diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20498 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 11696,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.3728413581848145,
+      "learning_rate": 4.998290013679891e-05,
+      "loss": 10.0117,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.599367380142212,
+      "learning_rate": 4.996580027359781e-05,
+      "loss": 9.1745,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4951720237731934,
+      "learning_rate": 4.994870041039672e-05,
+      "loss": 8.7624,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4932913780212402,
+      "learning_rate": 4.9931600547195625e-05,
+      "loss": 8.5709,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3983781337738037,
+      "learning_rate": 4.991450068399453e-05,
+      "loss": 8.4411,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2340972423553467,
+      "learning_rate": 4.989740082079344e-05,
+      "loss": 8.1957,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1964590549468994,
+      "learning_rate": 4.988030095759234e-05,
+      "loss": 8.0598,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9686617851257324,
+      "learning_rate": 4.986320109439125e-05,
+      "loss": 7.8679,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8235087394714355,
+      "learning_rate": 4.984610123119015e-05,
+      "loss": 7.7621,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7598885297775269,
+      "learning_rate": 4.9829001367989056e-05,
+      "loss": 7.5417,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.94789719581604,
+      "learning_rate": 4.981190150478797e-05,
+      "loss": 7.3331,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9366331100463867,
+      "learning_rate": 4.979480164158687e-05,
+      "loss": 7.1039,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1228675842285156,
+      "learning_rate": 4.977770177838578e-05,
+      "loss": 6.9965,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9855167865753174,
+      "learning_rate": 4.976060191518468e-05,
+      "loss": 6.8865,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8035985231399536,
+      "learning_rate": 4.9743502051983585e-05,
+      "loss": 6.6322,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8589977025985718,
+      "learning_rate": 4.9726402188782486e-05,
+      "loss": 6.5616,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.745139241218567,
+      "learning_rate": 4.97093023255814e-05,
+      "loss": 6.3479,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.266196608543396,
+      "learning_rate": 4.969220246238031e-05,
+      "loss": 6.2343,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.005223035812378,
+      "learning_rate": 4.967510259917921e-05,
+      "loss": 6.1866,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5017377138137817,
+      "learning_rate": 4.9658002735978115e-05,
+      "loss": 5.972,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6136974096298218,
+      "learning_rate": 4.9640902872777016e-05,
+      "loss": 6.1335,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3865970373153687,
+      "learning_rate": 4.962380300957592e-05,
+      "loss": 5.944,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.283933162689209,
+      "learning_rate": 4.960670314637483e-05,
+      "loss": 5.8713,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1588549613952637,
+      "learning_rate": 4.958960328317374e-05,
+      "loss": 5.7731,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3081687688827515,
+      "learning_rate": 4.9572503419972645e-05,
+      "loss": 5.7429,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6212745904922485,
+      "learning_rate": 4.9555403556771546e-05,
+      "loss": 5.481,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.709843397140503,
+      "learning_rate": 4.953830369357045e-05,
+      "loss": 5.4855,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2814812660217285,
+      "learning_rate": 4.952120383036936e-05,
+      "loss": 5.4367,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5539968013763428,
+      "learning_rate": 4.950410396716826e-05,
+      "loss": 5.2652,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.500375509262085,
+      "learning_rate": 4.948700410396717e-05,
+      "loss": 5.5137,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7938899993896484,
+      "learning_rate": 4.9469904240766076e-05,
+      "loss": 5.4663,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6169573068618774,
+      "learning_rate": 4.945280437756498e-05,
+      "loss": 5.3137,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1904114484786987,
+      "learning_rate": 4.943570451436389e-05,
+      "loss": 5.1871,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5564723014831543,
+      "learning_rate": 4.941860465116279e-05,
+      "loss": 5.4058,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6187268495559692,
+      "learning_rate": 4.94015047879617e-05,
+      "loss": 5.1919,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2222367525100708,
