-06,
+      "loss": 1.1104,
+      "step": 610
+    },
+    {
+      "epoch": 0.3534022087638048,
+      "grad_norm": 0.8978621968763663,
+      "learning_rate": 7.0695553021664774e-06,
+      "loss": 1.0935,
+      "step": 620
+    },
+    {
+      "epoch": 0.35910224438902744,
+      "grad_norm": 0.923135430228548,
+      "learning_rate": 7.183580387685291e-06,
+      "loss": 1.0747,
+      "step": 630
+    },
+    {
+      "epoch": 0.3648022800142501,
+      "grad_norm": 0.9637043977541311,
+      "learning_rate": 7.297605473204105e-06,
+      "loss": 1.0766,
+      "step": 640
+    },
+    {
+      "epoch": 0.3705023156394727,
+      "grad_norm": 0.9634742025386968,
+      "learning_rate": 7.41163055872292e-06,
+      "loss": 1.0817,
+      "step": 650
+    },
+    {
+      "epoch": 0.3762023512646954,
+      "grad_norm": 1.0138264576502096,
+      "learning_rate": 7.525655644241734e-06,
+      "loss": 1.0708,
+      "step": 660
+    },
+    {
+      "epoch": 0.38190238688991807,
+      "grad_norm": 0.9754277408302766,
+      "learning_rate": 7.639680729760547e-06,
+      "loss": 1.0662,
+      "step": 670
+    },
+    {
+      "epoch": 0.3876024225151407,
+      "grad_norm": 0.9997112312036316,
+      "learning_rate": 7.753705815279363e-06,
+      "loss": 1.0743,
+      "step": 680
+    },
+    {
+      "epoch": 0.39330245814036335,
+      "grad_norm": 1.0137614677943894,
+      "learning_rate": 7.867730900798177e-06,
+      "loss": 1.055,
+      "step": 690
+    },
+    {
+      "epoch": 0.39900249376558605,
+      "grad_norm": 1.0243398381093383,
+      "learning_rate": 7.98175598631699e-06,
+      "loss": 1.0547,
+      "step": 700
+    },
+    {
+      "epoch": 0.4047025293908087,
+      "grad_norm": 1.077627151044799,
+      "learning_rate": 8.095781071835805e-06,
+      "loss": 1.0497,
+      "step": 710
+    },
+    {
+      "epoch": 0.41040256501603134,
+      "grad_norm": 1.0659844500266695,
+      "learning_rate": 8.209806157354619e-06,
+      "loss": 1.0603,
+      "step": 720
+    },
+    {
+      "epoch": 0.416102600641254,
+      "grad_norm": 1.0891087944465083,
+      "learning_rate": 8.323831242873433e-06,
+      "loss": 1.0428,
+      "step": 730
+    },
+    {
+      "epoch": 0.4218026362664767,
+      "grad_norm": 1.155107592465822,
+      "learning_rate": 8.437856328392247e-06,
+      "loss": 1.0424,
+      "step": 740
+    },
+    {
+      "epoch": 0.4275026718916993,
+      "grad_norm": 1.0357273266673794,
+      "learning_rate": 8.55188141391106e-06,
+      "loss": 1.0419,
+      "step": 750
+    },
+    {
+      "epoch": 0.43320270751692197,
+      "grad_norm": 1.1099226858511917,
+      "learning_rate": 8.665906499429876e-06,
+      "loss": 1.0368,
+      "step": 760
+    },
+    {
+      "epoch": 0.4389027431421446,
+      "grad_norm": 1.1550259480727676,
+      "learning_rate": 8.77993158494869e-06,
+      "loss": 1.0324,
+      "step": 770
+    },
+    {
+      "epoch": 0.4446027787673673,
+      "grad_norm": 1.1766907357775453,
+      "learning_rate": 8.893956670467504e-06,
+      "loss": 1.0344,
+      "step": 780
+    },
+    {
+      "epoch": 0.45030281439258996,
+      "grad_norm": 1.086671978162032,
+      "learning_rate": 9.007981755986318e-06,
+      "loss": 1.0273,
+      "step": 790
+    },
+    {
+      "epoch": 0.4560028500178126,
+      "grad_norm": 1.186181822112197,
+      "learning_rate": 9.122006841505132e-06,
+      "loss": 1.0247,
+      "step": 800
+    },
+    {
+      "epoch": 0.46170288564303524,
+      "grad_norm": 1.150017830725986,
+      "learning_rate": 9.236031927023946e-06,
+      "loss": 1.0234,
+      "step": 810
+    },
+    {
+      "epoch": 0.46740292126825794,
+      "grad_norm": 1.169372753028457,
+      "learning_rate": 9.35005701254276e-06,
+      "loss": 1.0206,
+      "step": 820
+    },
+    {
+      "epoch": 0.4731029568934806,
+      "grad_norm": 1.1106663677739594,
+      "learning_rate": 9.464082098061574e-06,
+      "loss": 1.0159,
+      "step": 830
+    },
+    {
+      "epoch": 0.47880299251870323,
+      "grad_norm": 1.2049682047754315,
+      "learning_rate": 9.578107183580388e-06,
+      "loss": 1.0073,
+      "step": 840
+    },
+    {
+      "epoch": 0.4845030281439259,
+      "grad_norm": 1.2123026532151817,
+      "learning_rate": 9.692132269099202e-06,
+      "loss": 1.0049,
+      "step": 850
+    },
+    {
+      "epoch": 0.4902030637691486,
+      "grad_norm": 1.2675857450988537,
+      "learning_rate": 9.806157354618016e-06,
+      "loss": 1.0008,
+      "step": 860
+    },
+    {
+      "epoch": 0.4959030993943712,
+      "grad_norm": 1.2939190886660976,
+      "learning_rate": 9.920182440136832e-06,
+      "loss": 0.997,
+      "step": 870
+    },
+    {
+      "epoch": 0.5016031350195939,
+      "grad_norm": 1.367015284250686,
+      "learning_rate": 1.0034207525655646e-05,
+      "loss": 0.9979,
+      "step": 880
+    },
+    {
+      "epoch": 0.5073031706448166,
+      "grad_norm": 1.24993826544231,
+      "learning_rate": 1.014823261117446e-05,
+      "loss": 0.9898,
+      "step": 890
+    },
+    {
+      "epoch": 0.5130032062700391,
+      "grad_norm": 1.188181097585842,
+      "learning_rate": 1.0262257696693273e-05,
+      "loss": 0.9821,
+      "step": 900
+    },
+    {
+      "epoch": 0.5187032418952618,
+      "grad_norm": 1.219257967517726,
+      "learning_rate": 1.0376282782212087e-05,
+      "loss": 0.9753,
+      "step": 910
+    },
+    {
+      "epoch": 0.5244032775204845,
+      "grad_norm": 1.3131103694864974,
+      "learning_rate": 1.0490307867730903e-05,
+      "loss": 0.9907,
+      "step": 920
+    },
+    {
+      "epoch": 0.5301033131457071,
+      "grad_norm": 1.3286474526232694,
+      "learning_rate": 1.0604332953249717e-05,
+      "loss": 0.9886,
+      "step": 930
+    },
+    {
+      "epoch": 0.5358033487709298,
+      "grad_norm": 1.3048668421776233,
+      "learning_rate": 1.0718358038768531e-05,
+      "loss": 0.9826,
+      "step": 940
+    },
+    {
+      "epoch": 0.5415033843961524,
+      "grad_norm": 1.3627327837670187,
+      "learning_rate": 1.0832383124287345e-05,
+      "loss": 0.9865,
+      "step": 950
+    },
+    {
+      "epoch": 0.5472034200213751,
+      "grad_norm": 1.3265339665857847,
+      "learning_rate": 1.0946408209806159e-05,
+      "loss": 0.9804,
+      "step": 960
+    },
+    {
+      "epoch": 0.5529034556465978,
+      "grad_norm": 1.3467055060741506,
+      "learning_rate": 1.1060433295324973e-05,
+      "loss": 0.9725,
+      "step": 970
+    },
+    {
+      "epoch": 0.5586034912718204,
+      "grad_norm": 1.3166304290161006,
+      "learning_rate": 1.1174458380843787e-05,
+      "loss": 0.9745,
+      "step": 980
+    },
+    {
+      "epoch": 0.5643035268970431,
+      "grad_norm": 1.2412125716176452,
+      "learning_rate": 1.12884834663626e-05,
+      "loss": 0.962,
+      "step": 990
+    },
+    {
+      "epoch": 0.5700035625222658,
+      "grad_norm": 1.3189399694165516,
+      "learning_rate": 1.1402508551881415e-05,
+      "loss": 0.9821,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5757035981474884,
+      "grad_norm": 1.3642955662034428,
+      "learning_rate": 1.1516533637400229e-05,
+      "loss": 0.9718,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5814036337727111,
+      "grad_norm": 1.322103414761333,
+      "learning_rate": 1.1630558722919043e-05,
+      "loss": 0.9831,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5871036693979338,
+      "grad_norm": 1.4293389786169786,
+      "learning_rate": 1.1744583808437857e-05,
+      "loss": 0.9661,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5928037050231564,
+      "grad_norm": 1.3823165883871669,
+      "learning_rate": 1.185860889395667e-05,
+      "loss": 0.973,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5985037406483791,
+      "grad_norm": 1.4800355218426224,
+      "learning_rate": 1.1972633979475485e-05,
+      "loss": 0.9701,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6042037762736017,
+      "grad_norm": 1.3491909788324237,
+      "learning_rate": 1.20866590649943e-05,
+      "loss": 0.9583,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6099038118988244,
+      "grad_norm": 1.4377241923522286,
+      "learning_rate": 1.2200684150513114e-05,
+      "loss": 0.9578,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6156038475240471,
+      "grad_norm": 1.4076537543124394,
+      "learning_rate": 1.2314709236031928e-05,
+      "loss": 0.9558,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6213038831492697,
+      "grad_norm": 1.4389533917231583,
+      "learning_rate": 1.2428734321550742e-05,
+      "loss": 0.9582,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6270039187744924,
+      "grad_norm": 1.414132752155285,
+      "learning_rate": 1.2542759407069556e-05,
+      "loss": 0.9552,
+      "step": 1100
+    },
+    {
+      "epoch": 0.632703954399715,
+      "grad_norm": 1.3992854580137006,
+      "learning_rate": 1.265678449258837e-05,
+      "loss": 0.9553,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6384039900249376,
+      "grad_norm": 1.3690113788333766,
+      "learning_rate": 1.2770809578107184e-05,
+      "loss": 0.9371,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6441040256501603,
+      "grad_norm": 1.4043265246321486,
+      "learning_rate": 1.2884834663625998e-05,
+      "loss": 0.9493,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6498040612753829,
+      "grad_norm": 1.3861708246884175,
+      "learning_rate": 1.2998859749144812e-05,
+      "loss": 0.9416,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6555040969006056,
+      "grad_norm": 1.4273466903674794,
+      "learning_rate": 1.3112884834663626e-05,
+      "loss": 0.9357,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6612041325258283,
+      "grad_norm": 1.4006529288929217,
+      "learning_rate": 1.322690992018244e-05,
+      "loss": 0.9391,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6669041681510509,
+      "grad_norm": 1.441422241182402,
+      "learning_rate": 1.3340935005701254e-05,
+      "loss": 0.9352,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6726042037762736,
+      "grad_norm": 1.574058398645153,
+      "learning_rate": 1.3454960091220068e-05,
+      "loss": 0.9272,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6783042394014963,
+      "grad_norm": 1.4756460944297158,
+      "learning_rate": 1.3568985176738885e-05,
+      "loss": 0.9366,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6840042750267189,
+      "grad_norm": 1.5008253996689045,
+      "learning_rate": 1.3683010262257699e-05,
+      "loss": 0.9337,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6897043106519416,
+      "grad_norm": 1.499817814739839,
+      "learning_rate": 1.3797035347776513e-05,
+      "loss": 0.921,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6954043462771642,
+      "grad_norm": 1.5055646516816286,
+      "learning_rate": 1.3911060433295327e-05,
+      "loss": 0.9332,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7011043819023869,
+      "grad_norm": 1.5564454315844756,
+      "learning_rate": 1.4025085518814141e-05,
+      "loss": 0.9303,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7068044175276096,
+      "grad_norm": 1.6439995518569874,
+      "learning_rate": 1.4139110604332955e-05,
+      "loss": 0.93,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7125044531528322,
+      "grad_norm": 1.6715644098081408,
+      "learning_rate": 1.4253135689851769e-05,
+      "loss": 0.9216,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7182044887780549,
+      "grad_norm": 1.530693928930533,
+      "learning_rate": 1.4367160775370583e-05,
+      "loss": 0.9202,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7239045244032776,
+      "grad_norm": 1.501811382005542,
+      "learning_rate": 1.4481185860889397e-05,
+      "loss": 0.9241,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7296045600285002,
+      "grad_norm": 1.5857176435296712,
+      "learning_rate": 1.459521094640821e-05,
+      "loss": 0.9182,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7353045956537229,
+      "grad_norm": 1.6016630245706045,
+      "learning_rate": 1.4709236031927025e-05,
+      "loss": 0.9315,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7410046312789454,
+      "grad_norm": 1.6390065509439025,
+      "learning_rate": 1.482326111744584e-05,
+      "loss": 0.9112,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7467046669041681,
+      "grad_norm": 1.6435383861134847,
+      "learning_rate": 1.4937286202964654e-05,
+      "loss": 0.9272,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7524047025293908,
+      "grad_norm": 1.6250668417933374,
+      "learning_rate": 1.5051311288483468e-05,
+      "loss": 0.9297,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7581047381546134,
+      "grad_norm": 1.5301736614268264,
+      "learning_rate": 1.5165336374002282e-05,
+      "loss": 0.9342,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7638047737798361,
+      "grad_norm": 1.4990905871574132,
+      "learning_rate": 1.5279361459521094e-05,
+      "loss": 0.9112,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7695048094050588,
+      "grad_norm": 1.5663816234369212,
+      "learning_rate": 1.539338654503991e-05,
+      "loss": 0.9117,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7752048450302814,
+      "grad_norm": 1.4930234575277266,
+      "learning_rate": 1.5507411630558726e-05,
+      "loss": 0.9004,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7809048806555041,
+      "grad_norm": 1.6617358858909224,
+      "learning_rate": 1.5621436716077538e-05,
+      "loss": 0.9117,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7866049162807267,
+      "grad_norm": 1.6825594646269288,
+      "learning_rate": 1.5735461801596354e-05,
+      "loss": 0.9163,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7923049519059494,
+      "grad_norm": 1.755907748115666,
+      "learning_rate": 1.5849486887115166e-05,
+      "loss": 0.9172,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7980049875311721,
+      "grad_norm": 1.6075387122236047,
+      "learning_rate": 1.596351197263398e-05,
+      "loss": 0.9072,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8037050231563947,
+      "grad_norm": 1.6602708406746058,
+      "learning_rate": 1.6077537058152794e-05,
+      "loss": 0.9038,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8094050587816174,
+      "grad_norm": 1.6473906939147263,
+      "learning_rate": 1.619156214367161e-05,
+      "loss": 0.9078,
+      "step": 1420
+    },
+    {
+      "epoch": 0.8151050944068401,
+      "grad_norm": 1.7005886333455729,
+      "learning_rate": 1.6305587229190422e-05,
+      "loss": 0.9092,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8208051300320627,
+      "grad_norm": 1.721609161880879,
+      "learning_rate": 1.6419612314709237e-05,
+      "loss": 0.8968,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8265051656572854,
+      "grad_norm": 1.6866164274833975,
+      "learning_rate": 1.653363740022805e-05,
+      "loss": 0.8897,
+      "step": 1450
+    },
+    {
+      "epoch": 0.832205201282508,
+      "grad_norm": 1.699305012779236,
+      "learning_rate": 1.6647662485746865e-05,
+      "loss": 0.9091,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8379052369077307,
+      "grad_norm": 1.7192783800368083,
+      "learning_rate": 1.6761687571265678e-05,
+      "loss": 0.8915,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8436052725329534,
+      "grad_norm": 1.7393841580348268,
+      "learning_rate": 1.6875712656784493e-05,
+      "loss": 0.9033,
+      "step": 1480
+    },
+    {
+      "epoch": 0.849305308158176,
+      "grad_norm": 1.6214835909104202,
+      "learning_rate": 1.6989737742303306e-05,
+      "loss": 0.9074,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8550053437833987,
+      "grad_norm": 1.725366806565937,
+      "learning_rate": 1.710376282782212e-05,
+      "loss": 0.8935,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8607053794086214,
+      "grad_norm": 1.7651184492996312,
+      "learning_rate": 1.7217787913340937e-05,
+      "loss": 0.886,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8664054150338439,
+      "grad_norm": 1.6507462754164786,
+      "learning_rate": 1.7331812998859753e-05,
+      "loss": 0.89,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8721054506590666,
+      "grad_norm": 1.638093001715848,
+      "learning_rate": 1.7445838084378565e-05,
+      "loss": 0.8916,
+      "step": 1530
+    },
+    {
+      "epoch": 0.8778054862842892,
+      "grad_norm": 1.8092851839912834,
+      "learning_rate": 1.755986316989738e-05,
+      "loss": 0.8859,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8835055219095119,
+      "grad_norm": 1.723062489410676,
+      "learning_rate": 1.7673888255416193e-05,
+      "loss": 0.8919,
+      "step": 1550
+    },
+    {
+      "epoch": 0.8892055575347346,
+      "grad_norm": 1.6985104237734552,
+      "learning_rate": 1.778791334093501e-05,
+      "loss": 0.8913,
+      "step": 1560
+    },
+    {
+      "epoch": 0.8949055931599572,
+      "grad_norm": 1.7350889236903257,
+      "learning_rate": 1.790193842645382e-05,
+      "loss": 0.8915,
+      "step": 1570
+    },
+    {
+      "epoch": 0.9006056287851799,
+      "grad_norm": 1.7491394404658165,
+      "learning_rate": 1.8015963511972636e-05,
+      "loss": 0.8925,
+      "step": 1580
+    },
+    {
+      "epoch": 0.9063056644104026,
+      "grad_norm": 1.7500511217245198,
+      "learning_rate": 1.812998859749145e-05,
+      "loss": 0.8763,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9120057000356252,
+      "grad_norm": 1.6928141820764897,
+      "learning_rate": 1.8244013683010264e-05,
+      "loss": 0.883,
+      "step": 1600
+    },
+    {
+      "epoch": 0.9177057356608479,
+      "grad_norm": 1.80488872658001,
+      "learning_rate": 1.835803876852908e-05,
+      "loss": 0.8849,
+      "step": 1610
+    },
+    {
+      "epoch": 0.9234057712860705,
+      "grad_norm": 1.787335771913563,
+      "learning_rate": 1.8472063854047892e-05,
+      "loss": 0.8887,
+      "step": 1620
+    },
+    {
+      "epoch": 0.9291058069112932,
+      "grad_norm": 1.7052851457365026,
+      "learning_rate": 1.8586088939566708e-05,
+      "loss": 0.8876,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9348058425365159,
+      "grad_norm": 1.6517039110275196,
+      "learning_rate": 1.870011402508552e-05,
+      "loss": 0.8719,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9405058781617385,
+      "grad_norm": 1.803379001547972,
+      "learning_rate": 1.8814139110604336e-05,
+      "loss": 0.8941,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9462059137869612,
+      "grad_norm": 1.7412794922786006,
+      "learning_rate": 1.8928164196123148e-05,
+      "loss": 0.8849,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9519059494121839,
+      "grad_norm": 1.675503434473841,
+      "learning_rate": 1.9042189281641964e-05,
+      "loss": 0.8878,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9576059850374065,
+      "grad_norm": 1.8836350723671362,
+      "learning_rate": 1.9156214367160776e-05,
+      "loss": 0.8885,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9633060206626292,
+      "grad_norm": 1.714840122525118,
+      "learning_rate": 1.927023945267959e-05,
+      "loss": 0.8804,
+      "step": 1690
+    },
+    {
+      "epoch": 0.9690060562878517,
+      "grad_norm": 1.693212071429786,
+      "learning_rate": 1.9384264538198404e-05,
+      "loss": 0.8664,
+      "step": 1700
+    },
+    {
+      "epoch": 0.9747060919130744,
+      "grad_norm": 1.784064006948934,
+      "learning_rate": 1.949828962371722e-05,
+      "loss": 0.887,
+      "step": 1710
+    },
+    {
+      "epoch": 0.9804061275382971,
+      "grad_norm": 1.7674046204897094,
+      "learning_rate": 1.9612314709236032e-05,
+      "loss": 0.8792,
+      "step": 1720
+    },
+    {
+      "epoch": 0.9861061631635197,
+      "grad_norm": 1.761211622720898,
+      "learning_rate": 1.9726339794754847e-05,
+      "loss": 0.8757,
+      "step": 1730
+    },
+    {
+      "epoch": 0.9918061987887424,
+      "grad_norm": 1.8153929367013604,
+      "learning_rate": 1.9840364880273663e-05,
+      "loss": 0.8753,
+      "step": 1740
+    },
+    {
+      "epoch": 0.9975062344139651,
+      "grad_norm": 1.7814599157686348,
+      "learning_rate": 1.9954389965792475e-05,
+      "loss": 0.8631,
+      "step": 1750
+    },
+    {
+      "epoch": 1.0032062700391877,
+      "grad_norm": 1.8020950772155282,
+      "learning_rate": 1.999999287101006e-05,
+      "loss": 0.8597,
+      "step": 1760
+    },
+    {
+      "epoch": 1.0089063056644103,
+      "grad_norm": 1.8834044260626686,
+      "learning_rate": 1.9999949304997227e-05,
+      "loss": 0.8749,
+      "step": 1770
+    },
+    {
+      "epoch": 1.0146063412896331,
+      "grad_norm": 1.8317172645022874,
+      "learning_rate": 1.9999866133693866e-05,
+      "loss": 0.8701,
+      "step": 1780
+    },
+    {
+      "epoch": 1.0203063769148557,
+      "grad_norm": 1.8838111981312964,
+      "learning_rate": 1.999974335742938e-05,
+      "loss": 0.8618,
+      "step": 1790
+    },
+    {
+      "epoch": 1.0260064125400783,
+      "grad_norm": 1.757836158840337,
+      "learning_rate": 1.999958097669003e-05,
+      "loss": 0.8654,
+      "step": 1800
+    },
+    {
+      "epoch": 1.031706448165301,
+      "grad_norm": 1.6747076009804915,
+      "learning_rate": 1.9999378992118937e-05,
+      "loss": 0.8559,
+      "step": 1810
+    },
+    {
+      "epoch": 1.0374064837905237,
+      "grad_norm": 1.763665724140925,
+      "learning_rate": 1.9999137404516062e-05,
+      "loss": 0.8496,
+      "step": 1820
+    },
+    {
+      "epoch": 1.0431065194157463,
+      "grad_norm": 1.7083504545677566,
+      "learning_rate": 1.999885621483823e-05,
+      "loss": 0.85,
+      "step": 1830
+    },
+    {
+      "epoch": 1.048806555040969,
+      "grad_norm": 1.7603057444620083,
+      "learning_rate": 1.9998535424199112e-05,
+      "loss": 0.8579,
+      "step": 1840
+    },
+    {
+      "epoch": 1.0545065906661917,
+      "grad_norm": 2.013331632937321,
+      "learning_rate": 1.9998175033869205e-05,
+      "loss": 0.8644,
+      "step": 1850
+    },
+    {
+      "epoch": 1.0602066262914143,
+      "grad_norm": 1.737011009746836,
+      "learning_rate": 1.999777504527586e-05,
+      "loss": 0.8597,
+      "step": 1860
+    },
+    {
+      "epoch": 1.065906661916637,
+      "grad_norm": 1.8813873322026473,
+      "learning_rate": 1.9997335460003246e-05,
+      "loss": 0.8589,
+      "step": 1870
+    },
+    {
+      "epoch": 1.0716066975418597,
+      "grad_norm": 1.8349067798622685,
+      "learning_rate": 1.9996856279792368e-05,
+      "loss": 0.8526,
+      "step": 1880
+    },
+    {
+      "epoch": 1.0773067331670823,
+      "grad_norm": 1.714703400856248,
+      "learning_rate": 1.999633750654104e-05,
+      "loss": 0.8605,
+      "step": 1890
+    },
+    {
+      "epoch": 1.083006768792305,
+      "grad_norm": 1.853883163691029,
+      "learning_rate": 1.999577914230388e-05,
+      "loss": 0.8582,
+      "step": 1900
+    },
+    {
+      "epoch": 1.0887068044175277,
+      "grad_norm": 1.8709920052287101,
+      "learning_rate": 1.9995181189292334e-05,
+      "loss": 0.8624,
+      "step": 1910
+    },
+    {
+      "epoch": 1.0944068400427502,
+      "grad_norm": 1.8883523943990927,
+      "learning_rate": 1.999454364987461e-05,
+      "loss": 0.8518,
+      "step": 1920
+    },
+    {
+      "epoch": 1.1001068756679728,
+      "grad_norm": 1.9780681144848358,
+      "learning_rate": 1.9993866526575723e-05,
+      "loss": 0.8561,
+      "step": 1930
+    },
+    {
+      "epoch": 1.1058069112931956,
+      "grad_norm": 1.7852231544209909,
+      "learning_rate": 1.999314982207745e-05,
+      "loss": 0.8627,
+      "step": 1940
+    },
+    {
+      "epoch": 1.1115069469184182,
+      "grad_norm": 1.7550418237943235,
+      "learning_rate": 1.9992393539218334e-05,
+      "loss": 0.8468,
+      "step": 1950
+    },
+    {
+      "epoch": 1.1172069825436408,
+      "grad_norm": 1.914656706563883,
+      "learning_rate": 1.999159768099367e-05,
+      "loss": 0.8612,
+      "step": 1960
+    },
+    {
+      "epoch": 1.1229070181688636,
+      "grad_norm": 2.030038254737503,
+      "learning_rate": 1.9990762250555495e-05,
+      "loss": 0.8515,
+      "step": 1970
+    },
+    {
+      "epoch": 1.1286070537940862,
+      "grad_norm": 1.8141661775314604,
+      "learning_rate": 1.9989887251212575e-05,
+      "loss": 0.8403,
+      "step": 1980
+    },
+    {
+      "epoch": 1.1343070894193088,
+      "grad_norm": 1.7992375971001897,
+      "learning_rate": 1.9988972686430382e-05,
+      "loss": 0.8434,
+      "step": 1990
+    },
+    {
+      "epoch": 1.1400071250445316,
+      "grad_norm": 1.8154747951291372,
+      "learning_rate": 1.9988018559831093e-05,
+      "loss": 0.8573,
+      "step": 2000
+    },
+    {
+      "epoch": 1.1457071606697542,
+      "grad_norm": 1.8823541957530339,
+      "learning_rate": 1.998702487519358e-05,
+      "loss": 0.8446,
+      "step": 2010
+    },
+    {
+      "epoch": 1.1514071962949768,
+      "grad_norm": 1.892962359232292,
+      "learning_rate": 1.998599163645338e-05,
+      "loss": 0.8471,
+      "step": 2020
+    },
+    {
+      "epoch": 1.1571072319201996,
+      "grad_norm": 1.8796275830005886,
+      "learning_rate": 1.9984918847702684e-05,
+      "loss": 0.8475,
+      "step": 2030
+    },
+    {
+      "epoch": 1.1628072675454222,
+      "grad_norm": 2.1504877623718777,
+      "learning_rate": 1.9983806513190323e-05,
+      "loss": 0.8463,
+      "step": 2040
+    },
+    {
+      "epoch": 1.1685073031706448,
+      "grad_norm": 2.1256410895372375,
+      "learning_rate": 1.998265463732175e-05,
+      "loss": 0.8406,
+      "step": 2050
+    },
+    {
+      "epoch": 1.1742073387958674,
+      "grad_norm": 1.9618985154355109,
+      "learning_rate": 1.9981463224659034e-05,
+      "loss": 0.8486,
+      "step": 2060
+    },
+    {
+      "epoch": 1.1799073744210902,
+      "grad_norm": 2.070139508836993,
+      "learning_rate": 1.9980232279920814e-05,
+      "loss": 0.836,
+      "step": 2070
+    },
+    {
+      "epoch": 1.1856074100463128,
+      "grad_norm": 2.0226054921350083,
+      "learning_rate": 1.9978961807982312e-05,
+      "loss": 0.8432,
+      "step": 2080
+    },
+    {
+      "epoch": 1.1913074456715353,
+      "grad_norm": 1.9367240553204879,
+      "learning_rate": 1.9977651813875293e-05,
+      "loss": 0.8327,
+      "step": 2090
+    },
+    {
+      "epoch": 1.1970074812967582,
+      "grad_norm": 1.9072166052236994,
+      "learning_rate": 1.997630230278806e-05,
+      "loss": 0.8362,
+      "step": 2100
+    },
+    {
+      "epoch": 1.2027075169219807,
+      "grad_norm": 1.9432404950125435,
+      "learning_rate": 1.997491328006541e-05,
+      "loss": 0.8441,
+      "step": 2110
+    },
+    {
+      "epoch": 1.2084075525472033,
+      "grad_norm": 1.911231463693163,
+      "learning_rate": 1.9973484751208636e-05,
+      "loss": 0.8383,
+      "step": 2120
+    },
+    {
+      "epoch": 1.2141075881724261,
+      "grad_norm": 1.9483520712665299,
+      "learning_rate": 1.99720167218755e-05,
+      "loss": 0.8507,
+      "step": 2130
+    },
+    {
+      "epoch": 1.2198076237976487,
+      "grad_norm": 1.8921621676164715,
+      "learning_rate": 1.9970509197880204e-05,
+      "loss": 0.8356,
+      "step": 2140
+    },
+    {
+      "epoch": 1.2255076594228713,
+      "grad_norm": 1.9747176106356008,
+      "learning_rate": 1.9968962185193367e-05,
+      "loss": 0.8411,
+      "step": 2150
+    },
+    {
+      "epoch": 1.2312076950480941,
+      "grad_norm": 1.9777158344485668,
+      "learning_rate": 1.9967375689942013e-05,
+      "loss": 0.8319,
+      "step": 2160
+    },
+    {
+      "epoch": 1.2369077306733167,
+      "grad_norm": 1.9527547946588864,
+      "learning_rate": 1.9965749718409532e-05,
+      "loss": 0.8488,
+      "step": 2170
+    },
+    {
+      "epoch": 1.2426077662985393,
+      "grad_norm": 1.9037749552318648,
+      "learning_rate": 1.9964084277035668e-05,
+      "loss": 0.8452,
+      "step": 2180
+    },
+    {
+      "epoch": 1.2483078019237621,
+      "grad_norm": 1.8459581980245552,
+      "learning_rate": 1.996237937241648e-05,
+      "loss": 0.8414,
+      "step": 2190
+    },
+    {
+      "epoch": 1.2540078375489847,
+      "grad_norm": 2.0833571443694288,
+      "learning_rate": 1.9960635011304325e-05,
+      "loss": 0.8434,
+      "step": 2200
+    },
+    {
+      "epoch": 1.2597078731742073,
+      "grad_norm": 1.86911029908039,
+      "learning_rate": 1.9958851200607833e-05,
+      "loss": 0.8395,
+      "step": 2210
+    },
+    {
+      "epoch": 1.26540790879943,
+      "grad_norm": 1.9980243217779092,
+      "learning_rate": 1.9957027947391873e-05,
+      "loss": 0.8477,
+      "step": 2220
+    },
+    {
+      "epoch": 1.2711079444246527,
+      "grad_norm": 1.9828685448765224,
+      "learning_rate": 1.9955165258877534e-05,
+      "loss": 0.8354,
+      "step": 2230
+    },
+    {
+      "epoch": 1.2768079800498753,
+      "grad_norm": 1.9814305922927593,
+      "learning_rate": 1.9953263142442078e-05,
+      "loss": 0.8356,
+      "step": 2240
+    },
+    {
+      "epoch": 1.282508015675098,
+      "grad_norm": 2.094648535162255,
+      "learning_rate": 1.9951321605618932e-05,
+      "loss": 0.8259,
+      "step": 2250
+    },
+    {
+      "epoch": 1.2882080513003207,
+      "grad_norm": 1.9713099315168099,
+      "learning_rate": 1.9949340656097652e-05,
+      "loss": 0.8307,
+      "step": 2260
+    },
+    {
+      "epoch": 1.2939080869255433,
+      "grad_norm": 1.9388006652656193,
+      "learning_rate": 1.9947320301723882e-05,
+      "loss": 0.8431,
+      "step": 2270
+    },
+    {
+      "epoch": 1.299608122550766,
+      "grad_norm": 2.0437788441257663,
+      "learning_rate": 1.9945260550499337e-05,
+      "loss": 0.839,
+      "step": 2280
+    },
+    {
+      "epoch": 1.3053081581759887,
+      "grad_norm": 1.9200807256749128,
+      "learning_rate": 1.9943161410581765e-05,
+      "loss": 0.8401,
+      "step": 2290
+    },
+    {
+      "epoch": 1.3110081938012113,
+      "grad_norm": 1.8932159307338257,
+      "learning_rate": 1.994102289028491e-05,
+      "loss": 0.8297,
+      "step": 2300
+    },
+    {
+      "epoch": 1.3167082294264338,
+      "grad_norm": 1.8604003595627263,
+      "learning_rate": 1.993884499807848e-05,
+      "loss": 0.8322,
+      "step": 2310
+    },
+    {
+      "epoch": 1.3224082650516567,
+      "grad_norm": 2.067435850526306,
+      "learning_rate": 1.9936627742588136e-05,
+      "loss": 0.8331,
+      "step": 2320
+    },
+    {
+      "epoch": 1.3281083006768792,
+      "grad_norm": 1.9736131852274952,
+      "learning_rate": 1.9934371132595426e-05,
+      "loss": 0.8253,
+      "step": 2330
+    },
+    {
+      "epoch": 1.3338083363021018,
+      "grad_norm": 1.973431146884113,
+      "learning_rate": 1.9932075177037757e-05,
+      "loss": 0.8252,
+      "step": 2340
+    },
+    {
+      "epoch": 1.3395083719273244,
+      "grad_norm": 1.987364467843,
+      "learning_rate": 1.9929739885008375e-05,
+      "loss": 0.8218,
+      "step": 2350
+    },
+    {
+      "epoch": 1.3452084075525472,
+      "grad_norm": 2.0389811693345234,
+      "learning_rate": 1.9927365265756326e-05,
+      "loss": 0.834,
+      "step": 2360
+    },
+    {
+      "epoch": 1.3509084431777698,
+      "grad_norm": 1.913868930297226,
+      "learning_rate": 1.9924951328686398e-05,
+      "loss": 0.8324,
+      "step": 2370
+    },
+    {
+      "epoch": 1.3566084788029924,
+      "grad_norm": 1.8211826510498335,
+      "learning_rate": 1.9922498083359113e-05,
+      "loss": 0.8257,
+      "step": 2380
+    },
+    {
+      "epoch": 1.3623085144282152,
+      "grad_norm": 2.0385423896344537,
+      "learning_rate": 1.9920005539490666e-05,
+      "loss": 0.8274,
+      "step": 2390
+    },
+    {
+      "epoch": 1.3680085500534378,
+      "grad_norm": 1.9828362611522425,
+      "learning_rate": 1.9917473706952905e-05,
+      "loss": 0.8349,
+      "step": 2400
+    },
+    {
+      "epoch": 1.3737085856786604,
+      "grad_norm": 2.016395513885357,
+      "learning_rate": 1.9914902595773268e-05,
+      "loss": 0.8306,
+      "step": 2410
+    },
+    {
+      "epoch": 1.3794086213038832,
+      "grad_norm": 1.9533372967663756,
+      "learning_rate": 1.9912292216134775e-05,
+      "loss": 0.8298,
+      "step": 2420
+    },
+    {
+      "epoch": 1.3851086569291058,
+      "grad_norm": 1.9530067606486703,
+      "learning_rate": 1.990964257837596e-05,
+      "loss": 0.8148,
+      "step": 2430
+    },
+    {
+      "epoch": 1.3908086925543284,
+      "grad_norm": 2.0379623003529295,
+      "learning_rate": 1.9906953692990843e-05,
+      "loss": 0.8277,
+      "step": 2440
+    },
+    {
+      "epoch": 1.3965087281795512,
+      "grad_norm": 1.996917345031334,
+      "learning_rate": 1.990422557062889e-05,
+      "loss": 0.8248,
+      "step": 2450
+    },
+    {
+      "epoch": 1.4022087638047738,
+      "grad_norm": 2.1834776606325508,
+      "learning_rate": 1.9901458222094964e-05,
+      "loss": 0.8291,
+      "step": 2460
+    },
+    {
+      "epoch": 1.4079087994299964,
+      "grad_norm": 1.8262603600285912,
+      "learning_rate": 1.9898651658349276e-05,
+      "loss": 0.8294,
+      "step": 2470
+    },
+    {
+      "epoch": 1.4136088350552192,
+      "grad_norm": 1.9148153734335014,
+      "learning_rate": 1.9895805890507368e-05,
+      "loss": 0.827,
+      "step": 2480
+    },
+    {
+      "epoch": 1.4193088706804418,
+      "grad_norm": 1.8109824730555837,
+      "learning_rate": 1.9892920929840042e-05,
+      "loss": 0.8256,
+      "step": 2490
+    },
+    {
+      "epoch": 1.4250089063056643,
+      "grad_norm": 1.9038427847599968,
+      "learning_rate": 1.988999678777332e-05,
+      "loss": 0.8149,
+      "step": 2500
+    },
+    {
+      "epoch": 1.4307089419308872,
+      "grad_norm": 1.9183688798514287,
+      "learning_rate": 1.988703347588842e-05,
+      "loss": 0.8219,
+      "step": 2510
+    },
+    {
+      "epoch": 1.4364089775561097,
+      "grad_norm": 2.0062935185811988,
+      "learning_rate": 1.988403100592168e-05,
+      "loss": 0.8272,
+      "step": 2520
+    },
+    {
+      "epoch": 1.4421090131813323,
+      "grad_norm": 1.916360321048196,
+      "learning_rate": 1.988098938976453e-05,
+      "loss": 0.8185,
+      "step": 2530
+    },
+    {
+      "epoch": 1.4478090488065551,
+      "grad_norm": 1.9474228376010432,
+      "learning_rate": 1.9877908639463438e-05,
+      "loss": 0.8224,
+      "step": 2540
+    },
+    {
+      "epoch": 1.4535090844317777,
+      "grad_norm": 2.030260343191728,
+      "learning_rate": 1.987478876721987e-05,
+      "loss": 0.829,
+      "step": 2550
+    },
+    {
+      "epoch": 1.4592091200570003,
+      "grad_norm": 2.0669389591243332,
+      "learning_rate": 1.9871629785390234e-05,
+      "loss": 0.823,
+      "step": 2560
+    },
+    {
+      "epoch": 1.4649091556822231,
+      "grad_norm": 1.939164521109589,
+      "learning_rate": 1.986843170648583e-05,
+      "loss": 0.8192,
+      "step": 2570
+    },
+    {
+      "epoch": 1.4706091913074457,
+      "grad_norm": 1.9340560162810296,
+      "learning_rate": 1.9865194543172808e-05,
+      "loss": 0.813,
+      "step": 2580
+    },
+    {
+      "epoch": 1.4763092269326683,
+      "grad_norm": 1.9826253649714038,
+      "learning_rate": 1.986191830827211e-05,
+      "loss": 0.8206,
+      "step": 2590
+    },
+    {
+      "epoch": 1.4820092625578911,
+      "grad_norm": 2.086689008755194,
+      "learning_rate": 1.985860301475943e-05,
+      "loss": 0.8288,
+      "step": 2600
+    },
+    {
+      "epoch": 1.4877092981831137,
+      "grad_norm": 1.9286741203215845,
+      "learning_rate": 1.9855248675765146e-05,
+      "loss": 0.8212,
+      "step": 2610
+    },
+    {
+      "epoch": 1.4934093338083363,
+      "grad_norm": 1.9752592212958118,
+      "learning_rate": 1.9851855304574287e-05,
+      "loss": 0.8271,
+      "step": 2620
+    },
+    {
+      "epoch": 1.4991093694335589,
+      "grad_norm": 1.9086583701062574,
+      "learning_rate": 1.9848422914626462e-05,
+      "loss": 0.8287,
+      "step": 2630
+    },
+    {
+      "epoch": 1.5048094050587815,
+      "grad_norm": 1.8077255473050955,
+      "learning_rate": 1.984495151951582e-05,
+      "loss": 0.8171,
+      "step": 2640
+    },
+    {
+      "epoch": 1.5105094406840043,
+      "grad_norm": 1.9632613712236342,
+      "learning_rate": 1.9841441132990998e-05,
+      "loss": 0.8253,
+      "step": 2650
+    },
+    {
+      "epoch": 1.516209476309227,
+      "grad_norm": 1.9602391094516611,
+      "learning_rate": 1.983789176895505e-05,
+      "loss": 0.809,
+      "step": 2660
+    },
+    {
+      "epoch": 1.5219095119344495,
+      "grad_norm": 1.9977668186898982,
+      "learning_rate": 1.9834303441465402e-05,
+      "loss": 0.8264,
+      "step": 2670
+    },
+    {
+      "epoch": 1.5276095475596723,
+      "grad_norm": 2.004699978471638,
+      "learning_rate": 1.9830676164733808e-05,
+      "loss": 0.8128,
+      "step": 2680
+    },
+    {
+      "epoch": 1.5333095831848949,
+      "grad_norm": 2.0025333597676043,
+      "learning_rate": 1.9827009953126277e-05,
+      "loss": 0.8049,
+      "step": 2690
+    },
+    {
+      "epoch": 1.5390096188101174,
+      "grad_norm": 2.006348925420673,
+      "learning_rate": 1.982330482116301e-05,
+      "loss": 0.8144,
+      "step": 2700
+    },
+    {
+      "epoch": 1.5447096544353403,
+      "grad_norm": 1.8911374175722429,
+      "learning_rate": 1.9819560783518378e-05,
+      "loss": 0.8044,
+      "step": 2710
+    },
+    {
+      "epoch": 1.5504096900605628,
+      "grad_norm": 2.027861640952926,
+      "learning_rate": 1.9815777855020818e-05,
+      "loss": 0.8171,
+      "step": 2720
+    },
+    {
+      "epoch": 1.5561097256857854,
+      "grad_norm": 1.972962804971287,