+      "learning_rate": 4.93844049247606e-05,
+      "loss": 5.1876,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5898586511611938,
+      "learning_rate": 4.936730506155951e-05,
+      "loss": 5.1217,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.446902871131897,
+      "learning_rate": 4.935020519835842e-05,
+      "loss": 5.0582,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.862309217453003,
+      "learning_rate": 4.933310533515732e-05,
+      "loss": 4.9028,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5455704927444458,
+      "learning_rate": 4.931600547195623e-05,
+      "loss": 5.0352,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.772558569908142,
+      "learning_rate": 4.929890560875513e-05,
+      "loss": 4.9707,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8154480457305908,
+      "learning_rate": 4.9281805745554036e-05,
+      "loss": 5.0588,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4536504745483398,
+      "learning_rate": 4.9264705882352944e-05,
+      "loss": 4.9997,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6166527271270752,
+      "learning_rate": 4.924760601915185e-05,
+      "loss": 4.9472,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1260979175567627,
+      "learning_rate": 4.923050615595076e-05,
+      "loss": 4.944,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4778213500976562,
+      "learning_rate": 4.921340629274966e-05,
+      "loss": 4.8657,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.255488395690918,
+      "learning_rate": 4.9196306429548566e-05,
+      "loss": 4.7728,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5698314905166626,
+      "learning_rate": 4.917920656634747e-05,
+      "loss": 4.9433,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.342402696609497,
+      "learning_rate": 4.9162106703146374e-05,
+      "loss": 4.6336,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4427978992462158,
+      "learning_rate": 4.914500683994528e-05,
+      "loss": 4.7112,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6528425216674805,
+      "learning_rate": 4.912790697674419e-05,
+      "loss": 4.876,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.706386923789978,
+      "learning_rate": 4.9110807113543096e-05,
+      "loss": 4.5971,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7773737907409668,
+      "learning_rate": 4.9093707250342e-05,
+      "loss": 4.7944,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7884222269058228,
+      "learning_rate": 4.9076607387140904e-05,
+      "loss": 4.5912,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5389587879180908,
+      "learning_rate": 4.905950752393981e-05,
+      "loss": 4.6656,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7006316184997559,
+      "learning_rate": 4.904240766073871e-05,
+      "loss": 4.6589,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6314700841903687,
+      "learning_rate": 4.9025307797537626e-05,
+      "loss": 4.5036,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7512930631637573,
+      "learning_rate": 4.900820793433653e-05,
+      "loss": 4.5105,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9109033346176147,
+      "learning_rate": 4.8991108071135434e-05,
+      "loss": 4.4449,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.547317385673523,
+      "learning_rate": 4.897400820793434e-05,
+      "loss": 4.4906,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5143060684204102,
+      "learning_rate": 4.895690834473324e-05,
+      "loss": 4.3545,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6648136377334595,
+      "learning_rate": 4.893980848153215e-05,
+      "loss": 4.4093,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5027951002120972,
+      "learning_rate": 4.892270861833106e-05,
+      "loss": 4.3444,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9429397583007812,
+      "learning_rate": 4.8905608755129964e-05,
+      "loss": 4.4937,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4009640216827393,
+      "learning_rate": 4.888850889192887e-05,
+      "loss": 4.3625,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7445279359817505,
+      "learning_rate": 4.887140902872777e-05,
+      "loss": 4.2759,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4381403923034668,
+      "learning_rate": 4.885430916552668e-05,
+      "loss": 4.574,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5582979917526245,
+      "learning_rate": 4.883720930232558e-05,
+      "loss": 4.3831,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.461166501045227,
+      "learning_rate": 4.882010943912449e-05,
+      "loss": 4.5049,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.756549596786499,
+      "learning_rate": 4.8803009575923394e-05,
+      "loss": 4.1334,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9006497859954834,
+      "learning_rate": 4.87859097127223e-05,
+      "loss": 4.3406,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4551759958267212,