+      "learning_rate": 1.9811956050652803e-05,
+      "loss": 0.8145,
+      "step": 2730
+    },
+    {
+      "epoch": 1.5618097613110082,
+      "grad_norm": 2.126768344817225,
+      "learning_rate": 1.9808095385550777e-05,
+      "loss": 0.8229,
+      "step": 2740
+    },
+    {
+      "epoch": 1.5675097969362308,
+      "grad_norm": 1.9270909648882355,
+      "learning_rate": 1.98041958750051e-05,
+      "loss": 0.8166,
+      "step": 2750
+    },
+    {
+      "epoch": 1.5732098325614534,
+      "grad_norm": 2.0062992830698794,
+      "learning_rate": 1.980025753445997e-05,
+      "loss": 0.8113,
+      "step": 2760
+    },
+    {
+      "epoch": 1.5789098681866762,
+      "grad_norm": 1.8604867203192628,
+      "learning_rate": 1.979628037951338e-05,
+      "loss": 0.818,
+      "step": 2770
+    },
+    {
+      "epoch": 1.5846099038118988,
+      "grad_norm": 2.0028993599947658,
+      "learning_rate": 1.9792264425917048e-05,
+      "loss": 0.8144,
+      "step": 2780
+    },
+    {
+      "epoch": 1.5903099394371214,
+      "grad_norm": 1.9586165011975456,
+      "learning_rate": 1.9788209689576356e-05,
+      "loss": 0.8135,
+      "step": 2790
+    },
+    {
+      "epoch": 1.5960099750623442,
+      "grad_norm": 2.0709951025768967,
+      "learning_rate": 1.9784116186550282e-05,
+      "loss": 0.8125,
+      "step": 2800
+    },
+    {
+      "epoch": 1.6017100106875668,
+      "grad_norm": 1.9884880289538327,
+      "learning_rate": 1.977998393305135e-05,
+      "loss": 0.8142,
+      "step": 2810
+    },
+    {
+      "epoch": 1.6074100463127894,
+      "grad_norm": 2.1537453722600888,
+      "learning_rate": 1.977581294544555e-05,
+      "loss": 0.8226,
+      "step": 2820
+    },
+    {
+      "epoch": 1.6131100819380122,
+      "grad_norm": 1.9578569344740548,
+      "learning_rate": 1.9771603240252287e-05,
+      "loss": 0.8222,
+      "step": 2830
+    },
+    {
+      "epoch": 1.6188101175632348,
+      "grad_norm": 1.9762705157898854,
+      "learning_rate": 1.97673548341443e-05,
+      "loss": 0.7992,
+      "step": 2840
+    },
+    {
+      "epoch": 1.6245101531884574,
+      "grad_norm": 1.9396847013255831,
+      "learning_rate": 1.9763067743947618e-05,
+      "loss": 0.8145,
+      "step": 2850
+    },
+    {
+      "epoch": 1.6302101888136802,
+      "grad_norm": 2.049404293109037,
+      "learning_rate": 1.9758741986641466e-05,
+      "loss": 0.8206,
+      "step": 2860
+    },
+    {
+      "epoch": 1.6359102244389028,
+      "grad_norm": 1.971677311667184,
+      "learning_rate": 1.9754377579358222e-05,
+      "loss": 0.8108,
+      "step": 2870
+    },
+    {
+      "epoch": 1.6416102600641254,
+      "grad_norm": 2.053189987118361,
+      "learning_rate": 1.974997453938333e-05,
+      "loss": 0.8131,
+      "step": 2880
+    },
+    {
+      "epoch": 1.6473102956893482,
+      "grad_norm": 1.9151820373034578,
+      "learning_rate": 1.974553288415525e-05,
+      "loss": 0.8231,
+      "step": 2890
+    },
+    {
+      "epoch": 1.6530103313145708,
+      "grad_norm": 2.11765172957531,
+      "learning_rate": 1.974105263126538e-05,
+      "loss": 0.8266,
+      "step": 2900
+    },
+    {
+      "epoch": 1.6587103669397933,
+      "grad_norm": 1.9800541115222154,
+      "learning_rate": 1.9736533798457976e-05,
+      "loss": 0.8157,
+      "step": 2910
+    },
+    {
+      "epoch": 1.6644104025650162,
+      "grad_norm": 2.064806833520072,
+      "learning_rate": 1.9731976403630096e-05,
+      "loss": 0.813,
+      "step": 2920
+    },
+    {
+      "epoch": 1.6701104381902387,
+      "grad_norm": 2.1962096297446214,
+      "learning_rate": 1.972738046483153e-05,
+      "loss": 0.8019,
+      "step": 2930
+    },
+    {
+      "epoch": 1.6758104738154613,
+      "grad_norm": 1.8787057397799634,
+      "learning_rate": 1.972274600026472e-05,
+      "loss": 0.816,
+      "step": 2940
+    },
+    {
+      "epoch": 1.6815105094406841,
+      "grad_norm": 2.1077697612705575,
+      "learning_rate": 1.9718073028284686e-05,
+      "loss": 0.8182,
+      "step": 2950
+    },
+    {
+      "epoch": 1.6872105450659065,
+      "grad_norm": 1.8352913460041305,
+      "learning_rate": 1.971336156739897e-05,
+      "loss": 0.8171,
+      "step": 2960
+    },
+    {
+      "epoch": 1.6929105806911293,
+      "grad_norm": 1.9189762589507913,
+      "learning_rate": 1.9708611636267538e-05,
+      "loss": 0.8136,
+      "step": 2970
+    },
+    {
+      "epoch": 1.6986106163163521,
+      "grad_norm": 1.7917414242527039,
+      "learning_rate": 1.9703823253702728e-05,
+      "loss": 0.8137,
+      "step": 2980
+    },
+    {
+      "epoch": 1.7043106519415745,
+      "grad_norm": 1.9742494605498844,
+      "learning_rate": 1.9698996438669163e-05,
+      "loss": 0.8145,
+      "step": 2990
+    },
+    {
+      "epoch": 1.7100106875667973,
+      "grad_norm": 2.1685413353256564,
+      "learning_rate": 1.969413121028368e-05,
+      "loss": 0.8183,
+      "step": 3000
+    },
+    {
+      "epoch": 1.7157107231920201,
+      "grad_norm": 1.9866035446983392,
+      "learning_rate": 1.9689227587815263e-05,
+      "loss": 0.8097,
+      "step": 3010
+    },
+    {
+      "epoch": 1.7214107588172425,
+      "grad_norm": 2.1414779890164404,
+      "learning_rate": 1.968428559068494e-05,
+      "loss": 0.8078,
+      "step": 3020
+    },
+    {
+      "epoch": 1.7271107944424653,
+      "grad_norm": 1.9774269101559483,
+      "learning_rate": 1.967930523846574e-05,
+      "loss": 0.8117,
+      "step": 3030
+    },
+    {
+      "epoch": 1.7328108300676879,
+      "grad_norm": 2.1161328813493263,
+      "learning_rate": 1.9674286550882593e-05,
+      "loss": 0.8007,
+      "step": 3040
+    },
+    {
+      "epoch": 1.7385108656929105,
+      "grad_norm": 2.043347202651497,
+      "learning_rate": 1.966922954781225e-05,
+      "loss": 0.8103,
+      "step": 3050
+    },
+    {
+      "epoch": 1.7442109013181333,
+      "grad_norm": 1.9187301402180963,
+      "learning_rate": 1.9664134249283226e-05,
+      "loss": 0.8119,
+      "step": 3060
+    },
+    {
+      "epoch": 1.7499109369433559,
+      "grad_norm": 1.9921052617040738,
+      "learning_rate": 1.96590006754757e-05,
+      "loss": 0.8055,
+      "step": 3070
+    },
+    {
+      "epoch": 1.7556109725685785,
+      "grad_norm": 1.9688286652209337,
+      "learning_rate": 1.9653828846721447e-05,
+      "loss": 0.8093,
+      "step": 3080
+    },
+    {
+      "epoch": 1.7613110081938013,
+      "grad_norm": 1.979593441899211,
+      "learning_rate": 1.964861878350374e-05,
+      "loss": 0.805,
+      "step": 3090
+    },
+    {
+      "epoch": 1.7670110438190239,
+      "grad_norm": 2.089479723603799,
+      "learning_rate": 1.96433705064573e-05,
+      "loss": 0.8012,
+      "step": 3100
+    },
+    {
+      "epoch": 1.7727110794442464,
+      "grad_norm": 2.0650887106817355,
+      "learning_rate": 1.963808403636818e-05,
+      "loss": 0.8117,
+      "step": 3110
+    },
+    {
+      "epoch": 1.7784111150694693,
+      "grad_norm": 2.036314415447528,
+      "learning_rate": 1.9632759394173705e-05,
+      "loss": 0.8053,
+      "step": 3120
+    },
+    {
+      "epoch": 1.7841111506946918,
+      "grad_norm": 2.0270600386737327,
+      "learning_rate": 1.962739660096239e-05,
+      "loss": 0.8116,
+      "step": 3130
+    },
+    {
+      "epoch": 1.7898111863199144,
+      "grad_norm": 2.019257712794015,
+      "learning_rate": 1.9621995677973827e-05,
+      "loss": 0.8076,
+      "step": 3140
+    },
+    {
+      "epoch": 1.7955112219451372,
+      "grad_norm": 2.00454036884464,
+      "learning_rate": 1.9616556646598647e-05,
+      "loss": 0.8129,
+      "step": 3150
+    },
+    {
+      "epoch": 1.8012112575703598,
+      "grad_norm": 2.039343370559229,
+      "learning_rate": 1.9611079528378395e-05,
+      "loss": 0.7991,
+      "step": 3160
+    },
+    {
+      "epoch": 1.8069112931955824,
+      "grad_norm": 2.1059222924259346,
+      "learning_rate": 1.9605564345005473e-05,
+      "loss": 0.7973,
+      "step": 3170
+    },
+    {
+      "epoch": 1.8126113288208052,
+      "grad_norm": 2.053612989554001,
+      "learning_rate": 1.9600011118323034e-05,
+      "loss": 0.7968,
+      "step": 3180
+    },
+    {
+      "epoch": 1.8183113644460278,
+      "grad_norm": 2.097986114345962,
+      "learning_rate": 1.9594419870324902e-05,
+      "loss": 0.7988,
+      "step": 3190
+    },
+    {
+      "epoch": 1.8240114000712504,
+      "grad_norm": 2.02925434718376,
+      "learning_rate": 1.958879062315549e-05,
+      "loss": 0.8106,
+      "step": 3200
+    },
+    {
+      "epoch": 1.8297114356964732,
+      "grad_norm": 2.01590406369675,
+      "learning_rate": 1.958312339910971e-05,
+      "loss": 0.806,
+      "step": 3210
+    },
+    {
+      "epoch": 1.8354114713216958,
+      "grad_norm": 1.9619147724985244,
+      "learning_rate": 1.957741822063288e-05,
+      "loss": 0.7976,
+      "step": 3220
+    },
+    {
+      "epoch": 1.8411115069469184,
+      "grad_norm": 2.110272560272491,
+      "learning_rate": 1.9571675110320643e-05,
+      "loss": 0.7943,
+      "step": 3230
+    },
+    {
+      "epoch": 1.8468115425721412,
+      "grad_norm": 2.0255760286052427,
+      "learning_rate": 1.9565894090918865e-05,
+      "loss": 0.8021,
+      "step": 3240
+    },
+    {
+      "epoch": 1.8525115781973638,
+      "grad_norm": 1.99496778091942,
+      "learning_rate": 1.956007518532356e-05,
+      "loss": 0.802,
+      "step": 3250
+    },
+    {
+      "epoch": 1.8582116138225864,
+      "grad_norm": 1.994786516261079,
+      "learning_rate": 1.9554218416580787e-05,
+      "loss": 0.8038,
+      "step": 3260
+    },
+    {
+      "epoch": 1.8639116494478092,
+      "grad_norm": 2.205732684895112,
+      "learning_rate": 1.9548323807886568e-05,
+      "loss": 0.8009,
+      "step": 3270
+    },
+    {
+      "epoch": 1.8696116850730315,
+      "grad_norm": 2.084099308100839,
+      "learning_rate": 1.954239138258679e-05,
+      "loss": 0.7997,
+      "step": 3280
+    },
+    {
+      "epoch": 1.8753117206982544,
+      "grad_norm": 2.061563413281635,
+      "learning_rate": 1.9536421164177115e-05,
+      "loss": 0.8007,
+      "step": 3290
+    },
+    {
+      "epoch": 1.8810117563234772,
+      "grad_norm": 2.029993447580369,
+      "learning_rate": 1.953041317630289e-05,
+      "loss": 0.8025,
+      "step": 3300
+    },
+    {
+      "epoch": 1.8867117919486995,
+      "grad_norm": 2.0214089034575022,
+      "learning_rate": 1.9524367442759038e-05,
+      "loss": 0.7985,
+      "step": 3310
+    },
+    {
+      "epoch": 1.8924118275739223,
+      "grad_norm": 1.935706084813715,
+      "learning_rate": 1.951828398748999e-05,
+      "loss": 0.7947,
+      "step": 3320
+    },
+    {
+      "epoch": 1.8981118631991452,
+      "grad_norm": 2.0906036255368834,
+      "learning_rate": 1.951216283458957e-05,
+      "loss": 0.7969,
+      "step": 3330
+    },
+    {
+      "epoch": 1.9038118988243675,
+      "grad_norm": 1.9338535149871947,
+      "learning_rate": 1.95060040083009e-05,
+      "loss": 0.7985,
+      "step": 3340
+    },
+    {
+      "epoch": 1.9095119344495903,
+      "grad_norm": 2.0268254380290665,
+      "learning_rate": 1.9499807533016314e-05,
+      "loss": 0.795,
+      "step": 3350
+    },
+    {
+      "epoch": 1.915211970074813,
+      "grad_norm": 1.9268507753120423,
+      "learning_rate": 1.9493573433277263e-05,
+      "loss": 0.7978,
+      "step": 3360
+    },
+    {
+      "epoch": 1.9209120057000355,
+      "grad_norm": 1.9415598197390873,
+      "learning_rate": 1.9487301733774205e-05,
+      "loss": 0.8048,
+      "step": 3370
+    },
+    {
+      "epoch": 1.9266120413252583,
+      "grad_norm": 2.15628713825871,
+      "learning_rate": 1.9480992459346506e-05,
+      "loss": 0.8002,
+      "step": 3380
+    },
+    {
+      "epoch": 1.932312076950481,
+      "grad_norm": 1.9609970358293323,
+      "learning_rate": 1.9474645634982363e-05,
+      "loss": 0.8063,
+      "step": 3390
+    },
+    {
+      "epoch": 1.9380121125757035,
+      "grad_norm": 1.926711494554939,
+      "learning_rate": 1.9468261285818686e-05,
+      "loss": 0.8002,
+      "step": 3400
+    },
+    {
+      "epoch": 1.9437121482009263,
+      "grad_norm": 2.0382657822085504,
+      "learning_rate": 1.9461839437141003e-05,
+      "loss": 0.7874,
+      "step": 3410
+    },
+    {
+      "epoch": 1.949412183826149,
+      "grad_norm": 2.025504138910519,
+      "learning_rate": 1.945538011438336e-05,
+      "loss": 0.7924,
+      "step": 3420
+    },
+    {
+      "epoch": 1.9551122194513715,
+      "grad_norm": 1.9730129265429237,
+      "learning_rate": 1.9448883343128222e-05,
+      "loss": 0.8058,
+      "step": 3430
+    },
+    {
+      "epoch": 1.9608122550765943,
+      "grad_norm": 2.053130371157395,
+      "learning_rate": 1.944234914910637e-05,
+      "loss": 0.7957,
+      "step": 3440
+    },
+    {
+      "epoch": 1.9665122907018169,
+      "grad_norm": 1.9496554034048525,
+      "learning_rate": 1.9435777558196804e-05,
+      "loss": 0.7956,
+      "step": 3450
+    },
+    {
+      "epoch": 1.9722123263270395,
+      "grad_norm": 2.017642077204114,
+      "learning_rate": 1.9429168596426635e-05,
+      "loss": 0.8089,
+      "step": 3460
+    },
+    {
+      "epoch": 1.9779123619522623,
+      "grad_norm": 2.0984701006035986,
+      "learning_rate": 1.9422522289970968e-05,
+      "loss": 0.7965,
+      "step": 3470
+    },
+    {
+      "epoch": 1.9836123975774849,
+      "grad_norm": 1.8720820855279117,
+      "learning_rate": 1.9415838665152837e-05,
+      "loss": 0.793,
+      "step": 3480
+    },
+    {
+      "epoch": 1.9893124332027075,
+      "grad_norm": 2.0863714679102254,
+      "learning_rate": 1.940911774844307e-05,
+      "loss": 0.806,
+      "step": 3490
+    },
+    {
+      "epoch": 1.9950124688279303,
+      "grad_norm": 2.004116422399095,
+      "learning_rate": 1.9402359566460175e-05,
+      "loss": 0.803,
+      "step": 3500
+    },
+    {
+      "epoch": 2.0007125044531526,
+      "grad_norm": 2.1353029003603106,
+      "learning_rate": 1.9395564145970275e-05,
+      "loss": 0.7912,
+      "step": 3510
+    },
+    {
+      "epoch": 2.0064125400783754,
+      "grad_norm": 1.9160373192820845,
+      "learning_rate": 1.9388731513886962e-05,
+      "loss": 0.772,
+      "step": 3520
+    },
+    {
+      "epoch": 2.0121125757035982,
+      "grad_norm": 2.059335935801628,
+      "learning_rate": 1.9381861697271208e-05,
+      "loss": 0.7845,
+      "step": 3530
+    },
+    {
+      "epoch": 2.0178126113288206,
+      "grad_norm": 1.9026716738700493,
+      "learning_rate": 1.9374954723331267e-05,
+      "loss": 0.7835,
+      "step": 3540
+    },
+    {
+      "epoch": 2.0235126469540434,
+      "grad_norm": 1.943020845917838,
+      "learning_rate": 1.9368010619422542e-05,
+      "loss": 0.7776,
+      "step": 3550
+    },
+    {
+      "epoch": 2.0292126825792662,
+      "grad_norm": 2.067353687772421,
+      "learning_rate": 1.93610294130475e-05,
+      "loss": 0.7837,
+      "step": 3560
+    },
+    {
+      "epoch": 2.0349127182044886,
+      "grad_norm": 2.118868836061161,
+      "learning_rate": 1.9354011131855554e-05,
+      "loss": 0.7818,
+      "step": 3570
+    },
+    {
+      "epoch": 2.0406127538297114,
+      "grad_norm": 2.1143965982692343,
+      "learning_rate": 1.934695580364295e-05,
+      "loss": 0.779,
+      "step": 3580
+    },
+    {
+      "epoch": 2.0463127894549342,
+      "grad_norm": 2.040166299899192,
+      "learning_rate": 1.9339863456352658e-05,
+      "loss": 0.7794,
+      "step": 3590
+    },
+    {
+      "epoch": 2.0520128250801566,
+      "grad_norm": 2.070863013108917,
+      "learning_rate": 1.9332734118074274e-05,
+      "loss": 0.7741,
+      "step": 3600
+    },
+    {
+      "epoch": 2.0577128607053794,
+      "grad_norm": 2.165533789996265,
+      "learning_rate": 1.9325567817043888e-05,
+      "loss": 0.7891,
+      "step": 3610
+    },
+    {
+      "epoch": 2.063412896330602,
+      "grad_norm": 2.0639753331696435,
+      "learning_rate": 1.931836458164399e-05,
+      "loss": 0.7745,
+      "step": 3620
+    },
+    {
+      "epoch": 2.0691129319558246,
+      "grad_norm": 2.2393200256888615,
+      "learning_rate": 1.9311124440403347e-05,
+      "loss": 0.7744,
+      "step": 3630
+    },
+    {
+      "epoch": 2.0748129675810474,
+      "grad_norm": 2.245493711349147,
+      "learning_rate": 1.9303847421996895e-05,
+      "loss": 0.7846,
+      "step": 3640
+    },
+    {
+      "epoch": 2.08051300320627,
+      "grad_norm": 2.12712542521556,
+      "learning_rate": 1.929653355524562e-05,
+      "loss": 0.7767,
+      "step": 3650
+    },
+    {
+      "epoch": 2.0862130388314926,
+      "grad_norm": 2.2156281019048354,
+      "learning_rate": 1.928918286911645e-05,
+      "loss": 0.7786,
+      "step": 3660
+    },
+    {
+      "epoch": 2.0919130744567154,
+      "grad_norm": 1.991829895411575,
+      "learning_rate": 1.9281795392722146e-05,
+      "loss": 0.788,
+      "step": 3670
+    },
+    {
+      "epoch": 2.097613110081938,
+      "grad_norm": 2.017400209394898,
+      "learning_rate": 1.9274371155321167e-05,
+      "loss": 0.7828,
+      "step": 3680
+    },
+    {
+      "epoch": 2.1033131457071605,
+      "grad_norm": 2.050803561945968,
+      "learning_rate": 1.9266910186317566e-05,
+      "loss": 0.784,
+      "step": 3690
+    },
+    {
+      "epoch": 2.1090131813323834,
+      "grad_norm": 1.9316917498870407,
+      "learning_rate": 1.925941251526088e-05,
+      "loss": 0.7681,
+      "step": 3700
+    },
+    {
+      "epoch": 2.114713216957606,
+      "grad_norm": 1.9848496136357934,
+      "learning_rate": 1.9251878171846008e-05,
+      "loss": 0.7695,
+      "step": 3710
+    },
+    {
+      "epoch": 2.1204132525828285,
+      "grad_norm": 2.071110017635957,
+      "learning_rate": 1.924430718591308e-05,
+      "loss": 0.7841,
+      "step": 3720
+    },
+    {
+      "epoch": 2.1261132882080513,
+      "grad_norm": 1.9736490227320278,
+      "learning_rate": 1.9236699587447363e-05,
+      "loss": 0.7768,
+      "step": 3730
+    },
+    {
+      "epoch": 2.131813323833274,
+      "grad_norm": 1.9869752975975226,
+      "learning_rate": 1.922905540657912e-05,
+      "loss": 0.7785,
+      "step": 3740
+    },
+    {
+      "epoch": 2.1375133594584965,
+      "grad_norm": 2.0695040821912647,
+      "learning_rate": 1.922137467358351e-05,
+      "loss": 0.7761,
+      "step": 3750
+    },
+    {
+      "epoch": 2.1432133950837193,
+      "grad_norm": 1.9299059979513908,
+      "learning_rate": 1.921365741888045e-05,
+      "loss": 0.7856,
+      "step": 3760
+    },
+    {
+      "epoch": 2.1489134307089417,
+      "grad_norm": 2.0076180790761153,
+      "learning_rate": 1.920590367303451e-05,
+      "loss": 0.7807,
+      "step": 3770
+    },
+    {
+      "epoch": 2.1546134663341645,
+      "grad_norm": 1.968046382602847,
+      "learning_rate": 1.9198113466754775e-05,
+      "loss": 0.7772,
+      "step": 3780
+    },
+    {
+      "epoch": 2.1603135019593873,
+      "grad_norm": 2.0927046946378285,
+      "learning_rate": 1.9190286830894744e-05,
+      "loss": 0.7753,
+      "step": 3790
+    },
+    {
+      "epoch": 2.16601353758461,
+      "grad_norm": 2.1166596531741004,
+      "learning_rate": 1.9182423796452196e-05,
+      "loss": 0.782,
+      "step": 3800
+    },
+    {
+      "epoch": 2.1717135732098325,
+      "grad_norm": 2.112843242331277,
+      "learning_rate": 1.9174524394569058e-05,
+      "loss": 0.7792,
+      "step": 3810
+    },
+    {
+      "epoch": 2.1774136088350553,
+      "grad_norm": 2.1683506945600723,
+      "learning_rate": 1.9166588656531305e-05,
+      "loss": 0.7726,
+      "step": 3820
+    },
+    {
+      "epoch": 2.1831136444602777,
+      "grad_norm": 2.054735148916789,
+      "learning_rate": 1.9158616613768812e-05,
+      "loss": 0.7743,
+      "step": 3830
+    },
+    {
+      "epoch": 2.1888136800855005,
+      "grad_norm": 2.1344363416734846,
+      "learning_rate": 1.915060829785525e-05,
+      "loss": 0.7771,
+      "step": 3840
+    },
+    {
+      "epoch": 2.1945137157107233,
+      "grad_norm": 1.984560409021633,
+      "learning_rate": 1.914256374050795e-05,
+      "loss": 0.778,
+      "step": 3850
+    },
+    {
+      "epoch": 2.2002137513359457,
+      "grad_norm": 1.972160935139131,
+      "learning_rate": 1.9134482973587773e-05,
+      "loss": 0.7782,
+      "step": 3860
+    },
+    {
+      "epoch": 2.2059137869611685,
+      "grad_norm": 1.9214738024805038,
+      "learning_rate": 1.912636602909899e-05,
+      "loss": 0.7771,
+      "step": 3870
+    },
+    {
+      "epoch": 2.2116138225863913,
+      "grad_norm": 1.9622083874450693,
+      "learning_rate": 1.9118212939189165e-05,
+      "loss": 0.7839,
+      "step": 3880
+    },
+    {
+      "epoch": 2.2173138582116136,
+      "grad_norm": 2.0208665741876035,
+      "learning_rate": 1.9110023736149007e-05,
+      "loss": 0.7681,
+      "step": 3890
+    },
+    {
+      "epoch": 2.2230138938368365,
+      "grad_norm": 2.226589059621567,
+      "learning_rate": 1.910179845241226e-05,
+      "loss": 0.779,
+      "step": 3900
+    },
+    {
+      "epoch": 2.2287139294620593,
+      "grad_norm": 2.1811770256146805,
+      "learning_rate": 1.9093537120555564e-05,
+      "loss": 0.7811,
+      "step": 3910
+    },
+    {
+      "epoch": 2.2344139650872816,
+      "grad_norm": 2.1292841346824325,
+      "learning_rate": 1.9085239773298324e-05,
+      "loss": 0.7859,
+      "step": 3920
+    },
+    {
+      "epoch": 2.2401140007125044,
+      "grad_norm": 1.9822686380206471,
+      "learning_rate": 1.9076906443502602e-05,
+      "loss": 0.7673,
+      "step": 3930
+    },
+    {
+      "epoch": 2.2458140363377272,
+      "grad_norm": 2.0426647477884043,
+      "learning_rate": 1.906853716417295e-05,
+      "loss": 0.7844,
+      "step": 3940
+    },
+    {
+      "epoch": 2.2515140719629496,
+      "grad_norm": 2.3068456301695677,
+      "learning_rate": 1.906013196845631e-05,
+      "loss": 0.7751,
+      "step": 3950
+    },
+    {
+      "epoch": 2.2572141075881724,
+      "grad_norm": 1.9668105468661568,
+      "learning_rate": 1.9051690889641884e-05,
+      "loss": 0.7792,
+      "step": 3960
+    },
+    {
+      "epoch": 2.2629141432133952,
+      "grad_norm": 2.0782727922468456,
+      "learning_rate": 1.904321396116097e-05,
+      "loss": 0.7707,
+      "step": 3970
+    },
+    {
+      "epoch": 2.2686141788386176,
+      "grad_norm": 2.0489858306617816,
+      "learning_rate": 1.903470121658686e-05,
+      "loss": 0.7848,
+      "step": 3980
+    },
+    {
+      "epoch": 2.2743142144638404,
+      "grad_norm": 2.0742582452117206,
+      "learning_rate": 1.90261526896347e-05,
+      "loss": 0.7766,
+      "step": 3990
+    },
+    {
+      "epoch": 2.280014250089063,
+      "grad_norm": 2.0272360660939164,
+      "learning_rate": 1.901756841416135e-05,
+      "loss": 0.7793,
+      "step": 4000
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 2.3308634299194333,
+      "learning_rate": 1.900894842416525e-05,
+      "loss": 0.7686,
+      "step": 4010
+    },
+    {
+      "epoch": 2.2914143213395084,
+      "grad_norm": 1.9994773525348128,
+      "learning_rate": 1.9000292753786305e-05,
+      "loss": 0.7725,
+      "step": 4020
+    },
+    {
+      "epoch": 2.297114356964731,
+      "grad_norm": 2.2336189138329114,
+      "learning_rate": 1.8991601437305715e-05,
+      "loss": 0.7773,
+      "step": 4030
+    },
+    {
+      "epoch": 2.3028143925899536,
+      "grad_norm": 2.031657276083233,
+      "learning_rate": 1.8982874509145866e-05,
+      "loss": 0.7754,
+      "step": 4040
+    },
+    {
+      "epoch": 2.3085144282151764,
+      "grad_norm": 2.216895543159883,
+      "learning_rate": 1.8974112003870186e-05,
+      "loss": 0.7761,
+      "step": 4050
+    },
+    {
+      "epoch": 2.314214463840399,
+      "grad_norm": 2.209356529326098,
+      "learning_rate": 1.896531395618301e-05,
+      "loss": 0.7704,
+      "step": 4060
+    },
+    {
+      "epoch": 2.3199144994656216,
+      "grad_norm": 2.156183494015324,
+      "learning_rate": 1.8956480400929438e-05,
+      "loss": 0.7787,
+      "step": 4070
+    },
+    {
+      "epoch": 2.3256145350908444,
+      "grad_norm": 1.984794235924156,
+      "learning_rate": 1.8947611373095196e-05,
+      "loss": 0.7753,
+      "step": 4080
+    },
+    {
+      "epoch": 2.331314570716067,
+      "grad_norm": 1.9518963935475173,
+      "learning_rate": 1.893870690780651e-05,
+      "loss": 0.7809,
+      "step": 4090
+    },
+    {
+      "epoch": 2.3370146063412895,
+      "grad_norm": 2.033804508497074,
+      "learning_rate": 1.892976704032994e-05,
+      "loss": 0.7716,
+      "step": 4100
+    },
+    {
+      "epoch": 2.3427146419665124,
+      "grad_norm": 2.068960781647104,
+      "learning_rate": 1.892079180607229e-05,
+      "loss": 0.7754,
+      "step": 4110
+    },
+    {
+      "epoch": 2.3484146775917347,
+      "grad_norm": 2.077601045276333,
+      "learning_rate": 1.8911781240580402e-05,
+      "loss": 0.7761,
+      "step": 4120
+    },
+    {
+      "epoch": 2.3541147132169575,
+      "grad_norm": 2.0915913435617743,
+      "learning_rate": 1.8902735379541064e-05,
+      "loss": 0.7685,
+      "step": 4130
+    },
+    {
+      "epoch": 2.3598147488421803,
+      "grad_norm": 2.0680783097084112,
+      "learning_rate": 1.889365425878086e-05,
+      "loss": 0.7799,
+      "step": 4140
+    },
+    {
+      "epoch": 2.365514784467403,
+      "grad_norm": 1.9963116614489376,
+      "learning_rate": 1.888453791426601e-05,
+      "loss": 0.7735,
+      "step": 4150
+    },
+    {
+      "epoch": 2.3712148200926255,
+      "grad_norm": 2.1464671268790934,
+      "learning_rate": 1.8875386382102245e-05,
+      "loss": 0.7718,
+      "step": 4160
+    },
+    {
+      "epoch": 2.3769148557178483,
+      "grad_norm": 2.175040646679394,
+      "learning_rate": 1.8866199698534658e-05,
+      "loss": 0.7788,
+      "step": 4170
+    },
+    {
+      "epoch": 2.3826148913430707,
+      "grad_norm": 2.0706867787823096,
+      "learning_rate": 1.885697789994756e-05,
+      "loss": 0.7627,
+      "step": 4180
+    },
+    {
+      "epoch": 2.3883149269682935,
+      "grad_norm": 2.194047609109707,
+      "learning_rate": 1.8847721022864336e-05,
+      "loss": 0.7793,
+      "step": 4190
+    },
+    {
+      "epoch": 2.3940149625935163,
+      "grad_norm": 2.122030501728222,
+      "learning_rate": 1.883842910394731e-05,
+      "loss": 0.7683,
+      "step": 4200
+    },
+    {
+      "epoch": 2.3997149982187387,
+      "grad_norm": 2.023047252118629,
+      "learning_rate": 1.8829102179997572e-05,
+      "loss": 0.7831,
+      "step": 4210
+    },
+    {
+      "epoch": 2.4054150338439615,
+      "grad_norm": 2.0437679196020393,
+      "learning_rate": 1.8819740287954876e-05,
+      "loss": 0.7695,
+      "step": 4220
+    },
+    {
+      "epoch": 2.4111150694691843,
+      "grad_norm": 2.0799095278671493,
+      "learning_rate": 1.881034346489744e-05,
+      "loss": 0.7665,
+      "step": 4230
+    },
+    {
+      "epoch": 2.4168151050944067,
+      "grad_norm": 2.1746368931589086,
+      "learning_rate": 1.880091174804186e-05,
+      "loss": 0.7612,
+      "step": 4240
+    },
+    {
+      "epoch": 2.4225151407196295,
+      "grad_norm": 2.282337806044003,
+      "learning_rate": 1.8791445174742894e-05,
+      "loss": 0.7766,
+      "step": 4250
+    },
+    {
+      "epoch": 2.4282151763448523,
+      "grad_norm": 2.071429297247938,
+      "learning_rate": 1.8781943782493392e-05,
+      "loss": 0.7721,
+      "step": 4260
+    },
+    {
+      "epoch": 2.4339152119700747,
+      "grad_norm": 2.0230879554093217,
+      "learning_rate": 1.8772407608924067e-05,
+      "loss": 0.7741,
+      "step": 4270
+    },
+    {
+      "epoch": 2.4396152475952975,
+      "grad_norm": 2.061387786789968,
+      "learning_rate": 1.8762836691803417e-05,
+      "loss": 0.7789,
+      "step": 4280
+    },
+    {
+      "epoch": 2.4453152832205203,
+      "grad_norm": 2.035005479237207,
+      "learning_rate": 1.8753231069037522e-05,
+      "loss": 0.7741,
+      "step": 4290
+    },
+    {
+      "epoch": 2.4510153188457426,
+      "grad_norm": 2.0348766862238947,
+      "learning_rate": 1.874359077866992e-05,
+      "loss": 0.7697,
+      "step": 4300
+    },
+    {
+      "epoch": 2.4567153544709655,
+      "grad_norm": 2.071413736213706,
+      "learning_rate": 1.8733915858881462e-05,
+      "loss": 0.7724,
+      "step": 4310
+    },
+    {
+      "epoch": 2.4624153900961883,
+      "grad_norm": 2.2246204070019595,
+      "learning_rate": 1.872420634799014e-05,
+      "loss": 0.7733,
+      "step": 4320
+    },
+    {
+      "epoch": 2.4681154257214106,
+      "grad_norm": 2.0913317389738153,
+      "learning_rate": 1.8714462284450948e-05,
+      "loss": 0.7668,
+      "step": 4330
+    },
+    {
+      "epoch": 2.4738154613466334,
+      "grad_norm": 2.143431885380468,
+      "learning_rate": 1.8704683706855728e-05,
+      "loss": 0.7758,
+      "step": 4340
+    },
+    {
+      "epoch": 2.4795154969718562,
+      "grad_norm": 2.0657978163391983,
+      "learning_rate": 1.869487065393302e-05,
+      "loss": 0.7719,
+      "step": 4350
+    },
+    {
+      "epoch": 2.4852155325970786,
+      "grad_norm": 2.2092417572395195,
+      "learning_rate": 1.86850231645479e-05,
+      "loss": 0.7751,
+      "step": 4360
+    },
+    {
+      "epoch": 2.4909155682223014,
+      "grad_norm": 2.0194258303784953,
+      "learning_rate": 1.8675141277701834e-05,
+      "loss": 0.7736,
+      "step": 4370
+    },
+    {
+      "epoch": 2.4966156038475242,
+      "grad_norm": 2.206101479863975,
+      "learning_rate": 1.866522503253252e-05,
+      "loss": 0.7672,
+      "step": 4380
+    },
+    {
+      "epoch": 2.5023156394727466,
+      "grad_norm": 2.195672996869424,
+      "learning_rate": 1.8655274468313732e-05,
+      "loss": 0.7691,
+      "step": 4390
+    },
+    {
+      "epoch": 2.5080156750979694,
+      "grad_norm": 2.068463197978191,
+      "learning_rate": 1.8645289624455175e-05,
+      "loss": 0.7696,
+      "step": 4400
+    },
+    {
+      "epoch": 2.5137157107231918,
+      "grad_norm": 2.0587484850743194,
+      "learning_rate": 1.8635270540502307e-05,
+      "loss": 0.7646,
+      "step": 4410
+    },
+    {
+      "epoch": 2.5194157463484146,
+      "grad_norm": 2.06954821802478,
+      "learning_rate": 1.8625217256136206e-05,
+      "loss": 0.7711,
+      "step": 4420
+    },
+    {
+      "epoch": 2.5251157819736374,
+      "grad_norm": 1.9723121174345877,
+      "learning_rate": 1.8615129811173398e-05,
+      "loss": 0.7805,
+      "step": 4430
+    },
+    {
+      "epoch": 2.53081581759886,
+      "grad_norm": 2.0838359377420193,
+      "learning_rate": 1.8605008245565704e-05,
+      "loss": 0.7732,
+      "step": 4440
+    },
+    {
+      "epoch": 2.5365158532240826,
+      "grad_norm": 1.995626097636077,
+      "learning_rate": 1.8594852599400083e-05,
+      "loss": 0.7645,
+      "step": 4450
+    },
+    {
+      "epoch": 2.5422158888493054,
+      "grad_norm": 2.0491043104993194,
+      "learning_rate": 1.8584662912898464e-05,
+      "loss": 0.7618,
+      "step": 4460
+    },
+    {
+      "epoch": 2.5479159244745277,
+      "grad_norm": 2.066323954109494,
+      "learning_rate": 1.857443922641761e-05,
+      "loss": 0.7721,
+      "step": 4470
+    },
+    {
+      "epoch": 2.5536159600997506,
+      "grad_norm": 1.978390527887617,
+      "learning_rate": 1.856418158044893e-05,
+      "loss": 0.7761,
+      "step": 4480
+    },
+    {
+      "epoch": 2.5593159957249734,
+      "grad_norm": 2.0613022758343122,
+      "learning_rate": 1.8553890015618333e-05,
+      "loss": 0.7617,
+      "step": 4490
+    },
+    {
+      "epoch": 2.565016031350196,
+      "grad_norm": 1.9964157030100178,
+      "learning_rate": 1.8543564572686072e-05,
+      "loss": 0.7691,
+      "step": 4500
+    },
+    {
+      "epoch": 2.5707160669754185,
+      "grad_norm": 2.0340098122596255,
+      "learning_rate": 1.8533205292546567e-05,
+      "loss": 0.7651,
+      "step": 4510
+    },
+    {
+      "epoch": 2.5764161026006414,
+      "grad_norm": 1.9963620850267063,
+      "learning_rate": 1.8522812216228254e-05,
+      "loss": 0.7706,
+      "step": 4520
+    },
+    {
+      "epoch": 2.5821161382258637,
+      "grad_norm": 2.107374755843351,
+      "learning_rate": 1.851238538489343e-05,
+      "loss": 0.7778,
+      "step": 4530
+    },
+    {
+      "epoch": 2.5878161738510865,
+      "grad_norm": 2.2279853889692656,
+      "learning_rate": 1.8501924839838062e-05,
+      "loss": 0.7698,
+      "step": 4540
+    },
+    {
+      "epoch": 2.5935162094763093,
+      "grad_norm": 2.057671470082954,
+      "learning_rate": 1.8491430622491665e-05,
+      "loss": 0.7605,
+      "step": 4550
+    },
+    {
+      "epoch": 2.599216245101532,
+      "grad_norm": 2.169407311007384,
+      "learning_rate": 1.8480902774417094e-05,
+      "loss": 0.7661,
+      "step": 4560
+    },
+    {
+      "epoch": 2.6049162807267545,
+      "grad_norm": 2.153643481958755,
+      "learning_rate": 1.8470341337310407e-05,
+      "loss": 0.7619,
+      "step": 4570
+    },
+    {
+      "epoch": 2.6106163163519773,
+      "grad_norm": 2.0441154909224064,
+      "learning_rate": 1.8459746353000704e-05,
+      "loss": 0.7615,
+      "step": 4580
+    },
+    {
+      "epoch": 2.6163163519771997,
+      "grad_norm": 2.1378232482633157,
+      "learning_rate": 1.8449117863449932e-05,
+      "loss": 0.7688,
+      "step": 4590
+    },
+    {
+      "epoch": 2.6220163876024225,
+      "grad_norm": 2.019999621488862,
+      "learning_rate": 1.843845591075275e-05,
+      "loss": 0.7682,
+      "step": 4600
+    },
+    {
+      "epoch": 2.6277164232276453,
+      "grad_norm": 2.0517533117948568,
+      "learning_rate": 1.8427760537136342e-05,
+      "loss": 0.7576,
+      "step": 4610
+    },
+    {
+      "epoch": 2.6334164588528677,
+      "grad_norm": 2.05638099267677,
+      "learning_rate": 1.8417031784960267e-05,
+      "loss": 0.7655,
+      "step": 4620
+    },
+    {
+      "epoch": 2.6391164944780905,
+      "grad_norm": 2.1453529544420373,
+      "learning_rate": 1.840626969671627e-05,
+      "loss": 0.7752,
+      "step": 4630
+    },
+    {
+      "epoch": 2.6448165301033133,
+      "grad_norm": 1.9624857487175413,
+      "learning_rate": 1.8395474315028134e-05,
+      "loss": 0.774,
+      "step": 4640
+    },
+    {
+      "epoch": 2.6505165657285357,
+      "grad_norm": 1.9796322862726417,
+      "learning_rate": 1.838464568265149e-05,
+      "loss": 0.7722,
+      "step": 4650
+    },
+    {
+      "epoch": 2.6562166013537585,
+      "grad_norm": 2.0862799346516905,
+      "learning_rate": 1.837378384247368e-05,
+      "loss": 0.7705,
+      "step": 4660
+    },
+    {
+      "epoch": 2.6619166369789813,
+      "grad_norm": 2.0408955441271974,
+      "learning_rate": 1.8362888837513548e-05,
+      "loss": 0.7633,
+      "step": 4670
+    },
+    {
+      "epoch": 2.6676166726042037,
+      "grad_norm": 2.135018499958134,
+      "learning_rate": 1.83519607109213e-05,
+      "loss": 0.7612,
+      "step": 4680
+    },
+    {
+      "epoch": 2.6733167082294265,
+      "grad_norm": 2.049209470183393,
+      "learning_rate": 1.834099950597832e-05,
+      "loss": 0.7627,
+      "step": 4690
+    },
+    {
+      "epoch": 2.679016743854649,
+      "grad_norm": 2.0472546095057553,
+      "learning_rate": 1.8330005266096992e-05,
+      "loss": 0.7661,
+      "step": 4700
+    },
+    {
+      "epoch": 2.6847167794798716,
+      "grad_norm": 2.0189392429541306,
+      "learning_rate": 1.8318978034820544e-05,
+      "loss": 0.7581,
+      "step": 4710
+    },
+    {
+      "epoch": 2.6904168151050945,
+      "grad_norm": 2.1622485501687256,
+      "learning_rate": 1.830791785582288e-05,
+      "loss": 0.7629,
+      "step": 4720
+    },
+    {
+      "epoch": 2.6961168507303173,
+      "grad_norm": 2.006323646975851,
+      "learning_rate": 1.8296824772908365e-05,
+      "loss": 0.7625,
+      "step": 4730
+    },
+    {
+      "epoch": 2.7018168863555396,
+      "grad_norm": 1.9594010919164644,
+      "learning_rate": 1.828569883001171e-05,
+      "loss": 0.7667,
+      "step": 4740
+    },
+    {
+      "epoch": 2.7075169219807624,
+      "grad_norm": 2.1349030657070474,
+      "learning_rate": 1.827454007119775e-05,
+      "loss": 0.7657,
+      "step": 4750
+    },
+    {
+      "epoch": 2.713216957605985,
+      "grad_norm": 2.0990532850307084,
+      "learning_rate": 1.8263348540661306e-05,
+      "loss": 0.7691,
+      "step": 4760
+    },
+    {
+      "epoch": 2.7189169932312076,
+      "grad_norm": 2.1746655424409345,
+      "learning_rate": 1.8252124282726984e-05,
+      "loss": 0.7635,
+      "step": 4770
+    },
+    {
+      "epoch": 2.7246170288564304,
+      "grad_norm": 2.152646261140533,
+      "learning_rate": 1.8240867341849e-05,
+      "loss": 0.7581,
+      "step": 4780
+    },
+    {
+      "epoch": 2.7303170644816532,
+      "grad_norm": 2.174162632472622,
+      "learning_rate": 1.8229577762611033e-05,
+      "loss": 0.7689,
+      "step": 4790
+    },
+    {
+      "epoch": 2.7360171001068756,
+      "grad_norm": 1.9316453580875002,
+      "learning_rate": 1.8218255589726007e-05,
+      "loss": 0.7546,
+      "step": 4800
+    },
+    {
+      "epoch": 2.7417171357320984,
+      "grad_norm": 2.1886515928391352,
+      "learning_rate": 1.820690086803595e-05,
+      "loss": 0.7674,
+      "step": 4810
+    },
+    {
+      "epoch": 2.7474171713573208,
+      "grad_norm": 2.052737570252945,
+      "learning_rate": 1.819551364251179e-05,
+      "loss": 0.7622,
+      "step": 4820
+    },
+    {
+      "epoch": 2.7531172069825436,
+      "grad_norm": 2.1180173414843906,
+      "learning_rate": 1.81840939582532e-05,
+      "loss": 0.7542,
+      "step": 4830
+    },
+    {
+      "epoch": 2.7588172426077664,
+      "grad_norm": 2.0722878980296127,
+      "learning_rate": 1.8172641860488393e-05,
+      "loss": 0.7626,
+      "step": 4840
+    },
+    {
+      "epoch": 2.764517278232989,
+      "grad_norm": 2.0043370083102685,
+      "learning_rate": 1.816115739457397e-05,
+      "loss": 0.7671,
+      "step": 4850
+    },
+    {
+      "epoch": 2.7702173138582116,
+      "grad_norm": 1.9835911307540945,
+      "learning_rate": 1.8149640605994722e-05,
+      "loss": 0.7722,
+      "step": 4860
+    },
+    {
+      "epoch": 2.7759173494834344,
+      "grad_norm": 1.9879722805165094,
+      "learning_rate": 1.8138091540363453e-05,
+      "loss": 0.7666,
+      "step": 4870
+    },
+    {
+      "epoch": 2.7816173851086567,
+      "grad_norm": 2.0295425663572026,
+      "learning_rate": 1.8126510243420807e-05,
+      "loss": 0.7553,
+      "step": 4880
+    },
+    {
+      "epoch": 2.7873174207338796,
+      "grad_norm": 2.146037273264629,
+      "learning_rate": 1.811489676103508e-05,
+      "loss": 0.7649,
+      "step": 4890
+    },
+    {
+      "epoch": 2.7930174563591024,
+      "grad_norm": 2.029168818823305,
+      "learning_rate": 1.8103251139202038e-05,
+      "loss": 0.7632,
+      "step": 4900
+    },
+    {
+      "epoch": 2.798717491984325,
+      "grad_norm": 2.0772701412050933,
+      "learning_rate": 1.8091573424044742e-05,
+      "loss": 0.7661,
+      "step": 4910
+    },
+    {
+      "epoch": 2.8044175276095475,
+      "grad_norm": 1.9717828970651405,
+      "learning_rate": 1.8079863661813352e-05,
+      "loss": 0.7709,
+      "step": 4920
+    },
+    {
+      "epoch": 2.8101175632347704,
+      "grad_norm": 2.174183368464108,
+      "learning_rate": 1.8068121898884955e-05,
+      "loss": 0.756,
+      "step": 4930
+    },
+    {
+      "epoch": 2.8158175988599927,
+      "grad_norm": 2.1486780441649547,
+      "learning_rate": 1.8056348181763387e-05,
+      "loss": 0.7537,
+      "step": 4940
+    },
+    {
+      "epoch": 2.8215176344852155,
+      "grad_norm": 1.9709706612211082,
+      "learning_rate": 1.8044542557079032e-05,
+      "loss": 0.763,
+      "step": 4950
+    },
+    {
+      "epoch": 2.8272176701104383,
+      "grad_norm": 2.044406248858559,
+      "learning_rate": 1.8032705071588638e-05,
+      "loss": 0.7667,
+      "step": 4960
+    },
+    {
+      "epoch": 2.8329177057356607,
+      "grad_norm": 1.9551236687248852,
+      "learning_rate": 1.8020835772175158e-05,
+      "loss": 0.7632,
+      "step": 4970
+    },
+    {
+      "epoch": 2.8386177413608835,
+      "grad_norm": 1.9293855795874906,
+      "learning_rate": 1.8008934705847533e-05,
+      "loss": 0.7636,
+      "step": 4980
+    },
+    {
+      "epoch": 2.8443177769861063,
+      "grad_norm": 2.158009404352588,
+      "learning_rate": 1.7997001919740514e-05,
+      "loss": 0.7606,
+      "step": 4990
+    },
+    {
+      "epoch": 2.8500178126113287,
+      "grad_norm": 2.124078613588665,
+      "learning_rate": 1.7985037461114497e-05,
+      "loss": 0.7615,
+      "step": 5000
+    },
+    {
+      "epoch": 2.8557178482365515,
+      "grad_norm": 2.021489305585605,
+      "learning_rate": 1.7973041377355303e-05,
+      "loss": 0.7549,
+      "step": 5010
+    },
+    {
+      "epoch": 2.8614178838617743,
+      "grad_norm": 2.060442251766284,
+      "learning_rate": 1.7961013715974008e-05,
+      "loss": 0.7661,
+      "step": 5020
+    },
+    {
+      "epoch": 2.8671179194869967,
+      "grad_norm": 2.144181800187938,
+      "learning_rate": 1.7948954524606764e-05,
+      "loss": 0.756,
+      "step": 5030
+    },
+    {
+      "epoch": 2.8728179551122195,
+      "grad_norm": 2.01230696124865,
+      "learning_rate": 1.7936863851014585e-05,
+      "loss": 0.7624,
+      "step": 5040
+    },
+    {
+      "epoch": 2.878517990737442,
+      "grad_norm": 1.9818525793021162,
+      "learning_rate": 1.7924741743083177e-05,
+      "loss": 0.7731,
+      "step": 5050
+    },
+    {
+      "epoch": 2.8842180263626647,
+      "grad_norm": 1.9797990283611266,
+      "learning_rate": 1.7912588248822744e-05,
+      "loss": 0.7478,
+      "step": 5060
+    },
+    {
+      "epoch": 2.8899180619878875,
+      "grad_norm": 1.918218840573413,
+      "learning_rate": 1.79004034163678e-05,
+      "loss": 0.7632,
+      "step": 5070
+    },
+    {
+      "epoch": 2.8956180976131103,
+      "grad_norm": 2.2113523658613956,
+      "learning_rate": 1.7888187293976974e-05,
+      "loss": 0.7634,
+      "step": 5080
+    },
+    {
+      "epoch": 2.9013181332383327,
+      "grad_norm": 2.0564005525418674,
+      "learning_rate": 1.7875939930032817e-05,
+      "loss": 0.7689,
+      "step": 5090
+    },
+    {
+      "epoch": 2.9070181688635555,
+      "grad_norm": 2.111212825140381,
+      "learning_rate": 1.786366137304161e-05,
+      "loss": 0.7721,
+      "step": 5100
+    },
+    {
+      "epoch": 2.912718204488778,
+      "grad_norm": 2.0529909877524974,
+      "learning_rate": 1.7851351671633192e-05,
+      "loss": 0.7583,
+      "step": 5110
+    },
+    {
+      "epoch": 2.9184182401140006,
+      "grad_norm": 1.9683300396637016,
+      "learning_rate": 1.7839010874560732e-05,
+      "loss": 0.7587,
+      "step": 5120
+    },
+    {
+      "epoch": 2.9241182757392234,
+      "grad_norm": 2.1382803877159113,
+      "learning_rate": 1.782663903070057e-05,
+      "loss": 0.7571,
+      "step": 5130
+    },
+    {
+      "epoch": 2.9298183113644463,
+      "grad_norm": 1.9734164028059396,
+      "learning_rate": 1.7814236189051995e-05,
+      "loss": 0.7591,
+      "step": 5140
+    },
+    {
+      "epoch": 2.9355183469896686,
+      "grad_norm": 1.9907600950849165,
+      "learning_rate": 1.780180239873707e-05,
+      "loss": 0.7583,
+      "step": 5150
+    },
+    {
+      "epoch": 2.9412183826148914,
+      "grad_norm": 2.0056029041693,
+      "learning_rate": 1.7789337709000435e-05,
+      "loss": 0.7664,
+      "step": 5160
+    },
+    {
+      "epoch": 2.946918418240114,
+      "grad_norm": 2.0284767043959673,
+      "learning_rate": 1.777684216920911e-05,
+      "loss": 0.7551,
+      "step": 5170
+    },
+    {
+      "epoch": 2.9526184538653366,
+      "grad_norm": 2.0770316896500054,
+      "learning_rate": 1.776431582885229e-05,
+      "loss": 0.7574,
+      "step": 5180
+    },
+    {
+      "epoch": 2.9583184894905594,
+      "grad_norm": 2.101104699087831,
+      "learning_rate": 1.775175873754116e-05,
+      "loss": 0.7648,
+      "step": 5190
+    },
+    {
+      "epoch": 2.9640185251157822,
+      "grad_norm": 2.132674479499415,
+      "learning_rate": 1.77391709450087e-05,
+      "loss": 0.7638,
+      "step": 5200
+    },
+    {
+      "epoch": 2.9697185607410046,
+      "grad_norm": 2.018973293495472,
+      "learning_rate": 1.772655250110948e-05,
+      "loss": 0.7627,
+      "step": 5210
+    },
+    {
+      "epoch": 2.9754185963662274,
+      "grad_norm": 1.9980958508890732,
+      "learning_rate": 1.771390345581947e-05,
+      "loss": 0.7736,
+      "step": 5220
+    },
+    {
+      "epoch": 2.9811186319914498,
+      "grad_norm": 2.1161985121549605,
+      "learning_rate": 1.7701223859235828e-05,
+      "loss": 0.7565,
+      "step": 5230
+    },
+    {
+      "epoch": 2.9868186676166726,
+      "grad_norm": 2.258396400625071,
+      "learning_rate": 1.7688513761576726e-05,
+      "loss": 0.7582,
+      "step": 5240
+    },
+    {
+      "epoch": 2.9925187032418954,
+      "grad_norm": 2.2236728251205875,
+      "learning_rate": 1.7675773213181124e-05,
+      "loss": 0.7639,
+      "step": 5250
+    },
+    {
+      "epoch": 2.9982187388671178,
+      "grad_norm": 2.0402341818317917,
+      "learning_rate": 1.7663002264508598e-05,
+      "loss": 0.7658,
+      "step": 5260
+    },
+    {
+      "epoch": 3.0039187744923406,
+      "grad_norm": 2.1178170815499,
+      "learning_rate": 1.765020096613911e-05,
+      "loss": 0.7438,
+      "step": 5270
+    },
+    {
+      "epoch": 3.0096188101175634,
+      "grad_norm": 2.053090963514229,
+      "learning_rate": 1.763736936877284e-05,
+      "loss": 0.7439,
+      "step": 5280
+    },
+    {
+      "epoch": 3.0153188457427857,
+      "grad_norm": 2.276361717966091,
+      "learning_rate": 1.762450752322995e-05,
+      "loss": 0.7484,
+      "step": 5290
+    },
+    {
+      "epoch": 3.0210188813680086,
+      "grad_norm": 2.0728033921919606,
+      "learning_rate": 1.7611615480450413e-05,
+      "loss": 0.7481,
+      "step": 5300
+    },
+    {
+      "epoch": 3.0267189169932314,
+      "grad_norm": 2.031463911733619,
+      "learning_rate": 1.7598693291493804e-05,
+      "loss": 0.7517,
+      "step": 5310
+    },
+    {
+      "epoch": 3.0324189526184537,
+      "grad_norm": 2.0727239453344173,
+      "learning_rate": 1.7585741007539083e-05,
+      "loss": 0.7393,
+      "step": 5320
+    },
+    {
+      "epoch": 3.0381189882436765,
+      "grad_norm": 2.245371926184325,
+      "learning_rate": 1.7572758679884406e-05,
+      "loss": 0.7366,
+      "step": 5330
+    },
+    {
+      "epoch": 3.0438190238688994,
+      "grad_norm": 2.081261044660946,
+      "learning_rate": 1.7559746359946925e-05,
+      "loss": 0.7449,
+      "step": 5340
+    },
+    {
+      "epoch": 3.0495190594941217,
+      "grad_norm": 2.0012713994421487,
+      "learning_rate": 1.7546704099262565e-05,
+      "loss": 0.7397,
+      "step": 5350
+    },
+    {
+      "epoch": 3.0552190951193445,
+      "grad_norm": 2.1632139989474886,
+      "learning_rate": 1.7533631949485847e-05,
+      "loss": 0.7436,
+      "step": 5360
+    },
+    {
+      "epoch": 3.0609191307445673,
+      "grad_norm": 2.112371468079618,
+      "learning_rate": 1.7520529962389655e-05,
+      "loss": 0.741,
+      "step": 5370
+    },
+    {
+      "epoch": 3.0666191663697897,
+      "grad_norm": 2.045319617108696,
+      "learning_rate": 1.7507398189865057e-05,
+      "loss": 0.7421,
+      "step": 5380
+    },
+    {
+      "epoch": 3.0723192019950125,
+      "grad_norm": 1.9909398875734512,
+      "learning_rate": 1.7494236683921084e-05,
+      "loss": 0.7399,
+      "step": 5390
+    },
+    {
+      "epoch": 3.0780192376202353,
+      "grad_norm": 2.1478493144485165,
+      "learning_rate": 1.7481045496684525e-05,
+      "loss": 0.7425,
+      "step": 5400
+    },
+    {
+      "epoch": 3.0837192732454577,
+      "grad_norm": 1.9881565887713344,
+      "learning_rate": 1.7467824680399728e-05,
+      "loss": 0.7397,
+      "step": 5410
+    },
+    {
+      "epoch": 3.0894193088706805,
+      "grad_norm": 2.115917226840997,
+      "learning_rate": 1.7454574287428382e-05,
+      "loss": 0.7357,
+      "step": 5420
+    },
+    {
+      "epoch": 3.0951193444959033,
+      "grad_norm": 2.003745833255236,
+      "learning_rate": 1.744129437024932e-05,
+      "loss": 0.7505,
+      "step": 5430
+    },
+    {
+      "epoch": 3.1008193801211257,
+      "grad_norm": 2.078865802580488,
+      "learning_rate": 1.7427984981458305e-05,
+      "loss": 0.7419,
+      "step": 5440
+    },
+    {
+      "epoch": 3.1065194157463485,
+      "grad_norm": 2.1288364598170664,
+      "learning_rate": 1.7414646173767833e-05,
+      "loss": 0.7298,
+      "step": 5450
+    },
+    {
+      "epoch": 3.112219451371571,
+      "grad_norm": 2.072021887298528,
+      "learning_rate": 1.74012780000069e-05,
+      "loss": 0.748,
+      "step": 5460
+    },
+    {
+      "epoch": 3.1179194869967937,
+      "grad_norm": 2.177237216421149,
+      "learning_rate": 1.7387880513120815e-05,
+      "loss": 0.7452,
+      "step": 5470
+    },
+    {
+      "epoch": 3.1236195226220165,
+      "grad_norm": 2.1683763979667843,
+      "learning_rate": 1.7374453766170987e-05,
+      "loss": 0.7436,
+      "step": 5480
+    },
+    {
+      "epoch": 3.129319558247239,
+      "grad_norm": 2.0904041608117216,
+      "learning_rate": 1.73609978123347e-05,
+      "loss": 0.7361,
+      "step": 5490
+    },
+    {
+      "epoch": 3.1350195938724617,
+      "grad_norm": 2.047553895215376,
+      "learning_rate": 1.734751270490493e-05,
+      "loss": 0.748,
+      "step": 5500
+    },
+    {
+      "epoch": 3.1407196294976845,
+      "grad_norm": 2.005927750946996,
+      "learning_rate": 1.7333998497290097e-05,
+      "loss": 0.7421,
+      "step": 5510
+    },
+    {
+      "epoch": 3.146419665122907,
+      "grad_norm": 2.1634130072879008,
+      "learning_rate": 1.7320455243013896e-05,
+      "loss": 0.751,
+      "step": 5520
+    },
+    {
+      "epoch": 3.1521197007481296,
+      "grad_norm": 2.0045457707116494,
+      "learning_rate": 1.730688299571504e-05,
+      "loss": 0.7378,
+      "step": 5530
+    },
+    {
+      "epoch": 3.1578197363733524,
+      "grad_norm": 2.136212824258042,
+      "learning_rate": 1.729328180914709e-05,
+      "loss": 0.7448,
+      "step": 5540
+    },
+    {
+      "epoch": 3.163519771998575,
+      "grad_norm": 1.9657759781442752,
+      "learning_rate": 1.7279651737178204e-05,
+      "loss": 0.7345,
+      "step": 5550
+    },
+    {
+      "epoch": 3.1692198076237976,
+      "grad_norm": 2.046679397792288,
+      "learning_rate": 1.726599283379096e-05,
+      "loss": 0.73,
+      "step": 5560
+    },
+    {
+      "epoch": 3.1749198432490204,
+      "grad_norm": 2.160754138869591,
+      "learning_rate": 1.7252305153082118e-05,
+      "loss": 0.7304,
+      "step": 5570
+    },
+    {
+      "epoch": 3.180619878874243,
+      "grad_norm": 2.149734469063933,
+      "learning_rate": 1.7238588749262396e-05,
+      "loss": 0.7405,
+      "step": 5580
+    },
+    {
+      "epoch": 3.1863199144994656,
+      "grad_norm": 2.1182922399585213,
+      "learning_rate": 1.72248436766563e-05,
+      "loss": 0.7462,
+      "step": 5590
+    },
+    {
+      "epoch": 3.1920199501246884,
+      "grad_norm": 2.0991759289002823,
+      "learning_rate": 1.7211069989701855e-05,
+      "loss": 0.7524,
+      "step": 5600
+    },
+    {
+      "epoch": 3.197719985749911,
+      "grad_norm": 2.088052832265774,
+      "learning_rate": 1.7197267742950435e-05,
+      "loss": 0.7562,
+      "step": 5610
+    },
+    {
+      "epoch": 3.2034200213751336,
+      "grad_norm": 2.116606374909782,
+      "learning_rate": 1.718343699106651e-05,
+      "loss": 0.7341,
+      "step": 5620
+    },
+    {
+      "epoch": 3.2091200570003564,
+      "grad_norm": 2.274297318117651,
+      "learning_rate": 1.7169577788827448e-05,
+      "loss": 0.751,
+      "step": 5630
+    },
+    {
+      "epoch": 3.2148200926255788,
+      "grad_norm": 2.053843021870526,
+      "learning_rate": 1.7155690191123313e-05,
+      "loss": 0.7317,
+      "step": 5640
+    },
+    {
+      "epoch": 3.2205201282508016,
+      "grad_norm": 2.014911666896998,
+      "learning_rate": 1.7141774252956606e-05,
+      "loss": 0.7426,
+      "step": 5650
+    },
+    {
+      "epoch": 3.2262201638760244,
+      "grad_norm": 2.071543993545443,
+      "learning_rate": 1.712783002944209e-05,
+      "loss": 0.7472,
+      "step": 5660
+    },
+    {
+      "epoch": 3.2319201995012468,
+      "grad_norm": 2.0226426858864324,
+      "learning_rate": 1.7113857575806544e-05,
+      "loss": 0.7368,
+      "step": 5670
+    },
+    {
+      "epoch": 3.2376202351264696,
+      "grad_norm": 2.2081035387522667,
+      "learning_rate": 1.709985694738856e-05,
+      "loss": 0.7412,
+      "step": 5680
+    },
+    {
+      "epoch": 3.2433202707516924,
+      "grad_norm": 2.0173621023828967,
+      "learning_rate": 1.7085828199638315e-05,
+      "loss": 0.7318,
+      "step": 5690
+    },
+    {
+      "epoch": 3.2490203063769147,
+      "grad_norm": 2.0533146117503325,
+      "learning_rate": 1.707177138811735e-05,
+      "loss": 0.7388,
+      "step": 5700
+    },
+    {
+      "epoch": 3.2547203420021376,
+      "grad_norm": 2.042845781990925,
+      "learning_rate": 1.7057686568498363e-05,
+      "loss": 0.7369,
+      "step": 5710
+    },
+    {
+      "epoch": 3.2604203776273604,
+      "grad_norm": 2.106781789823166,
+      "learning_rate": 1.7043573796564966e-05,
+      "loss": 0.74,
+      "step": 5720
+    },
+    {
+      "epoch": 3.2661204132525827,
+      "grad_norm": 2.068664681703829,
+      "learning_rate": 1.7029433128211495e-05,
+      "loss": 0.7356,
+      "step": 5730
+    },
+    {
+      "epoch": 3.2718204488778055,
+      "grad_norm": 2.046446640636462,
+      "learning_rate": 1.7015264619442758e-05,
+      "loss": 0.737,
+      "step": 5740
+    },
+    {
+      "epoch": 3.277520484503028,
+      "grad_norm": 2.0057975588623336,
+      "learning_rate": 1.7001068326373827e-05,
+      "loss": 0.7352,
+      "step": 5750
+    },
+    {
+      "epoch": 3.2832205201282507,
+      "grad_norm": 2.00011969394344,
+      "learning_rate": 1.698684430522982e-05,
+      "loss": 0.7485,
+      "step": 5760
+    },
+    {
+      "epoch": 3.2889205557534735,
+      "grad_norm": 1.9793845313476237,
+      "learning_rate": 1.6972592612345673e-05,
+      "loss": 0.7428,
+      "step": 5770
+    },
+    {
+      "epoch": 3.2946205913786963,
+      "grad_norm": 2.0457258452237674,
+      "learning_rate": 1.6958313304165915e-05,
+      "loss": 0.7457,
+      "step": 5780
+    },
+    {
+      "epoch": 3.3003206270039187,
+      "grad_norm": 1.9844119915920377,
+      "learning_rate": 1.694400643724445e-05,
+      "loss": 0.7288,
+      "step": 5790
+    },
+    {
+      "epoch": 3.3060206626291415,
+      "grad_norm": 2.1178149861323115,
+      "learning_rate": 1.6929672068244325e-05,
+      "loss": 0.734,
+      "step": 5800
+    },
+    {
+      "epoch": 3.311720698254364,
+      "grad_norm": 2.2494684992145864,
+      "learning_rate": 1.691531025393751e-05,
+      "loss": 0.7428,
+      "step": 5810
+    },
+    {
+      "epoch": 3.3174207338795867,
+      "grad_norm": 2.103045025019933,
+      "learning_rate": 1.690092105120468e-05,
+      "loss": 0.7336,
+      "step": 5820
+    },
+    {
+      "epoch": 3.3231207695048095,
+      "grad_norm": 2.079229864541264,
+      "learning_rate": 1.688650451703498e-05,
+      "loss": 0.735,
+      "step": 5830
+    },
+    {
+      "epoch": 3.3288208051300323,
+      "grad_norm": 2.1262241995520688,
+      "learning_rate": 1.68720607085258e-05,
+      "loss": 0.7478,
+      "step": 5840
+    },
+    {
+      "epoch": 3.3345208407552547,
+      "grad_norm": 2.0672011419206746,
+      "learning_rate": 1.685758968288255e-05,
+      "loss": 0.7364,
+      "step": 5850
+    },
+    {
+      "epoch": 3.3402208763804775,
+      "grad_norm": 2.018766214415992,
+      "learning_rate": 1.684309149741845e-05,
+      "loss": 0.7459,
+      "step": 5860
+    },
+    {
+      "epoch": 3.3459209120057,
+      "grad_norm": 2.024319172839621,
+      "learning_rate": 1.6828566209554254e-05,
+      "loss": 0.7293,
+      "step": 5870
+    },
+    {
+      "epoch": 3.3516209476309227,
+      "grad_norm": 2.2232243377101732,
+      "learning_rate": 1.68140138768181e-05,
+      "loss": 0.745,
+      "step": 5880
+    },
+    {
+      "epoch": 3.3573209832561455,
+      "grad_norm": 1.9892262862948757,
+      "learning_rate": 1.6799434556845206e-05,
+      "loss": 0.74,
+      "step": 5890
+    },
+    {
+      "epoch": 3.363021018881368,
+      "grad_norm": 1.996203658365478,
+      "learning_rate": 1.678482830737769e-05,
+      "loss": 0.7358,
+      "step": 5900
+    },
+    {
+      "epoch": 3.3687210545065907,
+      "grad_norm": 2.0492262539429476,
+      "learning_rate": 1.6770195186264318e-05,
+      "loss": 0.7481,
+      "step": 5910
+    },
+    {
+      "epoch": 3.3744210901318135,
+      "grad_norm": 2.126332808306086,
+      "learning_rate": 1.6755535251460282e-05,
+      "loss": 0.7356,
+      "step": 5920
+    },
+    {
+      "epoch": 3.380121125757036,
+      "grad_norm": 2.1510518131386216,
+      "learning_rate": 1.674084856102698e-05,
+      "loss": 0.7369,
+      "step": 5930
+    },
+    {
+      "epoch": 3.3858211613822586,
+      "grad_norm": 1.9898038869629846,
+      "learning_rate": 1.6726135173131767e-05,
+      "loss": 0.7502,
+      "step": 5940
+    },
+    {
+      "epoch": 3.3915211970074814,
+      "grad_norm": 2.0793774775395177,
+      "learning_rate": 1.671139514604774e-05,
+      "loss": 0.743,
+      "step": 5950
+    },
+    {
+      "epoch": 3.397221232632704,
+      "grad_norm": 2.0911243940655373,
+      "learning_rate": 1.6696628538153498e-05,
+      "loss": 0.7507,
+      "step": 5960
+    },
+    {
+      "epoch": 3.4029212682579266,
+      "grad_norm": 1.9431819413883284,
+      "learning_rate": 1.668183540793292e-05,
+      "loss": 0.7361,
+      "step": 5970
+    },
+    {
+      "epoch": 3.4086213038831494,
+      "grad_norm": 2.010182507808355,
+      "learning_rate": 1.6667015813974928e-05,
+      "loss": 0.7401,
+      "step": 5980
+    },
+    {
+      "epoch": 3.414321339508372,
+      "grad_norm": 2.161496770233213,
+      "learning_rate": 1.6652169814973246e-05,
+      "loss": 0.7407,
+      "step": 5990
+    },
+    {
+      "epoch": 3.4200213751335946,
+      "grad_norm": 2.026683695690911,
+      "learning_rate": 1.6637297469726182e-05,
+      "loss": 0.745,
+      "step": 6000
+    },
+    {
+      "epoch": 3.4257214107588174,
+      "grad_norm": 2.1544959902671623,
+      "learning_rate": 1.6622398837136397e-05,
+      "loss": 0.7443,
+      "step": 6010
+    },
+    {
+      "epoch": 3.43142144638404,
+      "grad_norm": 2.0254574449350002,
+      "learning_rate": 1.660747397621065e-05,
+      "loss": 0.7441,
+      "step": 6020
+    },
+    {
+      "epoch": 3.4371214820092626,
+      "grad_norm": 2.1063512225032213,
+      "learning_rate": 1.6592522946059594e-05,
+      "loss": 0.7358,
+      "step": 6030
+    },
+    {
+      "epoch": 3.4428215176344854,
+      "grad_norm": 2.109309750652788,
+      "learning_rate": 1.657754580589751e-05,
+      "loss": 0.735,
+      "step": 6040
+    },
+    {
+      "epoch": 3.4485215532597078,
+      "grad_norm": 2.1362388241431223,
+      "learning_rate": 1.65625426150421e-05,
+      "loss": 0.7528,
+      "step": 6050
+    },
+    {
+      "epoch": 3.4542215888849306,
+      "grad_norm": 1.9534769975619974,
+      "learning_rate": 1.6547513432914242e-05,
+      "loss": 0.7418,
+      "step": 6060
+    },
+    {
+      "epoch": 3.4599216245101534,
+      "grad_norm": 2.191011795660284,
+      "learning_rate": 1.6532458319037748e-05,
+      "loss": 0.7453,
+      "step": 6070
+    },
+    {
+      "epoch": 3.4656216601353758,
+      "grad_norm": 2.0481077058223627,
+      "learning_rate": 1.6517377333039134e-05,
+      "loss": 0.7355,
+      "step": 6080
+    },
+    {
+      "epoch": 3.4713216957605986,
+      "grad_norm": 2.0539224712223794,
+      "learning_rate": 1.650227053464739e-05,
+      "loss": 0.7311,
+      "step": 6090
+    },
+    {
+      "epoch": 3.477021731385821,
+      "grad_norm": 2.1197486154272345,
+      "learning_rate": 1.6487137983693732e-05,
+      "loss": 0.7416,
+      "step": 6100
+    },
+    {
+      "epoch": 3.4827217670110437,
+      "grad_norm": 1.956319564532167,
+      "learning_rate": 1.647197974011137e-05,
+      "loss": 0.7424,
+      "step": 6110
+    },
+    {
+      "epoch": 3.4884218026362666,
+      "grad_norm": 2.0396586196184296,
+      "learning_rate": 1.645679586393527e-05,
+      "loss": 0.7387,
+      "step": 6120
+    },
+    {
+      "epoch": 3.4941218382614894,
+      "grad_norm": 2.160261201715842,
+      "learning_rate": 1.6441586415301928e-05,
+      "loss": 0.7331,
+      "step": 6130
+    },
+    {
+      "epoch": 3.4998218738867117,
+      "grad_norm": 2.0696714508439307,
+      "learning_rate": 1.6426351454449102e-05,
+      "loss": 0.7382,
+      "step": 6140
+    },
+    {
+      "epoch": 3.5055219095119345,
+      "grad_norm": 2.0931254540417004,
+      "learning_rate": 1.641109104171561e-05,
+      "loss": 0.7386,
+      "step": 6150
+    },
+    {
+      "epoch": 3.511221945137157,
+      "grad_norm": 2.007931269061243,
+      "learning_rate": 1.6395805237541066e-05,
+      "loss": 0.7434,
+      "step": 6160
+    },
+    {
+      "epoch": 3.5169219807623797,
+      "grad_norm": 2.1952277405235647,
+      "learning_rate": 1.6380494102465644e-05,
+      "loss": 0.7403,
+      "step": 6170
+    },
+    {
+      "epoch": 3.5226220163876025,
+      "grad_norm": 2.0817367988428224,
+      "learning_rate": 1.6365157697129853e-05,
+      "loss": 0.7295,
+      "step": 6180
+    },
+    {
+      "epoch": 3.5283220520128253,
+      "grad_norm": 1.917307285187369,
+      "learning_rate": 1.6349796082274275e-05,
+      "loss": 0.7342,
+      "step": 6190
+    },
+    {
+      "epoch": 3.5340220876380477,
+      "grad_norm": 2.020375761125305,
+      "learning_rate": 1.6334409318739344e-05,
+      "loss": 0.737,
+      "step": 6200
+    },
+    {
+      "epoch": 3.5397221232632705,
+      "grad_norm": 1.9659177210333798,
+      "learning_rate": 1.631899746746509e-05,
+      "loss": 0.747,
+      "step": 6210
+    },
+    {
+      "epoch": 3.545422158888493,
+      "grad_norm": 2.135868919125441,
+      "learning_rate": 1.630356058949091e-05,
+      "loss": 0.7344,
+      "step": 6220
+    },
+    {
+      "epoch": 3.5511221945137157,
+      "grad_norm": 2.0069181029404195,
+      "learning_rate": 1.628809874595531e-05,
+      "loss": 0.7371,
+      "step": 6230
+    },
+    {
+      "epoch": 3.5568222301389385,
+      "grad_norm": 2.0680129880723492,
+      "learning_rate": 1.6272611998095694e-05,
+      "loss": 0.733,
+      "step": 6240
+    },
+    {
+      "epoch": 3.562522265764161,
+      "grad_norm": 1.968388651205215,
+      "learning_rate": 1.6257100407248075e-05,
+      "loss": 0.741,
+      "step": 6250
+    },
+    {
+      "epoch": 3.5682223013893837,
+      "grad_norm": 2.050635559497989,
+      "learning_rate": 1.6241564034846883e-05,
+      "loss": 0.7364,
+      "step": 6260
+    },
+    {
+      "epoch": 3.5739223370146065,
+      "grad_norm": 2.171816362983528,
+      "learning_rate": 1.622600294242467e-05,
+      "loss": 0.7358,
+      "step": 6270
+    },
+    {
+      "epoch": 3.579622372639829,
+      "grad_norm": 2.0930915534084744,
+      "learning_rate": 1.6210417191611917e-05,
+      "loss": 0.7519,
+      "step": 6280
+    },
+    {
+      "epoch": 3.5853224082650517,
+      "grad_norm": 2.0689257909003604,
+      "learning_rate": 1.6194806844136755e-05,
+      "loss": 0.7376,
+      "step": 6290
+    },
+    {
+      "epoch": 3.5910224438902745,
+      "grad_norm": 2.0357101666755772,
+      "learning_rate": 1.617917196182473e-05,
+      "loss": 0.7295,
+      "step": 6300
+    },
+    {
+      "epoch": 3.596722479515497,
+      "grad_norm": 1.9928260487092255,
+      "learning_rate": 1.616351260659856e-05,
+      "loss": 0.7433,
+      "step": 6310
+    },
+    {
+      "epoch": 3.6024225151407196,
+      "grad_norm": 1.9848046393102028,
+      "learning_rate": 1.6147828840477893e-05,
+      "loss": 0.7385,
+      "step": 6320
+    },
+    {
+      "epoch": 3.608122550765942,
+      "grad_norm": 2.01146751576876,
+      "learning_rate": 1.6132120725579057e-05,
+      "loss": 0.7346,
+      "step": 6330
+    },
+    {
+      "epoch": 3.613822586391165,
+      "grad_norm": 2.1500721922467694,
+      "learning_rate": 1.611638832411481e-05,
+      "loss": 0.7308,
+      "step": 6340
+    },
+    {
+      "epoch": 3.6195226220163876,
+      "grad_norm": 2.108969917572423,
+      "learning_rate": 1.61006316983941e-05,
+      "loss": 0.7402,
+      "step": 6350
+    },
+    {
+      "epoch": 3.6252226576416104,
+      "grad_norm": 2.1537890950662146,
+      "learning_rate": 1.6084850910821822e-05,
+      "loss": 0.7434,
+      "step": 6360
+    },
+    {
+      "epoch": 3.630922693266833,
+      "grad_norm": 2.0905163971990803,
+      "learning_rate": 1.6069046023898554e-05,
+      "loss": 0.7377,
+      "step": 6370
+    },
+    {
+      "epoch": 3.6366227288920556,
+      "grad_norm": 2.1833272882788486,
+      "learning_rate": 1.6053217100220332e-05,
+      "loss": 0.7387,
+      "step": 6380
+    },
+    {
+      "epoch": 3.642322764517278,
+      "grad_norm": 1.9978536000336802,
+      "learning_rate": 1.6037364202478386e-05,
+      "loss": 0.7436,
+      "step": 6390
+    },
+    {
+      "epoch": 3.648022800142501,
+      "grad_norm": 1.9857203809905162,
+      "learning_rate": 1.6021487393458893e-05,
+      "loss": 0.741,
+      "step": 6400
+    },
+    {
+      "epoch": 3.6537228357677236,
+      "grad_norm": 1.9785607872258348,
+      "learning_rate": 1.600558673604274e-05,
+      "loss": 0.7274,
+      "step": 6410
+    },
+    {
+      "epoch": 3.6594228713929464,
+      "grad_norm": 2.0892691974189086,
+      "learning_rate": 1.598966229320526e-05,
+      "loss": 0.7357,
+      "step": 6420
+    },
+    {
+      "epoch": 3.665122907018169,
+      "grad_norm": 2.1213516026120955,
+      "learning_rate": 1.5973714128015987e-05,
+      "loss": 0.7361,
+      "step": 6430
+    },
+    {
+      "epoch": 3.6708229426433916,
+      "grad_norm": 1.958345380862227,
+      "learning_rate": 1.595774230363842e-05,
+      "loss": 0.7328,
+      "step": 6440
+    },
+    {
+      "epoch": 3.676522978268614,
+      "grad_norm": 2.019606461449015,
+      "learning_rate": 1.5941746883329745e-05,
+      "loss": 0.7342,
+      "step": 6450
+    },
+    {
+      "epoch": 3.6822230138938368,
+      "grad_norm": 2.0150832864171186,
+      "learning_rate": 1.5925727930440617e-05,
+      "loss": 0.7331,
+      "step": 6460
+    },
+    {
+      "epoch": 3.6879230495190596,
+      "grad_norm": 2.055175500413531,
+      "learning_rate": 1.5909685508414884e-05,
+      "loss": 0.7435,
+      "step": 6470
+    },
+    {
+      "epoch": 3.6936230851442824,
+      "grad_norm": 2.06353886004344,
+      "learning_rate": 1.589361968078935e-05,
+      "loss": 0.738,
+      "step": 6480
+    },
+    {
+      "epoch": 3.6993231207695048,
+      "grad_norm": 2.008466348281505,
+      "learning_rate": 1.587753051119351e-05,
+      "loss": 0.744,
+      "step": 6490
+    },
+    {
+      "epoch": 3.7050231563947276,
+      "grad_norm": 2.045367700599523,
+      "learning_rate": 1.586141806334931e-05,
+      "loss": 0.7338,
+      "step": 6500
+    },
+    {
+      "epoch": 3.71072319201995,
+      "grad_norm": 2.0550785176203594,
+      "learning_rate": 1.5845282401070893e-05,
+      "loss": 0.7381,
+      "step": 6510
+    },
+    {
+      "epoch": 3.7164232276451727,
+      "grad_norm": 2.0872693575576386,
+      "learning_rate": 1.5829123588264348e-05,
+      "loss": 0.7305,
+      "step": 6520
+    },
+    {
+      "epoch": 3.7221232632703956,
+      "grad_norm": 1.9117270867086693,
+      "learning_rate": 1.5812941688927435e-05,
+      "loss": 0.732,
+      "step": 6530
+    },
+    {
+      "epoch": 3.7278232988956184,
+      "grad_norm": 2.1528521470214215,
+      "learning_rate": 1.579673676714937e-05,
+      "loss": 0.7357,
+      "step": 6540
+    },
+    {
+      "epoch": 3.7335233345208407,
+      "grad_norm": 2.1473073455155305,
+      "learning_rate": 1.5780508887110543e-05,
+      "loss": 0.7359,
+      "step": 6550
+    },
+    {
+      "epoch": 3.7392233701460635,
+      "grad_norm": 2.0521636868365336,
+      "learning_rate": 1.5764258113082266e-05,
+      "loss": 0.733,
+      "step": 6560
+    },
+    {
+      "epoch": 3.744923405771286,
+      "grad_norm": 2.262153989248003,
+      "learning_rate": 1.5747984509426528e-05,
+      "loss": 0.7177,
+      "step": 6570
+    },
+    {
+      "epoch": 3.7506234413965087,
+      "grad_norm": 2.013177191446144,
+      "learning_rate": 1.5731688140595737e-05,
+      "loss": 0.7336,
+      "step": 6580
+    },
+    {
+      "epoch": 3.7563234770217315,
+      "grad_norm": 2.0304305461425494,
+      "learning_rate": 1.5715369071132462e-05,
+      "loss": 0.7237,
+      "step": 6590
+    },
+    {
+      "epoch": 3.762023512646954,
+      "grad_norm": 2.0998682733697356,
+      "learning_rate": 1.569902736566918e-05,
+      "loss": 0.7311,
+      "step": 6600
+    },
+    {
+      "epoch": 3.7677235482721767,
+      "grad_norm": 2.052004412180822,
+      "learning_rate": 1.5682663088928017e-05,
+      "loss": 0.7254,
+      "step": 6610
+    },
+    {
+      "epoch": 3.7734235838973995,
+      "grad_norm": 2.083215696482381,
+      "learning_rate": 1.5666276305720497e-05,
+      "loss": 0.7347,
+      "step": 6620
+    },
+    {
+      "epoch": 3.779123619522622,
+      "grad_norm": 2.1017569504703646,
+      "learning_rate": 1.564986708094728e-05,
+      "loss": 0.7287,
+      "step": 6630
+    },
+    {
+      "epoch": 3.7848236551478447,
+      "grad_norm": 2.1295534422279347,
+      "learning_rate": 1.5633435479597906e-05,
+      "loss": 0.7382,
+      "step": 6640
+    },
+    {
+      "epoch": 3.7905236907730675,
+      "grad_norm": 2.130311845169098,
+      "learning_rate": 1.561698156675054e-05,
+      "loss": 0.7284,
+      "step": 6650
+    },
+    {
+      "epoch": 3.79622372639829,
+      "grad_norm": 2.062585126683027,
+      "learning_rate": 1.5600505407571706e-05,
+      "loss": 0.7428,
+      "step": 6660
+    },
+    {
+      "epoch": 3.8019237620235127,
+      "grad_norm": 2.133891320255428,
+      "learning_rate": 1.558400706731605e-05,
+      "loss": 0.7382,
+      "step": 6670
+    },
+    {
+      "epoch": 3.807623797648735,
+      "grad_norm": 2.0965977555975353,
+      "learning_rate": 1.5567486611326058e-05,
+      "loss": 0.731,
+      "step": 6680
+    },
+    {
+      "epoch": 3.813323833273958,
+      "grad_norm": 2.119618624800062,
+      "learning_rate": 1.555094410503181e-05,
+      "loss": 0.7333,
+      "step": 6690
+    },
+    {
+      "epoch": 3.8190238688991807,
+      "grad_norm": 1.988282420666724,
+      "learning_rate": 1.5534379613950704e-05,
+      "loss": 0.7327,
+      "step": 6700
+    },
+    {
+      "epoch": 3.8247239045244035,
+      "grad_norm": 2.1425468665186562,
+      "learning_rate": 1.5517793203687232e-05,
+      "loss": 0.7411,
+      "step": 6710
+    },
+    {
+      "epoch": 3.830423940149626,
+      "grad_norm": 2.2461283085130828,
+      "learning_rate": 1.5501184939932685e-05,
+      "loss": 0.7392,
+      "step": 6720
+    },
+    {
+      "epoch": 3.8361239757748486,
+      "grad_norm": 2.1980537869605907,
+      "learning_rate": 1.54845548884649e-05,
+      "loss": 0.7159,
+      "step": 6730
+    },
+    {
+      "epoch": 3.841824011400071,
+      "grad_norm": 2.099025441032915,
+      "learning_rate": 1.5467903115148023e-05,
+      "loss": 0.7358,
+      "step": 6740
+    },
+    {
+      "epoch": 3.847524047025294,
+      "grad_norm": 2.144448215533206,
+      "learning_rate": 1.5451229685932212e-05,
+      "loss": 0.732,
+      "step": 6750
+    },
+    {
+      "epoch": 3.8532240826505166,
+      "grad_norm": 2.0678206225034526,
+      "learning_rate": 1.5434534666853406e-05,
+      "loss": 0.7294,
+      "step": 6760
+    },
+    {
+      "epoch": 3.8589241182757394,
+      "grad_norm": 2.3507788100600915,
+      "learning_rate": 1.541781812403305e-05,
+      "loss": 0.7368,
+      "step": 6770
+    },
+    {
+      "epoch": 3.864624153900962,
+      "grad_norm": 2.062589684642445,
+      "learning_rate": 1.540108012367783e-05,
+      "loss": 0.7359,
+      "step": 6780
+    },
+    {
+      "epoch": 3.8703241895261846,
+      "grad_norm": 2.075536540767243,
+      "learning_rate": 1.538432073207942e-05,
+      "loss": 0.7237,
+      "step": 6790
+    },
+    {
+      "epoch": 3.876024225151407,
+      "grad_norm": 2.03594095541657,
+      "learning_rate": 1.536754001561422e-05,
+      "loss": 0.7227,
+      "step": 6800
+    },
+    {
+      "epoch": 3.88172426077663,
+      "grad_norm": 1.999572280341755,
+      "learning_rate": 1.535073804074307e-05,
+      "loss": 0.7384,
+      "step": 6810
+    },
+    {
+      "epoch": 3.8874242964018526,
+      "grad_norm": 2.14843223648859,
+      "learning_rate": 1.5333914874011025e-05,
+      "loss": 0.7278,
+      "step": 6820
+    },
+    {
+      "epoch": 3.8931243320270754,
+      "grad_norm": 2.0127815978780084,
+      "learning_rate": 1.5317070582047066e-05,
+      "loss": 0.7352,
+      "step": 6830
+    },
+    {
+      "epoch": 3.898824367652298,
+      "grad_norm": 2.1449298447570624,
+      "learning_rate": 1.530020523156383e-05,
+      "loss": 0.7296,
+      "step": 6840
+    },
+    {
+      "epoch": 3.9045244032775206,
+      "grad_norm": 1.9861876450849654,
+      "learning_rate": 1.5283318889357367e-05,
+      "loss": 0.72,
+      "step": 6850
+    },
+    {
+      "epoch": 3.910224438902743,
+      "grad_norm": 2.1397460467468887,
+      "learning_rate": 1.5266411622306873e-05,
+      "loss": 0.7379,
+      "step": 6860
+    },
+    {
+      "epoch": 3.9159244745279658,
+      "grad_norm": 2.1057354943710958,
+      "learning_rate": 1.5249483497374403e-05,
+      "loss": 0.7317,
+      "step": 6870
+    },
+    {
+      "epoch": 3.9216245101531886,
+      "grad_norm": 1.9498837345150901,
+      "learning_rate": 1.5232534581604633e-05,
+      "loss": 0.74,
+      "step": 6880
+    },
+    {
+      "epoch": 3.9273245457784114,
+      "grad_norm": 2.7851238097023225,
+      "learning_rate": 1.5215564942124573e-05,
+      "loss": 0.7305,
+      "step": 6890
+    },
+    {
+      "epoch": 3.9330245814036338,
+      "grad_norm": 2.19368206819208,
+      "learning_rate": 1.5198574646143311e-05,
+      "loss": 0.7415,
+      "step": 6900
+    },
+    {
+      "epoch": 3.9387246170288566,
+      "grad_norm": 2.1593430708374304,
+      "learning_rate": 1.5181563760951754e-05,
+      "loss": 0.7343,
+      "step": 6910
+    },
+    {
+      "epoch": 3.944424652654079,
+      "grad_norm": 2.1260855448365366,
+      "learning_rate": 1.516453235392235e-05,
+      "loss": 0.7305,
+      "step": 6920
+    },
+    {
+      "epoch": 3.9501246882793017,
+      "grad_norm": 1.9324032551297767,
+      "learning_rate": 1.5147480492508817e-05,
+      "loss": 0.7301,
+      "step": 6930
+    },
+    {
+      "epoch": 3.9558247239045246,
+      "grad_norm": 2.1164826662513105,
+      "learning_rate": 1.5130408244245893e-05,
+      "loss": 0.7441,
+      "step": 6940
+    },
+    {
+      "epoch": 3.961524759529747,
+      "grad_norm": 2.036890857491215,
+      "learning_rate": 1.5113315676749056e-05,
+      "loss": 0.7335,
+      "step": 6950
+    },
+    {
+      "epoch": 3.9672247951549697,
+      "grad_norm": 2.1496177354465345,
+      "learning_rate": 1.5096202857714261e-05,
+      "loss": 0.736,
+      "step": 6960