+      "learning_rate": 4.876880984952121e-05,
+      "loss": 4.4688,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5335747003555298,
+      "learning_rate": 4.875170998632011e-05,
+      "loss": 4.2096,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4802557229995728,
+      "learning_rate": 4.873461012311902e-05,
+      "loss": 4.0918,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9993939399719238,
+      "learning_rate": 4.8717510259917924e-05,
+      "loss": 4.3636,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.706895351409912,
+      "learning_rate": 4.8700410396716825e-05,
+      "loss": 4.0891,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4630484580993652,
+      "learning_rate": 4.868331053351574e-05,
+      "loss": 4.3254,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4353913068771362,
+      "learning_rate": 4.866621067031464e-05,
+      "loss": 4.1935,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5042275190353394,
+      "learning_rate": 4.864911080711355e-05,
+      "loss": 4.0913,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.792472004890442,
+      "learning_rate": 4.863201094391245e-05,
+      "loss": 4.2159,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4059948921203613,
+      "learning_rate": 4.8614911080711355e-05,
+      "loss": 4.2934,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8408161401748657,
+      "learning_rate": 4.859781121751026e-05,
+      "loss": 4.1482,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.716046690940857,
+      "learning_rate": 4.858071135430917e-05,
+      "loss": 4.0273,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8415429592132568,
+      "learning_rate": 4.856361149110808e-05,
+      "loss": 4.0837,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4880731105804443,
+      "learning_rate": 4.854651162790698e-05,
+      "loss": 4.1957,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8150016069412231,
+      "learning_rate": 4.8529411764705885e-05,
+      "loss": 4.225,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7828190326690674,
+      "learning_rate": 4.851231190150479e-05,
+      "loss": 4.1053,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8549565076828003,
+      "learning_rate": 4.849521203830369e-05,
+      "loss": 3.8068,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4295508861541748,
+      "learning_rate": 4.84781121751026e-05,
+      "loss": 3.9358,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4277971982955933,
+      "learning_rate": 4.846101231190151e-05,
+      "loss": 4.1512,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4937466382980347,
+      "learning_rate": 4.8443912448700415e-05,
+      "loss": 4.0162,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3076913356781006,
+      "learning_rate": 4.842681258549932e-05,
+      "loss": 4.0336,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.673946738243103,
+      "learning_rate": 4.840971272229822e-05,
+      "loss": 4.0915,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5978224277496338,
+      "learning_rate": 4.839261285909713e-05,
+      "loss": 4.137,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.214573383331299,
+      "learning_rate": 4.837551299589603e-05,
+      "loss": 3.8231,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4263286590576172,
+      "learning_rate": 4.835841313269494e-05,
+      "loss": 3.9326,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.694931149482727,
+      "learning_rate": 4.8341313269493845e-05,
+      "loss": 4.0673,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.558498740196228,
+      "learning_rate": 4.832421340629275e-05,
+      "loss": 3.9454,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5231866836547852,
+      "learning_rate": 4.830711354309166e-05,
+      "loss": 4.0214,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6748324632644653,
+      "learning_rate": 4.829001367989056e-05,
+      "loss": 3.9983,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3733506202697754,
+      "learning_rate": 4.827291381668947e-05,
+      "loss": 3.8526,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6695655584335327,
+      "learning_rate": 4.8255813953488375e-05,
+      "loss": 3.8948,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3915668725967407,
+      "learning_rate": 4.8238714090287276e-05,
+      "loss": 3.8511,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5736514329910278,
+      "learning_rate": 4.822161422708619e-05,
+      "loss": 3.9446,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.562687635421753,
+      "learning_rate": 4.820451436388509e-05,
+      "loss": 3.6682,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4215232133865356,
+      "learning_rate": 4.8187414500684e-05,
+      "loss": 3.6618,
+      "step": 424
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8750267028808594,
+      "learning_rate": 4.8170314637482905e-05,
+      "loss": 3.8246,
+      "step": 428
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5843207836151123,