+    },
+    {
+      "epoch": 3.9729248307801925,
+      "grad_norm": 2.0232753736758253,
+      "learning_rate": 1.5079069854917666e-05,
+      "loss": 0.7306,
+      "step": 6970
+    },
+    {
+      "epoch": 3.978624866405415,
+      "grad_norm": 2.005780628746836,
+      "learning_rate": 1.5061916736215372e-05,
+      "loss": 0.7335,
+      "step": 6980
+    },
+    {
+      "epoch": 3.9843249020306377,
+      "grad_norm": 1.9480858963457948,
+      "learning_rate": 1.5044743569543147e-05,
+      "loss": 0.7261,
+      "step": 6990
+    },
+    {
+      "epoch": 3.9900249376558605,
+      "grad_norm": 2.0513896063025703,
+      "learning_rate": 1.5027550422916164e-05,
+      "loss": 0.7319,
+      "step": 7000
+    },
+    {
+      "epoch": 3.995724973281083,
+      "grad_norm": 2.0474148046725933,
+      "learning_rate": 1.5010337364428723e-05,
+      "loss": 0.7394,
+      "step": 7010
+    },
+    {
+      "epoch": 4.001425008906305,
+      "grad_norm": 1.9251825898575523,
+      "learning_rate": 1.4993104462253987e-05,
+      "loss": 0.7275,
+      "step": 7020
+    },
+    {
+      "epoch": 4.007125044531528,
+      "grad_norm": 1.9744452135994894,
+      "learning_rate": 1.4975851784643713e-05,
+      "loss": 0.7152,
+      "step": 7030
+    },
+    {
+      "epoch": 4.012825080156751,
+      "grad_norm": 2.1433515486563293,
+      "learning_rate": 1.4958579399927977e-05,
+      "loss": 0.7165,
+      "step": 7040
+    },
+    {
+      "epoch": 4.018525115781974,
+      "grad_norm": 2.0710887445851345,
+      "learning_rate": 1.4941287376514908e-05,
+      "loss": 0.7102,
+      "step": 7050
+    },
+    {
+      "epoch": 4.0242251514071965,
+      "grad_norm": 2.0375282509919046,
+      "learning_rate": 1.4923975782890415e-05,
+      "loss": 0.7132,
+      "step": 7060
+    },
+    {
+      "epoch": 4.029925187032419,
+      "grad_norm": 2.097469944887364,
+      "learning_rate": 1.4906644687617915e-05,
+      "loss": 0.7147,
+      "step": 7070
+    },
+    {
+      "epoch": 4.035625222657641,
+      "grad_norm": 2.0530341428336993,
+      "learning_rate": 1.4889294159338061e-05,
+      "loss": 0.7158,
+      "step": 7080
+    },
+    {
+      "epoch": 4.041325258282864,
+      "grad_norm": 2.142877870616465,
+      "learning_rate": 1.4871924266768474e-05,
+      "loss": 0.7045,
+      "step": 7090
+    },
+    {
+      "epoch": 4.047025293908087,
+      "grad_norm": 2.2153688485348186,
+      "learning_rate": 1.4854535078703466e-05,
+      "loss": 0.7176,
+      "step": 7100
+    },
+    {
+      "epoch": 4.05272532953331,
+      "grad_norm": 2.0916124170647645,
+      "learning_rate": 1.483712666401377e-05,
+      "loss": 0.7142,
+      "step": 7110
+    },
+    {
+      "epoch": 4.0584253651585325,
+      "grad_norm": 2.062529383171578,
+      "learning_rate": 1.4819699091646272e-05,
+      "loss": 0.7241,
+      "step": 7120
+    },
+    {
+      "epoch": 4.064125400783755,
+      "grad_norm": 2.059615898158389,
+      "learning_rate": 1.4802252430623725e-05,
+      "loss": 0.7157,
+      "step": 7130
+    },
+    {
+      "epoch": 4.069825436408977,
+      "grad_norm": 2.048208123372868,
+      "learning_rate": 1.4784786750044486e-05,
+      "loss": 0.7154,
+      "step": 7140
+    },
+    {
+      "epoch": 4.0755254720342,
+      "grad_norm": 2.167028041354583,
+      "learning_rate": 1.4767302119082243e-05,
+      "loss": 0.7118,
+      "step": 7150
+    },
+    {
+      "epoch": 4.081225507659423,
+      "grad_norm": 2.055142808371912,
+      "learning_rate": 1.4749798606985735e-05,
+      "loss": 0.7051,
+      "step": 7160
+    },
+    {
+      "epoch": 4.086925543284646,
+      "grad_norm": 2.11856154255257,
+      "learning_rate": 1.4732276283078484e-05,
+      "loss": 0.7143,
+      "step": 7170
+    },
+    {
+      "epoch": 4.0926255789098684,
+      "grad_norm": 2.103529783943278,
+      "learning_rate": 1.4714735216758512e-05,
+      "loss": 0.7151,
+      "step": 7180
+    },
+    {
+      "epoch": 4.098325614535091,
+      "grad_norm": 2.171001048115003,
+      "learning_rate": 1.4697175477498074e-05,
+      "loss": 0.7058,
+      "step": 7190
+    },
+    {
+      "epoch": 4.104025650160313,
+      "grad_norm": 2.019947806571963,
+      "learning_rate": 1.4679597134843382e-05,
+      "loss": 0.7207,
+      "step": 7200
+    },
+    {
+      "epoch": 4.109725685785536,
+      "grad_norm": 2.0337415519897255,
+      "learning_rate": 1.4662000258414324e-05,
+      "loss": 0.7229,
+      "step": 7210
+    },
+    {
+      "epoch": 4.115425721410759,
+      "grad_norm": 2.1682062034587393,
+      "learning_rate": 1.4644384917904195e-05,
+      "loss": 0.7111,
+      "step": 7220
+    },
+    {
+      "epoch": 4.121125757035982,
+      "grad_norm": 2.045117187428103,
+      "learning_rate": 1.4626751183079415e-05,
+      "loss": 0.7205,
+      "step": 7230
+    },
+    {
+      "epoch": 4.126825792661204,
+      "grad_norm": 2.037782713872047,
+      "learning_rate": 1.460909912377926e-05,
+      "loss": 0.7108,
+      "step": 7240
+    },
+    {
+      "epoch": 4.132525828286427,
+      "grad_norm": 2.1443340636277664,
+      "learning_rate": 1.4591428809915573e-05,
+      "loss": 0.7157,
+      "step": 7250
+    },
+    {
+      "epoch": 4.138225863911649,
+      "grad_norm": 2.094115051994579,
+      "learning_rate": 1.4573740311472506e-05,
+      "loss": 0.7122,
+      "step": 7260
+    },
+    {
+      "epoch": 4.143925899536872,
+      "grad_norm": 2.1551146535510606,
+      "learning_rate": 1.4556033698506224e-05,
+      "loss": 0.7211,
+      "step": 7270
+    },
+    {
+      "epoch": 4.149625935162095,
+      "grad_norm": 2.0963416387595397,
+      "learning_rate": 1.4538309041144636e-05,
+      "loss": 0.7222,
+      "step": 7280
+    },
+    {
+      "epoch": 4.155325970787318,
+      "grad_norm": 2.151941102245385,
+      "learning_rate": 1.4520566409587118e-05,
+      "loss": 0.7139,
+      "step": 7290
+    },
+    {
+      "epoch": 4.16102600641254,
+      "grad_norm": 2.2244832792640348,
+      "learning_rate": 1.4502805874104237e-05,
+      "loss": 0.7097,
+      "step": 7300
+    },
+    {
+      "epoch": 4.166726042037762,
+      "grad_norm": 2.1294605571769893,
+      "learning_rate": 1.4485027505037464e-05,
+      "loss": 0.7055,
+      "step": 7310
+    },
+    {
+      "epoch": 4.172426077662985,
+      "grad_norm": 2.1285692780070793,
+      "learning_rate": 1.4467231372798905e-05,
+      "loss": 0.7246,
+      "step": 7320
+    },
+    {
+      "epoch": 4.178126113288208,
+      "grad_norm": 1.963831315280734,
+      "learning_rate": 1.4449417547871014e-05,
+      "loss": 0.7241,
+      "step": 7330
+    },
+    {
+      "epoch": 4.183826148913431,
+      "grad_norm": 2.1043850167694593,
+      "learning_rate": 1.443158610080632e-05,
+      "loss": 0.7273,
+      "step": 7340
+    },
+    {
+      "epoch": 4.1895261845386536,
+      "grad_norm": 2.050815944061624,
+      "learning_rate": 1.441373710222715e-05,
+      "loss": 0.7065,
+      "step": 7350
+    },
+    {
+      "epoch": 4.195226220163876,
+      "grad_norm": 2.151648220360366,
+      "learning_rate": 1.439587062282533e-05,
+      "loss": 0.7081,
+      "step": 7360
+    },
+    {
+      "epoch": 4.200926255789098,
+      "grad_norm": 1.9835522892363455,
+      "learning_rate": 1.437798673336194e-05,
+      "loss": 0.7213,
+      "step": 7370
+    },
+    {
+      "epoch": 4.206626291414321,
+      "grad_norm": 2.0768211194392894,
+      "learning_rate": 1.4360085504666994e-05,
+      "loss": 0.7115,
+      "step": 7380
+    },
+    {
+      "epoch": 4.212326327039544,
+      "grad_norm": 2.1521860372568713,
+      "learning_rate": 1.4342167007639196e-05,
+      "loss": 0.7073,
+      "step": 7390
+    },
+    {
+      "epoch": 4.218026362664767,
+      "grad_norm": 2.05947948843524,
+      "learning_rate": 1.4324231313245629e-05,
+      "loss": 0.7124,
+      "step": 7400
+    },
+    {
+      "epoch": 4.2237263982899895,
+      "grad_norm": 2.069971597086405,
+      "learning_rate": 1.430627849252149e-05,
+      "loss": 0.7051,
+      "step": 7410
+    },
+    {
+      "epoch": 4.229426433915212,
+      "grad_norm": 2.02026677977584,
+      "learning_rate": 1.4288308616569811e-05,
+      "loss": 0.7127,
+      "step": 7420
+    },
+    {
+      "epoch": 4.235126469540434,
+      "grad_norm": 2.276712198730305,
+      "learning_rate": 1.4270321756561169e-05,
+      "loss": 0.7189,
+      "step": 7430
+    },
+    {
+      "epoch": 4.240826505165657,
+      "grad_norm": 2.0748662348756524,
+      "learning_rate": 1.4252317983733406e-05,
+      "loss": 0.7076,
+      "step": 7440
+    },
+    {
+      "epoch": 4.24652654079088,
+      "grad_norm": 2.2204058148387618,
+      "learning_rate": 1.4234297369391345e-05,
+      "loss": 0.7144,
+      "step": 7450
+    },
+    {
+      "epoch": 4.252226576416103,
+      "grad_norm": 2.122075457505512,
+      "learning_rate": 1.4216259984906522e-05,
+      "loss": 0.7106,
+      "step": 7460
+    },
+    {
+      "epoch": 4.2579266120413255,
+      "grad_norm": 2.1374680801145756,
+      "learning_rate": 1.4198205901716877e-05,
+      "loss": 0.7251,
+      "step": 7470
+    },
+    {
+      "epoch": 4.263626647666548,
+      "grad_norm": 2.052440861765808,
+      "learning_rate": 1.4180135191326498e-05,
+      "loss": 0.7147,
+      "step": 7480
+    },
+    {
+      "epoch": 4.26932668329177,
+      "grad_norm": 2.1734320912245457,
+      "learning_rate": 1.4162047925305318e-05,
+      "loss": 0.7142,
+      "step": 7490
+    },
+    {
+      "epoch": 4.275026718916993,
+      "grad_norm": 2.2361008396279543,
+      "learning_rate": 1.4143944175288846e-05,
+      "loss": 0.7152,
+      "step": 7500
+    },
+    {
+      "epoch": 4.280726754542216,
+      "grad_norm": 2.0505673843747227,
+      "learning_rate": 1.4125824012977871e-05,
+      "loss": 0.7135,
+      "step": 7510
+    },
+    {
+      "epoch": 4.286426790167439,
+      "grad_norm": 2.107861616724096,
+      "learning_rate": 1.4107687510138193e-05,
+      "loss": 0.7124,
+      "step": 7520
+    },
+    {
+      "epoch": 4.2921268257926615,
+      "grad_norm": 2.076372674933872,
+      "learning_rate": 1.408953473860031e-05,
+      "loss": 0.719,
+      "step": 7530
+    },
+    {
+      "epoch": 4.297826861417883,
+      "grad_norm": 2.1240116212346734,
+      "learning_rate": 1.4071365770259175e-05,
+      "loss": 0.7076,
+      "step": 7540
+    },
+    {
+      "epoch": 4.303526897043106,
+      "grad_norm": 2.0692593377515944,
+      "learning_rate": 1.4053180677073877e-05,
+      "loss": 0.7181,
+      "step": 7550
+    },
+    {
+      "epoch": 4.309226932668329,
+      "grad_norm": 2.0202823988722702,
+      "learning_rate": 1.403497953106737e-05,
+      "loss": 0.7278,
+      "step": 7560
+    },
+    {
+      "epoch": 4.314926968293552,
+      "grad_norm": 2.1221365119576254,
+      "learning_rate": 1.4016762404326189e-05,
+      "loss": 0.7179,
+      "step": 7570
+    },
+    {
+      "epoch": 4.320627003918775,
+      "grad_norm": 2.048229882442656,
+      "learning_rate": 1.399852936900016e-05,
+      "loss": 0.7209,
+      "step": 7580
+    },
+    {
+      "epoch": 4.326327039543997,
+      "grad_norm": 2.1165056652873697,
+      "learning_rate": 1.3980280497302113e-05,
+      "loss": 0.7174,
+      "step": 7590
+    },
+    {
+      "epoch": 4.33202707516922,
+      "grad_norm": 2.0241250324247955,
+      "learning_rate": 1.39620158615076e-05,
+      "loss": 0.7197,
+      "step": 7600
+    },
+    {
+      "epoch": 4.337727110794442,
+      "grad_norm": 2.198418819325663,
+      "learning_rate": 1.3943735533954612e-05,
+      "loss": 0.7134,
+      "step": 7610
+    },
+    {
+      "epoch": 4.343427146419665,
+      "grad_norm": 2.0800472200100857,
+      "learning_rate": 1.392543958704328e-05,
+      "loss": 0.7173,
+      "step": 7620
+    },
+    {
+      "epoch": 4.349127182044888,
+      "grad_norm": 1.985410025478677,
+      "learning_rate": 1.3907128093235604e-05,
+      "loss": 0.7125,
+      "step": 7630
+    },
+    {
+      "epoch": 4.354827217670111,
+      "grad_norm": 1.9651488876336438,
+      "learning_rate": 1.3888801125055156e-05,
+      "loss": 0.7149,
+      "step": 7640
+    },
+    {
+      "epoch": 4.360527253295333,
+      "grad_norm": 2.043158401774305,
+      "learning_rate": 1.3870458755086793e-05,
+      "loss": 0.7023,
+      "step": 7650
+    },
+    {
+      "epoch": 4.366227288920555,
+      "grad_norm": 2.063696570625022,
+      "learning_rate": 1.3852101055976367e-05,
+      "loss": 0.7072,
+      "step": 7660
+    },
+    {
+      "epoch": 4.371927324545778,
+      "grad_norm": 2.2378999921887535,
+      "learning_rate": 1.3833728100430455e-05,
+      "loss": 0.7185,
+      "step": 7670
+    },
+    {
+      "epoch": 4.377627360171001,
+      "grad_norm": 2.2324714292824925,
+      "learning_rate": 1.3815339961216046e-05,
+      "loss": 0.7144,
+      "step": 7680
+    },
+    {
+      "epoch": 4.383327395796224,
+      "grad_norm": 2.010917436631881,
+      "learning_rate": 1.3796936711160269e-05,
+      "loss": 0.7162,
+      "step": 7690
+    },
+    {
+      "epoch": 4.389027431421447,
+      "grad_norm": 2.0127926011934245,
+      "learning_rate": 1.3778518423150101e-05,
+      "loss": 0.7119,
+      "step": 7700
+    },
+    {
+      "epoch": 4.394727467046669,
+      "grad_norm": 2.123951318878207,
+      "learning_rate": 1.3760085170132076e-05,
+      "loss": 0.7098,
+      "step": 7710
+    },
+    {
+      "epoch": 4.400427502671891,
+      "grad_norm": 2.0649137209507504,
+      "learning_rate": 1.3741637025112e-05,
+      "loss": 0.7142,
+      "step": 7720
+    },
+    {
+      "epoch": 4.406127538297114,
+      "grad_norm": 2.1529758269316117,
+      "learning_rate": 1.3723174061154652e-05,
+      "loss": 0.7104,
+      "step": 7730
+    },
+    {
+      "epoch": 4.411827573922337,
+      "grad_norm": 2.215341160229645,
+      "learning_rate": 1.3704696351383516e-05,
+      "loss": 0.716,
+      "step": 7740
+    },
+    {
+      "epoch": 4.41752760954756,
+      "grad_norm": 2.063609107382071,
+      "learning_rate": 1.3686203968980465e-05,
+      "loss": 0.7295,
+      "step": 7750
+    },
+    {
+      "epoch": 4.4232276451727826,
+      "grad_norm": 2.168009767132644,
+      "learning_rate": 1.3667696987185486e-05,
+      "loss": 0.7153,
+      "step": 7760
+    },
+    {
+      "epoch": 4.428927680798005,
+      "grad_norm": 2.1839100011866517,
+      "learning_rate": 1.3649175479296393e-05,
+      "loss": 0.7121,
+      "step": 7770
+    },
+    {
+      "epoch": 4.434627716423227,
+      "grad_norm": 2.1865961035302273,
+      "learning_rate": 1.3630639518668528e-05,
+      "loss": 0.7179,
+      "step": 7780
+    },
+    {
+      "epoch": 4.44032775204845,
+      "grad_norm": 2.1902585443901548,
+      "learning_rate": 1.3612089178714473e-05,
+      "loss": 0.7206,
+      "step": 7790
+    },
+    {
+      "epoch": 4.446027787673673,
+      "grad_norm": 2.0700382912342796,
+      "learning_rate": 1.3593524532903757e-05,
+      "loss": 0.7262,
+      "step": 7800
+    },
+    {
+      "epoch": 4.451727823298896,
+      "grad_norm": 1.991885261826358,
+      "learning_rate": 1.357494565476258e-05,
+      "loss": 0.7133,
+      "step": 7810
+    },
+    {
+      "epoch": 4.4574278589241185,
+      "grad_norm": 2.075547581598166,
+      "learning_rate": 1.3556352617873492e-05,
+      "loss": 0.7216,
+      "step": 7820
+    },
+    {
+      "epoch": 4.463127894549341,
+      "grad_norm": 2.069830636023337,
+      "learning_rate": 1.3537745495875138e-05,
+      "loss": 0.7069,
+      "step": 7830
+    },
+    {
+      "epoch": 4.468827930174563,
+      "grad_norm": 2.1761013806481775,
+      "learning_rate": 1.3519124362461937e-05,
+      "loss": 0.7175,
+      "step": 7840
+    },
+    {
+      "epoch": 4.474527965799786,
+      "grad_norm": 2.1477462457811294,
+      "learning_rate": 1.3500489291383798e-05,
+      "loss": 0.7019,
+      "step": 7850
+    },
+    {
+      "epoch": 4.480228001425009,
+      "grad_norm": 2.091553579148087,
+      "learning_rate": 1.348184035644584e-05,
+      "loss": 0.7056,
+      "step": 7860
+    },
+    {
+      "epoch": 4.485928037050232,
+      "grad_norm": 2.175016873764588,
+      "learning_rate": 1.3463177631508079e-05,
+      "loss": 0.7065,
+      "step": 7870
+    },
+    {
+      "epoch": 4.4916280726754545,
+      "grad_norm": 2.0416242410436483,
+      "learning_rate": 1.3444501190485164e-05,
+      "loss": 0.7064,
+      "step": 7880
+    },
+    {
+      "epoch": 4.497328108300676,
+      "grad_norm": 2.047781511840451,
+      "learning_rate": 1.3425811107346052e-05,
+      "loss": 0.7087,
+      "step": 7890
+    },
+    {
+      "epoch": 4.503028143925899,
+      "grad_norm": 2.120940746397696,
+      "learning_rate": 1.3407107456113737e-05,
+      "loss": 0.7087,
+      "step": 7900
+    },
+    {
+      "epoch": 4.508728179551122,
+      "grad_norm": 2.182832991914863,
+      "learning_rate": 1.3388390310864945e-05,
+      "loss": 0.7146,
+      "step": 7910
+    },
+    {
+      "epoch": 4.514428215176345,
+      "grad_norm": 2.0731174610042014,
+      "learning_rate": 1.3369659745729854e-05,
+      "loss": 0.7129,
+      "step": 7920
+    },
+    {
+      "epoch": 4.520128250801568,
+      "grad_norm": 2.2501441238665967,
+      "learning_rate": 1.3350915834891786e-05,
+      "loss": 0.7174,
+      "step": 7930
+    },
+    {
+      "epoch": 4.5258282864267905,
+      "grad_norm": 2.0216070729039166,
+      "learning_rate": 1.333215865258692e-05,
+      "loss": 0.7129,
+      "step": 7940
+    },
+    {
+      "epoch": 4.531528322052013,
+      "grad_norm": 1.9716477987035963,
+      "learning_rate": 1.3313388273103999e-05,
+      "loss": 0.7152,
+      "step": 7950
+    },
+    {
+      "epoch": 4.537228357677235,
+      "grad_norm": 2.0783658817431285,
+      "learning_rate": 1.3294604770784035e-05,
+      "loss": 0.7132,
+      "step": 7960
+    },
+    {
+      "epoch": 4.542928393302458,
+      "grad_norm": 2.261698541426297,
+      "learning_rate": 1.3275808220020006e-05,
+      "loss": 0.7131,
+      "step": 7970
+    },
+    {
+      "epoch": 4.548628428927681,
+      "grad_norm": 2.1347742048994487,
+      "learning_rate": 1.3256998695256578e-05,
+      "loss": 0.7175,
+      "step": 7980
+    },
+    {
+      "epoch": 4.554328464552904,
+      "grad_norm": 2.140168139545353,
+      "learning_rate": 1.32381762709898e-05,
+      "loss": 0.7203,
+      "step": 7990
+    },
+    {
+      "epoch": 4.560028500178126,
+      "grad_norm": 2.1547186591948493,
+      "learning_rate": 1.3219341021766803e-05,
+      "loss": 0.7064,
+      "step": 8000
+    },
+    {
+      "epoch": 4.565728535803348,
+      "grad_norm": 2.1008597696102562,
+      "learning_rate": 1.3200493022185525e-05,
+      "loss": 0.7195,
+      "step": 8010
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 2.184478573038733,
+      "learning_rate": 1.3181632346894388e-05,
+      "loss": 0.7171,
+      "step": 8020
+    },
+    {
+      "epoch": 4.577128607053794,
+      "grad_norm": 2.0573951158453125,
+      "learning_rate": 1.3162759070592024e-05,
+      "loss": 0.7108,
+      "step": 8030
+    },
+    {
+      "epoch": 4.582828642679017,
+      "grad_norm": 2.2271090428071085,
+      "learning_rate": 1.314387326802697e-05,
+      "loss": 0.7104,
+      "step": 8040
+    },
+    {
+      "epoch": 4.58852867830424,
+      "grad_norm": 2.122074817755831,
+      "learning_rate": 1.312497501399738e-05,
+      "loss": 0.712,
+      "step": 8050
+    },
+    {
+      "epoch": 4.594228713929462,
+      "grad_norm": 2.0920821005791455,
+      "learning_rate": 1.3106064383350715e-05,
+      "loss": 0.7187,
+      "step": 8060
+    },
+    {
+      "epoch": 4.599928749554684,
+      "grad_norm": 2.116249375789271,
+      "learning_rate": 1.308714145098345e-05,
+      "loss": 0.7225,
+      "step": 8070
+    },
+    {
+      "epoch": 4.605628785179907,
+      "grad_norm": 2.136803566912254,
+      "learning_rate": 1.3068206291840799e-05,
+      "loss": 0.7096,
+      "step": 8080
+    },
+    {
+      "epoch": 4.61132882080513,
+      "grad_norm": 2.124886312634362,
+      "learning_rate": 1.3049258980916387e-05,
+      "loss": 0.717,
+      "step": 8090
+    },
+    {
+      "epoch": 4.617028856430353,
+      "grad_norm": 2.121467934682898,
+      "learning_rate": 1.3030299593251964e-05,
+      "loss": 0.7188,
+      "step": 8100
+    },
+    {
+      "epoch": 4.622728892055576,
+      "grad_norm": 1.9399148435216957,
+      "learning_rate": 1.3011328203937121e-05,
+      "loss": 0.7071,
+      "step": 8110
+    },
+    {
+      "epoch": 4.628428927680798,
+      "grad_norm": 2.154407536044687,
+      "learning_rate": 1.2992344888108981e-05,
+      "loss": 0.7141,
+      "step": 8120
+    },
+    {
+      "epoch": 4.63412896330602,
+      "grad_norm": 2.032093437381081,
+      "learning_rate": 1.297334972095189e-05,
+      "loss": 0.713,
+      "step": 8130
+    },
+    {
+      "epoch": 4.639828998931243,
+      "grad_norm": 2.0495210294018436,
+      "learning_rate": 1.2954342777697152e-05,
+      "loss": 0.7026,
+      "step": 8140
+    },
+    {
+      "epoch": 4.645529034556466,
+      "grad_norm": 2.1069899723792616,
+      "learning_rate": 1.2935324133622688e-05,
+      "loss": 0.7108,
+      "step": 8150
+    },
+    {
+      "epoch": 4.651229070181689,
+      "grad_norm": 2.1144338292156792,
+      "learning_rate": 1.291629386405278e-05,
+      "loss": 0.7091,
+      "step": 8160
+    },
+    {
+      "epoch": 4.6569291058069116,
+      "grad_norm": 2.0212838390369585,
+      "learning_rate": 1.2897252044357745e-05,
+      "loss": 0.7193,
+      "step": 8170
+    },
+    {
+      "epoch": 4.662629141432134,
+      "grad_norm": 2.100544352290128,
+      "learning_rate": 1.2878198749953642e-05,
+      "loss": 0.7222,
+      "step": 8180
+    },
+    {
+      "epoch": 4.668329177057356,
+      "grad_norm": 2.004292862457442,
+      "learning_rate": 1.285913405630198e-05,
+      "loss": 0.71,
+      "step": 8190
+    },
+    {
+      "epoch": 4.674029212682579,
+      "grad_norm": 2.066805788673636,
+      "learning_rate": 1.2840058038909415e-05,
+      "loss": 0.7058,
+      "step": 8200
+    },
+    {
+      "epoch": 4.679729248307802,
+      "grad_norm": 2.1905106670861714,
+      "learning_rate": 1.2820970773327456e-05,
+      "loss": 0.7139,
+      "step": 8210
+    },
+    {
+      "epoch": 4.685429283933025,
+      "grad_norm": 2.113303850774233,
+      "learning_rate": 1.2801872335152152e-05,
+      "loss": 0.7075,
+      "step": 8220
+    },
+    {
+      "epoch": 4.6911293195582475,
+      "grad_norm": 2.1795819458323464,
+      "learning_rate": 1.2782762800023806e-05,
+      "loss": 0.7069,
+      "step": 8230
+    },
+    {
+      "epoch": 4.6968293551834694,
+      "grad_norm": 2.1289949423759658,
+      "learning_rate": 1.2763642243626668e-05,
+      "loss": 0.7087,
+      "step": 8240
+    },
+    {
+      "epoch": 4.702529390808692,
+      "grad_norm": 2.3383021717285266,
+      "learning_rate": 1.2744510741688648e-05,
+      "loss": 0.7047,
+      "step": 8250
+    },
+    {
+      "epoch": 4.708229426433915,
+      "grad_norm": 2.0055553076818424,
+      "learning_rate": 1.2725368369980996e-05,
+      "loss": 0.7204,
+      "step": 8260
+    },
+    {
+      "epoch": 4.713929462059138,
+      "grad_norm": 2.1654348631523486,
+      "learning_rate": 1.270621520431801e-05,
+      "loss": 0.7111,
+      "step": 8270
+    },
+    {
+      "epoch": 4.719629497684361,
+      "grad_norm": 2.1644826955051664,
+      "learning_rate": 1.2687051320556751e-05,
+      "loss": 0.7067,
+      "step": 8280
+    },
+    {
+      "epoch": 4.7253295333095835,
+      "grad_norm": 2.089047757071869,
+      "learning_rate": 1.2667876794596721e-05,
+      "loss": 0.7161,
+      "step": 8290
+    },
+    {
+      "epoch": 4.731029568934806,
+      "grad_norm": 2.0039738679317507,
+      "learning_rate": 1.2648691702379568e-05,
+      "loss": 0.7004,
+      "step": 8300
+    },
+    {
+      "epoch": 4.736729604560028,
+      "grad_norm": 2.2125009851795343,
+      "learning_rate": 1.2629496119888795e-05,
+      "loss": 0.699,
+      "step": 8310
+    },
+    {
+      "epoch": 4.742429640185251,
+      "grad_norm": 2.187645275278001,
+      "learning_rate": 1.2610290123149454e-05,
+      "loss": 0.7087,
+      "step": 8320
+    },
+    {
+      "epoch": 4.748129675810474,
+      "grad_norm": 2.1075929722847757,
+      "learning_rate": 1.2591073788227827e-05,
+      "loss": 0.7161,
+      "step": 8330
+    },
+    {
+      "epoch": 4.753829711435697,
+      "grad_norm": 2.0657586426111414,
+      "learning_rate": 1.257184719123117e-05,
+      "loss": 0.7142,
+      "step": 8340
+    },
+    {
+      "epoch": 4.7595297470609195,
+      "grad_norm": 2.085585284929549,
+      "learning_rate": 1.2552610408307348e-05,
+      "loss": 0.7126,
+      "step": 8350
+    },
+    {
+      "epoch": 4.765229782686141,
+      "grad_norm": 2.017479984454758,
+      "learning_rate": 1.2533363515644595e-05,
+      "loss": 0.7135,
+      "step": 8360
+    },
+    {
+      "epoch": 4.770929818311364,
+      "grad_norm": 2.1270569310907774,
+      "learning_rate": 1.2514106589471169e-05,
+      "loss": 0.7187,
+      "step": 8370
+    },
+    {
+      "epoch": 4.776629853936587,
+      "grad_norm": 2.08061530850382,
+      "learning_rate": 1.2494839706055075e-05,
+      "loss": 0.7017,
+      "step": 8380
+    },
+    {
+      "epoch": 4.78232988956181,
+      "grad_norm": 2.0904751797462446,
+      "learning_rate": 1.2475562941703755e-05,
+      "loss": 0.7078,
+      "step": 8390
+    },
+    {
+      "epoch": 4.788029925187033,
+      "grad_norm": 2.079199224404677,
+      "learning_rate": 1.2456276372763776e-05,
+      "loss": 0.7135,
+      "step": 8400
+    },
+    {
+      "epoch": 4.793729960812255,
+      "grad_norm": 1.9760702520656304,
+      "learning_rate": 1.2436980075620543e-05,
+      "loss": 0.7227,
+      "step": 8410
+    },
+    {
+      "epoch": 4.799429996437477,
+      "grad_norm": 2.0407360391918266,
+      "learning_rate": 1.2417674126697989e-05,
+      "loss": 0.7125,
+      "step": 8420
+    },
+    {
+      "epoch": 4.8051300320627,
+      "grad_norm": 2.0872991926867575,
+      "learning_rate": 1.2398358602458275e-05,
+      "loss": 0.7094,
+      "step": 8430
+    },
+    {
+      "epoch": 4.810830067687923,
+      "grad_norm": 2.075605380407668,
+      "learning_rate": 1.2379033579401483e-05,
+      "loss": 0.7029,
+      "step": 8440
+    },
+    {
+      "epoch": 4.816530103313146,
+      "grad_norm": 2.0695319205546863,
+      "learning_rate": 1.2359699134065316e-05,
+      "loss": 0.7028,
+      "step": 8450
+    },
+    {
+      "epoch": 4.822230138938369,
+      "grad_norm": 2.142832073636643,
+      "learning_rate": 1.2340355343024793e-05,
+      "loss": 0.709,
+      "step": 8460
+    },
+    {
+      "epoch": 4.8279301745635905,
+      "grad_norm": 2.050091950243262,
+      "learning_rate": 1.2321002282891952e-05,
+      "loss": 0.7006,
+      "step": 8470
+    },
+    {
+      "epoch": 4.833630210188813,
+      "grad_norm": 2.024439862175209,
+      "learning_rate": 1.2301640030315537e-05,
+      "loss": 0.7033,
+      "step": 8480
+    },
+    {
+      "epoch": 4.839330245814036,
+      "grad_norm": 2.0787794513253544,
+      "learning_rate": 1.2282268661980697e-05,
+      "loss": 0.7032,
+      "step": 8490
+    },
+    {
+      "epoch": 4.845030281439259,
+      "grad_norm": 2.174009548757474,
+      "learning_rate": 1.2262888254608691e-05,
+      "loss": 0.6981,
+      "step": 8500
+    },
+    {
+      "epoch": 4.850730317064482,
+      "grad_norm": 2.0765315091697043,
+      "learning_rate": 1.2243498884956578e-05,
+      "loss": 0.7065,
+      "step": 8510
+    },
+    {
+      "epoch": 4.856430352689705,
+      "grad_norm": 2.038584770594007,
+      "learning_rate": 1.2224100629816905e-05,
+      "loss": 0.7143,
+      "step": 8520
+    },
+    {
+      "epoch": 4.862130388314927,
+      "grad_norm": 2.0005800600511168,
+      "learning_rate": 1.2204693566017417e-05,
+      "loss": 0.7185,
+      "step": 8530
+    },
+    {
+      "epoch": 4.867830423940149,
+      "grad_norm": 2.1842325091967703,
+      "learning_rate": 1.2185277770420739e-05,
+      "loss": 0.7134,
+      "step": 8540
+    },
+    {
+      "epoch": 4.873530459565372,
+      "grad_norm": 2.130842764618701,
+      "learning_rate": 1.2165853319924088e-05,
+      "loss": 0.7213,
+      "step": 8550
+    },
+    {
+      "epoch": 4.879230495190595,
+      "grad_norm": 2.1578893100012078,
+      "learning_rate": 1.2146420291458954e-05,
+      "loss": 0.7084,
+      "step": 8560
+    },
+    {
+      "epoch": 4.884930530815818,
+      "grad_norm": 2.1830068437308934,
+      "learning_rate": 1.21269787619908e-05,
+      "loss": 0.7061,
+      "step": 8570
+    },
+    {
+      "epoch": 4.8906305664410405,
+      "grad_norm": 2.0507704117164516,
+      "learning_rate": 1.2107528808518756e-05,
+      "loss": 0.7121,
+      "step": 8580
+    },
+    {
+      "epoch": 4.8963306020662625,
+      "grad_norm": 2.2296960181873975,
+      "learning_rate": 1.2088070508075325e-05,
+      "loss": 0.7134,
+      "step": 8590
+    },
+    {
+      "epoch": 4.902030637691485,
+      "grad_norm": 2.2294466692347052,
+      "learning_rate": 1.2068603937726057e-05,
+      "loss": 0.7018,
+      "step": 8600
+    },
+    {
+      "epoch": 4.907730673316708,
+      "grad_norm": 2.3162681168819486,
+      "learning_rate": 1.2049129174569261e-05,
+      "loss": 0.7129,
+      "step": 8610
+    },
+    {
+      "epoch": 4.913430708941931,
+      "grad_norm": 2.2283898023173556,
+      "learning_rate": 1.2029646295735694e-05,
+      "loss": 0.7033,
+      "step": 8620
+    },
+    {
+      "epoch": 4.919130744567154,
+      "grad_norm": 2.1019985515460213,
+      "learning_rate": 1.2010155378388253e-05,
+      "loss": 0.7102,
+      "step": 8630
+    },
+    {
+      "epoch": 4.9248307801923765,
+      "grad_norm": 2.104644017177413,
+      "learning_rate": 1.1990656499721673e-05,
+      "loss": 0.7059,
+      "step": 8640
+    },
+    {
+      "epoch": 4.930530815817599,
+      "grad_norm": 2.1752936170393844,
+      "learning_rate": 1.1971149736962229e-05,
+      "loss": 0.7019,
+      "step": 8650
+    },
+    {
+      "epoch": 4.936230851442821,
+      "grad_norm": 2.1210263601772774,
+      "learning_rate": 1.1951635167367403e-05,
+      "loss": 0.7094,
+      "step": 8660
+    },
+    {
+      "epoch": 4.941930887068044,
+      "grad_norm": 2.1174594655618657,
+      "learning_rate": 1.1932112868225613e-05,
+      "loss": 0.7186,
+      "step": 8670
+    },
+    {
+      "epoch": 4.947630922693267,
+      "grad_norm": 2.0085889940321304,
+      "learning_rate": 1.1912582916855883e-05,
+      "loss": 0.7129,
+      "step": 8680
+    },
+    {
+      "epoch": 4.95333095831849,
+      "grad_norm": 2.1847254708548682,
+      "learning_rate": 1.1893045390607542e-05,
+      "loss": 0.7101,
+      "step": 8690
+    },
+    {
+      "epoch": 4.9590309939437125,
+      "grad_norm": 2.1005181730831346,
+      "learning_rate": 1.1873500366859925e-05,
+      "loss": 0.7117,
+      "step": 8700
+    },
+    {
+      "epoch": 4.964731029568934,
+      "grad_norm": 2.197303113242465,
+      "learning_rate": 1.1853947923022057e-05,
+      "loss": 0.7124,
+      "step": 8710
+    },
+    {
+      "epoch": 4.970431065194157,
+      "grad_norm": 2.119459567257772,
+      "learning_rate": 1.1834388136532358e-05,
+      "loss": 0.7061,
+      "step": 8720
+    },
+    {
+      "epoch": 4.97613110081938,
+      "grad_norm": 2.106855530802309,
+      "learning_rate": 1.1814821084858315e-05,
+      "loss": 0.7056,
+      "step": 8730
+    },
+    {
+      "epoch": 4.981831136444603,
+      "grad_norm": 2.1509371468107608,
+      "learning_rate": 1.1795246845496205e-05,
+      "loss": 0.6997,
+      "step": 8740
+    },
+    {
+      "epoch": 4.987531172069826,
+      "grad_norm": 2.0890300669870783,
+      "learning_rate": 1.1775665495970756e-05,
+      "loss": 0.7088,
+      "step": 8750
+    },
+    {
+      "epoch": 4.9932312076950485,
+      "grad_norm": 2.132056711702782,
+      "learning_rate": 1.1756077113834873e-05,
+      "loss": 0.705,
+      "step": 8760
+    },
+    {
+      "epoch": 4.99893124332027,
+      "grad_norm": 2.153283209947663,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 0.7137,
+      "step": 8770
+    },
+    {
+      "epoch": 5.004631278945493,
+      "grad_norm": 2.1163261553034682,
+      "learning_rate": 1.1716879562082343e-05,
+      "loss": 0.6993,
+      "step": 8780
+    },
+    {
+      "epoch": 5.010331314570716,
+      "grad_norm": 2.1583328617449897,
+      "learning_rate": 1.1697270547709527e-05,
+      "loss": 0.6859,
+      "step": 8790
+    },
+    {
+      "epoch": 5.016031350195939,
+      "grad_norm": 2.1781625538447145,
+      "learning_rate": 1.1677654811213316e-05,
+      "loss": 0.6986,
+      "step": 8800
+    },
+    {
+      "epoch": 5.021731385821162,
+      "grad_norm": 2.12558766978219,
+      "learning_rate": 1.16580324302828e-05,
+      "loss": 0.6982,
+      "step": 8810
+    },
+    {
+      "epoch": 5.027431421446384,
+      "grad_norm": 2.176681743446634,
+      "learning_rate": 1.1638403482633383e-05,
+      "loss": 0.7033,
+      "step": 8820
+    },
+    {
+      "epoch": 5.033131457071606,
+      "grad_norm": 2.079820413393339,
+      "learning_rate": 1.1618768046006476e-05,
+      "loss": 0.673,
+      "step": 8830
+    },
+    {
+      "epoch": 5.038831492696829,
+      "grad_norm": 2.0700636259608696,
+      "learning_rate": 1.1599126198169196e-05,
+      "loss": 0.6934,
+      "step": 8840
+    },
+    {
+      "epoch": 5.044531528322052,
+      "grad_norm": 2.151973353252644,
+      "learning_rate": 1.1579478016914038e-05,
+      "loss": 0.6989,
+      "step": 8850
+    },
+    {
+      "epoch": 5.050231563947275,
+      "grad_norm": 2.129485987400305,
+      "learning_rate": 1.1559823580058591e-05,
+      "loss": 0.6962,
+      "step": 8860
+    },
+    {
+      "epoch": 5.055931599572498,
+      "grad_norm": 2.1772520549019405,
+      "learning_rate": 1.1540162965445224e-05,
+      "loss": 0.6789,
+      "step": 8870
+    },
+    {
+      "epoch": 5.06163163519772,
+      "grad_norm": 2.1068284615452515,
+      "learning_rate": 1.152049625094076e-05,
+      "loss": 0.6963,
+      "step": 8880
+    },
+    {
+      "epoch": 5.067331670822942,
+      "grad_norm": 2.1446212460776137,
+      "learning_rate": 1.150082351443619e-05,
+      "loss": 0.6852,
+      "step": 8890
+    },
+    {
+      "epoch": 5.073031706448165,
+      "grad_norm": 2.1765018405494434,
+      "learning_rate": 1.1481144833846358e-05,
+      "loss": 0.6904,
+      "step": 8900
+    },
+    {
+      "epoch": 5.078731742073388,
+      "grad_norm": 2.2251286404486788,
+      "learning_rate": 1.146146028710964e-05,
+      "loss": 0.6896,
+      "step": 8910
+    },
+    {
+      "epoch": 5.084431777698611,
+      "grad_norm": 2.0897124016602113,
+      "learning_rate": 1.144176995218765e-05,
+      "loss": 0.6951,
+      "step": 8920
+    },
+    {
+      "epoch": 5.090131813323834,
+      "grad_norm": 2.1995547700271887,
+      "learning_rate": 1.1422073907064932e-05,
+      "loss": 0.6918,
+      "step": 8930
+    },
+    {
+      "epoch": 5.0958318489490555,
+      "grad_norm": 2.3209537414628434,
+      "learning_rate": 1.1402372229748635e-05,
+      "loss": 0.6855,
+      "step": 8940
+    },
+    {
+      "epoch": 5.101531884574278,
+      "grad_norm": 2.0555932442118343,
+      "learning_rate": 1.1382664998268222e-05,
+      "loss": 0.6899,
+      "step": 8950
+    },
+    {
+      "epoch": 5.107231920199501,
+      "grad_norm": 2.044771606602399,
+      "learning_rate": 1.1362952290675153e-05,
+      "loss": 0.6946,
+      "step": 8960
+    },
+    {
+      "epoch": 5.112931955824724,
+      "grad_norm": 2.061837650585162,
+      "learning_rate": 1.1343234185042575e-05,
+      "loss": 0.69,
+      "step": 8970
+    },
+    {
+      "epoch": 5.118631991449947,
+      "grad_norm": 2.151710810846666,
+      "learning_rate": 1.1323510759465012e-05,
+      "loss": 0.6932,
+      "step": 8980
+    },
+    {
+      "epoch": 5.1243320270751695,
+      "grad_norm": 2.1434546309764704,
+      "learning_rate": 1.1303782092058062e-05,
+      "loss": 0.695,
+      "step": 8990
+    },
+    {
+      "epoch": 5.1300320627003915,
+      "grad_norm": 2.1143478500205597,
+      "learning_rate": 1.1284048260958076e-05,
+      "loss": 0.6916,
+      "step": 9000
+    },
+    {
+      "epoch": 5.135732098325614,
+      "grad_norm": 2.1239759708768737,
+      "learning_rate": 1.126430934432187e-05,
+      "loss": 0.7005,
+      "step": 9010
+    },
+    {
+      "epoch": 5.141432133950837,
+      "grad_norm": 2.0615509246767463,
+      "learning_rate": 1.1244565420326388e-05,
+      "loss": 0.6972,
+      "step": 9020
+    },
+    {
+      "epoch": 5.14713216957606,
+      "grad_norm": 2.2157339698853917,
+      "learning_rate": 1.1224816567168413e-05,
+      "loss": 0.6872,
+      "step": 9030
+    },
+    {
+      "epoch": 5.152832205201283,
+      "grad_norm": 2.1414826849096045,
+      "learning_rate": 1.1205062863064247e-05,
+      "loss": 0.7017,
+      "step": 9040
+    },
+    {
+      "epoch": 5.1585322408265055,
+      "grad_norm": 2.0947592953491916,
+      "learning_rate": 1.1185304386249405e-05,
+      "loss": 0.6879,
+      "step": 9050
+    },
+    {
+      "epoch": 5.164232276451727,
+      "grad_norm": 2.241240477683422,
+      "learning_rate": 1.1165541214978306e-05,
+      "loss": 0.6969,
+      "step": 9060
+    },
+    {
+      "epoch": 5.16993231207695,
+      "grad_norm": 2.195267971963587,
+      "learning_rate": 1.1145773427523963e-05,
+      "loss": 0.6879,
+      "step": 9070
+    },
+    {
+      "epoch": 5.175632347702173,
+      "grad_norm": 2.2669294865119647,
+      "learning_rate": 1.1126001102177667e-05,
+      "loss": 0.7013,
+      "step": 9080
+    },
+    {
+      "epoch": 5.181332383327396,
+      "grad_norm": 2.123451159279259,
+      "learning_rate": 1.1106224317248682e-05,
+      "loss": 0.693,
+      "step": 9090
+    },
+    {
+      "epoch": 5.187032418952619,
+      "grad_norm": 2.236333085476668,
+      "learning_rate": 1.108644315106394e-05,
+      "loss": 0.7021,
+      "step": 9100
+    },
+    {
+      "epoch": 5.1927324545778415,
+      "grad_norm": 2.092129797924252,
+      "learning_rate": 1.1066657681967723e-05,
+      "loss": 0.6828,
+      "step": 9110
+    },
+    {
+      "epoch": 5.198432490203063,
+      "grad_norm": 2.130859878723095,
+      "learning_rate": 1.1046867988321349e-05,
+      "loss": 0.6917,
+      "step": 9120
+    },
+    {
+      "epoch": 5.204132525828286,
+      "grad_norm": 2.0855014986112743,
+      "learning_rate": 1.102707414850287e-05,
+      "loss": 0.7009,
+      "step": 9130
+    },
+    {
+      "epoch": 5.209832561453509,
+      "grad_norm": 2.221078070968992,
+      "learning_rate": 1.100727624090677e-05,
+      "loss": 0.6877,
+      "step": 9140
+    },
+    {
+      "epoch": 5.215532597078732,
+      "grad_norm": 2.1677759083316617,
+      "learning_rate": 1.0987474343943625e-05,
+      "loss": 0.6892,
+      "step": 9150
+    },
+    {
+      "epoch": 5.221232632703955,
+      "grad_norm": 2.1121478588367304,
+      "learning_rate": 1.0967668536039828e-05,
+      "loss": 0.686,
+      "step": 9160
+    },
+    {
+      "epoch": 5.2269326683291775,
+      "grad_norm": 2.140684147466254,
+      "learning_rate": 1.0947858895637255e-05,
+      "loss": 0.7028,
+      "step": 9170
+    },
+    {
+      "epoch": 5.232632703954399,
+      "grad_norm": 2.1807307861135676,
+      "learning_rate": 1.0928045501192952e-05,
+      "loss": 0.6941,
+      "step": 9180
+    },
+    {
+      "epoch": 5.238332739579622,
+      "grad_norm": 2.208380922343372,
+      "learning_rate": 1.0908228431178847e-05,
+      "loss": 0.7014,
+      "step": 9190
+    },
+    {
+      "epoch": 5.244032775204845,
+      "grad_norm": 2.1946065233509287,
+      "learning_rate": 1.0888407764081416e-05,
+      "loss": 0.6871,
+      "step": 9200
+    },
+    {
+      "epoch": 5.249732810830068,
+      "grad_norm": 2.135339588212707,
+      "learning_rate": 1.0868583578401391e-05,
+      "loss": 0.6838,
+      "step": 9210
+    },
+    {
+      "epoch": 5.255432846455291,
+      "grad_norm": 2.2183442641416327,
+      "learning_rate": 1.0848755952653426e-05,
+      "loss": 0.7054,
+      "step": 9220
+    },
+    {
+      "epoch": 5.261132882080513,
+      "grad_norm": 2.107836660256195,
+      "learning_rate": 1.0828924965365814e-05,
+      "loss": 0.6933,
+      "step": 9230
+    },
+    {
+      "epoch": 5.266832917705735,
+      "grad_norm": 2.2166875612836248,
+      "learning_rate": 1.0809090695080148e-05,
+      "loss": 0.686,
+      "step": 9240
+    },
+    {
+      "epoch": 5.272532953330958,
+      "grad_norm": 2.104650506590125,
+      "learning_rate": 1.0789253220351035e-05,
+      "loss": 0.6979,
+      "step": 9250
+    },
+    {
+      "epoch": 5.278232988956181,
+      "grad_norm": 2.0369346584749244,
+      "learning_rate": 1.0769412619745762e-05,
+      "loss": 0.6888,
+      "step": 9260
+    },
+    {
+      "epoch": 5.283933024581404,
+      "grad_norm": 2.182316976219354,
+      "learning_rate": 1.0749568971844011e-05,
+      "loss": 0.6962,
+      "step": 9270
+    },
+    {
+      "epoch": 5.289633060206627,
+      "grad_norm": 2.08731443365099,
+      "learning_rate": 1.0729722355237519e-05,
+      "loss": 0.6888,
+      "step": 9280
+    },
+    {
+      "epoch": 5.2953330958318485,
+      "grad_norm": 2.07024796554627,
+      "learning_rate": 1.0709872848529787e-05,
+      "loss": 0.6942,
+      "step": 9290
+    },
+    {
+      "epoch": 5.301033131457071,
+      "grad_norm": 2.135528373311544,
+      "learning_rate": 1.0690020530335764e-05,
+      "loss": 0.6944,
+      "step": 9300
+    },
+    {
+      "epoch": 5.306733167082294,
+      "grad_norm": 2.0853893054463732,
+      "learning_rate": 1.0670165479281522e-05,
+      "loss": 0.6955,
+      "step": 9310
+    },
+    {
+      "epoch": 5.312433202707517,
+      "grad_norm": 2.1518235426488115,
+      "learning_rate": 1.065030777400398e-05,
+      "loss": 0.6836,
+      "step": 9320
+    },
+    {
+      "epoch": 5.31813323833274,
+      "grad_norm": 2.2237139366167478,
+      "learning_rate": 1.0630447493150547e-05,
+      "loss": 0.7044,
+      "step": 9330
+    },
+    {
+      "epoch": 5.323833273957963,
+      "grad_norm": 2.213573275816285,
+      "learning_rate": 1.0610584715378843e-05,
+      "loss": 0.6893,
+      "step": 9340
+    },
+    {
+      "epoch": 5.3295333095831845,
+      "grad_norm": 2.1970195247994906,
+      "learning_rate": 1.0590719519356373e-05,
+      "loss": 0.7029,
+      "step": 9350
+    },
+    {
+      "epoch": 5.335233345208407,
+      "grad_norm": 2.1836535830369037,
+      "learning_rate": 1.0570851983760228e-05,
+      "loss": 0.6918,
+      "step": 9360
+    },
+    {
+      "epoch": 5.34093338083363,
+      "grad_norm": 2.2205086474880513,
+      "learning_rate": 1.0550982187276752e-05,
+      "loss": 0.6962,
+      "step": 9370
+    },
+    {
+      "epoch": 5.346633416458853,
+      "grad_norm": 2.0062208831981163,
+      "learning_rate": 1.0531110208601254e-05,
+      "loss": 0.7039,
+      "step": 9380
+    },
+    {
+      "epoch": 5.352333452084076,
+      "grad_norm": 2.197549239307711,
+      "learning_rate": 1.0511236126437682e-05,
+      "loss": 0.6922,
+      "step": 9390
+    },
+    {
+      "epoch": 5.3580334877092985,
+      "grad_norm": 2.1844139260865623,
+      "learning_rate": 1.0491360019498312e-05,
+      "loss": 0.6929,
+      "step": 9400
+    },
+    {
+      "epoch": 5.3637335233345205,
+      "grad_norm": 2.2073816657140957,
+      "learning_rate": 1.0471481966503446e-05,
+      "loss": 0.6905,
+      "step": 9410
+    },
+    {
+      "epoch": 5.369433558959743,
+      "grad_norm": 2.0734746928445227,
+      "learning_rate": 1.0451602046181084e-05,
+      "loss": 0.6809,
+      "step": 9420
+    },
+    {
+      "epoch": 5.375133594584966,
+      "grad_norm": 2.0981553256945515,
+      "learning_rate": 1.0431720337266632e-05,
+      "loss": 0.6954,
+      "step": 9430
+    },
+    {
+      "epoch": 5.380833630210189,
+      "grad_norm": 1.9956443908942936,
+      "learning_rate": 1.0411836918502573e-05,
+      "loss": 0.6869,
+      "step": 9440
+    },
+    {
+      "epoch": 5.386533665835412,
+      "grad_norm": 2.160809451240416,
+      "learning_rate": 1.0391951868638167e-05,
+      "loss": 0.6908,
+      "step": 9450
+    },
+    {
+      "epoch": 5.3922337014606345,
+      "grad_norm": 2.1171495609558932,
+      "learning_rate": 1.0372065266429124e-05,
+      "loss": 0.693,
+      "step": 9460
+    },
+    {
+      "epoch": 5.397933737085856,
+      "grad_norm": 2.1299349506902123,
+      "learning_rate": 1.0352177190637315e-05,
+      "loss": 0.6885,
+      "step": 9470
+    },
+    {
+      "epoch": 5.403633772711079,
+      "grad_norm": 2.2267815303969356,
+      "learning_rate": 1.0332287720030442e-05,
+      "loss": 0.6905,
+      "step": 9480
+    },
+    {
+      "epoch": 5.409333808336302,
+      "grad_norm": 2.173479864292032,
+      "learning_rate": 1.0312396933381728e-05,
+      "loss": 0.6978,
+      "step": 9490
+    },
+    {
+      "epoch": 5.415033843961525,
+      "grad_norm": 2.0878022848536077,
+      "learning_rate": 1.0292504909469612e-05,
+      "loss": 0.6881,
+      "step": 9500
+    },
+    {
+      "epoch": 5.420733879586748,
+      "grad_norm": 2.2809403605624077,
+      "learning_rate": 1.0272611727077426e-05,
+      "loss": 0.691,
+      "step": 9510
+    },
+    {
+      "epoch": 5.42643391521197,
+      "grad_norm": 2.09129513356794,
+      "learning_rate": 1.0252717464993105e-05,
+      "loss": 0.6909,
+      "step": 9520
+    },
+    {
+      "epoch": 5.432133950837192,
+      "grad_norm": 2.1587358103587326,
+      "learning_rate": 1.0232822202008845e-05,
+      "loss": 0.6877,
+      "step": 9530
+    },
+    {
+      "epoch": 5.437833986462415,
+      "grad_norm": 2.0969324770102014,
+      "learning_rate": 1.0212926016920816e-05,
+      "loss": 0.6854,
+      "step": 9540
+    },
+    {
+      "epoch": 5.443534022087638,
+      "grad_norm": 2.2219797664048375,
+      "learning_rate": 1.019302898852884e-05,
+      "loss": 0.6834,
+      "step": 9550
+    },
+    {
+      "epoch": 5.449234057712861,
+      "grad_norm": 2.245112065626013,
+      "learning_rate": 1.0173131195636068e-05,
+      "loss": 0.6935,
+      "step": 9560
+    },
+    {
+      "epoch": 5.454934093338084,
+      "grad_norm": 2.207368160265503,
+      "learning_rate": 1.0153232717048686e-05,
+      "loss": 0.6978,
+      "step": 9570
+    },
+    {
+      "epoch": 5.4606341289633065,
+      "grad_norm": 2.1020563924041133,
+      "learning_rate": 1.0133333631575606e-05,
+      "loss": 0.694,
+      "step": 9580
+    },
+    {
+      "epoch": 5.466334164588528,
+      "grad_norm": 2.125673123994935,
+      "learning_rate": 1.0113434018028124e-05,
+      "loss": 0.6774,
+      "step": 9590
+    },
+    {
+      "epoch": 5.472034200213751,
+      "grad_norm": 2.136834440229622,
+      "learning_rate": 1.0093533955219639e-05,
+      "loss": 0.6924,
+      "step": 9600
+    },
+    {
+      "epoch": 5.477734235838974,
+      "grad_norm": 2.158621618784402,
+      "learning_rate": 1.0073633521965334e-05,
+      "loss": 0.6998,
+      "step": 9610
+    },
+    {
+      "epoch": 5.483434271464197,
+      "grad_norm": 2.2025522074206116,
+      "learning_rate": 1.0053732797081843e-05,
+      "loss": 0.6927,
+      "step": 9620
+    },
+    {
+      "epoch": 5.48913430708942,
+      "grad_norm": 2.171821093820873,
+      "learning_rate": 1.003383185938697e-05,
+      "loss": 0.6846,
+      "step": 9630
+    },
+    {
+      "epoch": 5.4948343427146416,
+      "grad_norm": 2.384577160611531,
+      "learning_rate": 1.0013930787699358e-05,
+      "loss": 0.6924,
+      "step": 9640
+    },
+    {
+      "epoch": 5.500534378339864,
+      "grad_norm": 2.122322707377548,
+      "learning_rate": 9.994029660838175e-06,
+      "loss": 0.6988,
+      "step": 9650
+    },
+    {
+      "epoch": 5.506234413965087,
+      "grad_norm": 2.1255532435570084,
+      "learning_rate": 9.974128557622814e-06,
+      "loss": 0.6856,
+      "step": 9660
+    },
+    {
+      "epoch": 5.51193444959031,
+      "grad_norm": 2.2432791993701904,
+      "learning_rate": 9.95422755687257e-06,
+      "loss": 0.6862,
+      "step": 9670
+    },
+    {
+      "epoch": 5.517634485215533,
+      "grad_norm": 2.157704074197768,
+      "learning_rate": 9.934326737406338e-06,
+      "loss": 0.6937,
+      "step": 9680
+    },
+    {
+      "epoch": 5.523334520840756,
+      "grad_norm": 2.132689938935754,
+      "learning_rate": 9.91442617804229e-06,
+      "loss": 0.6912,
+      "step": 9690
+    },
+    {
+      "epoch": 5.529034556465978,
+      "grad_norm": 2.086298565136052,
+      "learning_rate": 9.894525957597566e-06,
+      "loss": 0.703,
+      "step": 9700
+    },
+    {
+      "epoch": 5.5347345920912,
+      "grad_norm": 2.1735739979051765,
+      "learning_rate": 9.87462615488797e-06,
+      "loss": 0.6823,
+      "step": 9710
+    },
+    {
+      "epoch": 5.540434627716423,
+      "grad_norm": 2.2046428225180508,
+      "learning_rate": 9.854726848727645e-06,
+      "loss": 0.6936,
+      "step": 9720
+    },
+    {
+      "epoch": 5.546134663341646,
+      "grad_norm": 2.119617171719753,
+      "learning_rate": 9.834828117928776e-06,
+      "loss": 0.6909,
+      "step": 9730
+    },
+    {
+      "epoch": 5.551834698966869,
+      "grad_norm": 2.1771878721565594,
+      "learning_rate": 9.81493004130126e-06,
+      "loss": 0.7087,
+      "step": 9740
+    },
+    {
+      "epoch": 5.557534734592092,
+      "grad_norm": 2.097542061204044,
+      "learning_rate": 9.795032697652408e-06,
+      "loss": 0.6869,
+      "step": 9750
+    },
+    {
+      "epoch": 5.5632347702173135,
+      "grad_norm": 2.060454397168932,
+      "learning_rate": 9.775136165786626e-06,
+      "loss": 0.6975,
+      "step": 9760
+    },
+    {
+      "epoch": 5.568934805842536,
+      "grad_norm": 2.220837079068152,
+      "learning_rate": 9.755240524505107e-06,
+      "loss": 0.7019,
+      "step": 9770
+    },
+    {
+      "epoch": 5.574634841467759,
+      "grad_norm": 2.161939035304091,
+      "learning_rate": 9.735345852605519e-06,
+      "loss": 0.6906,
+      "step": 9780
+    },
+    {
+      "epoch": 5.580334877092982,
+      "grad_norm": 2.1577219592337507,
+      "learning_rate": 9.715452228881683e-06,
+      "loss": 0.7011,
+      "step": 9790
+    },
+    {
+      "epoch": 5.586034912718205,
+      "grad_norm": 2.191829870908184,
+      "learning_rate": 9.695559732123275e-06,
+      "loss": 0.6822,
+      "step": 9800
+    },
+    {
+      "epoch": 5.5917349483434275,
+      "grad_norm": 2.2162513440456593,
+      "learning_rate": 9.675668441115503e-06,
+      "loss": 0.6941,
+      "step": 9810
+    },
+    {
+      "epoch": 5.5974349839686495,
+      "grad_norm": 2.108945629779053,
+      "learning_rate": 9.655778434638807e-06,
+      "loss": 0.6934,
+      "step": 9820
+    },
+    {
+      "epoch": 5.603135019593872,
+      "grad_norm": 2.1700938789600275,
+      "learning_rate": 9.635889791468533e-06,
+      "loss": 0.6963,
+      "step": 9830
+    },
+    {
+      "epoch": 5.608835055219095,
+      "grad_norm": 2.233530335715926,
+      "learning_rate": 9.616002590374628e-06,
+      "loss": 0.6995,
+      "step": 9840
+    },
+    {
+      "epoch": 5.614535090844318,
+      "grad_norm": 2.1198785161387286,
+      "learning_rate": 9.596116910121328e-06,
+      "loss": 0.6951,
+      "step": 9850
+    },
+    {
+      "epoch": 5.620235126469541,
+      "grad_norm": 2.12074082476983,
+      "learning_rate": 9.57623282946685e-06,
+      "loss": 0.6916,
+      "step": 9860
+    },
+    {
+      "epoch": 5.625935162094763,
+      "grad_norm": 2.1419743093669585,
+      "learning_rate": 9.556350427163073e-06,
+      "loss": 0.6943,
+      "step": 9870
+    },
+    {
+      "epoch": 5.631635197719985,
+      "grad_norm": 2.21297293287053,
+      "learning_rate": 9.536469781955224e-06,
+      "loss": 0.6797,
+      "step": 9880
+    },
+    {
+      "epoch": 5.637335233345208,
+      "grad_norm": 2.080452235855697,
+      "learning_rate": 9.516590972581579e-06,
+      "loss": 0.6842,
+      "step": 9890
+    },
+    {
+      "epoch": 5.643035268970431,
+      "grad_norm": 2.1622295794258495,
+      "learning_rate": 9.496714077773132e-06,
+      "loss": 0.703,
+      "step": 9900
+    },
+    {
+      "epoch": 5.648735304595654,
+      "grad_norm": 2.2116620471484416,
+      "learning_rate": 9.476839176253311e-06,
+      "loss": 0.6915,
+      "step": 9910
+    },
+    {
+      "epoch": 5.654435340220877,
+      "grad_norm": 2.1804184869804986,
+      "learning_rate": 9.456966346737638e-06,
+      "loss": 0.6946,
+      "step": 9920
+    },
+    {
+      "epoch": 5.6601353758460995,
+      "grad_norm": 2.1894753754867984,
+      "learning_rate": 9.437095667933427e-06,
+      "loss": 0.6936,
+      "step": 9930
+    },
+    {
+      "epoch": 5.665835411471321,
+      "grad_norm": 2.2271688400846186,
+      "learning_rate": 9.417227218539475e-06,
+      "loss": 0.6929,
+      "step": 9940
+    },
+    {
+      "epoch": 5.671535447096544,
+      "grad_norm": 2.1505497312386157,
+      "learning_rate": 9.397361077245762e-06,
+      "loss": 0.6865,
+      "step": 9950
+    },
+    {
+      "epoch": 5.677235482721767,
+      "grad_norm": 2.2384375215999195,
+      "learning_rate": 9.377497322733109e-06,
+      "loss": 0.6944,
+      "step": 9960
+    },
+    {
+      "epoch": 5.68293551834699,
+      "grad_norm": 2.2881024340389917,
+      "learning_rate": 9.357636033672892e-06,
+      "loss": 0.7028,
+      "step": 9970
+    },
+    {
+      "epoch": 5.688635553972213,
+      "grad_norm": 2.213265696372914,
+      "learning_rate": 9.337777288726722e-06,
+      "loss": 0.6857,
+      "step": 9980
+    },
+    {
+      "epoch": 5.694335589597435,
+      "grad_norm": 2.299796287206333,
+      "learning_rate": 9.317921166546139e-06,
+      "loss": 0.6923,
+      "step": 9990
+    },
+    {
+      "epoch": 5.700035625222657,
+      "grad_norm": 2.228394016754354,
+      "learning_rate": 9.298067745772286e-06,
+      "loss": 0.6904,
+      "step": 10000
+    },
+    {
+      "epoch": 5.70573566084788,
+      "grad_norm": 2.1624614203874604,
+      "learning_rate": 9.278217105035613e-06,
+      "loss": 0.689,
+      "step": 10010
+    },
+    {
+      "epoch": 5.711435696473103,
+      "grad_norm": 2.1817564157072034,
+      "learning_rate": 9.258369322955558e-06,
+      "loss": 0.6867,
+      "step": 10020
+    },
+    {
+      "epoch": 5.717135732098326,
+      "grad_norm": 2.0985906977408306,
+      "learning_rate": 9.238524478140231e-06,
+      "loss": 0.6988,
+      "step": 10030
+    },
+    {
+      "epoch": 5.722835767723549,
+      "grad_norm": 2.2148178236948817,
+      "learning_rate": 9.218682649186123e-06,
+      "loss": 0.6813,
+      "step": 10040
+    },
+    {
+      "epoch": 5.7285358033487705,
+      "grad_norm": 2.1637803609456503,
+      "learning_rate": 9.198843914677776e-06,
+      "loss": 0.6828,
+      "step": 10050
+    },
+    {
+      "epoch": 5.734235838973993,
+      "grad_norm": 2.0867931176845436,
+      "learning_rate": 9.17900835318746e-06,
+      "loss": 0.6947,
+      "step": 10060
+    },
+    {
+      "epoch": 5.739935874599216,
+      "grad_norm": 2.1688434329476047,
+      "learning_rate": 9.159176043274896e-06,
+      "loss": 0.6869,
+      "step": 10070
+    },
+    {
+      "epoch": 5.745635910224439,
+      "grad_norm": 2.154324487594934,
+      "learning_rate": 9.139347063486926e-06,
+      "loss": 0.6807,
+      "step": 10080
+    },
+    {
+      "epoch": 5.751335945849662,
+      "grad_norm": 2.2017177171263063,
+      "learning_rate": 9.119521492357196e-06,
+      "loss": 0.6905,
+      "step": 10090
+    },
+    {
+      "epoch": 5.757035981474885,
+      "grad_norm": 2.118326945408275,
+      "learning_rate": 9.099699408405854e-06,
+      "loss": 0.6914,
+      "step": 10100
+    },
+    {
+      "epoch": 5.7627360171001065,
+      "grad_norm": 2.119914772970993,
+      "learning_rate": 9.079880890139238e-06,
+      "loss": 0.6947,
+      "step": 10110
+    },
+    {
+      "epoch": 5.768436052725329,
+      "grad_norm": 2.115252049304522,
+      "learning_rate": 9.06006601604956e-06,
+      "loss": 0.6937,
+      "step": 10120
+    },
+    {
+      "epoch": 5.774136088350552,
+      "grad_norm": 2.0727940773951716,
+      "learning_rate": 9.040254864614608e-06,
+      "loss": 0.7006,
+      "step": 10130
+    },
+    {
+      "epoch": 5.779836123975775,
+      "grad_norm": 2.2180027506072326,
+      "learning_rate": 9.020447514297417e-06,
+      "loss": 0.6984,
+      "step": 10140
+    },
+    {
+      "epoch": 5.785536159600998,
+      "grad_norm": 2.1795734441999723,
+      "learning_rate": 9.000644043545974e-06,
+      "loss": 0.6908,
+      "step": 10150
+    },
+    {
+      "epoch": 5.791236195226221,
+      "grad_norm": 2.315220502416694,
+      "learning_rate": 8.980844530792889e-06,
+      "loss": 0.6913,
+      "step": 10160
+    },
+    {
+      "epoch": 5.7969362308514425,
+      "grad_norm": 2.2227835874849644,
+      "learning_rate": 8.96104905445512e-06,
+      "loss": 0.6887,
+      "step": 10170
+    },
+    {
+      "epoch": 5.802636266476665,
+      "grad_norm": 2.148778586167764,
+      "learning_rate": 8.941257692933613e-06,
+      "loss": 0.6937,
+      "step": 10180
+    },
+    {
+      "epoch": 5.808336302101888,
+      "grad_norm": 2.135091331207858,
+      "learning_rate": 8.92147052461303e-06,
+      "loss": 0.6893,
+      "step": 10190
+    },
+    {
+      "epoch": 5.814036337727111,
+      "grad_norm": 2.2959580471537495,
+      "learning_rate": 8.901687627861423e-06,
+      "loss": 0.6976,
+      "step": 10200
+    },
+    {
+      "epoch": 5.819736373352334,
+      "grad_norm": 2.134327621747421,
+      "learning_rate": 8.881909081029923e-06,
+      "loss": 0.6935,
+      "step": 10210
+    },
+    {
+      "epoch": 5.825436408977556,
+      "grad_norm": 2.2490208608775273,
+      "learning_rate": 8.862134962452444e-06,
+      "loss": 0.7015,
+      "step": 10220
+    },
+    {
+      "epoch": 5.8311364446027785,
+      "grad_norm": 2.1876216665258577,
+      "learning_rate": 8.84236535044535e-06,
+      "loss": 0.6956,
+      "step": 10230
+    },
+    {
+      "epoch": 5.836836480228001,
+      "grad_norm": 2.1348120705699674,
+      "learning_rate": 8.822600323307163e-06,
+      "loss": 0.6904,
+      "step": 10240
+    },
+    {
+      "epoch": 5.842536515853224,
+      "grad_norm": 2.276591677956022,
+      "learning_rate": 8.802839959318238e-06,
+      "loss": 0.6876,
+      "step": 10250
+    },
+    {
+      "epoch": 5.848236551478447,
+      "grad_norm": 2.1847877639758937,
+      "learning_rate": 8.783084336740474e-06,
+      "loss": 0.6939,
+      "step": 10260
+    },
+    {
+      "epoch": 5.85393658710367,
+      "grad_norm": 2.172811243873737,
+      "learning_rate": 8.763333533816985e-06,
+      "loss": 0.6881,
+      "step": 10270
+    },
+    {
+      "epoch": 5.8596366227288925,
+      "grad_norm": 2.0077396330477715,
+      "learning_rate": 8.743587628771793e-06,
+      "loss": 0.6843,
+      "step": 10280
+    },
+    {
+      "epoch": 5.865336658354114,
+      "grad_norm": 2.1776154252978968,
+      "learning_rate": 8.723846699809522e-06,
+      "loss": 0.6834,
+      "step": 10290
+    },
+    {
+      "epoch": 5.871036693979337,
+      "grad_norm": 2.197726726366027,
+      "learning_rate": 8.704110825115098e-06,
+      "loss": 0.6918,
+      "step": 10300
+    },
+    {
+      "epoch": 5.87673672960456,
+      "grad_norm": 2.185810249326048,
+      "learning_rate": 8.68438008285342e-06,
+      "loss": 0.6928,
+      "step": 10310
+    },
+    {
+      "epoch": 5.882436765229783,
+      "grad_norm": 2.2642176568146435,
+      "learning_rate": 8.664654551169061e-06,
+      "loss": 0.6844,
+      "step": 10320
+    },
+    {
+      "epoch": 5.888136800855006,
+      "grad_norm": 2.161758512793611,
+      "learning_rate": 8.644934308185959e-06,
+      "loss": 0.6915,
+      "step": 10330
+    },
+    {
+      "epoch": 5.893836836480228,
+      "grad_norm": 2.163900058783309,
+      "learning_rate": 8.6252194320071e-06,
+      "loss": 0.6826,
+      "step": 10340
+    },
+    {
+      "epoch": 5.89953687210545,
+      "grad_norm": 2.1666177712043115,
+      "learning_rate": 8.605510000714228e-06,
+      "loss": 0.6871,
+      "step": 10350
+    },
+    {
+      "epoch": 5.905236907730673,
+      "grad_norm": 2.186489891927606,
+      "learning_rate": 8.585806092367513e-06,
+      "loss": 0.6926,
+      "step": 10360
+    },
+    {
+      "epoch": 5.910936943355896,
+      "grad_norm": 2.0842711436314687,
+      "learning_rate": 8.566107785005251e-06,
+      "loss": 0.6819,
+      "step": 10370
+    },
+    {
+      "epoch": 5.916636978981119,
+      "grad_norm": 2.2450803050111814,
+      "learning_rate": 8.546415156643549e-06,
+      "loss": 0.692,
+      "step": 10380
+    },
+    {
+      "epoch": 5.922337014606342,
+      "grad_norm": 2.2474530881350674,
+      "learning_rate": 8.526728285276039e-06,
+      "loss": 0.6984,
+      "step": 10390
+    },
+    {
+      "epoch": 5.928037050231564,
+      "grad_norm": 2.2904276116665474,
+      "learning_rate": 8.507047248873539e-06,
+      "loss": 0.686,
+      "step": 10400
+    },
+    {
+      "epoch": 5.933737085856786,
+      "grad_norm": 2.112697139828902,
+      "learning_rate": 8.487372125383757e-06,
+      "loss": 0.6824,
+      "step": 10410
+    },
+    {
+      "epoch": 5.939437121482009,
+      "grad_norm": 2.1620065424570045,
+      "learning_rate": 8.467702992730992e-06,
+      "loss": 0.6952,
+      "step": 10420
+    },
+    {
+      "epoch": 5.945137157107232,
+      "grad_norm": 2.1255308293368116,
+      "learning_rate": 8.448039928815804e-06,
+      "loss": 0.691,
+      "step": 10430
+    },
+    {
+      "epoch": 5.950837192732455,
+      "grad_norm": 2.028997680326995,
+      "learning_rate": 8.42838301151473e-06,
+      "loss": 0.6969,
+      "step": 10440
+    },
+    {
+      "epoch": 5.956537228357677,
+      "grad_norm": 2.1235085422204536,
+      "learning_rate": 8.408732318679953e-06,
+      "loss": 0.711,
+      "step": 10450
+    },
+    {
+      "epoch": 5.9622372639828995,
+      "grad_norm": 2.0842748601067176,
+      "learning_rate": 8.389087928139008e-06,
+      "loss": 0.6955,
+      "step": 10460
+    },
+    {
+      "epoch": 5.967937299608122,
+      "grad_norm": 2.2061370472464668,
+      "learning_rate": 8.369449917694466e-06,
+      "loss": 0.6943,
+      "step": 10470
+    },
+    {
+      "epoch": 5.973637335233345,
+      "grad_norm": 2.154886985406157,
+      "learning_rate": 8.34981836512364e-06,
+      "loss": 0.6905,
+      "step": 10480
+    },
+    {
+      "epoch": 5.979337370858568,
+      "grad_norm": 2.1100063096741875,
+      "learning_rate": 8.330193348178254e-06,
+      "loss": 0.703,
+      "step": 10490
+    },
+    {
+      "epoch": 5.985037406483791,
+      "grad_norm": 2.260276094006089,
+      "learning_rate": 8.310574944584151e-06,
+      "loss": 0.6957,
+      "step": 10500
+    },
+    {
+      "epoch": 5.990737442109014,
+      "grad_norm": 2.1602419464603755,
+      "learning_rate": 8.290963232040984e-06,
+      "loss": 0.6918,
+      "step": 10510
+    },
+    {
+      "epoch": 5.9964374777342355,
+      "grad_norm": 2.2948853765089563,
+      "learning_rate": 8.271358288221897e-06,
+      "loss": 0.6945,
+      "step": 10520
+    },
+    {
+      "epoch": 6.002137513359458,
+      "grad_norm": 2.170085520328391,
+      "learning_rate": 8.251760190773243e-06,
+      "loss": 0.6872,
+      "step": 10530
+    },
+    {
+      "epoch": 6.007837548984681,
+      "grad_norm": 2.279696142110909,
+      "learning_rate": 8.232169017314247e-06,
+      "loss": 0.6774,
+      "step": 10540
+    },
+    {
+      "epoch": 6.013537584609904,
+      "grad_norm": 2.322165475925097,
+      "learning_rate": 8.212584845436713e-06,
+      "loss": 0.6681,
+      "step": 10550
+    },
+    {
+      "epoch": 6.019237620235127,
+      "grad_norm": 2.240989851797247,
+      "learning_rate": 8.193007752704714e-06,
+      "loss": 0.6697,
+      "step": 10560
+    },
+    {
+      "epoch": 6.024937655860349,
+      "grad_norm": 2.246153691155315,
+      "learning_rate": 8.173437816654292e-06,
+      "loss": 0.6725,
+      "step": 10570
+    },
+    {
+      "epoch": 6.0306376914855715,
+      "grad_norm": 2.1966286414290868,
+      "learning_rate": 8.153875114793137e-06,
+      "loss": 0.6894,
+      "step": 10580
+    },
+    {
+      "epoch": 6.036337727110794,
+      "grad_norm": 2.1294095450667108,
+      "learning_rate": 8.13431972460029e-06,
+      "loss": 0.6775,
+      "step": 10590
+    },
+    {
+      "epoch": 6.042037762736017,
+      "grad_norm": 2.2155021741663647,
+      "learning_rate": 8.11477172352584e-06,
+      "loss": 0.6709,
+      "step": 10600
+    },
+    {
+      "epoch": 6.04773779836124,
+      "grad_norm": 2.259295223173431,
+      "learning_rate": 8.095231188990597e-06,
+      "loss": 0.6823,
+      "step": 10610
+    },
+    {
+      "epoch": 6.053437833986463,
+      "grad_norm": 2.2718220988832867,
+      "learning_rate": 8.075698198385817e-06,
+      "loss": 0.681,
+      "step": 10620
+    },
+    {
+      "epoch": 6.059137869611685,
+      "grad_norm": 2.1822599854506817,
+      "learning_rate": 8.056172829072863e-06,
+      "loss": 0.6738,
+      "step": 10630
+    },
+    {
+      "epoch": 6.0648379052369075,
+      "grad_norm": 2.2108465993733857,
+      "learning_rate": 8.036655158382922e-06,
+      "loss": 0.6846,
+      "step": 10640
+    },
+    {
+      "epoch": 6.07053794086213,
+      "grad_norm": 2.3040922125703323,
+      "learning_rate": 8.017145263616683e-06,
+      "loss": 0.674,
+      "step": 10650
+    },
+    {
+      "epoch": 6.076237976487353,
+      "grad_norm": 2.106527392828737,
+      "learning_rate": 7.997643222044051e-06,
+      "loss": 0.6852,
+      "step": 10660
+    },
+    {
+      "epoch": 6.081938012112576,
+      "grad_norm": 2.1551369757776464,
+      "learning_rate": 7.978149110903816e-06,
+      "loss": 0.6688,
+      "step": 10670
+    },
+    {
+      "epoch": 6.087638047737799,
+      "grad_norm": 2.2730916010608686,
+      "learning_rate": 7.958663007403362e-06,
+      "loss": 0.6797,
+      "step": 10680
+    },
+    {
+      "epoch": 6.093338083363021,
+      "grad_norm": 2.292642755654638,
+      "learning_rate": 7.939184988718359e-06,
+      "loss": 0.6658,
+      "step": 10690
+    },
+    {
+      "epoch": 6.099038118988243,
+      "grad_norm": 2.192515472600447,
+      "learning_rate": 7.919715131992459e-06,
+      "loss": 0.676,
+      "step": 10700
+    },
+    {
+      "epoch": 6.104738154613466,
+      "grad_norm": 2.1376221361926935,
+      "learning_rate": 7.900253514336985e-06,
+      "loss": 0.6753,
+      "step": 10710
+    },
+    {
+      "epoch": 6.110438190238689,
+      "grad_norm": 2.1942538782779097,
+      "learning_rate": 7.88080021283063e-06,
+      "loss": 0.667,
+      "step": 10720
+    },
+    {
+      "epoch": 6.116138225863912,
+      "grad_norm": 2.246819786705824,
+      "learning_rate": 7.86135530451915e-06,
+      "loss": 0.6666,
+      "step": 10730
+    },
+    {
+      "epoch": 6.121838261489135,
+      "grad_norm": 2.202275918725809,
+      "learning_rate": 7.84191886641506e-06,
+      "loss": 0.6774,
+      "step": 10740
+    },
+    {
+      "epoch": 6.127538297114357,
+      "grad_norm": 2.1771264629986016,
+      "learning_rate": 7.822490975497326e-06,
+      "loss": 0.6766,
+      "step": 10750
+    },
+    {
+      "epoch": 6.133238332739579,
+      "grad_norm": 2.391203193647443,
+      "learning_rate": 7.80307170871107e-06,
+      "loss": 0.6845,
+      "step": 10760
+    },
+    {
+      "epoch": 6.138938368364802,
+      "grad_norm": 2.2744480554444513,
+      "learning_rate": 7.783661142967247e-06,
+      "loss": 0.682,
+      "step": 10770
+    },
+    {
+      "epoch": 6.144638403990025,
+      "grad_norm": 2.234849216259114,
+      "learning_rate": 7.764259355142354e-06,
+      "loss": 0.6716,
+      "step": 10780
+    },
+    {
+      "epoch": 6.150338439615248,
+      "grad_norm": 2.202045737877405,
+      "learning_rate": 7.744866422078133e-06,
+      "loss": 0.6846,
+      "step": 10790
+    },
+    {
+      "epoch": 6.156038475240471,
+      "grad_norm": 2.1917326270336144,
+      "learning_rate": 7.725482420581245e-06,
+      "loss": 0.6793,
+      "step": 10800
+    },
+    {
+      "epoch": 6.161738510865693,
+      "grad_norm": 2.2201548403178544,
+      "learning_rate": 7.70610742742298e-06,
+      "loss": 0.6787,
+      "step": 10810
+    },
+    {
+      "epoch": 6.167438546490915,
+      "grad_norm": 2.1974557147442306,
+      "learning_rate": 7.686741519338949e-06,
+      "loss": 0.6801,
+      "step": 10820
+    },
+    {
+      "epoch": 6.173138582116138,
+      "grad_norm": 2.262594067441599,
+      "learning_rate": 7.667384773028778e-06,
+      "loss": 0.6813,
+      "step": 10830
+    },
+    {
+      "epoch": 6.178838617741361,
+      "grad_norm": 2.14103947815712,
+      "learning_rate": 7.64803726515582e-06,
+      "loss": 0.6784,
+      "step": 10840
+    },
+    {
+      "epoch": 6.184538653366584,
+      "grad_norm": 2.1449864947544,
+      "learning_rate": 7.62869907234683e-06,
+      "loss": 0.6873,
+      "step": 10850
+    },
+    {
+      "epoch": 6.190238688991807,
+      "grad_norm": 2.202782593471053,
+      "learning_rate": 7.609370271191667e-06,
+      "loss": 0.6816,
+      "step": 10860
+    },
+    {
+      "epoch": 6.1959387246170285,
+      "grad_norm": 2.2955199701611906,
+      "learning_rate": 7.590050938242997e-06,
+      "loss": 0.6745,
+      "step": 10870
+    },
+    {
+      "epoch": 6.201638760242251,
+      "grad_norm": 2.3085316499548534,
+      "learning_rate": 7.57074115001599e-06,
+      "loss": 0.6805,
+      "step": 10880
+    },
+    {
+      "epoch": 6.207338795867474,
+      "grad_norm": 2.287598178157803,
+      "learning_rate": 7.551440982988011e-06,
+      "loss": 0.6695,
+      "step": 10890
+    },
+    {
+      "epoch": 6.213038831492697,
+      "grad_norm": 2.3348856472179125,
+      "learning_rate": 7.532150513598318e-06,
+      "loss": 0.671,
+      "step": 10900
+    },
+    {
+      "epoch": 6.21873886711792,
+      "grad_norm": 2.2779474366048786,
+      "learning_rate": 7.512869818247763e-06,
+      "loss": 0.6755,
+      "step": 10910
+    },
+    {
+      "epoch": 6.224438902743142,
+      "grad_norm": 2.2397875686297515,
+      "learning_rate": 7.493598973298485e-06,
+      "loss": 0.6838,
+      "step": 10920
+    },
+    {
+      "epoch": 6.2301389383683645,
+      "grad_norm": 2.2689325500785857,
+      "learning_rate": 7.47433805507362e-06,
+      "loss": 0.6861,
+      "step": 10930
+    },
+    {
+      "epoch": 6.235838973993587,
+      "grad_norm": 2.321132598662188,
+      "learning_rate": 7.4550871398569755e-06,
+      "loss": 0.6751,
+      "step": 10940
+    },
+    {
+      "epoch": 6.24153900961881,
+      "grad_norm": 2.146161666540449,
+      "learning_rate": 7.4358463038927464e-06,
+      "loss": 0.6846,
+      "step": 10950
+    },
+    {
+      "epoch": 6.247239045244033,
+      "grad_norm": 2.191499499785015,
+      "learning_rate": 7.416615623385205e-06,
+      "loss": 0.6689,
+      "step": 10960
+    },
+    {
+      "epoch": 6.252939080869256,
+      "grad_norm": 2.2642018013263905,
+      "learning_rate": 7.397395174498416e-06,
+      "loss": 0.6758,
+      "step": 10970
+    },
+    {
+      "epoch": 6.258639116494478,
+      "grad_norm": 2.259935424672811,
+      "learning_rate": 7.3781850333559065e-06,
+      "loss": 0.6769,
+      "step": 10980
+    },
+    {
+      "epoch": 6.2643391521197005,
+      "grad_norm": 2.308436223900447,
+      "learning_rate": 7.3589852760403845e-06,
+      "loss": 0.672,
+      "step": 10990
+    },
+    {
+      "epoch": 6.270039187744923,
+      "grad_norm": 2.2759428945966302,
+      "learning_rate": 7.3397959785934305e-06,
+      "loss": 0.6762,
+      "step": 11000
+    },
+    {
+      "epoch": 6.275739223370146,
+      "grad_norm": 2.2795748590311167,
+      "learning_rate": 7.3206172170152025e-06,
+      "loss": 0.6711,
+      "step": 11010
+    },
+    {
+      "epoch": 6.281439258995369,
+      "grad_norm": 2.21392152862467,
+      "learning_rate": 7.301449067264128e-06,
+      "loss": 0.6748,
+      "step": 11020
+    },
+    {
+      "epoch": 6.287139294620592,
+      "grad_norm": 2.1429505866001684,
+      "learning_rate": 7.282291605256604e-06,
+      "loss": 0.673,
+      "step": 11030
+    },
+    {
+      "epoch": 6.292839330245814,
+      "grad_norm": 2.239080481574686,
+      "learning_rate": 7.263144906866701e-06,
+      "loss": 0.6695,
+      "step": 11040
+    },
+    {
+      "epoch": 6.2985393658710365,
+      "grad_norm": 2.1917221329279086,
+      "learning_rate": 7.244009047925858e-06,
+      "loss": 0.6782,
+      "step": 11050
+    },
+    {
+      "epoch": 6.304239401496259,
+      "grad_norm": 2.2307933648937475,
+      "learning_rate": 7.224884104222585e-06,
+      "loss": 0.6746,
+      "step": 11060
+    },
+    {
+      "epoch": 6.309939437121482,
+      "grad_norm": 2.1416509854746377,
+      "learning_rate": 7.205770151502163e-06,
+      "loss": 0.6846,
+      "step": 11070
+    },
+    {
+      "epoch": 6.315639472746705,
+      "grad_norm": 2.173159302277719,
+      "learning_rate": 7.186667265466337e-06,
+      "loss": 0.668,
+      "step": 11080
+    },
+    {
+      "epoch": 6.321339508371928,
+      "grad_norm": 2.261191207971506,
+      "learning_rate": 7.1675755217730245e-06,
+      "loss": 0.6764,
+      "step": 11090
+    },
+    {
+      "epoch": 6.32703954399715,
+      "grad_norm": 2.3224880630444007,
+      "learning_rate": 7.148494996036022e-06,
+      "loss": 0.6776,
+      "step": 11100
+    },
+    {
+      "epoch": 6.332739579622372,
+      "grad_norm": 2.0716743785582055,
+      "learning_rate": 7.129425763824683e-06,
+      "loss": 0.6749,
+      "step": 11110
+    },
+    {
+      "epoch": 6.338439615247595,
+      "grad_norm": 2.261127297203104,
+      "learning_rate": 7.110367900663642e-06,
+      "loss": 0.6732,
+      "step": 11120
+    },
+    {
+      "epoch": 6.344139650872818,
+      "grad_norm": 2.2896807270604116,
+      "learning_rate": 7.091321482032501e-06,
+      "loss": 0.6766,
+      "step": 11130
+    },
+    {
+      "epoch": 6.349839686498041,
+      "grad_norm": 2.162182939499591,
+      "learning_rate": 7.072286583365533e-06,
+      "loss": 0.6753,
+      "step": 11140
+    },
+    {
+      "epoch": 6.355539722123264,
+      "grad_norm": 2.17833933141201,
+      "learning_rate": 7.053263280051394e-06,
+      "loss": 0.6821,
+      "step": 11150
+    },
+    {
+      "epoch": 6.361239757748486,
+      "grad_norm": 2.209348413867267,
+      "learning_rate": 7.034251647432811e-06,
+      "loss": 0.6628,
+      "step": 11160
+    },
+    {
+      "epoch": 6.366939793373708,
+      "grad_norm": 2.249600616326594,
+      "learning_rate": 7.01525176080629e-06,
+      "loss": 0.6708,
+      "step": 11170
+    },
+    {
+      "epoch": 6.372639828998931,
+      "grad_norm": 2.1575927428642876,
+      "learning_rate": 6.99626369542181e-06,
+      "loss": 0.6746,
+      "step": 11180
+    },
+    {
+      "epoch": 6.378339864624154,
+      "grad_norm": 2.2669648805317344,
+      "learning_rate": 6.977287526482541e-06,
+      "loss": 0.6877,
+      "step": 11190
+    },
+    {
+      "epoch": 6.384039900249377,
+      "grad_norm": 2.2249493885686378,
+      "learning_rate": 6.958323329144534e-06,
+      "loss": 0.6732,
+      "step": 11200
+    },
+    {
+      "epoch": 6.3897399358746,
+      "grad_norm": 2.3167895861205148,
+      "learning_rate": 6.939371178516423e-06,