+      "learning_rate": 4.8153214774281806e-05,
+      "loss": 3.807,
+      "step": 432
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6384755373001099,
+      "learning_rate": 4.813611491108071e-05,
+      "loss": 3.8329,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.646612286567688,
+      "learning_rate": 4.811901504787962e-05,
+      "loss": 3.9736,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.592463493347168,
+      "learning_rate": 4.810191518467853e-05,
+      "loss": 3.6107,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6230803728103638,
+      "learning_rate": 4.808481532147743e-05,
+      "loss": 3.8141,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7098625898361206,
+      "learning_rate": 4.8067715458276336e-05,
+      "loss": 3.8751,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.441146731376648,
+      "learning_rate": 4.805061559507524e-05,
+      "loss": 3.8461,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4387036561965942,
+      "learning_rate": 4.8033515731874144e-05,
+      "loss": 3.8432,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6620376110076904,
+      "learning_rate": 4.801641586867305e-05,
+      "loss": 3.7844,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5403114557266235,
+      "learning_rate": 4.799931600547196e-05,
+      "loss": 3.6556,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5642478466033936,
+      "learning_rate": 4.7982216142270866e-05,
+      "loss": 3.7526,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5027506351470947,
+      "learning_rate": 4.796511627906977e-05,
+      "loss": 3.7862,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5570485591888428,
+      "learning_rate": 4.7948016415868674e-05,
+      "loss": 3.7656,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2708029747009277,
+      "learning_rate": 4.793091655266758e-05,
+      "loss": 3.8742,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5446516275405884,
+      "learning_rate": 4.791381668946648e-05,
+      "loss": 3.6401,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5480107069015503,
+      "learning_rate": 4.789671682626539e-05,
+      "loss": 3.7441,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5571659803390503,
+      "learning_rate": 4.78796169630643e-05,
+      "loss": 3.6849,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9021155834197998,
+      "learning_rate": 4.7862517099863204e-05,
+      "loss": 3.7989,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5292832851409912,
+      "learning_rate": 4.784541723666211e-05,
+      "loss": 3.8129,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5750986337661743,
+      "learning_rate": 4.782831737346101e-05,
+      "loss": 3.648,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6995611190795898,
+      "learning_rate": 4.781121751025992e-05,
+      "loss": 3.7686,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7573764324188232,
+      "learning_rate": 4.7794117647058826e-05,
+      "loss": 3.6636,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5131741762161255,
+      "learning_rate": 4.7777017783857733e-05,
+      "loss": 3.5793,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6838548183441162,
+      "learning_rate": 4.775991792065664e-05,
+      "loss": 3.6336,
+      "step": 524
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9402954578399658,
+      "learning_rate": 4.774281805745554e-05,
+      "loss": 3.6733,
+      "step": 528
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7752444744110107,
+      "learning_rate": 4.772571819425445e-05,
+      "loss": 3.8021,
+      "step": 532
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4470579624176025,
+      "learning_rate": 4.7708618331053356e-05,
+      "loss": 3.6526,
+      "step": 536
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5560115575790405,
+      "learning_rate": 4.769151846785226e-05,
+      "loss": 3.7236,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6849740743637085,
+      "learning_rate": 4.7674418604651164e-05,
+      "loss": 3.7586,
+      "step": 544
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4917711019515991,
+      "learning_rate": 4.765731874145007e-05,
+      "loss": 3.7791,
+      "step": 548
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.791801929473877,
+      "learning_rate": 4.764021887824898e-05,
+      "loss": 3.5659,
+      "step": 552
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.531817078590393,
+      "learning_rate": 4.7623119015047886e-05,
+      "loss": 3.5811,
+      "step": 556
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8725793361663818,
+      "learning_rate": 4.7606019151846787e-05,
+      "loss": 3.7337,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5851105451583862,
+      "learning_rate": 4.7588919288645694e-05,
+      "loss": 3.4899,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4937329292297363,
+      "learning_rate": 4.7571819425444594e-05,
+      "loss": 3.5653,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8102850914001465,