+      "loss": 0.677,
+      "step": 11210
+    },
+    {
+      "epoch": 6.395439971499822,
+      "grad_norm": 2.202354910854256,
+      "learning_rate": 6.920431149659128e-06,
+      "loss": 0.6673,
+      "step": 11220
+    },
+    {
+      "epoch": 6.401140007125044,
+      "grad_norm": 2.289065268830389,
+      "learning_rate": 6.901503317585565e-06,
+      "loss": 0.6773,
+      "step": 11230
+    },
+    {
+      "epoch": 6.406840042750267,
+      "grad_norm": 2.3316974329710667,
+      "learning_rate": 6.882587757260349e-06,
+      "loss": 0.6767,
+      "step": 11240
+    },
+    {
+      "epoch": 6.41254007837549,
+      "grad_norm": 2.187793708950635,
+      "learning_rate": 6.86368454359948e-06,
+      "loss": 0.6848,
+      "step": 11250
+    },
+    {
+      "epoch": 6.418240114000713,
+      "grad_norm": 2.2413921903856924,
+      "learning_rate": 6.844793751470069e-06,
+      "loss": 0.6827,
+      "step": 11260
+    },
+    {
+      "epoch": 6.423940149625935,
+      "grad_norm": 2.1505171696058576,
+      "learning_rate": 6.825915455690015e-06,
+      "loss": 0.6712,
+      "step": 11270
+    },
+    {
+      "epoch": 6.4296401852511575,
+      "grad_norm": 2.335901213288781,
+      "learning_rate": 6.807049731027751e-06,
+      "loss": 0.6856,
+      "step": 11280
+    },
+    {
+      "epoch": 6.43534022087638,
+      "grad_norm": 2.3470142982475677,
+      "learning_rate": 6.788196652201899e-06,
+      "loss": 0.6775,
+      "step": 11290
+    },
+    {
+      "epoch": 6.441040256501603,
+      "grad_norm": 2.2525507766531327,
+      "learning_rate": 6.769356293881005e-06,
+      "loss": 0.6784,
+      "step": 11300
+    },
+    {
+      "epoch": 6.446740292126826,
+      "grad_norm": 2.260718619048699,
+      "learning_rate": 6.750528730683231e-06,
+      "loss": 0.6781,
+      "step": 11310
+    },
+    {
+      "epoch": 6.452440327752049,
+      "grad_norm": 2.0555539548278055,
+      "learning_rate": 6.731714037176071e-06,
+      "loss": 0.6872,
+      "step": 11320
+    },
+    {
+      "epoch": 6.458140363377271,
+      "grad_norm": 2.2640903532916643,
+      "learning_rate": 6.712912287876041e-06,
+      "loss": 0.6746,
+      "step": 11330
+    },
+    {
+      "epoch": 6.4638403990024935,
+      "grad_norm": 2.295093150078742,
+      "learning_rate": 6.6941235572483905e-06,
+      "loss": 0.6717,
+      "step": 11340
+    },
+    {
+      "epoch": 6.469540434627716,
+      "grad_norm": 2.1669622783413276,
+      "learning_rate": 6.6753479197068136e-06,
+      "loss": 0.6888,
+      "step": 11350
+    },
+    {
+      "epoch": 6.475240470252939,
+      "grad_norm": 2.239696811909756,
+      "learning_rate": 6.65658544961314e-06,
+      "loss": 0.6704,
+      "step": 11360
+    },
+    {
+      "epoch": 6.480940505878162,
+      "grad_norm": 2.325134420293425,
+      "learning_rate": 6.637836221277063e-06,
+      "loss": 0.681,
+      "step": 11370
+    },
+    {
+      "epoch": 6.486640541503385,
+      "grad_norm": 2.297069745780302,
+      "learning_rate": 6.619100308955817e-06,
+      "loss": 0.68,
+      "step": 11380
+    },
+    {
+      "epoch": 6.492340577128607,
+      "grad_norm": 2.2432527420421815,
+      "learning_rate": 6.600377786853903e-06,
+      "loss": 0.6787,
+      "step": 11390
+    },
+    {
+      "epoch": 6.4980406127538295,
+      "grad_norm": 2.2220378243498797,
+      "learning_rate": 6.581668729122788e-06,
+      "loss": 0.6819,
+      "step": 11400
+    },
+    {
+      "epoch": 6.503740648379052,
+      "grad_norm": 2.3093011410269684,
+      "learning_rate": 6.562973209860619e-06,
+      "loss": 0.6799,
+      "step": 11410
+    },
+    {
+      "epoch": 6.509440684004275,
+      "grad_norm": 2.2678503311756026,
+      "learning_rate": 6.544291303111918e-06,
+      "loss": 0.6789,
+      "step": 11420
+    },
+    {
+      "epoch": 6.515140719629498,
+      "grad_norm": 2.2357396617872434,
+      "learning_rate": 6.525623082867292e-06,
+      "loss": 0.6779,
+      "step": 11430
+    },
+    {
+      "epoch": 6.520840755254721,
+      "grad_norm": 2.2611435548190144,
+      "learning_rate": 6.506968623063145e-06,
+      "loss": 0.6828,
+      "step": 11440
+    },
+    {
+      "epoch": 6.526540790879943,
+      "grad_norm": 2.243445764732507,
+      "learning_rate": 6.488327997581383e-06,
+      "loss": 0.6665,
+      "step": 11450
+    },
+    {
+      "epoch": 6.5322408265051655,
+      "grad_norm": 2.347032248498623,
+      "learning_rate": 6.469701280249118e-06,
+      "loss": 0.675,
+      "step": 11460
+    },
+    {
+      "epoch": 6.537940862130388,
+      "grad_norm": 2.348855883099169,
+      "learning_rate": 6.4510885448383796e-06,
+      "loss": 0.681,
+      "step": 11470
+    },
+    {
+      "epoch": 6.543640897755611,
+      "grad_norm": 2.310343148383439,
+      "learning_rate": 6.432489865065821e-06,
+      "loss": 0.6743,
+      "step": 11480
+    },
+    {
+      "epoch": 6.549340933380834,
+      "grad_norm": 2.1760150064972605,
+      "learning_rate": 6.4139053145924234e-06,
+      "loss": 0.6714,
+      "step": 11490
+    },
+    {
+      "epoch": 6.555040969006056,
+      "grad_norm": 2.242725273011376,
+      "learning_rate": 6.395334967023219e-06,
+      "loss": 0.6763,
+      "step": 11500
+    },
+    {
+      "epoch": 6.560741004631279,
+      "grad_norm": 2.2683025520996885,
+      "learning_rate": 6.3767788959069765e-06,
+      "loss": 0.6829,
+      "step": 11510
+    },
+    {
+      "epoch": 6.566441040256501,
+      "grad_norm": 2.2405754012097443,
+      "learning_rate": 6.358237174735931e-06,
+      "loss": 0.6764,
+      "step": 11520
+    },
+    {
+      "epoch": 6.572141075881724,
+      "grad_norm": 2.132139690504168,
+      "learning_rate": 6.339709876945475e-06,
+      "loss": 0.6869,
+      "step": 11530
+    },
+    {
+      "epoch": 6.577841111506947,
+      "grad_norm": 2.1531564491426662,
+      "learning_rate": 6.321197075913883e-06,
+      "loss": 0.6776,
+      "step": 11540
+    },
+    {
+      "epoch": 6.58354114713217,
+      "grad_norm": 2.240267761454674,
+      "learning_rate": 6.302698844962019e-06,
+      "loss": 0.6768,
+      "step": 11550
+    },
+    {
+      "epoch": 6.589241182757393,
+      "grad_norm": 2.251819890328819,
+      "learning_rate": 6.2842152573530294e-06,
+      "loss": 0.6839,
+      "step": 11560
+    },
+    {
+      "epoch": 6.594941218382615,
+      "grad_norm": 2.287793168135825,
+      "learning_rate": 6.265746386292073e-06,
+      "loss": 0.6822,
+      "step": 11570
+    },
+    {
+      "epoch": 6.600641254007837,
+      "grad_norm": 2.207158562564074,
+      "learning_rate": 6.24729230492602e-06,
+      "loss": 0.6687,
+      "step": 11580
+    },
+    {
+      "epoch": 6.60634128963306,
+      "grad_norm": 2.210442842730923,
+      "learning_rate": 6.228853086343169e-06,
+      "loss": 0.6845,
+      "step": 11590
+    },
+    {
+      "epoch": 6.612041325258283,
+      "grad_norm": 2.3938104834418343,
+      "learning_rate": 6.210428803572949e-06,
+      "loss": 0.675,
+      "step": 11600
+    },
+    {
+      "epoch": 6.617741360883506,
+      "grad_norm": 2.267869603038391,
+      "learning_rate": 6.192019529585638e-06,
+      "loss": 0.6877,
+      "step": 11610
+    },
+    {
+      "epoch": 6.623441396508728,
+      "grad_norm": 2.177035041489817,
+      "learning_rate": 6.173625337292068e-06,
+      "loss": 0.6703,
+      "step": 11620
+    },
+    {
+      "epoch": 6.629141432133951,
+      "grad_norm": 2.095897873005044,
+      "learning_rate": 6.155246299543342e-06,
+      "loss": 0.6641,
+      "step": 11630
+    },
+    {
+      "epoch": 6.634841467759173,
+      "grad_norm": 2.278131320767482,
+      "learning_rate": 6.136882489130545e-06,
+      "loss": 0.6715,
+      "step": 11640
+    },
+    {
+      "epoch": 6.640541503384396,
+      "grad_norm": 2.2722217340570294,
+      "learning_rate": 6.1185339787844475e-06,
+      "loss": 0.6773,
+      "step": 11650
+    },
+    {
+      "epoch": 6.646241539009619,
+      "grad_norm": 2.1265165938539847,
+      "learning_rate": 6.100200841175228e-06,
+      "loss": 0.67,
+      "step": 11660
+    },
+    {
+      "epoch": 6.651941574634842,
+      "grad_norm": 2.249161145163528,
+      "learning_rate": 6.081883148912174e-06,
+      "loss": 0.6761,
+      "step": 11670
+    },
+    {
+      "epoch": 6.657641610260065,
+      "grad_norm": 2.3663245069496117,
+      "learning_rate": 6.06358097454341e-06,
+      "loss": 0.6764,
+      "step": 11680
+    },
+    {
+      "epoch": 6.6633416458852865,
+      "grad_norm": 2.241759785395175,
+      "learning_rate": 6.045294390555598e-06,
+      "loss": 0.6778,
+      "step": 11690
+    },
+    {
+      "epoch": 6.669041681510509,
+      "grad_norm": 2.1832541172871047,
+      "learning_rate": 6.027023469373654e-06,
+      "loss": 0.6683,
+      "step": 11700
+    },
+    {
+      "epoch": 6.674741717135732,
+      "grad_norm": 2.165879211709872,
+      "learning_rate": 6.0087682833604475e-06,
+      "loss": 0.6883,
+      "step": 11710
+    },
+    {
+      "epoch": 6.680441752760955,
+      "grad_norm": 2.158951360462029,
+      "learning_rate": 5.990528904816553e-06,
+      "loss": 0.6803,
+      "step": 11720
+    },
+    {
+      "epoch": 6.686141788386178,
+      "grad_norm": 2.2647654446771055,
+      "learning_rate": 5.972305405979919e-06,
+      "loss": 0.6711,
+      "step": 11730
+    },
+    {
+      "epoch": 6.6918418240114,
+      "grad_norm": 2.3092540339887795,
+      "learning_rate": 5.954097859025609e-06,
+      "loss": 0.6813,
+      "step": 11740
+    },
+    {
+      "epoch": 6.6975418596366225,
+      "grad_norm": 2.189512090931114,
+      "learning_rate": 5.9359063360655065e-06,
+      "loss": 0.6802,
+      "step": 11750
+    },
+    {
+      "epoch": 6.703241895261845,
+      "grad_norm": 2.2124034866664357,
+      "learning_rate": 5.9177309091480295e-06,
+      "loss": 0.6762,
+      "step": 11760
+    },
+    {
+      "epoch": 6.708941930887068,
+      "grad_norm": 2.2058246047742798,
+      "learning_rate": 5.899571650257856e-06,
+      "loss": 0.674,
+      "step": 11770
+    },
+    {
+      "epoch": 6.714641966512291,
+      "grad_norm": 2.2319022200132204,
+      "learning_rate": 5.88142863131562e-06,
+      "loss": 0.6759,
+      "step": 11780
+    },
+    {
+      "epoch": 6.720342002137514,
+      "grad_norm": 2.2625836265140147,
+      "learning_rate": 5.863301924177638e-06,
+      "loss": 0.6806,
+      "step": 11790
+    },
+    {
+      "epoch": 6.726042037762736,
+      "grad_norm": 2.217514359432777,
+      "learning_rate": 5.84519160063562e-06,
+      "loss": 0.6889,
+      "step": 11800
+    },
+    {
+      "epoch": 6.7317420733879585,
+      "grad_norm": 2.1712652173946685,
+      "learning_rate": 5.827097732416404e-06,
+      "loss": 0.6768,
+      "step": 11810
+    },
+    {
+      "epoch": 6.737442109013181,
+      "grad_norm": 2.1940538960439104,
+      "learning_rate": 5.809020391181635e-06,
+      "loss": 0.67,
+      "step": 11820
+    },
+    {
+      "epoch": 6.743142144638404,
+      "grad_norm": 2.3137907789358176,
+      "learning_rate": 5.790959648527513e-06,
+      "loss": 0.6825,
+      "step": 11830
+    },
+    {
+      "epoch": 6.748842180263627,
+      "grad_norm": 2.2737669042973683,
+      "learning_rate": 5.772915575984497e-06,
+      "loss": 0.6653,
+      "step": 11840
+    },
+    {
+      "epoch": 6.754542215888849,
+      "grad_norm": 2.1182723284435094,
+      "learning_rate": 5.754888245017019e-06,
+      "loss": 0.6715,
+      "step": 11850
+    },
+    {
+      "epoch": 6.760242251514072,
+      "grad_norm": 2.11533837427729,
+      "learning_rate": 5.736877727023217e-06,
+      "loss": 0.6687,
+      "step": 11860
+    },
+    {
+      "epoch": 6.7659422871392945,
+      "grad_norm": 2.20018065816376,
+      "learning_rate": 5.7188840933346265e-06,
+      "loss": 0.676,
+      "step": 11870
+    },
+    {
+      "epoch": 6.771642322764517,
+      "grad_norm": 2.1443107954137477,
+      "learning_rate": 5.700907415215922e-06,
+      "loss": 0.6783,
+      "step": 11880
+    },
+    {
+      "epoch": 6.77734235838974,
+      "grad_norm": 2.2913214558881303,
+      "learning_rate": 5.682947763864612e-06,
+      "loss": 0.6734,
+      "step": 11890
+    },
+    {
+      "epoch": 6.783042394014963,
+      "grad_norm": 2.287316672737734,
+      "learning_rate": 5.665005210410788e-06,
+      "loss": 0.6719,
+      "step": 11900
+    },
+    {
+      "epoch": 6.788742429640186,
+      "grad_norm": 2.2262316652328,
+      "learning_rate": 5.64707982591681e-06,
+      "loss": 0.6756,
+      "step": 11910
+    },
+    {
+      "epoch": 6.794442465265408,
+      "grad_norm": 2.2678002183176638,
+      "learning_rate": 5.629171681377049e-06,
+      "loss": 0.6682,
+      "step": 11920
+    },
+    {
+      "epoch": 6.80014250089063,
+      "grad_norm": 2.2387850423103735,
+      "learning_rate": 5.611280847717581e-06,
+      "loss": 0.6729,
+      "step": 11930
+    },
+    {
+      "epoch": 6.805842536515853,
+      "grad_norm": 2.226404135799429,
+      "learning_rate": 5.593407395795936e-06,
+      "loss": 0.6746,
+      "step": 11940
+    },
+    {
+      "epoch": 6.811542572141076,
+      "grad_norm": 2.17331328615478,
+      "learning_rate": 5.575551396400802e-06,
+      "loss": 0.6637,
+      "step": 11950
+    },
+    {
+      "epoch": 6.817242607766299,
+      "grad_norm": 2.2195107960727682,
+      "learning_rate": 5.557712920251741e-06,
+      "loss": 0.6861,
+      "step": 11960
+    },
+    {
+      "epoch": 6.822942643391521,
+      "grad_norm": 2.211970729931292,
+      "learning_rate": 5.539892037998911e-06,
+      "loss": 0.6742,
+      "step": 11970
+    },
+    {
+      "epoch": 6.828642679016744,
+      "grad_norm": 2.114732092115801,
+      "learning_rate": 5.5220888202227906e-06,
+      "loss": 0.6707,
+      "step": 11980
+    },
+    {
+      "epoch": 6.834342714641966,
+      "grad_norm": 2.2228110319573164,
+      "learning_rate": 5.504303337433905e-06,
+      "loss": 0.6833,
+      "step": 11990
+    },
+    {
+      "epoch": 6.840042750267189,
+      "grad_norm": 2.272654536271849,
+      "learning_rate": 5.48653566007253e-06,
+      "loss": 0.682,
+      "step": 12000
+    },
+    {
+      "epoch": 6.845742785892412,
+      "grad_norm": 2.3142083790744534,
+      "learning_rate": 5.468785858508423e-06,
+      "loss": 0.6661,
+      "step": 12010
+    },
+    {
+      "epoch": 6.851442821517635,
+      "grad_norm": 2.306620120881866,
+      "learning_rate": 5.451054003040541e-06,
+      "loss": 0.6825,
+      "step": 12020
+    },
+    {
+      "epoch": 6.857142857142857,
+      "grad_norm": 2.0840208651157,
+      "learning_rate": 5.4333401638967794e-06,
+      "loss": 0.6817,
+      "step": 12030
+    },
+    {
+      "epoch": 6.86284289276808,
+      "grad_norm": 2.211367100073465,
+      "learning_rate": 5.415644411233667e-06,
+      "loss": 0.6756,
+      "step": 12040
+    },
+    {
+      "epoch": 6.868542928393302,
+      "grad_norm": 2.2686426180873513,
+      "learning_rate": 5.3979668151360905e-06,
+      "loss": 0.6769,
+      "step": 12050
+    },
+    {
+      "epoch": 6.874242964018525,
+      "grad_norm": 2.226550828270745,
+      "learning_rate": 5.380307445617048e-06,
+      "loss": 0.6759,
+      "step": 12060
+    },
+    {
+      "epoch": 6.879942999643748,
+      "grad_norm": 2.3277448858046332,
+      "learning_rate": 5.362666372617331e-06,
+      "loss": 0.6758,
+      "step": 12070
+    },
+    {
+      "epoch": 6.885643035268971,
+      "grad_norm": 2.278894744962253,
+      "learning_rate": 5.345043666005287e-06,
+      "loss": 0.6658,
+      "step": 12080
+    },
+    {
+      "epoch": 6.891343070894193,
+      "grad_norm": 2.181303637950725,
+      "learning_rate": 5.327439395576503e-06,
+      "loss": 0.6705,
+      "step": 12090
+    },
+    {
+      "epoch": 6.8970431065194155,
+      "grad_norm": 2.2149757242569086,
+      "learning_rate": 5.309853631053563e-06,
+      "loss": 0.6748,
+      "step": 12100
+    },
+    {
+      "epoch": 6.902743142144638,
+      "grad_norm": 2.192223874753431,
+      "learning_rate": 5.2922864420857425e-06,
+      "loss": 0.6728,
+      "step": 12110
+    },
+    {
+      "epoch": 6.908443177769861,
+      "grad_norm": 2.2881232299842993,
+      "learning_rate": 5.274737898248767e-06,
+      "loss": 0.6847,
+      "step": 12120
+    },
+    {
+      "epoch": 6.914143213395084,
+      "grad_norm": 2.187704076050992,
+      "learning_rate": 5.257208069044501e-06,
+      "loss": 0.6755,
+      "step": 12130
+    },
+    {
+      "epoch": 6.919843249020307,
+      "grad_norm": 2.2275431904161938,
+      "learning_rate": 5.239697023900696e-06,
+      "loss": 0.6694,
+      "step": 12140
+    },
+    {
+      "epoch": 6.925543284645529,
+      "grad_norm": 2.2332210515648416,
+      "learning_rate": 5.222204832170705e-06,
+      "loss": 0.6787,
+      "step": 12150
+    },
+    {
+      "epoch": 6.9312433202707515,
+      "grad_norm": 2.180968213454905,
+      "learning_rate": 5.204731563133214e-06,
+      "loss": 0.6771,
+      "step": 12160
+    },
+    {
+      "epoch": 6.936943355895974,
+      "grad_norm": 2.2106229537355526,
+      "learning_rate": 5.187277285991963e-06,
+      "loss": 0.6743,
+      "step": 12170
+    },
+    {
+      "epoch": 6.942643391521197,
+      "grad_norm": 2.233420363116453,
+      "learning_rate": 5.169842069875474e-06,
+      "loss": 0.6817,
+      "step": 12180
+    },
+    {
+      "epoch": 6.94834342714642,
+      "grad_norm": 2.2651013060170366,
+      "learning_rate": 5.152425983836777e-06,
+      "loss": 0.6808,
+      "step": 12190
+    },
+    {
+      "epoch": 6.954043462771642,
+      "grad_norm": 2.172658855315481,
+      "learning_rate": 5.135029096853132e-06,
+      "loss": 0.6699,
+      "step": 12200
+    },
+    {
+      "epoch": 6.959743498396865,
+      "grad_norm": 2.2475701558030265,
+      "learning_rate": 5.117651477825776e-06,
+      "loss": 0.6749,
+      "step": 12210
+    },
+    {
+      "epoch": 6.9654435340220875,
+      "grad_norm": 2.1616444155760446,
+      "learning_rate": 5.100293195579613e-06,
+      "loss": 0.6681,
+      "step": 12220
+    },
+    {
+      "epoch": 6.97114356964731,
+      "grad_norm": 2.2189909840543285,
+      "learning_rate": 5.082954318862978e-06,
+      "loss": 0.6765,
+      "step": 12230
+    },
+    {
+      "epoch": 6.976843605272533,
+      "grad_norm": 2.2374869184686443,
+      "learning_rate": 5.0656349163473405e-06,
+      "loss": 0.6758,
+      "step": 12240
+    },
+    {
+      "epoch": 6.982543640897756,
+      "grad_norm": 2.24523811915557,
+      "learning_rate": 5.048335056627043e-06,
+      "loss": 0.6793,
+      "step": 12250
+    },
+    {
+      "epoch": 6.988243676522979,
+      "grad_norm": 2.1336155734190387,
+      "learning_rate": 5.031054808219038e-06,
+      "loss": 0.6733,
+      "step": 12260
+    },
+    {
+      "epoch": 6.993943712148201,
+      "grad_norm": 2.0935622253465223,
+      "learning_rate": 5.013794239562593e-06,
+      "loss": 0.6736,
+      "step": 12270
+    },
+    {
+      "epoch": 6.9996437477734235,
+      "grad_norm": 2.226381254274828,
+      "learning_rate": 4.996553419019039e-06,
+      "loss": 0.6801,
+      "step": 12280
+    },
+    {
+      "epoch": 7.005343783398646,
+      "grad_norm": 2.2742882385031793,
+      "learning_rate": 4.9793324148714935e-06,
+      "loss": 0.66,
+      "step": 12290
+    },
+    {
+      "epoch": 7.011043819023869,
+      "grad_norm": 2.4163699536407717,
+      "learning_rate": 4.962131295324588e-06,
+      "loss": 0.6675,
+      "step": 12300
+    },
+    {
+      "epoch": 7.016743854649092,
+      "grad_norm": 2.2283661151132548,
+      "learning_rate": 4.944950128504202e-06,
+      "loss": 0.6717,
+      "step": 12310
+    },
+    {
+      "epoch": 7.022443890274314,
+      "grad_norm": 2.2757432948656873,
+      "learning_rate": 4.9277889824571925e-06,
+      "loss": 0.6669,
+      "step": 12320
+    },
+    {
+      "epoch": 7.028143925899537,
+      "grad_norm": 2.200927526322527,
+      "learning_rate": 4.910647925151115e-06,
+      "loss": 0.6622,
+      "step": 12330
+    },
+    {
+      "epoch": 7.033843961524759,
+      "grad_norm": 2.2724813454996355,
+      "learning_rate": 4.893527024473979e-06,
+      "loss": 0.6572,
+      "step": 12340
+    },
+    {
+      "epoch": 7.039543997149982,
+      "grad_norm": 2.186235554279113,
+      "learning_rate": 4.876426348233948e-06,
+      "loss": 0.6626,
+      "step": 12350
+    },
+    {
+      "epoch": 7.045244032775205,
+      "grad_norm": 2.309704746767001,
+      "learning_rate": 4.85934596415909e-06,
+      "loss": 0.6636,
+      "step": 12360
+    },
+    {
+      "epoch": 7.050944068400428,
+      "grad_norm": 2.269715193602036,
+      "learning_rate": 4.842285939897107e-06,
+      "loss": 0.6621,
+      "step": 12370
+    },
+    {
+      "epoch": 7.05664410402565,
+      "grad_norm": 2.3094500491239187,
+      "learning_rate": 4.825246343015056e-06,
+      "loss": 0.657,
+      "step": 12380
+    },
+    {
+      "epoch": 7.062344139650873,
+      "grad_norm": 2.2173637465487848,
+      "learning_rate": 4.808227240999109e-06,
+      "loss": 0.6591,
+      "step": 12390
+    },
+    {
+      "epoch": 7.068044175276095,
+      "grad_norm": 2.3536529122634366,
+      "learning_rate": 4.791228701254251e-06,
+      "loss": 0.6561,
+      "step": 12400
+    },
+    {
+      "epoch": 7.073744210901318,
+      "grad_norm": 2.2288027952210956,
+      "learning_rate": 4.774250791104033e-06,
+      "loss": 0.6647,
+      "step": 12410
+    },
+    {
+      "epoch": 7.079444246526541,
+      "grad_norm": 2.1423835925047,
+      "learning_rate": 4.757293577790302e-06,
+      "loss": 0.6634,
+      "step": 12420
+    },
+    {
+      "epoch": 7.085144282151764,
+      "grad_norm": 2.2470079391106665,
+      "learning_rate": 4.740357128472936e-06,
+      "loss": 0.6785,
+      "step": 12430
+    },
+    {
+      "epoch": 7.090844317776986,
+      "grad_norm": 2.2574430392774785,
+      "learning_rate": 4.723441510229572e-06,
+      "loss": 0.6785,
+      "step": 12440
+    },
+    {
+      "epoch": 7.096544353402209,
+      "grad_norm": 2.3043198643832166,
+      "learning_rate": 4.70654679005535e-06,
+      "loss": 0.6596,
+      "step": 12450
+    },
+    {
+      "epoch": 7.102244389027431,
+      "grad_norm": 2.295680833918298,
+      "learning_rate": 4.689673034862637e-06,
+      "loss": 0.6594,
+      "step": 12460
+    },
+    {
+      "epoch": 7.107944424652654,
+      "grad_norm": 2.1605206188100206,
+      "learning_rate": 4.672820311480768e-06,
+      "loss": 0.6627,
+      "step": 12470
+    },
+    {
+      "epoch": 7.113644460277877,
+      "grad_norm": 2.1919151628756586,
+      "learning_rate": 4.655988686655787e-06,
+      "loss": 0.6667,
+      "step": 12480
+    },
+    {
+      "epoch": 7.1193444959031,
+      "grad_norm": 2.267528809716308,
+      "learning_rate": 4.639178227050169e-06,
+      "loss": 0.6582,
+      "step": 12490
+    },
+    {
+      "epoch": 7.125044531528322,
+      "grad_norm": 2.3241456885445535,
+      "learning_rate": 4.622388999242564e-06,
+      "loss": 0.6558,
+      "step": 12500
+    },
+    {
+      "epoch": 7.1307445671535445,
+      "grad_norm": 2.2495978263074763,
+      "learning_rate": 4.6056210697275315e-06,
+      "loss": 0.666,
+      "step": 12510
+    },
+    {
+      "epoch": 7.136444602778767,
+      "grad_norm": 2.3115077705620957,
+      "learning_rate": 4.588874504915287e-06,
+      "loss": 0.6659,
+      "step": 12520
+    },
+    {
+      "epoch": 7.14214463840399,
+      "grad_norm": 2.2516114253186887,
+      "learning_rate": 4.572149371131419e-06,
+      "loss": 0.6569,
+      "step": 12530
+    },
+    {
+      "epoch": 7.147844674029213,
+      "grad_norm": 2.331648891990796,
+      "learning_rate": 4.555445734616641e-06,
+      "loss": 0.6671,
+      "step": 12540
+    },
+    {
+      "epoch": 7.153544709654435,
+      "grad_norm": 2.3183931570473444,
+      "learning_rate": 4.538763661526527e-06,
+      "loss": 0.6711,
+      "step": 12550
+    },
+    {
+      "epoch": 7.159244745279658,
+      "grad_norm": 2.208243417910211,
+      "learning_rate": 4.522103217931247e-06,
+      "loss": 0.6503,
+      "step": 12560
+    },
+    {
+      "epoch": 7.1649447809048805,
+      "grad_norm": 2.155675760594788,
+      "learning_rate": 4.505464469815307e-06,
+      "loss": 0.6661,
+      "step": 12570
+    },
+    {
+      "epoch": 7.170644816530103,
+      "grad_norm": 2.2664894262413497,
+      "learning_rate": 4.488847483077285e-06,
+      "loss": 0.666,
+      "step": 12580
+    },
+    {
+      "epoch": 7.176344852155326,
+      "grad_norm": 2.2795499909428423,
+      "learning_rate": 4.472252323529575e-06,
+      "loss": 0.6674,
+      "step": 12590
+    },
+    {
+      "epoch": 7.182044887780549,
+      "grad_norm": 2.223281501378646,
+      "learning_rate": 4.455679056898116e-06,
+      "loss": 0.6597,
+      "step": 12600
+    },
+    {
+      "epoch": 7.187744923405771,
+      "grad_norm": 2.277136474302064,
+      "learning_rate": 4.439127748822153e-06,
+      "loss": 0.6648,
+      "step": 12610
+    },
+    {
+      "epoch": 7.193444959030994,
+      "grad_norm": 2.2711057573337072,
+      "learning_rate": 4.4225984648539525e-06,
+      "loss": 0.677,
+      "step": 12620
+    },
+    {
+      "epoch": 7.1991449946562165,
+      "grad_norm": 2.2561112509358727,
+      "learning_rate": 4.406091270458553e-06,
+      "loss": 0.6709,
+      "step": 12630
+    },
+    {
+      "epoch": 7.204845030281439,
+      "grad_norm": 2.2208121144419524,
+      "learning_rate": 4.389606231013512e-06,
+      "loss": 0.6582,
+      "step": 12640
+    },
+    {
+      "epoch": 7.210545065906662,
+      "grad_norm": 2.2549802019435083,
+      "learning_rate": 4.3731434118086324e-06,
+      "loss": 0.6633,
+      "step": 12650
+    },
+    {
+      "epoch": 7.216245101531885,
+      "grad_norm": 2.252019210437217,
+      "learning_rate": 4.356702878045728e-06,
+      "loss": 0.6662,
+      "step": 12660
+    },
+    {
+      "epoch": 7.221945137157107,
+      "grad_norm": 2.2006478078981693,
+      "learning_rate": 4.3402846948383334e-06,
+      "loss": 0.6622,
+      "step": 12670
+    },
+    {
+      "epoch": 7.22764517278233,
+      "grad_norm": 2.3860767372289704,
+      "learning_rate": 4.323888927211472e-06,
+      "loss": 0.6759,
+      "step": 12680
+    },
+    {
+      "epoch": 7.2333452084075525,
+      "grad_norm": 2.384532210236136,
+      "learning_rate": 4.307515640101387e-06,
+      "loss": 0.6648,
+      "step": 12690
+    },
+    {
+      "epoch": 7.239045244032775,
+      "grad_norm": 2.2186370735450662,
+      "learning_rate": 4.291164898355286e-06,
+      "loss": 0.6614,
+      "step": 12700
+    },
+    {
+      "epoch": 7.244745279657998,
+      "grad_norm": 2.275761327122912,
+      "learning_rate": 4.274836766731087e-06,
+      "loss": 0.6583,
+      "step": 12710
+    },
+    {
+      "epoch": 7.250445315283221,
+      "grad_norm": 2.3723166854023945,
+      "learning_rate": 4.2585313098971535e-06,
+      "loss": 0.6676,
+      "step": 12720
+    },
+    {
+      "epoch": 7.256145350908443,
+      "grad_norm": 2.1797372895464084,
+      "learning_rate": 4.242248592432048e-06,
+      "loss": 0.6629,
+      "step": 12730
+    },
+    {
+      "epoch": 7.261845386533666,
+      "grad_norm": 2.223664652891473,
+      "learning_rate": 4.225988678824279e-06,
+      "loss": 0.6704,
+      "step": 12740
+    },
+    {
+      "epoch": 7.267545422158888,
+      "grad_norm": 2.23232293309273,
+      "learning_rate": 4.209751633472029e-06,
+      "loss": 0.676,
+      "step": 12750
+    },
+    {
+      "epoch": 7.273245457784111,
+      "grad_norm": 2.245225811111464,
+      "learning_rate": 4.1935375206829156e-06,
+      "loss": 0.6634,
+      "step": 12760
+    },
+    {
+      "epoch": 7.278945493409334,
+      "grad_norm": 2.3450273180018466,
+      "learning_rate": 4.1773464046737275e-06,
+      "loss": 0.6745,
+      "step": 12770
+    },
+    {
+      "epoch": 7.284645529034557,
+      "grad_norm": 2.297072523505608,
+      "learning_rate": 4.161178349570173e-06,
+      "loss": 0.6689,
+      "step": 12780
+    },
+    {
+      "epoch": 7.290345564659779,
+      "grad_norm": 2.2269748343650537,
+      "learning_rate": 4.145033419406635e-06,
+      "loss": 0.6625,
+      "step": 12790
+    },
+    {
+      "epoch": 7.296045600285002,
+      "grad_norm": 2.3128792584540534,
+      "learning_rate": 4.128911678125902e-06,
+      "loss": 0.6645,
+      "step": 12800
+    },
+    {
+      "epoch": 7.301745635910224,
+      "grad_norm": 2.232593518867681,
+      "learning_rate": 4.112813189578917e-06,
+      "loss": 0.6658,
+      "step": 12810
+    },
+    {
+      "epoch": 7.307445671535447,
+      "grad_norm": 2.1477231869993267,
+      "learning_rate": 4.096738017524533e-06,
+      "loss": 0.6608,
+      "step": 12820
+    },
+    {
+      "epoch": 7.31314570716067,
+      "grad_norm": 2.3814575079547184,
+      "learning_rate": 4.080686225629267e-06,
+      "loss": 0.6643,
+      "step": 12830
+    },
+    {
+      "epoch": 7.318845742785893,
+      "grad_norm": 2.2391714897073522,
+      "learning_rate": 4.064657877467025e-06,
+      "loss": 0.6463,
+      "step": 12840
+    },
+    {
+      "epoch": 7.324545778411115,
+      "grad_norm": 2.2795231271731833,
+      "learning_rate": 4.048653036518869e-06,
+      "loss": 0.6593,
+      "step": 12850
+    },
+    {
+      "epoch": 7.330245814036338,
+      "grad_norm": 2.2963500847619196,
+      "learning_rate": 4.032671766172756e-06,
+      "loss": 0.6589,
+      "step": 12860
+    },
+    {
+      "epoch": 7.33594584966156,
+      "grad_norm": 2.186846566909248,
+      "learning_rate": 4.016714129723291e-06,
+      "loss": 0.6707,
+      "step": 12870
+    },
+    {
+      "epoch": 7.341645885286783,
+      "grad_norm": 2.3184883940263243,
+      "learning_rate": 4.00078019037148e-06,
+      "loss": 0.6701,
+      "step": 12880
+    },
+    {
+      "epoch": 7.347345920912006,
+      "grad_norm": 2.2939704186094976,
+      "learning_rate": 3.984870011224474e-06,
+      "loss": 0.6648,
+      "step": 12890
+    },
+    {
+      "epoch": 7.353045956537228,
+      "grad_norm": 2.1953034187690497,
+      "learning_rate": 3.968983655295317e-06,
+      "loss": 0.6736,
+      "step": 12900
+    },
+    {
+      "epoch": 7.358745992162451,
+      "grad_norm": 2.233747449114086,
+      "learning_rate": 3.953121185502699e-06,
+      "loss": 0.6745,
+      "step": 12910
+    },
+    {
+      "epoch": 7.3644460277876735,
+      "grad_norm": 2.348797145656744,
+      "learning_rate": 3.9372826646707215e-06,
+      "loss": 0.6576,
+      "step": 12920
+    },
+    {
+      "epoch": 7.370146063412896,
+      "grad_norm": 2.279768759531536,
+      "learning_rate": 3.921468155528614e-06,
+      "loss": 0.6681,
+      "step": 12930
+    },
+    {
+      "epoch": 7.375846099038119,
+      "grad_norm": 2.332254197841433,
+      "learning_rate": 3.90567772071052e-06,
+      "loss": 0.658,
+      "step": 12940
+    },
+    {
+      "epoch": 7.381546134663342,
+      "grad_norm": 2.2512646676554167,
+      "learning_rate": 3.8899114227552315e-06,
+      "loss": 0.6643,
+      "step": 12950
+    },
+    {
+      "epoch": 7.387246170288564,
+      "grad_norm": 2.229958740556661,
+      "learning_rate": 3.874169324105945e-06,
+      "loss": 0.6682,
+      "step": 12960
+    },
+    {
+      "epoch": 7.392946205913787,
+      "grad_norm": 2.2359413189154433,
+      "learning_rate": 3.8584514871100206e-06,
+      "loss": 0.6643,
+      "step": 12970
+    },
+    {
+      "epoch": 7.3986462415390095,
+      "grad_norm": 2.2773932967294597,
+      "learning_rate": 3.842757974018721e-06,
+      "loss": 0.6674,
+      "step": 12980
+    },
+    {
+      "epoch": 7.404346277164232,
+      "grad_norm": 2.26805369430861,
+      "learning_rate": 3.827088846986977e-06,
+      "loss": 0.6578,
+      "step": 12990
+    },
+    {
+      "epoch": 7.410046312789455,
+      "grad_norm": 2.2090255220562858,
+      "learning_rate": 3.8114441680731317e-06,
+      "loss": 0.6627,
+      "step": 13000
+    },
+    {
+      "epoch": 7.415746348414678,
+      "grad_norm": 2.419885181729568,
+      "learning_rate": 3.7958239992387113e-06,
+      "loss": 0.6672,
+      "step": 13010
+    },
+    {
+      "epoch": 7.4214463840399,
+      "grad_norm": 2.35136144391521,
+      "learning_rate": 3.7802284023481582e-06,
+      "loss": 0.6591,
+      "step": 13020
+    },
+    {
+      "epoch": 7.427146419665123,
+      "grad_norm": 2.214639339938116,
+      "learning_rate": 3.7646574391686007e-06,
+      "loss": 0.6593,
+      "step": 13030
+    },
+    {
+      "epoch": 7.4328464552903455,
+      "grad_norm": 2.255892281186678,
+      "learning_rate": 3.7491111713696026e-06,
+      "loss": 0.6728,
+      "step": 13040
+    },
+    {
+      "epoch": 7.438546490915568,
+      "grad_norm": 2.20115600566101,
+      "learning_rate": 3.733589660522923e-06,
+      "loss": 0.6656,
+      "step": 13050
+    },
+    {
+      "epoch": 7.444246526540791,
+      "grad_norm": 2.18933378089345,
+      "learning_rate": 3.718092968102267e-06,
+      "loss": 0.6544,
+      "step": 13060
+    },
+    {
+      "epoch": 7.449946562166014,
+      "grad_norm": 2.21831722919064,
+      "learning_rate": 3.702621155483046e-06,
+      "loss": 0.6719,
+      "step": 13070
+    },
+    {
+      "epoch": 7.455646597791236,
+      "grad_norm": 2.3224124638500188,
+      "learning_rate": 3.6871742839421344e-06,
+      "loss": 0.663,
+      "step": 13080
+    },
+    {
+      "epoch": 7.461346633416459,
+      "grad_norm": 2.285975538502232,
+      "learning_rate": 3.6717524146576234e-06,
+      "loss": 0.6627,
+      "step": 13090
+    },
+    {
+      "epoch": 7.4670466690416815,
+      "grad_norm": 2.2816850224350693,
+      "learning_rate": 3.6563556087085894e-06,
+      "loss": 0.6607,
+      "step": 13100
+    },
+    {
+      "epoch": 7.472746704666904,
+      "grad_norm": 2.223032277100294,
+      "learning_rate": 3.640983927074836e-06,
+      "loss": 0.6718,
+      "step": 13110
+    },
+    {
+      "epoch": 7.478446740292127,
+      "grad_norm": 2.3196995262580264,
+      "learning_rate": 3.6256374306366635e-06,
+      "loss": 0.6601,
+      "step": 13120
+    },
+    {
+      "epoch": 7.48414677591735,
+      "grad_norm": 2.179693956322987,
+      "learning_rate": 3.6103161801746224e-06,
+      "loss": 0.6634,
+      "step": 13130
+    },
+    {
+      "epoch": 7.489846811542572,
+      "grad_norm": 2.3196268076269093,
+      "learning_rate": 3.5950202363692822e-06,
+      "loss": 0.6626,
+      "step": 13140
+    },
+    {
+      "epoch": 7.495546847167795,
+      "grad_norm": 2.1820384798845858,
+      "learning_rate": 3.5797496598009794e-06,
+      "loss": 0.6632,
+      "step": 13150
+    },
+    {
+      "epoch": 7.501246882793017,
+      "grad_norm": 2.2375359061986417,
+      "learning_rate": 3.564504510949581e-06,
+      "loss": 0.6693,
+      "step": 13160
+    },
+    {
+      "epoch": 7.50694691841824,
+      "grad_norm": 2.2623386175960385,
+      "learning_rate": 3.54928485019425e-06,
+      "loss": 0.6685,
+      "step": 13170
+    },
+    {
+      "epoch": 7.512646954043463,
+      "grad_norm": 2.3114477674940273,
+      "learning_rate": 3.534090737813198e-06,
+      "loss": 0.669,
+      "step": 13180
+    },
+    {
+      "epoch": 7.518346989668686,
+      "grad_norm": 2.2773591124406862,
+      "learning_rate": 3.518922233983455e-06,
+      "loss": 0.6648,
+      "step": 13190
+    },
+    {
+      "epoch": 7.524047025293908,
+      "grad_norm": 2.2249861523523027,
+      "learning_rate": 3.503779398780626e-06,
+      "loss": 0.6674,
+      "step": 13200
+    },
+    {
+      "epoch": 7.529747060919131,
+      "grad_norm": 2.370282321028903,
+      "learning_rate": 3.4886622921786517e-06,
+      "loss": 0.6679,
+      "step": 13210
+    },
+    {
+      "epoch": 7.535447096544353,
+      "grad_norm": 2.2532698509194544,
+      "learning_rate": 3.4735709740495748e-06,
+      "loss": 0.6731,
+      "step": 13220
+    },
+    {
+      "epoch": 7.541147132169576,
+      "grad_norm": 2.30291311854667,
+      "learning_rate": 3.4585055041633076e-06,
+      "loss": 0.6609,
+      "step": 13230
+    },
+    {
+      "epoch": 7.546847167794799,
+      "grad_norm": 2.1904288998748833,
+      "learning_rate": 3.4434659421873807e-06,
+      "loss": 0.6629,
+      "step": 13240
+    },
+    {
+      "epoch": 7.552547203420021,
+      "grad_norm": 2.2214360575027823,
+      "learning_rate": 3.428452347686717e-06,
+      "loss": 0.6712,
+      "step": 13250
+    },
+    {
+      "epoch": 7.558247239045244,
+      "grad_norm": 2.294107190530825,
+      "learning_rate": 3.4134647801233976e-06,
+      "loss": 0.6684,
+      "step": 13260
+    },
+    {
+      "epoch": 7.563947274670467,
+      "grad_norm": 2.23145548647493,
+      "learning_rate": 3.3985032988564147e-06,
+      "loss": 0.6651,
+      "step": 13270
+    },
+    {
+      "epoch": 7.569647310295689,
+      "grad_norm": 2.3301816146667638,
+      "learning_rate": 3.3835679631414588e-06,
+      "loss": 0.6653,
+      "step": 13280
+    },
+    {
+      "epoch": 7.575347345920912,
+      "grad_norm": 2.4013703507360225,
+      "learning_rate": 3.3686588321306546e-06,
+      "loss": 0.6603,
+      "step": 13290
+    },
+    {
+      "epoch": 7.581047381546135,
+      "grad_norm": 2.3636437939957324,
+      "learning_rate": 3.35377596487235e-06,
+      "loss": 0.6591,
+      "step": 13300
+    },
+    {
+      "epoch": 7.586747417171358,
+      "grad_norm": 2.4085206590723485,
+      "learning_rate": 3.338919420310871e-06,
+      "loss": 0.6793,
+      "step": 13310
+    },
+    {
+      "epoch": 7.59244745279658,
+      "grad_norm": 2.234619347891406,
+      "learning_rate": 3.3240892572862924e-06,
+      "loss": 0.6711,
+      "step": 13320
+    },
+    {
+      "epoch": 7.5981474884218025,
+      "grad_norm": 2.231117043894081,
+      "learning_rate": 3.3092855345342047e-06,
+      "loss": 0.6655,