+      "learning_rate": 4.75547195622435e-05,
+      "loss": 3.517,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4572982788085938,
+      "learning_rate": 4.753761969904241e-05,
+      "loss": 3.723,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5815645456314087,
+      "learning_rate": 4.7520519835841317e-05,
+      "loss": 3.6615,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6406457424163818,
+      "learning_rate": 4.7503419972640224e-05,
+      "loss": 3.6796,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8512486219406128,
+      "learning_rate": 4.7486320109439124e-05,
+      "loss": 3.5586,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6738507747650146,
+      "learning_rate": 4.746922024623803e-05,
+      "loss": 3.4724,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.766518235206604,
+      "learning_rate": 4.745212038303693e-05,
+      "loss": 3.7213,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.681229591369629,
+      "learning_rate": 4.7435020519835846e-05,
+      "loss": 3.6538,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7900785207748413,
+      "learning_rate": 4.7417920656634754e-05,
+      "loss": 3.7952,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5610637664794922,
+      "learning_rate": 4.7400820793433654e-05,
+      "loss": 3.5255,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3365826606750488,
+      "learning_rate": 4.738372093023256e-05,
+      "loss": 3.5621,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8946609497070312,
+      "learning_rate": 4.736662106703146e-05,
+      "loss": 3.5664,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.728330135345459,
+      "learning_rate": 4.734952120383037e-05,
+      "loss": 3.7577,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7583847045898438,
+      "learning_rate": 4.733242134062928e-05,
+      "loss": 3.5722,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9061710834503174,
+      "learning_rate": 4.7315321477428184e-05,
+      "loss": 3.3436,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7074024677276611,
+      "learning_rate": 4.729822161422709e-05,
+      "loss": 3.5426,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.649793028831482,
+      "learning_rate": 4.728112175102599e-05,
+      "loss": 3.5826,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7260462045669556,
+      "learning_rate": 4.72640218878249e-05,
+      "loss": 3.5353,
+      "step": 640
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.745391607284546,
+      "learning_rate": 4.724692202462381e-05,
+      "loss": 3.3534,
+      "step": 644
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.374226450920105,
+      "learning_rate": 4.722982216142271e-05,
+      "loss": 3.386,
+      "step": 648
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8732798099517822,
+      "learning_rate": 4.7212722298221615e-05,
+      "loss": 3.6073,
+      "step": 652
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4075376987457275,
+      "learning_rate": 4.719562243502052e-05,
+      "loss": 3.5393,
+      "step": 656
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.54508638381958,
+      "learning_rate": 4.717852257181943e-05,
+      "loss": 3.475,
+      "step": 660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5737498998641968,
+      "learning_rate": 4.716142270861834e-05,
+      "loss": 3.5497,
+      "step": 664
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8623074293136597,
+      "learning_rate": 4.714432284541724e-05,
+      "loss": 3.4603,
+      "step": 668
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7199251651763916,
+      "learning_rate": 4.7127222982216145e-05,
+      "loss": 3.5801,
+      "step": 672
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5866843461990356,
+      "learning_rate": 4.7110123119015045e-05,
+      "loss": 3.3454,
+      "step": 676
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9907779693603516,
+      "learning_rate": 4.709302325581396e-05,
+      "loss": 3.6247,
+      "step": 680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3720687627792358,
+      "learning_rate": 4.707592339261287e-05,
+      "loss": 3.3614,
+      "step": 684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.739660620689392,
+      "learning_rate": 4.705882352941177e-05,
+      "loss": 3.4211,
+      "step": 688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6425236463546753,
+      "learning_rate": 4.7041723666210675e-05,
+      "loss": 3.423,
+      "step": 692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5457091331481934,
+      "learning_rate": 4.7024623803009575e-05,
+      "loss": 3.3258,
+      "step": 696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5979949235916138,
+      "learning_rate": 4.700752393980848e-05,
+      "loss": 3.2838,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6761040687561035,
+      "learning_rate": 4.699042407660739e-05,
+      "loss": 3.334,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.552573323249817,
+      "learning_rate": 4.69733242134063e-05,
+      "loss": 3.5355,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6743354797363281,
+      "learning_rate": 4.6956224350205205e-05,
+      "loss": 3.4076,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.555662155151367,
+      "learning_rate": 4.6939124487004105e-05,
+      "loss": 3.7446,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5942860841751099,
+      "learning_rate": 4.692202462380301e-05,
+      "loss": 3.553,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9089189767837524,
+      "learning_rate": 4.690492476060191e-05,
+      "loss": 3.3584,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0768373012542725,
+      "learning_rate": 4.688782489740082e-05,
+      "loss": 3.4295,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3060362339019775,
+      "learning_rate": 4.687072503419973e-05,
+      "loss": 3.415,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9307421445846558,
+      "learning_rate": 4.6853625170998635e-05,
+      "loss": 3.4901,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7142446041107178,
+      "learning_rate": 4.683652530779754e-05,
+      "loss": 3.6096,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7756657600402832,
+      "learning_rate": 4.681942544459644e-05,
+      "loss": 3.4534,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5459442138671875,
+      "learning_rate": 4.680232558139535e-05,
+      "loss": 3.3813,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1961963176727295,
+      "learning_rate": 4.678522571819426e-05,
+      "loss": 3.3467,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7320613861083984,
+      "learning_rate": 4.676812585499316e-05,
+      "loss": 3.3927,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6255303621292114,
+      "learning_rate": 4.6751025991792066e-05,
+      "loss": 3.4644,
+      "step": 760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5142388343811035,
+      "learning_rate": 4.673392612859097e-05,
+      "loss": 3.4657,
+      "step": 764
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5093744993209839,
+      "learning_rate": 4.671682626538988e-05,
+      "loss": 3.3144,
+      "step": 768
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4730446338653564,
+      "learning_rate": 4.669972640218879e-05,
+      "loss": 3.3382,
+      "step": 772
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7650116682052612,
+      "learning_rate": 4.668262653898769e-05,
+      "loss": 3.4519,
+      "step": 776
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.637071132659912,
+      "learning_rate": 4.6665526675786596e-05,
+      "loss": 3.3723,
+      "step": 780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4582788944244385,
+      "learning_rate": 4.6648426812585496e-05,
+      "loss": 3.4839,
+      "step": 784
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4654922485351562,
+      "learning_rate": 4.663132694938441e-05,
+      "loss": 3.1257,
+      "step": 788
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4668978452682495,
+      "learning_rate": 4.661422708618332e-05,
+      "loss": 3.2506,
+      "step": 792
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5190536975860596,
+      "learning_rate": 4.659712722298222e-05,
+      "loss": 3.4055,
+      "step": 796
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9057687520980835,
+      "learning_rate": 4.6580027359781126e-05,
+      "loss": 3.4804,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4573813676834106,
+      "learning_rate": 4.6562927496580026e-05,
+      "loss": 3.3792,
+      "step": 804
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3757171630859375,
+      "learning_rate": 4.6545827633378933e-05,
+      "loss": 3.4649,
+      "step": 808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7024015188217163,
+      "learning_rate": 4.652872777017784e-05,
+      "loss": 3.3149,
+      "step": 812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5424153804779053,
+      "learning_rate": 4.651162790697675e-05,
+      "loss": 3.3,
+      "step": 816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4667370319366455,
+      "learning_rate": 4.6494528043775655e-05,
+      "loss": 3.2845,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5216193199157715,
+      "learning_rate": 4.6477428180574556e-05,
+      "loss": 3.3062,
+      "step": 824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5499827861785889,
+      "learning_rate": 4.6460328317373463e-05,
+      "loss": 3.263,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.644148588180542,
+      "learning_rate": 4.6443228454172364e-05,
+      "loss": 3.4374,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6420782804489136,
+      "learning_rate": 4.642612859097127e-05,
+      "loss": 3.4049,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4325385093688965,
+      "learning_rate": 4.640902872777018e-05,
+      "loss": 3.3499,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6447879076004028,