+      "step": 13330
+    },
+    {
+      "epoch": 7.603847524047025,
+      "grad_norm": 2.268181240569829,
+      "learning_rate": 3.294508310685478e-06,
+      "loss": 0.661,
+      "step": 13340
+    },
+    {
+      "epoch": 7.609547559672248,
+      "grad_norm": 2.33558451140121,
+      "learning_rate": 3.2797576442660293e-06,
+      "loss": 0.6635,
+      "step": 13350
+    },
+    {
+      "epoch": 7.615247595297471,
+      "grad_norm": 2.301128072633845,
+      "learning_rate": 3.265033593696606e-06,
+      "loss": 0.6627,
+      "step": 13360
+    },
+    {
+      "epoch": 7.620947630922693,
+      "grad_norm": 2.1001706479882993,
+      "learning_rate": 3.25033621729253e-06,
+      "loss": 0.6675,
+      "step": 13370
+    },
+    {
+      "epoch": 7.626647666547916,
+      "grad_norm": 2.2899545362349767,
+      "learning_rate": 3.2356655732634825e-06,
+      "loss": 0.667,
+      "step": 13380
+    },
+    {
+      "epoch": 7.6323477021731385,
+      "grad_norm": 2.236065324768278,
+      "learning_rate": 3.2210217197132685e-06,
+      "loss": 0.6647,
+      "step": 13390
+    },
+    {
+      "epoch": 7.638047737798361,
+      "grad_norm": 2.423950515329974,
+      "learning_rate": 3.2064047146395894e-06,
+      "loss": 0.6582,
+      "step": 13400
+    },
+    {
+      "epoch": 7.643747773423584,
+      "grad_norm": 2.113006486576792,
+      "learning_rate": 3.191814615933816e-06,
+      "loss": 0.6694,
+      "step": 13410
+    },
+    {
+      "epoch": 7.649447809048807,
+      "grad_norm": 2.272493144954798,
+      "learning_rate": 3.1772514813807474e-06,
+      "loss": 0.664,
+      "step": 13420
+    },
+    {
+      "epoch": 7.655147844674029,
+      "grad_norm": 2.2477799449220157,
+      "learning_rate": 3.1627153686583954e-06,
+      "loss": 0.6665,
+      "step": 13430
+    },
+    {
+      "epoch": 7.660847880299252,
+      "grad_norm": 2.359330416347929,
+      "learning_rate": 3.1482063353377468e-06,
+      "loss": 0.6608,
+      "step": 13440
+    },
+    {
+      "epoch": 7.6665479159244745,
+      "grad_norm": 2.2768436025456107,
+      "learning_rate": 3.1337244388825428e-06,
+      "loss": 0.6662,
+      "step": 13450
+    },
+    {
+      "epoch": 7.672247951549697,
+      "grad_norm": 2.222221818207048,
+      "learning_rate": 3.1192697366490443e-06,
+      "loss": 0.6691,
+      "step": 13460
+    },
+    {
+      "epoch": 7.67794798717492,
+      "grad_norm": 2.318255375011537,
+      "learning_rate": 3.104842285885811e-06,
+      "loss": 0.6597,
+      "step": 13470
+    },
+    {
+      "epoch": 7.683648022800142,
+      "grad_norm": 2.2622741956334798,
+      "learning_rate": 3.0904421437334685e-06,
+      "loss": 0.647,
+      "step": 13480
+    },
+    {
+      "epoch": 7.689348058425365,
+      "grad_norm": 2.3412395862257624,
+      "learning_rate": 3.0760693672244858e-06,
+      "loss": 0.6608,
+      "step": 13490
+    },
+    {
+      "epoch": 7.695048094050588,
+      "grad_norm": 2.2700240775817644,
+      "learning_rate": 3.061724013282956e-06,
+      "loss": 0.6575,
+      "step": 13500
+    },
+    {
+      "epoch": 7.7007481296758105,
+      "grad_norm": 2.2489411090100275,
+      "learning_rate": 3.047406138724355e-06,
+      "loss": 0.6635,
+      "step": 13510
+    },
+    {
+      "epoch": 7.706448165301033,
+      "grad_norm": 2.40198054648677,
+      "learning_rate": 3.0331158002553296e-06,
+      "loss": 0.6638,
+      "step": 13520
+    },
+    {
+      "epoch": 7.712148200926256,
+      "grad_norm": 2.276918761843776,
+      "learning_rate": 3.018853054473463e-06,
+      "loss": 0.665,
+      "step": 13530
+    },
+    {
+      "epoch": 7.717848236551479,
+      "grad_norm": 2.2169633358638414,
+      "learning_rate": 3.0046179578670664e-06,
+      "loss": 0.6632,
+      "step": 13540
+    },
+    {
+      "epoch": 7.723548272176701,
+      "grad_norm": 2.183177198102162,
+      "learning_rate": 2.9904105668149374e-06,
+      "loss": 0.668,
+      "step": 13550
+    },
+    {
+      "epoch": 7.729248307801924,
+      "grad_norm": 2.3235798898652975,
+      "learning_rate": 2.9762309375861462e-06,
+      "loss": 0.6633,
+      "step": 13560
+    },
+    {
+      "epoch": 7.734948343427146,
+      "grad_norm": 2.325494714539688,
+      "learning_rate": 2.9620791263398142e-06,
+      "loss": 0.6619,
+      "step": 13570
+    },
+    {
+      "epoch": 7.740648379052369,
+      "grad_norm": 2.2715717713625696,
+      "learning_rate": 2.9479551891248746e-06,
+      "loss": 0.6591,
+      "step": 13580
+    },
+    {
+      "epoch": 7.746348414677592,
+      "grad_norm": 2.32880260951295,
+      "learning_rate": 2.9338591818798856e-06,
+      "loss": 0.6642,
+      "step": 13590
+    },
+    {
+      "epoch": 7.752048450302814,
+      "grad_norm": 2.1873728881485874,
+      "learning_rate": 2.919791160432772e-06,
+      "loss": 0.6644,
+      "step": 13600
+    },
+    {
+      "epoch": 7.757748485928037,
+      "grad_norm": 2.2332074048709165,
+      "learning_rate": 2.9057511805006246e-06,
+      "loss": 0.6637,
+      "step": 13610
+    },
+    {
+      "epoch": 7.76344852155326,
+      "grad_norm": 2.2648757192147073,
+      "learning_rate": 2.89173929768947e-06,
+      "loss": 0.6672,
+      "step": 13620
+    },
+    {
+      "epoch": 7.769148557178482,
+      "grad_norm": 2.258292024109965,
+      "learning_rate": 2.877755567494066e-06,
+      "loss": 0.6566,
+      "step": 13630
+    },
+    {
+      "epoch": 7.774848592803705,
+      "grad_norm": 2.2498733750901283,
+      "learning_rate": 2.863800045297659e-06,
+      "loss": 0.6713,
+      "step": 13640
+    },
+    {
+      "epoch": 7.780548628428928,
+      "grad_norm": 2.257203818446692,
+      "learning_rate": 2.8498727863717803e-06,
+      "loss": 0.6689,
+      "step": 13650
+    },
+    {
+      "epoch": 7.786248664054151,
+      "grad_norm": 2.2968049679993947,
+      "learning_rate": 2.835973845876022e-06,
+      "loss": 0.6716,
+      "step": 13660
+    },
+    {
+      "epoch": 7.791948699679373,
+      "grad_norm": 2.3196500348038955,
+      "learning_rate": 2.8221032788578206e-06,
+      "loss": 0.6732,
+      "step": 13670
+    },
+    {
+      "epoch": 7.797648735304596,
+      "grad_norm": 2.216899403748187,
+      "learning_rate": 2.808261140252242e-06,
+      "loss": 0.6718,
+      "step": 13680
+    },
+    {
+      "epoch": 7.803348770929818,
+      "grad_norm": 2.254643395065768,
+      "learning_rate": 2.7944474848817572e-06,
+      "loss": 0.6619,
+      "step": 13690
+    },
+    {
+      "epoch": 7.809048806555041,
+      "grad_norm": 2.2820083783430105,
+      "learning_rate": 2.780662367456021e-06,
+      "loss": 0.6655,
+      "step": 13700
+    },
+    {
+      "epoch": 7.814748842180264,
+      "grad_norm": 2.2433132656609933,
+      "learning_rate": 2.7669058425716676e-06,
+      "loss": 0.6602,
+      "step": 13710
+    },
+    {
+      "epoch": 7.820448877805486,
+      "grad_norm": 2.335507982400234,
+      "learning_rate": 2.753177964712096e-06,
+      "loss": 0.6596,
+      "step": 13720
+    },
+    {
+      "epoch": 7.826148913430709,
+      "grad_norm": 2.2275175940366965,
+      "learning_rate": 2.7394787882472374e-06,
+      "loss": 0.6644,
+      "step": 13730
+    },
+    {
+      "epoch": 7.8318489490559315,
+      "grad_norm": 2.295329033852509,
+      "learning_rate": 2.7258083674333545e-06,
+      "loss": 0.6616,
+      "step": 13740
+    },
+    {
+      "epoch": 7.837548984681154,
+      "grad_norm": 2.192580449354095,
+      "learning_rate": 2.7121667564128173e-06,
+      "loss": 0.6535,
+      "step": 13750
+    },
+    {
+      "epoch": 7.843249020306377,
+      "grad_norm": 2.342528376405033,
+      "learning_rate": 2.698554009213902e-06,
+      "loss": 0.6774,
+      "step": 13760
+    },
+    {
+      "epoch": 7.8489490559316,
+      "grad_norm": 2.2950904697540024,
+      "learning_rate": 2.684970179750559e-06,
+      "loss": 0.6659,
+      "step": 13770
+    },
+    {
+      "epoch": 7.854649091556822,
+      "grad_norm": 2.328359619364796,
+      "learning_rate": 2.6714153218222127e-06,
+      "loss": 0.6699,
+      "step": 13780
+    },
+    {
+      "epoch": 7.860349127182045,
+      "grad_norm": 2.215674623073645,
+      "learning_rate": 2.6578894891135455e-06,
+      "loss": 0.6633,
+      "step": 13790
+    },
+    {
+      "epoch": 7.8660491628072675,
+      "grad_norm": 2.3862089040460206,
+      "learning_rate": 2.6443927351942798e-06,
+      "loss": 0.6624,
+      "step": 13800
+    },
+    {
+      "epoch": 7.87174919843249,
+      "grad_norm": 2.243717235079729,
+      "learning_rate": 2.630925113518974e-06,
+      "loss": 0.6643,
+      "step": 13810
+    },
+    {
+      "epoch": 7.877449234057713,
+      "grad_norm": 2.2927993968105946,
+      "learning_rate": 2.617486677426806e-06,
+      "loss": 0.6627,
+      "step": 13820
+    },
+    {
+      "epoch": 7.883149269682935,
+      "grad_norm": 2.322818052755111,
+      "learning_rate": 2.6040774801413616e-06,
+      "loss": 0.6582,
+      "step": 13830
+    },
+    {
+      "epoch": 7.888849305308158,
+      "grad_norm": 2.2736195651698066,
+      "learning_rate": 2.590697574770421e-06,
+      "loss": 0.6633,
+      "step": 13840
+    },
+    {
+      "epoch": 7.894549340933381,
+      "grad_norm": 2.171993807344973,
+      "learning_rate": 2.5773470143057657e-06,
+      "loss": 0.6605,
+      "step": 13850
+    },
+    {
+      "epoch": 7.9002493765586035,
+      "grad_norm": 2.2587649223942456,
+      "learning_rate": 2.564025851622941e-06,
+      "loss": 0.6654,
+      "step": 13860
+    },
+    {
+      "epoch": 7.905949412183826,
+      "grad_norm": 2.333016160998036,
+      "learning_rate": 2.550734139481067e-06,
+      "loss": 0.6674,
+      "step": 13870
+    },
+    {
+      "epoch": 7.911649447809049,
+      "grad_norm": 2.369469386300691,
+      "learning_rate": 2.5374719305226226e-06,
+      "loss": 0.6709,
+      "step": 13880
+    },
+    {
+      "epoch": 7.917349483434272,
+      "grad_norm": 2.3460895427142434,
+      "learning_rate": 2.524239277273235e-06,
+      "loss": 0.671,
+      "step": 13890
+    },
+    {
+      "epoch": 7.923049519059494,
+      "grad_norm": 2.3133515698467852,
+      "learning_rate": 2.511036232141484e-06,
+      "loss": 0.6658,
+      "step": 13900
+    },
+    {
+      "epoch": 7.928749554684717,
+      "grad_norm": 2.1978929439065147,
+      "learning_rate": 2.497862847418674e-06,
+      "loss": 0.6588,
+      "step": 13910
+    },
+    {
+      "epoch": 7.9344495903099395,
+      "grad_norm": 2.339957351608019,
+      "learning_rate": 2.4847191752786437e-06,
+      "loss": 0.6611,
+      "step": 13920
+    },
+    {
+      "epoch": 7.940149625935162,
+      "grad_norm": 2.2748253461664456,
+      "learning_rate": 2.4716052677775524e-06,
+      "loss": 0.6594,
+      "step": 13930
+    },
+    {
+      "epoch": 7.945849661560385,
+      "grad_norm": 2.2765085031372365,
+      "learning_rate": 2.4585211768536754e-06,
+      "loss": 0.6682,
+      "step": 13940
+    },
+    {
+      "epoch": 7.951549697185607,
+      "grad_norm": 2.143922983187977,
+      "learning_rate": 2.445466954327196e-06,
+      "loss": 0.6636,
+      "step": 13950
+    },
+    {
+      "epoch": 7.95724973281083,
+      "grad_norm": 2.3081123161562926,
+      "learning_rate": 2.4324426519000056e-06,
+      "loss": 0.6693,
+      "step": 13960
+    },
+    {
+      "epoch": 7.962949768436053,
+      "grad_norm": 2.3648144828229434,
+      "learning_rate": 2.419448321155493e-06,
+      "loss": 0.6661,
+      "step": 13970
+    },
+    {
+      "epoch": 7.968649804061275,
+      "grad_norm": 2.319378028427634,
+      "learning_rate": 2.4064840135583413e-06,
+      "loss": 0.6511,
+      "step": 13980
+    },
+    {
+      "epoch": 7.974349839686498,
+      "grad_norm": 2.3294256788629664,
+      "learning_rate": 2.3935497804543317e-06,
+      "loss": 0.66,
+      "step": 13990
+    },
+    {
+      "epoch": 7.980049875311721,
+      "grad_norm": 2.3273735482235773,
+      "learning_rate": 2.380645673070129e-06,
+      "loss": 0.6638,
+      "step": 14000
+    },
+    {
+      "epoch": 7.985749910936943,
+      "grad_norm": 2.3047174470726413,
+      "learning_rate": 2.3677717425130832e-06,
+      "loss": 0.6664,
+      "step": 14010
+    },
+    {
+      "epoch": 7.991449946562166,
+      "grad_norm": 2.2591163992818006,
+      "learning_rate": 2.3549280397710273e-06,
+      "loss": 0.6536,
+      "step": 14020
+    },
+    {
+      "epoch": 7.997149982187389,
+      "grad_norm": 2.2848907363838156,
+      "learning_rate": 2.3421146157120813e-06,
+      "loss": 0.674,
+      "step": 14030
+    },
+    {
+      "epoch": 8.00285001781261,
+      "grad_norm": 2.395804129166372,
+      "learning_rate": 2.329331521084439e-06,
+      "loss": 0.663,
+      "step": 14040
+    },
+    {
+      "epoch": 8.008550053437833,
+      "grad_norm": 2.349509693484751,
+      "learning_rate": 2.3165788065161742e-06,
+      "loss": 0.6537,
+      "step": 14050
+    },
+    {
+      "epoch": 8.014250089063056,
+      "grad_norm": 2.317112213637775,
+      "learning_rate": 2.303856522515039e-06,
+      "loss": 0.6553,
+      "step": 14060
+    },
+    {
+      "epoch": 8.019950124688279,
+      "grad_norm": 2.2344274209565165,
+      "learning_rate": 2.291164719468265e-06,
+      "loss": 0.646,
+      "step": 14070
+    },
+    {
+      "epoch": 8.025650160313502,
+      "grad_norm": 2.2402337776074766,
+      "learning_rate": 2.2785034476423608e-06,
+      "loss": 0.6645,
+      "step": 14080
+    },
+    {
+      "epoch": 8.031350195938725,
+      "grad_norm": 2.2434977798959985,
+      "learning_rate": 2.2658727571829176e-06,
+      "loss": 0.6647,
+      "step": 14090
+    },
+    {
+      "epoch": 8.037050231563947,
+      "grad_norm": 2.2912602059287193,
+      "learning_rate": 2.2532726981144028e-06,
+      "loss": 0.6608,
+      "step": 14100
+    },
+    {
+      "epoch": 8.04275026718917,
+      "grad_norm": 2.2971956830820885,
+      "learning_rate": 2.2407033203399687e-06,
+      "loss": 0.6525,
+      "step": 14110
+    },
+    {
+      "epoch": 8.048450302814393,
+      "grad_norm": 2.5095251658904014,
+      "learning_rate": 2.2281646736412575e-06,
+      "loss": 0.6555,
+      "step": 14120
+    },
+    {
+      "epoch": 8.054150338439616,
+      "grad_norm": 2.346002360483435,
+      "learning_rate": 2.215656807678194e-06,
+      "loss": 0.6531,
+      "step": 14130
+    },
+    {
+      "epoch": 8.059850374064839,
+      "grad_norm": 2.3550136302279174,
+      "learning_rate": 2.203179771988796e-06,
+      "loss": 0.652,
+      "step": 14140
+    },
+    {
+      "epoch": 8.065550409690061,
+      "grad_norm": 2.298658799397616,
+      "learning_rate": 2.1907336159889712e-06,
+      "loss": 0.6641,
+      "step": 14150
+    },
+    {
+      "epoch": 8.071250445315282,
+      "grad_norm": 2.274958209439512,
+      "learning_rate": 2.1783183889723415e-06,
+      "loss": 0.6556,
+      "step": 14160
+    },
+    {
+      "epoch": 8.076950480940505,
+      "grad_norm": 2.354922282381701,
+      "learning_rate": 2.1659341401100165e-06,
+      "loss": 0.6625,
+      "step": 14170
+    },
+    {
+      "epoch": 8.082650516565728,
+      "grad_norm": 2.2464704977690815,
+      "learning_rate": 2.1535809184504255e-06,
+      "loss": 0.6495,
+      "step": 14180
+    },
+    {
+      "epoch": 8.088350552190951,
+      "grad_norm": 2.427423409659305,
+      "learning_rate": 2.141258772919108e-06,
+      "loss": 0.6603,
+      "step": 14190
+    },
+    {
+      "epoch": 8.094050587816174,
+      "grad_norm": 2.3827561888368,
+      "learning_rate": 2.128967752318527e-06,
+      "loss": 0.6551,
+      "step": 14200
+    },
+    {
+      "epoch": 8.099750623441397,
+      "grad_norm": 2.4054911753771884,
+      "learning_rate": 2.116707905327874e-06,
+      "loss": 0.6511,
+      "step": 14210
+    },
+    {
+      "epoch": 8.10545065906662,
+      "grad_norm": 2.2407461145856686,
+      "learning_rate": 2.1044792805028756e-06,
+      "loss": 0.6607,
+      "step": 14220
+    },
+    {
+      "epoch": 8.111150694691842,
+      "grad_norm": 2.378679457250059,
+      "learning_rate": 2.0922819262756e-06,
+      "loss": 0.6453,
+      "step": 14230
+    },
+    {
+      "epoch": 8.116850730317065,
+      "grad_norm": 2.3368426825434816,
+      "learning_rate": 2.080115890954266e-06,
+      "loss": 0.6517,
+      "step": 14240
+    },
+    {
+      "epoch": 8.122550765942288,
+      "grad_norm": 2.2661181727697532,
+      "learning_rate": 2.06798122272306e-06,
+      "loss": 0.6485,
+      "step": 14250
+    },
+    {
+      "epoch": 8.12825080156751,
+      "grad_norm": 2.2638492639891625,
+      "learning_rate": 2.0558779696419274e-06,
+      "loss": 0.6633,
+      "step": 14260
+    },
+    {
+      "epoch": 8.133950837192732,
+      "grad_norm": 2.370400165547832,
+      "learning_rate": 2.043806179646399e-06,
+      "loss": 0.665,
+      "step": 14270
+    },
+    {
+      "epoch": 8.139650872817954,
+      "grad_norm": 2.2408509531522856,
+      "learning_rate": 2.03176590054739e-06,
+      "loss": 0.6639,
+      "step": 14280
+    },
+    {
+      "epoch": 8.145350908443177,
+      "grad_norm": 2.1835237645071515,
+      "learning_rate": 2.019757180031017e-06,
+      "loss": 0.6618,
+      "step": 14290
+    },
+    {
+      "epoch": 8.1510509440684,
+      "grad_norm": 2.379295837422735,
+      "learning_rate": 2.0077800656584102e-06,
+      "loss": 0.6589,
+      "step": 14300
+    },
+    {
+      "epoch": 8.156750979693623,
+      "grad_norm": 2.3318576043677277,
+      "learning_rate": 1.9958346048655188e-06,
+      "loss": 0.6497,
+      "step": 14310
+    },
+    {
+      "epoch": 8.162451015318846,
+      "grad_norm": 2.3660969704194783,
+      "learning_rate": 1.9839208449629265e-06,
+      "loss": 0.6612,
+      "step": 14320
+    },
+    {
+      "epoch": 8.168151050944068,
+      "grad_norm": 2.2599900871895224,
+      "learning_rate": 1.9720388331356643e-06,
+      "loss": 0.6544,
+      "step": 14330
+    },
+    {
+      "epoch": 8.173851086569291,
+      "grad_norm": 2.2625667227411386,
+      "learning_rate": 1.960188616443025e-06,
+      "loss": 0.6577,
+      "step": 14340
+    },
+    {
+      "epoch": 8.179551122194514,
+      "grad_norm": 2.1949842609024497,
+      "learning_rate": 1.9483702418183725e-06,
+      "loss": 0.64,
+      "step": 14350
+    },
+    {
+      "epoch": 8.185251157819737,
+      "grad_norm": 2.2246089702453533,
+      "learning_rate": 1.9365837560689626e-06,
+      "loss": 0.6573,
+      "step": 14360
+    },
+    {
+      "epoch": 8.19095119344496,
+      "grad_norm": 2.355732389959134,
+      "learning_rate": 1.924829205875746e-06,
+      "loss": 0.6627,
+      "step": 14370
+    },
+    {
+      "epoch": 8.196651229070183,
+      "grad_norm": 2.290300180364329,
+      "learning_rate": 1.9131066377932017e-06,
+      "loss": 0.661,
+      "step": 14380
+    },
+    {
+      "epoch": 8.202351264695404,
+      "grad_norm": 2.3069453870221093,
+      "learning_rate": 1.901416098249136e-06,
+      "loss": 0.6467,
+      "step": 14390
+    },
+    {
+      "epoch": 8.208051300320626,
+      "grad_norm": 2.287176978146281,
+      "learning_rate": 1.8897576335445023e-06,
+      "loss": 0.6597,
+      "step": 14400
+    },
+    {
+      "epoch": 8.21375133594585,
+      "grad_norm": 2.3455594812602687,
+      "learning_rate": 1.8781312898532256e-06,
+      "loss": 0.6534,
+      "step": 14410
+    },
+    {
+      "epoch": 8.219451371571072,
+      "grad_norm": 2.2756337406814207,
+      "learning_rate": 1.8665371132220068e-06,
+      "loss": 0.6431,
+      "step": 14420
+    },
+    {
+      "epoch": 8.225151407196295,
+      "grad_norm": 2.249136593496119,
+      "learning_rate": 1.8549751495701584e-06,
+      "loss": 0.6552,
+      "step": 14430
+    },
+    {
+      "epoch": 8.230851442821518,
+      "grad_norm": 2.279678512876365,
+      "learning_rate": 1.8434454446894023e-06,
+      "loss": 0.6562,
+      "step": 14440
+    },
+    {
+      "epoch": 8.23655147844674,
+      "grad_norm": 2.449260711946442,
+      "learning_rate": 1.8319480442437043e-06,
+      "loss": 0.6656,
+      "step": 14450
+    },
+    {
+      "epoch": 8.242251514071963,
+      "grad_norm": 2.2927685740460997,
+      "learning_rate": 1.8204829937690748e-06,
+      "loss": 0.647,
+      "step": 14460
+    },
+    {
+      "epoch": 8.247951549697186,
+      "grad_norm": 2.2528090490901493,
+      "learning_rate": 1.8090503386734181e-06,
+      "loss": 0.6562,
+      "step": 14470
+    },
+    {
+      "epoch": 8.253651585322409,
+      "grad_norm": 2.338568674068967,
+      "learning_rate": 1.7976501242363242e-06,
+      "loss": 0.6632,
+      "step": 14480
+    },
+    {
+      "epoch": 8.259351620947632,
+      "grad_norm": 2.291192489078667,
+      "learning_rate": 1.7862823956089014e-06,
+      "loss": 0.6603,
+      "step": 14490
+    },
+    {
+      "epoch": 8.265051656572854,
+      "grad_norm": 2.2914805164289413,
+      "learning_rate": 1.774947197813598e-06,
+      "loss": 0.653,
+      "step": 14500
+    },
+    {
+      "epoch": 8.270751692198075,
+      "grad_norm": 2.286679795683027,
+      "learning_rate": 1.763644575744019e-06,
+      "loss": 0.6568,
+      "step": 14510
+    },
+    {
+      "epoch": 8.276451727823298,
+      "grad_norm": 2.2870797249748516,
+      "learning_rate": 1.7523745741647602e-06,
+      "loss": 0.6526,
+      "step": 14520
+    },
+    {
+      "epoch": 8.282151763448521,
+      "grad_norm": 2.3049667918026002,
+      "learning_rate": 1.7411372377112146e-06,
+      "loss": 0.6552,
+      "step": 14530
+    },
+    {
+      "epoch": 8.287851799073744,
+      "grad_norm": 2.360017011300003,
+      "learning_rate": 1.7299326108894033e-06,
+      "loss": 0.6571,
+      "step": 14540
+    },
+    {
+      "epoch": 8.293551834698967,
+      "grad_norm": 2.2514456956710065,
+      "learning_rate": 1.7187607380757998e-06,
+      "loss": 0.6634,
+      "step": 14550
+    },
+    {
+      "epoch": 8.29925187032419,
+      "grad_norm": 2.287113507409648,
+      "learning_rate": 1.7076216635171594e-06,
+      "loss": 0.6518,
+      "step": 14560
+    },
+    {
+      "epoch": 8.304951905949412,
+      "grad_norm": 2.2783805652974376,
+      "learning_rate": 1.6965154313303367e-06,
+      "loss": 0.6652,
+      "step": 14570
+    },
+    {
+      "epoch": 8.310651941574635,
+      "grad_norm": 2.222279328728929,
+      "learning_rate": 1.6854420855021026e-06,
+      "loss": 0.6661,
+      "step": 14580
+    },
+    {
+      "epoch": 8.316351977199858,
+      "grad_norm": 2.252278786396242,
+      "learning_rate": 1.6744016698889897e-06,
+      "loss": 0.6517,
+      "step": 14590
+    },
+    {
+      "epoch": 8.32205201282508,
+      "grad_norm": 2.345685753280762,
+      "learning_rate": 1.6633942282171056e-06,
+      "loss": 0.6551,
+      "step": 14600
+    },
+    {
+      "epoch": 8.327752048450304,
+      "grad_norm": 2.338706727686837,
+      "learning_rate": 1.6524198040819683e-06,
+      "loss": 0.6543,
+      "step": 14610
+    },
+    {
+      "epoch": 8.333452084075525,
+      "grad_norm": 2.384430342462321,
+      "learning_rate": 1.6414784409483197e-06,
+      "loss": 0.6569,
+      "step": 14620
+    },
+    {
+      "epoch": 8.339152119700747,
+      "grad_norm": 2.3701656962457673,
+      "learning_rate": 1.6305701821499686e-06,
+      "loss": 0.6535,
+      "step": 14630
+    },
+    {
+      "epoch": 8.34485215532597,
+      "grad_norm": 2.290965795306861,
+      "learning_rate": 1.6196950708896053e-06,
+      "loss": 0.6643,
+      "step": 14640
+    },
+    {
+      "epoch": 8.350552190951193,
+      "grad_norm": 2.3311296972099442,
+      "learning_rate": 1.6088531502386484e-06,
+      "loss": 0.6509,
+      "step": 14650
+    },
+    {
+      "epoch": 8.356252226576416,
+      "grad_norm": 2.335399650649135,
+      "learning_rate": 1.598044463137054e-06,
+      "loss": 0.6601,
+      "step": 14660
+    },
+    {
+      "epoch": 8.361952262201639,
+      "grad_norm": 2.455475427043487,
+      "learning_rate": 1.58726905239316e-06,
+      "loss": 0.657,
+      "step": 14670
+    },
+    {
+      "epoch": 8.367652297826861,
+      "grad_norm": 2.3282079850959754,
+      "learning_rate": 1.5765269606835054e-06,
+      "loss": 0.6679,
+      "step": 14680
+    },
+    {
+      "epoch": 8.373352333452084,
+      "grad_norm": 2.311135221618032,
+      "learning_rate": 1.5658182305526815e-06,
+      "loss": 0.6565,
+      "step": 14690
+    },
+    {
+      "epoch": 8.379052369077307,
+      "grad_norm": 2.2488393528174426,
+      "learning_rate": 1.5551429044131305e-06,
+      "loss": 0.6595,
+      "step": 14700
+    },
+    {
+      "epoch": 8.38475240470253,
+      "grad_norm": 2.267703234475785,
+      "learning_rate": 1.544501024545011e-06,
+      "loss": 0.6591,
+      "step": 14710
+    },
+    {
+      "epoch": 8.390452440327753,
+      "grad_norm": 2.2332246831041456,
+      "learning_rate": 1.5338926330960102e-06,
+      "loss": 0.6619,
+      "step": 14720
+    },
+    {
+      "epoch": 8.396152475952976,
+      "grad_norm": 2.365659533928728,
+      "learning_rate": 1.5233177720811798e-06,
+      "loss": 0.6543,
+      "step": 14730
+    },
+    {
+      "epoch": 8.401852511578197,
+      "grad_norm": 2.3371976515238586,
+      "learning_rate": 1.512776483382783e-06,
+      "loss": 0.6609,
+      "step": 14740
+    },
+    {
+      "epoch": 8.40755254720342,
+      "grad_norm": 2.374043422057661,
+      "learning_rate": 1.5022688087501092e-06,
+      "loss": 0.6643,
+      "step": 14750
+    },
+    {
+      "epoch": 8.413252582828642,
+      "grad_norm": 2.4076504196305817,
+      "learning_rate": 1.491794789799319e-06,
+      "loss": 0.6641,
+      "step": 14760
+    },
+    {
+      "epoch": 8.418952618453865,
+      "grad_norm": 2.2023957766134474,
+      "learning_rate": 1.4813544680132763e-06,
+      "loss": 0.6536,
+      "step": 14770
+    },
+    {
+      "epoch": 8.424652654079088,
+      "grad_norm": 2.202310327507326,
+      "learning_rate": 1.4709478847413948e-06,
+      "loss": 0.6467,
+      "step": 14780
+    },
+    {
+      "epoch": 8.43035268970431,
+      "grad_norm": 2.2051414307499106,
+      "learning_rate": 1.4605750811994557e-06,
+      "loss": 0.662,
+      "step": 14790
+    },
+    {
+      "epoch": 8.436052725329533,
+      "grad_norm": 2.434564818817124,
+      "learning_rate": 1.4502360984694563e-06,
+      "loss": 0.6532,
+      "step": 14800
+    },
+    {
+      "epoch": 8.441752760954756,
+      "grad_norm": 2.2946949223375737,
+      "learning_rate": 1.4399309774994475e-06,
+      "loss": 0.6584,
+      "step": 14810
+    },
+    {
+      "epoch": 8.447452796579979,
+      "grad_norm": 2.228248217670367,
+      "learning_rate": 1.4296597591033656e-06,
+      "loss": 0.6614,
+      "step": 14820
+    },
+    {
+      "epoch": 8.453152832205202,
+      "grad_norm": 2.2820771284655637,
+      "learning_rate": 1.4194224839608761e-06,
+      "loss": 0.6451,
+      "step": 14830
+    },
+    {
+      "epoch": 8.458852867830425,
+      "grad_norm": 2.393253761036466,
+      "learning_rate": 1.4092191926172106e-06,
+      "loss": 0.6543,
+      "step": 14840
+    },
+    {
+      "epoch": 8.464552903455647,
+      "grad_norm": 2.282212085947841,
+      "learning_rate": 1.3990499254830047e-06,
+      "loss": 0.6595,
+      "step": 14850
+    },
+    {
+      "epoch": 8.470252939080869,
+      "grad_norm": 2.365145023219437,
+      "learning_rate": 1.3889147228341394e-06,
+      "loss": 0.664,
+      "step": 14860
+    },
+    {
+      "epoch": 8.475952974706091,
+      "grad_norm": 2.189934931833651,
+      "learning_rate": 1.3788136248115869e-06,
+      "loss": 0.6629,
+      "step": 14870
+    },
+    {
+      "epoch": 8.481653010331314,
+      "grad_norm": 2.4011038278551173,
+      "learning_rate": 1.3687466714212393e-06,
+      "loss": 0.6577,
+      "step": 14880
+    },
+    {
+      "epoch": 8.487353045956537,
+      "grad_norm": 2.4106598851131475,
+      "learning_rate": 1.3587139025337615e-06,
+      "loss": 0.658,
+      "step": 14890
+    },
+    {
+      "epoch": 8.49305308158176,
+      "grad_norm": 2.3592819788417523,
+      "learning_rate": 1.348715357884427e-06,
+      "loss": 0.6579,
+      "step": 14900
+    },
+    {
+      "epoch": 8.498753117206983,
+      "grad_norm": 2.3305573077262505,
+      "learning_rate": 1.3387510770729595e-06,
+      "loss": 0.665,
+      "step": 14910
+    },
+    {
+      "epoch": 8.504453152832205,
+      "grad_norm": 2.3672792720086564,
+      "learning_rate": 1.3288210995633888e-06,
+      "loss": 0.6547,
+      "step": 14920
+    },
+    {
+      "epoch": 8.510153188457428,
+      "grad_norm": 2.280075759106149,
+      "learning_rate": 1.3189254646838766e-06,
+      "loss": 0.6652,
+      "step": 14930
+    },
+    {
+      "epoch": 8.515853224082651,
+      "grad_norm": 2.3369049806095297,
+      "learning_rate": 1.3090642116265695e-06,
+      "loss": 0.6568,
+      "step": 14940
+    },
+    {
+      "epoch": 8.521553259707874,
+      "grad_norm": 2.316623374340676,
+      "learning_rate": 1.2992373794474466e-06,
+      "loss": 0.6551,
+      "step": 14950
+    },
+    {
+      "epoch": 8.527253295333097,
+      "grad_norm": 2.303245948854903,
+      "learning_rate": 1.289445007066158e-06,
+      "loss": 0.6494,
+      "step": 14960
+    },
+    {
+      "epoch": 8.532953330958318,
+      "grad_norm": 2.359287289036829,
+      "learning_rate": 1.2796871332658756e-06,
+      "loss": 0.6558,
+      "step": 14970
+    },
+    {
+      "epoch": 8.53865336658354,
+      "grad_norm": 2.25661886393714,
+      "learning_rate": 1.26996379669314e-06,
+      "loss": 0.6602,
+      "step": 14980
+    },
+    {
+      "epoch": 8.544353402208763,
+      "grad_norm": 2.309228332149793,
+      "learning_rate": 1.260275035857701e-06,
+      "loss": 0.6609,
+      "step": 14990
+    },
+    {
+      "epoch": 8.550053437833986,
+      "grad_norm": 2.174364791553763,
+      "learning_rate": 1.2506208891323711e-06,
+      "loss": 0.6567,
+      "step": 15000
+    },
+    {
+      "epoch": 8.555753473459209,
+      "grad_norm": 2.43153962366673,
+      "learning_rate": 1.2410013947528766e-06,
+      "loss": 0.6589,
+      "step": 15010
+    },
+    {
+      "epoch": 8.561453509084432,
+      "grad_norm": 2.272819737875179,
+      "learning_rate": 1.2314165908176956e-06,
+      "loss": 0.6559,
+      "step": 15020
+    },
+    {
+      "epoch": 8.567153544709655,
+      "grad_norm": 2.179261588787464,
+      "learning_rate": 1.221866515287915e-06,
+      "loss": 0.6575,
+      "step": 15030
+    },
+    {
+      "epoch": 8.572853580334877,
+      "grad_norm": 2.221719946180341,
+      "learning_rate": 1.2123512059870756e-06,
+      "loss": 0.6535,
+      "step": 15040
+    },
+    {
+      "epoch": 8.5785536159601,
+      "grad_norm": 2.3048878158808206,
+      "learning_rate": 1.202870700601032e-06,
+      "loss": 0.6605,
+      "step": 15050
+    },
+    {
+      "epoch": 8.584253651585323,
+      "grad_norm": 2.365355709443227,
+      "learning_rate": 1.1934250366777899e-06,
+      "loss": 0.6649,
+      "step": 15060
+    },
+    {
+      "epoch": 8.589953687210546,
+      "grad_norm": 2.269976756998104,
+      "learning_rate": 1.1840142516273644e-06,
+      "loss": 0.6587,
+      "step": 15070
+    },
+    {
+      "epoch": 8.595653722835767,
+      "grad_norm": 2.4599719284501274,
+      "learning_rate": 1.1746383827216334e-06,
+      "loss": 0.6598,
+      "step": 15080
+    },
+    {
+      "epoch": 8.60135375846099,
+      "grad_norm": 2.270776618032317,
+      "learning_rate": 1.165297467094184e-06,
+      "loss": 0.6604,
+      "step": 15090
+    },
+    {
+      "epoch": 8.607053794086212,
+      "grad_norm": 2.5066430755058726,
+      "learning_rate": 1.1559915417401746e-06,
+      "loss": 0.6616,
+      "step": 15100
+    },
+    {
+      "epoch": 8.612753829711435,
+      "grad_norm": 2.2437923308781786,
+      "learning_rate": 1.146720643516177e-06,
+      "loss": 0.6556,
+      "step": 15110
+    },
+    {
+      "epoch": 8.618453865336658,
+      "grad_norm": 2.1955644450862613,
+      "learning_rate": 1.1374848091400403e-06,
+      "loss": 0.656,
+      "step": 15120
+    },
+    {
+      "epoch": 8.62415390096188,
+      "grad_norm": 2.2820502927208577,
+      "learning_rate": 1.1282840751907387e-06,
+      "loss": 0.6586,
+      "step": 15130
+    },
+    {
+      "epoch": 8.629853936587104,
+      "grad_norm": 2.2244689899988437,
+      "learning_rate": 1.1191184781082342e-06,
+      "loss": 0.6555,
+      "step": 15140
+    },
+    {
+      "epoch": 8.635553972212326,
+      "grad_norm": 2.2241919830845953,
+      "learning_rate": 1.1099880541933228e-06,
+      "loss": 0.6593,
+      "step": 15150
+    },
+    {
+      "epoch": 8.64125400783755,
+      "grad_norm": 2.2162843709409437,
+      "learning_rate": 1.100892839607497e-06,
+      "loss": 0.6502,
+      "step": 15160
+    },
+    {
+      "epoch": 8.646954043462772,
+      "grad_norm": 2.3805084624468624,
+      "learning_rate": 1.0918328703727998e-06,
+      "loss": 0.6616,
+      "step": 15170
+    },
+    {
+      "epoch": 8.652654079087995,
+      "grad_norm": 2.33050488774571,
+      "learning_rate": 1.0828081823716862e-06,
+      "loss": 0.6489,
+      "step": 15180
+    },
+    {
+      "epoch": 8.658354114713218,
+      "grad_norm": 2.396488801272341,
+      "learning_rate": 1.0738188113468762e-06,
+      "loss": 0.6563,
+      "step": 15190
+    },
+    {
+      "epoch": 8.66405415033844,
+      "grad_norm": 2.1970815029045685,
+      "learning_rate": 1.0648647929012157e-06,
+      "loss": 0.6626,
+      "step": 15200
+    },
+    {
+      "epoch": 8.669754185963662,
+      "grad_norm": 2.356974951580113,
+      "learning_rate": 1.0559461624975343e-06,
+      "loss": 0.6599,
+      "step": 15210
+    },
+    {
+      "epoch": 8.675454221588884,
+      "grad_norm": 2.231841861511598,
+      "learning_rate": 1.0470629554585043e-06,
+      "loss": 0.656,
+      "step": 15220
+    },
+    {
+      "epoch": 8.681154257214107,
+      "grad_norm": 2.3405687246859013,
+      "learning_rate": 1.0382152069665063e-06,
+      "loss": 0.6534,
+      "step": 15230
+    },
+    {
+      "epoch": 8.68685429283933,
+      "grad_norm": 2.3057634631991344,
+      "learning_rate": 1.0294029520634806e-06,
+      "loss": 0.6569,
+      "step": 15240
+    },
+    {
+      "epoch": 8.692554328464553,
+      "grad_norm": 2.3591092566045826,
+      "learning_rate": 1.020626225650797e-06,
+      "loss": 0.6516,
+      "step": 15250
+    },
+    {
+      "epoch": 8.698254364089776,
+      "grad_norm": 2.2774547172739794,
+      "learning_rate": 1.0118850624891097e-06,
+      "loss": 0.6611,
+      "step": 15260
+    },
+    {
+      "epoch": 8.703954399714998,
+      "grad_norm": 2.3044521216927687,
+      "learning_rate": 1.0031794971982278e-06,
+      "loss": 0.657,
+      "step": 15270
+    },
+    {
+      "epoch": 8.709654435340221,
+      "grad_norm": 2.328406536786459,
+      "learning_rate": 9.945095642569692e-07,
+      "loss": 0.6549,
+      "step": 15280
+    },
+    {
+      "epoch": 8.715354470965444,
+      "grad_norm": 2.258894872917727,
+      "learning_rate": 9.858752980030295e-07,
+      "loss": 0.658,
+      "step": 15290
+    },
+    {
+      "epoch": 8.721054506590667,
+      "grad_norm": 2.221407186169279,
+      "learning_rate": 9.772767326328435e-07,
+      "loss": 0.6627,
+      "step": 15300
+    },
+    {
+      "epoch": 8.72675454221589,
+      "grad_norm": 2.371525100568372,
+      "learning_rate": 9.687139022014502e-07,
+      "loss": 0.6614,
+      "step": 15310
+    },
+    {
+      "epoch": 8.73245457784111,
+      "grad_norm": 2.3645791912994274,
+      "learning_rate": 9.601868406223647e-07,
+      "loss": 0.6589,
+      "step": 15320
+    },
+    {
+      "epoch": 8.738154613466333,
+      "grad_norm": 2.253478405026816,
+      "learning_rate": 9.516955816674311e-07,
+      "loss": 0.6512,
+      "step": 15330
+    },
+    {
+      "epoch": 8.743854649091556,
+      "grad_norm": 2.253371950211027,
+      "learning_rate": 9.432401589666984e-07,
+      "loss": 0.6547,
+      "step": 15340
+    },
+    {
+      "epoch": 8.749554684716779,
+      "grad_norm": 2.2348557907031155,
+      "learning_rate": 9.348206060082799e-07,
+      "loss": 0.6517,
+      "step": 15350
+    },
+    {
+      "epoch": 8.755254720342002,
+      "grad_norm": 2.2444613696453306,
+      "learning_rate": 9.264369561382336e-07,
+      "loss": 0.6548,
+      "step": 15360
+    },
+    {
+      "epoch": 8.760954755967225,
+      "grad_norm": 2.337184508845707,
+      "learning_rate": 9.180892425604149e-07,
+      "loss": 0.6598,
+      "step": 15370
+    },
+    {
+      "epoch": 8.766654791592448,
+      "grad_norm": 2.280658542347082,
+      "learning_rate": 9.097774983363527e-07,
+      "loss": 0.6623,
+      "step": 15380
+    },
+    {
+      "epoch": 8.77235482721767,
+      "grad_norm": 2.3828437996657983,
+      "learning_rate": 9.01501756385117e-07,
+      "loss": 0.6563,
+      "step": 15390
+    },
+    {
+      "epoch": 8.778054862842893,
+      "grad_norm": 2.349125904984566,
+      "learning_rate": 8.932620494831945e-07,
+      "loss": 0.6652,
+      "step": 15400
+    },
+    {
+      "epoch": 8.783754898468116,
+      "grad_norm": 2.347074312107561,
+      "learning_rate": 8.850584102643478e-07,
+      "loss": 0.6536,
+      "step": 15410
+    },
+    {
+      "epoch": 8.789454934093339,
+      "grad_norm": 2.4067679774208464,
+      "learning_rate": 8.768908712194913e-07,
+      "loss": 0.6548,
+      "step": 15420
+    },
+    {
+      "epoch": 8.79515496971856,
+      "grad_norm": 2.3226632236683464,
+      "learning_rate": 8.687594646965669e-07,
+      "loss": 0.6535,
+      "step": 15430
+    },
+    {
+      "epoch": 8.800855005343783,
+      "grad_norm": 2.315292189470015,
+      "learning_rate": 8.606642229004059e-07,
+      "loss": 0.6576,
+      "step": 15440
+    },
+    {
+      "epoch": 8.806555040969005,
+      "grad_norm": 2.3198254368211724,