+      "learning_rate": 4.6391928864569086e-05,
+      "loss": 3.4098,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4675137996673584,
+      "learning_rate": 4.637482900136799e-05,
+      "loss": 3.2778,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5847750902175903,
+      "learning_rate": 4.6357729138166894e-05,
+      "loss": 3.1695,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.514772653579712,
+      "learning_rate": 4.63406292749658e-05,
+      "loss": 3.1556,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.511367917060852,
+      "learning_rate": 4.632352941176471e-05,
+      "loss": 3.3307,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7640196084976196,
+      "learning_rate": 4.630642954856361e-05,
+      "loss": 3.2985,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7074265480041504,
+      "learning_rate": 4.628932968536252e-05,
+      "loss": 3.2598,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.52577805519104,
+      "learning_rate": 4.6272229822161424e-05,
+      "loss": 3.0872,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6908501386642456,
+      "learning_rate": 4.625512995896033e-05,
+      "loss": 3.1931,
+      "step": 876
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6583012342453003,
+      "learning_rate": 4.623803009575924e-05,
+      "loss": 3.1849,
+      "step": 880
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4067027568817139,
+      "learning_rate": 4.622093023255814e-05,
+      "loss": 3.1912,
+      "step": 884
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.388487458229065,
+      "learning_rate": 4.6203830369357046e-05,
+      "loss": 3.2845,
+      "step": 888
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7910948991775513,
+      "learning_rate": 4.6186730506155954e-05,
+      "loss": 3.1146,
+      "step": 892
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4454721212387085,
+      "learning_rate": 4.616963064295486e-05,
+      "loss": 3.2114,
+      "step": 896
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4809476137161255,
+      "learning_rate": 4.615253077975377e-05,
+      "loss": 3.3329,
+      "step": 900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6668082475662231,
+      "learning_rate": 4.613543091655267e-05,
+      "loss": 3.2517,
+      "step": 904
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.72063410282135,
+      "learning_rate": 4.6118331053351576e-05,
+      "loss": 3.0895,
+      "step": 908
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5995526313781738,
+      "learning_rate": 4.610123119015048e-05,
+      "loss": 3.2703,
+      "step": 912
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.453747034072876,
+      "learning_rate": 4.6084131326949384e-05,
+      "loss": 3.3619,
+      "step": 916
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7982878684997559,
+      "learning_rate": 4.606703146374829e-05,
+      "loss": 3.2657,
+      "step": 920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3937926292419434,
+      "learning_rate": 4.60499316005472e-05,
+      "loss": 3.1727,
+      "step": 924
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5705368518829346,
+      "learning_rate": 4.6032831737346106e-05,
+      "loss": 3.0657,
+      "step": 928
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6821500062942505,
+      "learning_rate": 4.601573187414501e-05,
+      "loss": 3.3191,
+      "step": 932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5288244485855103,
+      "learning_rate": 4.5998632010943914e-05,
+      "loss": 3.0876,
+      "step": 936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7748781442642212,
+      "learning_rate": 4.598153214774282e-05,
+      "loss": 3.18,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7024669647216797,
+      "learning_rate": 4.596443228454172e-05,
+      "loss": 3.3601,
+      "step": 944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7017521858215332,
+      "learning_rate": 4.5947332421340636e-05,
+      "loss": 2.9751,
+      "step": 948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7405025959014893,
+      "learning_rate": 4.593023255813954e-05,
+      "loss": 3.2073,
+      "step": 952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6610547304153442,
+      "learning_rate": 4.5913132694938444e-05,
+      "loss": 3.2363,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4785616397857666,
+      "learning_rate": 4.5896032831737345e-05,
+      "loss": 3.3254,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.453533411026001,
+      "learning_rate": 4.587893296853625e-05,
+      "loss": 3.21,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4303959608078003,
+      "learning_rate": 4.586183310533516e-05,
+      "loss": 3.3226,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6272943019866943,
+      "learning_rate": 4.584473324213407e-05,
+      "loss": 3.5944,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.823716402053833,
+      "learning_rate": 4.5827633378932974e-05,