+      "learning_rate": 8.526051778926181e-07,
+      "loss": 0.6542,
+      "step": 15450
+    },
+    {
+      "epoch": 8.812255076594228,
+      "grad_norm": 2.2794027212990184,
+      "learning_rate": 8.445823615914405e-07,
+      "loss": 0.6521,
+      "step": 15460
+    },
+    {
+      "epoch": 8.817955112219451,
+      "grad_norm": 2.3108872183589018,
+      "learning_rate": 8.365958057716339e-07,
+      "loss": 0.6491,
+      "step": 15470
+    },
+    {
+      "epoch": 8.823655147844674,
+      "grad_norm": 2.3739240284864165,
+      "learning_rate": 8.286455420643424e-07,
+      "loss": 0.6709,
+      "step": 15480
+    },
+    {
+      "epoch": 8.829355183469897,
+      "grad_norm": 2.2853027691895034,
+      "learning_rate": 8.207316019569811e-07,
+      "loss": 0.6671,
+      "step": 15490
+    },
+    {
+      "epoch": 8.83505521909512,
+      "grad_norm": 2.255831984475305,
+      "learning_rate": 8.128540167930942e-07,
+      "loss": 0.6555,
+      "step": 15500
+    },
+    {
+      "epoch": 8.840755254720342,
+      "grad_norm": 2.3936588178116907,
+      "learning_rate": 8.050128177722482e-07,
+      "loss": 0.6479,
+      "step": 15510
+    },
+    {
+      "epoch": 8.846455290345565,
+      "grad_norm": 2.4135299551433977,
+      "learning_rate": 7.972080359498946e-07,
+      "loss": 0.6602,
+      "step": 15520
+    },
+    {
+      "epoch": 8.852155325970788,
+      "grad_norm": 2.358391743128982,
+      "learning_rate": 7.894397022372535e-07,
+      "loss": 0.6522,
+      "step": 15530
+    },
+    {
+      "epoch": 8.85785536159601,
+      "grad_norm": 2.363252827528486,
+      "learning_rate": 7.817078474011974e-07,
+      "loss": 0.6524,
+      "step": 15540
+    },
+    {
+      "epoch": 8.863555397221234,
+      "grad_norm": 2.4081698427293876,
+      "learning_rate": 7.740125020641143e-07,
+      "loss": 0.6525,
+      "step": 15550
+    },
+    {
+      "epoch": 8.869255432846455,
+      "grad_norm": 2.3027191036174477,
+      "learning_rate": 7.663536967037977e-07,
+      "loss": 0.6563,
+      "step": 15560
+    },
+    {
+      "epoch": 8.874955468471677,
+      "grad_norm": 2.2800033610965613,
+      "learning_rate": 7.587314616533226e-07,
+      "loss": 0.6604,
+      "step": 15570
+    },
+    {
+      "epoch": 8.8806555040969,
+      "grad_norm": 2.2376325505940806,
+      "learning_rate": 7.511458271009254e-07,
+      "loss": 0.6509,
+      "step": 15580
+    },
+    {
+      "epoch": 8.886355539722123,
+      "grad_norm": 2.341968038078447,
+      "learning_rate": 7.435968230898838e-07,
+      "loss": 0.65,
+      "step": 15590
+    },
+    {
+      "epoch": 8.892055575347346,
+      "grad_norm": 2.2842053769053385,
+      "learning_rate": 7.360844795184007e-07,
+      "loss": 0.6594,
+      "step": 15600
+    },
+    {
+      "epoch": 8.897755610972569,
+      "grad_norm": 2.340094367909099,
+      "learning_rate": 7.286088261394797e-07,
+      "loss": 0.6592,
+      "step": 15610
+    },
+    {
+      "epoch": 8.903455646597791,
+      "grad_norm": 2.2678395556771136,
+      "learning_rate": 7.211698925608134e-07,
+      "loss": 0.6699,
+      "step": 15620
+    },
+    {
+      "epoch": 8.909155682223014,
+      "grad_norm": 2.3621110355578336,
+      "learning_rate": 7.137677082446659e-07,
+      "loss": 0.6596,
+      "step": 15630
+    },
+    {
+      "epoch": 8.914855717848237,
+      "grad_norm": 2.272216215850836,
+      "learning_rate": 7.064023025077516e-07,
+      "loss": 0.6435,
+      "step": 15640
+    },
+    {
+      "epoch": 8.92055575347346,
+      "grad_norm": 2.4488683011179617,
+      "learning_rate": 6.990737045211204e-07,
+      "loss": 0.6607,
+      "step": 15650
+    },
+    {
+      "epoch": 8.926255789098683,
+      "grad_norm": 2.308404823149232,
+      "learning_rate": 6.917819433100436e-07,
+      "loss": 0.6544,
+      "step": 15660
+    },
+    {
+      "epoch": 8.931955824723904,
+      "grad_norm": 2.421245275018334,
+      "learning_rate": 6.845270477539034e-07,
+      "loss": 0.6585,
+      "step": 15670
+    },
+    {
+      "epoch": 8.937655860349127,
+      "grad_norm": 2.3489968102572534,
+      "learning_rate": 6.773090465860677e-07,
+      "loss": 0.654,
+      "step": 15680
+    },
+    {
+      "epoch": 8.94335589597435,
+      "grad_norm": 2.2797356549942593,
+      "learning_rate": 6.701279683937844e-07,
+      "loss": 0.6533,
+      "step": 15690
+    },
+    {
+      "epoch": 8.949055931599572,
+      "grad_norm": 2.3453879162469633,
+      "learning_rate": 6.629838416180679e-07,
+      "loss": 0.6567,
+      "step": 15700
+    },
+    {
+      "epoch": 8.954755967224795,
+      "grad_norm": 2.3713385483483496,
+      "learning_rate": 6.558766945535822e-07,
+      "loss": 0.6597,
+      "step": 15710
+    },
+    {
+      "epoch": 8.960456002850018,
+      "grad_norm": 2.2925429331052034,
+      "learning_rate": 6.488065553485334e-07,
+      "loss": 0.6563,
+      "step": 15720
+    },
+    {
+      "epoch": 8.96615603847524,
+      "grad_norm": 2.2600327088475236,
+      "learning_rate": 6.417734520045537e-07,
+      "loss": 0.6586,
+      "step": 15730
+    },
+    {
+      "epoch": 8.971856074100463,
+      "grad_norm": 2.300420354697253,
+      "learning_rate": 6.34777412376596e-07,
+      "loss": 0.6543,
+      "step": 15740
+    },
+    {
+      "epoch": 8.977556109725686,
+      "grad_norm": 2.30692835763649,
+      "learning_rate": 6.278184641728169e-07,
+      "loss": 0.6444,
+      "step": 15750
+    },
+    {
+      "epoch": 8.983256145350909,
+      "grad_norm": 2.359181343602721,
+      "learning_rate": 6.208966349544754e-07,
+      "loss": 0.6622,
+      "step": 15760
+    },
+    {
+      "epoch": 8.988956180976132,
+      "grad_norm": 2.3613948749177043,
+      "learning_rate": 6.140119521358146e-07,
+      "loss": 0.646,
+      "step": 15770
+    },
+    {
+      "epoch": 8.994656216601353,
+      "grad_norm": 2.407625610224588,
+      "learning_rate": 6.071644429839585e-07,
+      "loss": 0.6581,
+      "step": 15780
+    },
+    {
+      "epoch": 9.000356252226576,
+      "grad_norm": 2.393181765255249,
+      "learning_rate": 6.003541346188036e-07,
+      "loss": 0.6535,
+      "step": 15790
+    },
+    {
+      "epoch": 9.006056287851798,
+      "grad_norm": 2.283188570163856,
+      "learning_rate": 5.935810540129128e-07,
+      "loss": 0.6461,
+      "step": 15800
+    },
+    {
+      "epoch": 9.011756323477021,
+      "grad_norm": 2.264655867396922,
+      "learning_rate": 5.868452279914039e-07,
+      "loss": 0.6532,
+      "step": 15810
+    },
+    {
+      "epoch": 9.017456359102244,
+      "grad_norm": 2.3281868349695345,
+      "learning_rate": 5.801466832318458e-07,
+      "loss": 0.6554,
+      "step": 15820
+    },
+    {
+      "epoch": 9.023156394727467,
+      "grad_norm": 2.375263134251282,
+      "learning_rate": 5.734854462641548e-07,
+      "loss": 0.649,
+      "step": 15830
+    },
+    {
+      "epoch": 9.02885643035269,
+      "grad_norm": 2.290766353741812,
+      "learning_rate": 5.66861543470486e-07,
+      "loss": 0.652,
+      "step": 15840
+    },
+    {
+      "epoch": 9.034556465977913,
+      "grad_norm": 2.3498529360472316,
+      "learning_rate": 5.602750010851332e-07,
+      "loss": 0.6498,
+      "step": 15850
+    },
+    {
+      "epoch": 9.040256501603135,
+      "grad_norm": 2.2284167118477938,
+      "learning_rate": 5.537258451944206e-07,
+      "loss": 0.6462,
+      "step": 15860
+    },
+    {
+      "epoch": 9.045956537228358,
+      "grad_norm": 2.2655672578940695,
+      "learning_rate": 5.472141017366029e-07,
+      "loss": 0.6529,
+      "step": 15870
+    },
+    {
+      "epoch": 9.051656572853581,
+      "grad_norm": 2.249704353631801,
+      "learning_rate": 5.407397965017569e-07,
+      "loss": 0.6579,
+      "step": 15880
+    },
+    {
+      "epoch": 9.057356608478804,
+      "grad_norm": 2.299461894717784,
+      "learning_rate": 5.343029551316892e-07,
+      "loss": 0.6475,
+      "step": 15890
+    },
+    {
+      "epoch": 9.063056644104025,
+      "grad_norm": 2.2344281001437616,
+      "learning_rate": 5.27903603119827e-07,
+      "loss": 0.6593,
+      "step": 15900
+    },
+    {
+      "epoch": 9.068756679729248,
+      "grad_norm": 2.326794202315961,
+      "learning_rate": 5.215417658111166e-07,
+      "loss": 0.6513,
+      "step": 15910
+    },
+    {
+      "epoch": 9.07445671535447,
+      "grad_norm": 2.2784333346869,
+      "learning_rate": 5.152174684019285e-07,
+      "loss": 0.6504,
+      "step": 15920
+    },
+    {
+      "epoch": 9.080156750979693,
+      "grad_norm": 2.3541870558342093,
+      "learning_rate": 5.089307359399498e-07,
+      "loss": 0.6496,
+      "step": 15930
+    },
+    {
+      "epoch": 9.085856786604916,
+      "grad_norm": 2.3054372491712023,
+      "learning_rate": 5.02681593324098e-07,
+      "loss": 0.6581,
+      "step": 15940
+    },
+    {
+      "epoch": 9.091556822230139,
+      "grad_norm": 2.3799751500353894,
+      "learning_rate": 4.964700653044086e-07,
+      "loss": 0.65,
+      "step": 15950
+    },
+    {
+      "epoch": 9.097256857855362,
+      "grad_norm": 2.215324072504103,
+      "learning_rate": 4.902961764819414e-07,
+      "loss": 0.651,
+      "step": 15960
+    },
+    {
+      "epoch": 9.102956893480584,
+      "grad_norm": 2.2845100469483928,
+      "learning_rate": 4.84159951308687e-07,
+      "loss": 0.6549,
+      "step": 15970
+    },
+    {
+      "epoch": 9.108656929105807,
+      "grad_norm": 2.3153851871079962,
+      "learning_rate": 4.780614140874685e-07,
+      "loss": 0.6602,
+      "step": 15980
+    },
+    {
+      "epoch": 9.11435696473103,
+      "grad_norm": 2.3349945480874483,
+      "learning_rate": 4.720005889718393e-07,
+      "loss": 0.654,
+      "step": 15990
+    },
+    {
+      "epoch": 9.120057000356253,
+      "grad_norm": 2.2495628939704653,
+      "learning_rate": 4.6597749996599716e-07,
+      "loss": 0.6563,
+      "step": 16000
+    },
+    {
+      "epoch": 9.125757035981476,
+      "grad_norm": 2.2943503815780217,
+      "learning_rate": 4.5999217092468127e-07,
+      "loss": 0.6552,
+      "step": 16010
+    },
+    {
+      "epoch": 9.131457071606697,
+      "grad_norm": 2.3123395747706774,
+      "learning_rate": 4.540446255530806e-07,
+      "loss": 0.6536,
+      "step": 16020
+    },
+    {
+      "epoch": 9.13715710723192,
+      "grad_norm": 2.356164507527656,
+      "learning_rate": 4.481348874067426e-07,
+      "loss": 0.6496,
+      "step": 16030
+    },
+    {
+      "epoch": 9.142857142857142,
+      "grad_norm": 2.315756183759805,
+      "learning_rate": 4.422629798914757e-07,
+      "loss": 0.6557,
+      "step": 16040
+    },
+    {
+      "epoch": 9.148557178482365,
+      "grad_norm": 2.412495445910888,
+      "learning_rate": 4.3642892626325595e-07,
+      "loss": 0.6613,
+      "step": 16050
+    },
+    {
+      "epoch": 9.154257214107588,
+      "grad_norm": 2.277413770697624,
+      "learning_rate": 4.3063274962813926e-07,
+      "loss": 0.6552,
+      "step": 16060
+    },
+    {
+      "epoch": 9.15995724973281,
+      "grad_norm": 2.3240365339718316,
+      "learning_rate": 4.2487447294217056e-07,
+      "loss": 0.6644,
+      "step": 16070
+    },
+    {
+      "epoch": 9.165657285358034,
+      "grad_norm": 2.1543175867284448,
+      "learning_rate": 4.1915411901128577e-07,
+      "loss": 0.6597,
+      "step": 16080
+    },
+    {
+      "epoch": 9.171357320983256,
+      "grad_norm": 2.3289956989255236,
+      "learning_rate": 4.1347171049122894e-07,
+      "loss": 0.6537,
+      "step": 16090
+    },
+    {
+      "epoch": 9.17705735660848,
+      "grad_norm": 2.441818257544798,
+      "learning_rate": 4.0782726988745634e-07,
+      "loss": 0.6544,
+      "step": 16100
+    },
+    {
+      "epoch": 9.182757392233702,
+      "grad_norm": 2.1992411470753814,
+      "learning_rate": 4.0222081955505367e-07,
+      "loss": 0.6612,
+      "step": 16110
+    },
+    {
+      "epoch": 9.188457427858925,
+      "grad_norm": 2.260053918486078,
+      "learning_rate": 3.966523816986434e-07,
+      "loss": 0.6628,
+      "step": 16120
+    },
+    {
+      "epoch": 9.194157463484148,
+      "grad_norm": 2.3191018874769624,
+      "learning_rate": 3.911219783722953e-07,
+      "loss": 0.6578,
+      "step": 16130
+    },
+    {
+      "epoch": 9.199857499109369,
+      "grad_norm": 2.190947251083796,
+      "learning_rate": 3.85629631479445e-07,
+      "loss": 0.6494,
+      "step": 16140
+    },
+    {
+      "epoch": 9.205557534734591,
+      "grad_norm": 2.3462795173662854,
+      "learning_rate": 3.801753627728011e-07,
+      "loss": 0.6538,
+      "step": 16150
+    },
+    {
+      "epoch": 9.211257570359814,
+      "grad_norm": 2.3600995397447497,
+      "learning_rate": 3.7475919385426384e-07,
+      "loss": 0.6484,
+      "step": 16160
+    },
+    {
+      "epoch": 9.216957605985037,
+      "grad_norm": 2.319158550292185,
+      "learning_rate": 3.6938114617483646e-07,
+      "loss": 0.655,
+      "step": 16170
+    },
+    {
+      "epoch": 9.22265764161026,
+      "grad_norm": 2.3069140811983413,
+      "learning_rate": 3.6404124103453954e-07,
+      "loss": 0.6551,
+      "step": 16180
+    },
+    {
+      "epoch": 9.228357677235483,
+      "grad_norm": 2.335764090264895,
+      "learning_rate": 3.587394995823301e-07,
+      "loss": 0.6588,
+      "step": 16190
+    },
+    {
+      "epoch": 9.234057712860706,
+      "grad_norm": 2.3001727472033227,
+      "learning_rate": 3.5347594281601837e-07,
+      "loss": 0.6569,
+      "step": 16200
+    },
+    {
+      "epoch": 9.239757748485928,
+      "grad_norm": 2.3662929871693272,
+      "learning_rate": 3.482505915821766e-07,
+      "loss": 0.6579,
+      "step": 16210
+    },
+    {
+      "epoch": 9.245457784111151,
+      "grad_norm": 2.2676447187387345,
+      "learning_rate": 3.430634665760668e-07,
+      "loss": 0.65,
+      "step": 16220
+    },
+    {
+      "epoch": 9.251157819736374,
+      "grad_norm": 2.2594944840854008,
+      "learning_rate": 3.379145883415502e-07,
+      "loss": 0.6534,
+      "step": 16230
+    },
+    {
+      "epoch": 9.256857855361597,
+      "grad_norm": 2.2559662121895627,
+      "learning_rate": 3.328039772710123e-07,
+      "loss": 0.6572,
+      "step": 16240
+    },
+    {
+      "epoch": 9.262557890986818,
+      "grad_norm": 2.2928810036817895,
+      "learning_rate": 3.277316536052821e-07,
+      "loss": 0.6572,
+      "step": 16250
+    },
+    {
+      "epoch": 9.26825792661204,
+      "grad_norm": 2.298962085856491,
+      "learning_rate": 3.2269763743354445e-07,
+      "loss": 0.6466,
+      "step": 16260
+    },
+    {
+      "epoch": 9.273957962237263,
+      "grad_norm": 2.3393731734236938,
+      "learning_rate": 3.1770194869326864e-07,
+      "loss": 0.6632,
+      "step": 16270
+    },
+    {
+      "epoch": 9.279657997862486,
+      "grad_norm": 2.389534898394802,
+      "learning_rate": 3.1274460717012346e-07,
+      "loss": 0.6521,
+      "step": 16280
+    },
+    {
+      "epoch": 9.285358033487709,
+      "grad_norm": 2.319294326677638,
+      "learning_rate": 3.0782563249790567e-07,
+      "loss": 0.6517,
+      "step": 16290
+    },
+    {
+      "epoch": 9.291058069112932,
+      "grad_norm": 2.4153981864437184,
+      "learning_rate": 3.0294504415845585e-07,
+      "loss": 0.6584,
+      "step": 16300
+    },
+    {
+      "epoch": 9.296758104738155,
+      "grad_norm": 2.3825627392396673,
+      "learning_rate": 2.98102861481584e-07,
+      "loss": 0.6456,
+      "step": 16310
+    },
+    {
+      "epoch": 9.302458140363377,
+      "grad_norm": 2.3879172528078825,
+      "learning_rate": 2.932991036449917e-07,
+      "loss": 0.6613,
+      "step": 16320
+    },
+    {
+      "epoch": 9.3081581759886,
+      "grad_norm": 2.346463251210415,
+      "learning_rate": 2.8853378967419686e-07,
+      "loss": 0.655,
+      "step": 16330
+    },
+    {
+      "epoch": 9.313858211613823,
+      "grad_norm": 2.2927763648573016,
+      "learning_rate": 2.8380693844246355e-07,
+      "loss": 0.6502,
+      "step": 16340
+    },
+    {
+      "epoch": 9.319558247239046,
+      "grad_norm": 2.3550355480798304,
+      "learning_rate": 2.7911856867071427e-07,
+      "loss": 0.6409,
+      "step": 16350
+    },
+    {
+      "epoch": 9.325258282864269,
+      "grad_norm": 2.234142791432648,
+      "learning_rate": 2.744686989274692e-07,
+      "loss": 0.6592,
+      "step": 16360
+    },
+    {
+      "epoch": 9.33095831848949,
+      "grad_norm": 2.2655908560680627,
+      "learning_rate": 2.698573476287658e-07,
+      "loss": 0.6581,
+      "step": 16370
+    },
+    {
+      "epoch": 9.336658354114713,
+      "grad_norm": 2.304446959403882,
+      "learning_rate": 2.652845330380882e-07,
+      "loss": 0.6515,
+      "step": 16380
+    },
+    {
+      "epoch": 9.342358389739935,
+      "grad_norm": 2.2890950540257777,
+      "learning_rate": 2.6075027326629253e-07,
+      "loss": 0.639,
+      "step": 16390
+    },
+    {
+      "epoch": 9.348058425365158,
+      "grad_norm": 2.2796528945861656,
+      "learning_rate": 2.562545862715382e-07,
+      "loss": 0.6417,
+      "step": 16400
+    },
+    {
+      "epoch": 9.353758460990381,
+      "grad_norm": 2.456419871618796,
+      "learning_rate": 2.517974898592124e-07,
+      "loss": 0.6574,
+      "step": 16410
+    },
+    {
+      "epoch": 9.359458496615604,
+      "grad_norm": 2.297770693420203,
+      "learning_rate": 2.4737900168186667e-07,
+      "loss": 0.6549,
+      "step": 16420
+    },
+    {
+      "epoch": 9.365158532240827,
+      "grad_norm": 2.363808835548694,
+      "learning_rate": 2.429991392391395e-07,
+      "loss": 0.6415,
+      "step": 16430
+    },
+    {
+      "epoch": 9.37085856786605,
+      "grad_norm": 2.2713206751436865,
+      "learning_rate": 2.386579198776917e-07,
+      "loss": 0.652,
+      "step": 16440
+    },
+    {
+      "epoch": 9.376558603491272,
+      "grad_norm": 2.303657933588446,
+      "learning_rate": 2.343553607911353e-07,
+      "loss": 0.6512,
+      "step": 16450
+    },
+    {
+      "epoch": 9.382258639116495,
+      "grad_norm": 2.2112371877259402,
+      "learning_rate": 2.300914790199682e-07,
+      "loss": 0.6601,
+      "step": 16460
+    },
+    {
+      "epoch": 9.387958674741718,
+      "grad_norm": 2.3239289332415716,
+      "learning_rate": 2.2586629145150195e-07,
+      "loss": 0.6557,
+      "step": 16470
+    },
+    {
+      "epoch": 9.393658710366939,
+      "grad_norm": 2.262292683671461,
+      "learning_rate": 2.2167981481980073e-07,
+      "loss": 0.6476,
+      "step": 16480
+    },
+    {
+      "epoch": 9.399358745992162,
+      "grad_norm": 2.347838181209718,
+      "learning_rate": 2.1753206570561015e-07,
+      "loss": 0.6503,
+      "step": 16490
+    },
+    {
+      "epoch": 9.405058781617385,
+      "grad_norm": 2.285652751528127,
+      "learning_rate": 2.1342306053629414e-07,
+      "loss": 0.6379,
+      "step": 16500
+    },
+    {
+      "epoch": 9.410758817242607,
+      "grad_norm": 2.232250790375402,
+      "learning_rate": 2.0935281558577048e-07,
+      "loss": 0.659,
+      "step": 16510
+    },
+    {
+      "epoch": 9.41645885286783,
+      "grad_norm": 2.3688051039592795,
+      "learning_rate": 2.0532134697444417e-07,
+      "loss": 0.6543,
+      "step": 16520
+    },
+    {
+      "epoch": 9.422158888493053,
+      "grad_norm": 2.2868901408436613,
+      "learning_rate": 2.0132867066914418e-07,
+      "loss": 0.6632,
+      "step": 16530
+    },
+    {
+      "epoch": 9.427858924118276,
+      "grad_norm": 2.3010337804794125,
+      "learning_rate": 1.9737480248306128e-07,
+      "loss": 0.657,
+      "step": 16540
+    },
+    {
+      "epoch": 9.433558959743499,
+      "grad_norm": 2.268005007344826,
+      "learning_rate": 1.9345975807568473e-07,
+      "loss": 0.6472,
+      "step": 16550
+    },
+    {
+      "epoch": 9.439258995368721,
+      "grad_norm": 2.259506030712857,
+      "learning_rate": 1.8958355295274012e-07,
+      "loss": 0.6545,
+      "step": 16560
+    },
+    {
+      "epoch": 9.444959030993944,
+      "grad_norm": 2.3732707468529233,
+      "learning_rate": 1.857462024661294e-07,
+      "loss": 0.6336,
+      "step": 16570
+    },
+    {
+      "epoch": 9.450659066619167,
+      "grad_norm": 2.2545076985099515,
+      "learning_rate": 1.8194772181386655e-07,
+      "loss": 0.6443,
+      "step": 16580
+    },
+    {
+      "epoch": 9.45635910224439,
+      "grad_norm": 2.2871428515450685,
+      "learning_rate": 1.781881260400209e-07,
+      "loss": 0.6504,
+      "step": 16590
+    },
+    {
+      "epoch": 9.46205913786961,
+      "grad_norm": 2.2933586803868686,
+      "learning_rate": 1.7446743003465606e-07,
+      "loss": 0.6561,
+      "step": 16600
+    },
+    {
+      "epoch": 9.467759173494834,
+      "grad_norm": 2.3486885562113704,
+      "learning_rate": 1.707856485337722e-07,
+      "loss": 0.6542,
+      "step": 16610
+    },
+    {
+      "epoch": 9.473459209120056,
+      "grad_norm": 2.3141845660170546,
+      "learning_rate": 1.6714279611924512e-07,
+      "loss": 0.6548,
+      "step": 16620
+    },
+    {
+      "epoch": 9.47915924474528,
+      "grad_norm": 2.3378979544356175,
+      "learning_rate": 1.6353888721877154e-07,
+      "loss": 0.6549,
+      "step": 16630
+    },
+    {
+      "epoch": 9.484859280370502,
+      "grad_norm": 2.3555369385409595,
+      "learning_rate": 1.5997393610580837e-07,
+      "loss": 0.6508,
+      "step": 16640
+    },
+    {
+      "epoch": 9.490559315995725,
+      "grad_norm": 2.3050119556793534,
+      "learning_rate": 1.564479568995203e-07,
+      "loss": 0.6548,
+      "step": 16650
+    },
+    {
+      "epoch": 9.496259351620948,
+      "grad_norm": 2.2004369929766594,
+      "learning_rate": 1.5296096356472223e-07,
+      "loss": 0.6529,
+      "step": 16660
+    },
+    {
+      "epoch": 9.50195938724617,
+      "grad_norm": 2.337622517571759,
+      "learning_rate": 1.495129699118214e-07,
+      "loss": 0.6486,
+      "step": 16670
+    },
+    {
+      "epoch": 9.507659422871393,
+      "grad_norm": 2.2515692465257438,
+      "learning_rate": 1.461039895967653e-07,
+      "loss": 0.6591,
+      "step": 16680
+    },
+    {
+      "epoch": 9.513359458496616,
+      "grad_norm": 2.257958729425762,
+      "learning_rate": 1.4273403612099057e-07,
+      "loss": 0.6589,
+      "step": 16690
+    },
+    {
+      "epoch": 9.519059494121839,
+      "grad_norm": 2.2551941842725474,
+      "learning_rate": 1.3940312283136192e-07,
+      "loss": 0.6573,
+      "step": 16700
+    },
+    {
+      "epoch": 9.524759529747062,
+      "grad_norm": 2.2839034425055753,
+      "learning_rate": 1.3611126292012444e-07,
+      "loss": 0.6528,
+      "step": 16710
+    },
+    {
+      "epoch": 9.530459565372283,
+      "grad_norm": 2.2513087425795963,
+      "learning_rate": 1.3285846942485247e-07,
+      "loss": 0.642,
+      "step": 16720
+    },
+    {
+      "epoch": 9.536159600997506,
+      "grad_norm": 2.373914409653791,
+      "learning_rate": 1.2964475522839304e-07,
+      "loss": 0.6513,
+      "step": 16730
+    },
+    {
+      "epoch": 9.541859636622728,
+      "grad_norm": 2.3653513175384746,
+      "learning_rate": 1.2647013305882138e-07,
+      "loss": 0.6521,
+      "step": 16740
+    },
+    {
+      "epoch": 9.547559672247951,
+      "grad_norm": 2.346784431083733,
+      "learning_rate": 1.2333461548938109e-07,
+      "loss": 0.6556,
+      "step": 16750
+    },
+    {
+      "epoch": 9.553259707873174,
+      "grad_norm": 2.2471608192504795,
+      "learning_rate": 1.2023821493844623e-07,
+      "loss": 0.6442,
+      "step": 16760
+    },
+    {
+      "epoch": 9.558959743498397,
+      "grad_norm": 2.41300169259645,
+      "learning_rate": 1.1718094366946264e-07,
+      "loss": 0.6393,
+      "step": 16770
+    },
+    {
+      "epoch": 9.56465977912362,
+      "grad_norm": 2.228473749196598,
+      "learning_rate": 1.1416281379090343e-07,
+      "loss": 0.65,
+      "step": 16780
+    },
+    {
+      "epoch": 9.570359814748842,
+      "grad_norm": 2.2588046626210154,
+      "learning_rate": 1.1118383725622018e-07,
+      "loss": 0.6435,
+      "step": 16790
+    },
+    {
+      "epoch": 9.576059850374065,
+      "grad_norm": 2.272713399884446,
+      "learning_rate": 1.0824402586379512e-07,
+      "loss": 0.6551,
+      "step": 16800
+    },
+    {
+      "epoch": 9.581759885999288,
+      "grad_norm": 2.3364321112703874,
+      "learning_rate": 1.0534339125689686e-07,
+      "loss": 0.6567,
+      "step": 16810
+    },
+    {
+      "epoch": 9.58745992162451,
+      "grad_norm": 2.40037730077152,
+      "learning_rate": 1.0248194492363028e-07,
+      "loss": 0.6511,
+      "step": 16820
+    },
+    {
+      "epoch": 9.593159957249732,
+      "grad_norm": 2.4025249286013217,
+      "learning_rate": 9.965969819689558e-08,
+      "loss": 0.656,
+      "step": 16830
+    },
+    {
+      "epoch": 9.598859992874955,
+      "grad_norm": 2.247494854846253,
+      "learning_rate": 9.687666225433823e-08,
+      "loss": 0.6529,
+      "step": 16840
+    },
+    {
+      "epoch": 9.604560028500178,
+      "grad_norm": 2.255657436136958,
+      "learning_rate": 9.413284811830903e-08,
+      "loss": 0.652,
+      "step": 16850
+    },
+    {
+      "epoch": 9.6102600641254,
+      "grad_norm": 2.264893980259995,
+      "learning_rate": 9.142826665581972e-08,
+      "loss": 0.6439,
+      "step": 16860
+    },
+    {
+      "epoch": 9.615960099750623,
+      "grad_norm": 2.266295402923885,
+      "learning_rate": 8.876292857849633e-08,
+      "loss": 0.6484,
+      "step": 16870
+    },
+    {
+      "epoch": 9.621660135375846,
+      "grad_norm": 2.243963802741696,
+      "learning_rate": 8.613684444254256e-08,
+      "loss": 0.6562,
+      "step": 16880
+    },
+    {
+      "epoch": 9.627360171001069,
+      "grad_norm": 2.2887376758264057,
+      "learning_rate": 8.35500246486931e-08,
+      "loss": 0.6523,
+      "step": 16890
+    },
+    {
+      "epoch": 9.633060206626292,
+      "grad_norm": 2.3370474229044733,
+      "learning_rate": 8.100247944217488e-08,
+      "loss": 0.6578,
+      "step": 16900
+    },
+    {
+      "epoch": 9.638760242251514,
+      "grad_norm": 2.328501011399279,
+      "learning_rate": 7.849421891266585e-08,
+      "loss": 0.6588,
+      "step": 16910
+    },
+    {
+      "epoch": 9.644460277876737,
+      "grad_norm": 2.3212148370181094,
+      "learning_rate": 7.602525299425623e-08,
+      "loss": 0.6483,
+      "step": 16920
+    },
+    {
+      "epoch": 9.65016031350196,
+      "grad_norm": 2.33295833370495,
+      "learning_rate": 7.359559146540518e-08,
+      "loss": 0.6587,
+      "step": 16930
+    },
+    {
+      "epoch": 9.655860349127183,
+      "grad_norm": 2.4077865683158968,
+      "learning_rate": 7.120524394890748e-08,
+      "loss": 0.6544,
+      "step": 16940
+    },
+    {
+      "epoch": 9.661560384752406,
+      "grad_norm": 2.356327360124305,
+      "learning_rate": 6.885421991185027e-08,
+      "loss": 0.6523,
+      "step": 16950
+    },
+    {
+      "epoch": 9.667260420377627,
+      "grad_norm": 2.407074508590822,
+      "learning_rate": 6.654252866558186e-08,
+      "loss": 0.6407,
+      "step": 16960
+    },
+    {
+      "epoch": 9.67296045600285,
+      "grad_norm": 2.3457387427111227,
+      "learning_rate": 6.427017936566859e-08,
+      "loss": 0.6548,
+      "step": 16970
+    },
+    {
+      "epoch": 9.678660491628072,
+      "grad_norm": 2.265353341929187,
+      "learning_rate": 6.203718101186141e-08,
+      "loss": 0.6543,
+      "step": 16980
+    },
+    {
+      "epoch": 9.684360527253295,
+      "grad_norm": 2.3296743025735216,
+      "learning_rate": 5.984354244805924e-08,
+      "loss": 0.658,
+      "step": 16990
+    },
+    {
+      "epoch": 9.690060562878518,
+      "grad_norm": 2.299627411976298,
+      "learning_rate": 5.768927236227684e-08,
+      "loss": 0.6614,
+      "step": 17000
+    },
+    {
+      "epoch": 9.69576059850374,
+      "grad_norm": 2.3220359703166666,
+      "learning_rate": 5.5574379286604805e-08,
+      "loss": 0.6502,
+      "step": 17010
+    },
+    {
+      "epoch": 9.701460634128964,
+      "grad_norm": 2.2422049453185373,
+      "learning_rate": 5.349887159718181e-08,
+      "loss": 0.6548,
+      "step": 17020
+    },
+    {
+      "epoch": 9.707160669754186,
+      "grad_norm": 2.364441388396548,
+      "learning_rate": 5.146275751415908e-08,
+      "loss": 0.6551,
+      "step": 17030
+    },
+    {
+      "epoch": 9.71286070537941,
+      "grad_norm": 2.3231425171152202,
+      "learning_rate": 4.9466045101664864e-08,
+      "loss": 0.6532,
+      "step": 17040
+    },
+    {
+      "epoch": 9.718560741004632,
+      "grad_norm": 2.457726448339882,
+      "learning_rate": 4.750874226777891e-08,
+      "loss": 0.6569,
+      "step": 17050
+    },
+    {
+      "epoch": 9.724260776629855,
+      "grad_norm": 2.352927243366894,
+      "learning_rate": 4.5590856764492486e-08,
+      "loss": 0.6481,
+      "step": 17060
+    },
+    {
+      "epoch": 9.729960812255076,
+      "grad_norm": 2.3064337676695805,
+      "learning_rate": 4.37123961876873e-08,
+      "loss": 0.6484,
+      "step": 17070
+    },
+    {
+      "epoch": 9.735660847880299,
+      "grad_norm": 2.2681358596665904,
+      "learning_rate": 4.187336797709884e-08,
+      "loss": 0.6547,
+      "step": 17080
+    },
+    {
+      "epoch": 9.741360883505521,
+      "grad_norm": 2.3506977886549754,
+      "learning_rate": 4.007377941628754e-08,
+      "loss": 0.6528,
+      "step": 17090
+    },
+    {
+      "epoch": 9.747060919130744,
+      "grad_norm": 2.3589241159878775,
+      "learning_rate": 3.8313637632613196e-08,
+      "loss": 0.6545,
+      "step": 17100
+    },
+    {
+      "epoch": 9.752760954755967,
+      "grad_norm": 2.3427788634759934,
+      "learning_rate": 3.659294959720283e-08,
+      "loss": 0.6599,
+      "step": 17110
+    },
+    {
+      "epoch": 9.75846099038119,
+      "grad_norm": 2.3522287518440455,
+      "learning_rate": 3.491172212492733e-08,
+      "loss": 0.6463,
+      "step": 17120
+    },
+    {
+      "epoch": 9.764161026006413,
+      "grad_norm": 2.351901096283413,
+      "learning_rate": 3.326996187436926e-08,
+      "loss": 0.6538,
+      "step": 17130
+    },
+    {
+      "epoch": 9.769861061631635,
+      "grad_norm": 2.2918549555739154,
+      "learning_rate": 3.1667675347801795e-08,
+      "loss": 0.6571,
+      "step": 17140
+    },
+    {
+      "epoch": 9.775561097256858,
+      "grad_norm": 2.377678545336172,
+      "learning_rate": 3.0104868891159825e-08,
+      "loss": 0.6555,
+      "step": 17150
+    },
+    {
+      "epoch": 9.781261132882081,
+      "grad_norm": 2.2424384418651893,
+      "learning_rate": 2.8581548694013304e-08,
+      "loss": 0.6623,
+      "step": 17160
+    },
+    {
+      "epoch": 9.786961168507304,
+      "grad_norm": 2.2658459936780235,
+      "learning_rate": 2.709772078954842e-08,
+      "loss": 0.6466,
+      "step": 17170
+    },
+    {
+      "epoch": 9.792661204132525,
+      "grad_norm": 2.2725019855816937,
+      "learning_rate": 2.565339105453757e-08,
+      "loss": 0.653,
+      "step": 17180
+    },
+    {
+      "epoch": 9.798361239757748,
+      "grad_norm": 2.3682684509359877,
+      "learning_rate": 2.4248565209320507e-08,
+      "loss": 0.651,
+      "step": 17190
+    },
+    {
+      "epoch": 9.80406127538297,
+      "grad_norm": 2.2230547848834714,
+      "learning_rate": 2.2883248817777703e-08,
+      "loss": 0.6567,
+      "step": 17200
+    },
+    {
+      "epoch": 9.809761311008193,
+      "grad_norm": 2.4480084011558034,
+      "learning_rate": 2.1557447287312572e-08,
+      "loss": 0.6516,
+      "step": 17210
+    },
+    {
+      "epoch": 9.815461346633416,
+      "grad_norm": 2.3857177484205594,
+      "learning_rate": 2.0271165868828157e-08,
+      "loss": 0.6609,
+      "step": 17220
+    },
+    {
+      "epoch": 9.821161382258639,
+      "grad_norm": 2.294663835248534,
+      "learning_rate": 1.9024409656703824e-08,
+      "loss": 0.6652,
+      "step": 17230
+    },
+    {
+      "epoch": 9.826861417883862,
+      "grad_norm": 2.252195410951251,
+      "learning_rate": 1.7817183588778596e-08,
+      "loss": 0.6531,
+      "step": 17240
+    },
+    {
+      "epoch": 9.832561453509085,
+      "grad_norm": 2.3618394206239466,
+      "learning_rate": 1.6649492446332292e-08,
+      "loss": 0.6479,
+      "step": 17250
+    },
+    {
+      "epoch": 9.838261489134307,
+      "grad_norm": 2.2660526958358624,
+      "learning_rate": 1.5521340854061097e-08,
+      "loss": 0.654,
+      "step": 17260
+    },
+    {
+      "epoch": 9.84396152475953,
+      "grad_norm": 2.5207873951096644,
+      "learning_rate": 1.4432733280065336e-08,
+      "loss": 0.6469,
+      "step": 17270
+    },
+    {
+      "epoch": 9.849661560384753,
+      "grad_norm": 2.383762606625419,
+      "learning_rate": 1.338367403583063e-08,
+      "loss": 0.6663,
+      "step": 17280
+    },
+    {
+      "epoch": 9.855361596009976,
+      "grad_norm": 2.374411011614862,
+      "learning_rate": 1.2374167276205663e-08,
+      "loss": 0.6425,
+      "step": 17290
+    },
+    {
+      "epoch": 9.861061631635197,
+      "grad_norm": 2.2493191496711678,
+      "learning_rate": 1.1404216999391093e-08,
+      "loss": 0.6497,
+      "step": 17300
+    },
+    {
+      "epoch": 9.86676166726042,
+      "grad_norm": 2.3796088769072177,
+      "learning_rate": 1.0473827046925122e-08,
+      "loss": 0.6605,
+      "step": 17310
+    },
+    {
+      "epoch": 9.872461702885643,
+      "grad_norm": 2.2180769836324776,
+      "learning_rate": 9.583001103661283e-09,
+      "loss": 0.6522,
+      "step": 17320
+    },
+    {
+      "epoch": 9.878161738510865,
+      "grad_norm": 2.3072964999263403,
+      "learning_rate": 8.731742697758449e-09,
+      "loss": 0.6512,
+      "step": 17330
+    },
+    {
+      "epoch": 9.883861774136088,
+      "grad_norm": 2.283779982842626,
+      "learning_rate": 7.92005520066974e-09,
+      "loss": 0.6636,
+      "step": 17340
+    },
+    {
+      "epoch": 9.889561809761311,
+      "grad_norm": 2.297326593288268,
+      "learning_rate": 7.147941827121419e-09,
+      "loss": 0.6404,
+      "step": 17350
+    },
+    {
+      "epoch": 9.895261845386534,
+      "grad_norm": 2.3893193554640257,
+      "learning_rate": 6.415405635107341e-09,
+      "loss": 0.6488,
+      "step": 17360
+    },
+    {
+      "epoch": 9.900961881011757,
+      "grad_norm": 2.3494192289938107,
+      "learning_rate": 5.722449525873419e-09,
+      "loss": 0.6595,
+      "step": 17370
+    },
+    {
+      "epoch": 9.90666191663698,
+      "grad_norm": 2.2589001422532387,
+      "learning_rate": 5.069076243905402e-09,
+      "loss": 0.6578,
+      "step": 17380
+    },
+    {
+      "epoch": 9.912361952262202,
+      "grad_norm": 2.2649210783000293,
+      "learning_rate": 4.455288376921108e-09,
+      "loss": 0.6498,
+      "step": 17390
+    },
+    {
+      "epoch": 9.918061987887425,
+      "grad_norm": 2.4144073651514524,
+      "learning_rate": 3.881088355855989e-09,
+      "loss": 0.6547,
+      "step": 17400
+    },
+    {
+      "epoch": 9.923762023512648,
+      "grad_norm": 2.278865809364427,
+      "learning_rate": 3.346478454859803e-09,
+      "loss": 0.651,
+      "step": 17410
+    },
+    {
+      "epoch": 9.929462059137869,
+      "grad_norm": 2.3417238771811055,
+      "learning_rate": 2.8514607912799586e-09,
+      "loss": 0.6503,
+      "step": 17420
+    },
+    {
+      "epoch": 9.935162094763092,
+      "grad_norm": 2.331772674843219,
+      "learning_rate": 2.3960373256581846e-09,
+      "loss": 0.6525,
+      "step": 17430
+    },
+    {
+      "epoch": 9.940862130388314,
+      "grad_norm": 2.2924527253475575,
+      "learning_rate": 1.9802098617216494e-09,
+      "loss": 0.6562,
+      "step": 17440
+    },
+    {
+      "epoch": 9.946562166013537,
+      "grad_norm": 2.320131850003266,
+      "learning_rate": 1.6039800463762966e-09,
+      "loss": 0.6554,
+      "step": 17450
+    },
+    {
+      "epoch": 9.95226220163876,
+      "grad_norm": 2.258058768319489,
+      "learning_rate": 1.2673493696979677e-09,
+      "loss": 0.6471,
+      "step": 17460
+    },
+    {
+      "epoch": 9.957962237263983,
+      "grad_norm": 2.376801772568866,
+      "learning_rate": 9.703191649279574e-10,
+      "loss": 0.6512,
+      "step": 17470
+    },
+    {
+      "epoch": 9.963662272889206,
+      "grad_norm": 2.321621141337843,
+      "learning_rate": 7.128906084707954e-10,
+      "loss": 0.6526,
+      "step": 17480
+    },
+    {
+      "epoch": 9.969362308514429,
+      "grad_norm": 2.354903966389974,
+      "learning_rate": 4.950647198842529e-10,
+      "loss": 0.6455,
+      "step": 17490
+    },
+    {
+      "epoch": 9.975062344139651,
+      "grad_norm": 2.261998747744868,
+      "learning_rate": 3.1684236187823345e-10,
+      "loss": 0.645,
+      "step": 17500
+    },
+    {
+      "epoch": 9.980762379764874,
+      "grad_norm": 2.2877903683587695,
+      "learning_rate": 1.7822424031144203e-10,
+      "loss": 0.6454,
+      "step": 17510
+    },
+    {
+      "epoch": 9.986462415390097,
+      "grad_norm": 2.3510418373932804,
+      "learning_rate": 7.921090418805399e-11,
+      "loss": 0.6525,
+      "step": 17520
+    },
+    {
+      "epoch": 9.992162451015318,
+      "grad_norm": 2.4334827508597705,
+      "learning_rate": 1.980274565438478e-11,
+      "loss": 0.6616,
+      "step": 17530
+    },
+    {
+      "epoch": 9.99786248664054,
+      "grad_norm": 2.416970594856183,
+      "learning_rate": 0.0,
+      "loss": 0.6481,
+      "step": 17540
+    },
+    {
+      "epoch": 9.99786248664054,
+      "step": 17540,
+      "total_flos": 2.4261309001996698e+17,
+      "train_loss": 0.7485582114081563,
+      "train_runtime": 249180.9208,
+      "train_samples_per_second": 27.035,
+      "train_steps_per_second": 0.07
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 17540,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.4261309001996698e+17,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}