+      "loss": 3.234,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7006231546401978,
+      "learning_rate": 4.5810533515731875e-05,
+      "loss": 3.0123,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3645117282867432,
+      "learning_rate": 4.579343365253078e-05,
+      "loss": 3.2243,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7724872827529907,
+      "learning_rate": 4.577633378932969e-05,
+      "loss": 3.1705,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6377599239349365,
+      "learning_rate": 4.575923392612859e-05,
+      "loss": 3.2647,
+      "step": 992
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4828346967697144,
+      "learning_rate": 4.57421340629275e-05,
+      "loss": 3.0613,
+      "step": 996
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.515553593635559,
+      "learning_rate": 4.5725034199726405e-05,
+      "loss": 3.0368,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.498844861984253,
+      "learning_rate": 4.570793433652531e-05,
+      "loss": 3.102,
+      "step": 1004
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4339860677719116,
+      "learning_rate": 4.569083447332422e-05,
+      "loss": 3.1874,
+      "step": 1008
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5611178874969482,
+      "learning_rate": 4.567373461012312e-05,
+      "loss": 3.075,
+      "step": 1012
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6089402437210083,
+      "learning_rate": 4.565663474692203e-05,
+      "loss": 3.1544,
+      "step": 1016
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8874709606170654,
+      "learning_rate": 4.563953488372093e-05,
+      "loss": 3.2806,
+      "step": 1020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5624805688858032,
+      "learning_rate": 4.5622435020519835e-05,
+      "loss": 3.1442,
+      "step": 1024
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.968634009361267,
+      "learning_rate": 4.560533515731875e-05,
+      "loss": 3.2743,
+      "step": 1028
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4064279794692993,
+      "learning_rate": 4.558823529411765e-05,
+      "loss": 3.0726,
+      "step": 1032
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.577048897743225,
+      "learning_rate": 4.557113543091656e-05,
+      "loss": 3.1129,
+      "step": 1036
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.548323392868042,
+      "learning_rate": 4.555403556771546e-05,
+      "loss": 3.144,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5542961359024048,
+      "learning_rate": 4.5536935704514365e-05,
+      "loss": 3.182,
+      "step": 1044
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5506356954574585,
+      "learning_rate": 4.551983584131327e-05,
+      "loss": 3.2186,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6962274312973022,
+      "learning_rate": 4.550273597811218e-05,
+      "loss": 3.0385,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4382833242416382,
+      "learning_rate": 4.548563611491109e-05,
+      "loss": 3.1268,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.269392728805542,
+      "learning_rate": 4.546853625170999e-05,
+      "loss": 3.2579,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5414201021194458,
+      "learning_rate": 4.5451436388508895e-05,
+      "loss": 3.1501,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.936946153640747,
+      "learning_rate": 4.54343365253078e-05,
+      "loss": 3.1711,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7430529594421387,
+      "learning_rate": 4.54172366621067e-05,
+      "loss": 3.1299,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3500404357910156,
+      "learning_rate": 4.540013679890561e-05,
+      "loss": 3.0935,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5768132209777832,
+      "learning_rate": 4.538303693570452e-05,
+      "loss": 3.1091,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4829493761062622,
+      "learning_rate": 4.5365937072503425e-05,
+      "loss": 3.0862,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5560483932495117,
+      "learning_rate": 4.5348837209302326e-05,
+      "loss": 3.1024,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6295199394226074,
+      "learning_rate": 4.533173734610123e-05,
+      "loss": 2.9539,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1724135875701904,
+      "learning_rate": 4.531463748290014e-05,
+      "loss": 3.1936,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7400479316711426,
+      "learning_rate": 4.529753761969904e-05,
+      "loss": 3.1873,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.554962396621704,
+      "learning_rate": 4.528043775649795e-05,
+      "loss": 3.0643,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3722119331359863,
+      "learning_rate": 4.5263337893296855e-05,
+      "loss": 3.0088,
+      "step": 1108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.395577311515808,
+      "learning_rate": 4.524623803009576e-05,
+      "loss": 3.0584,
+      "step": 1112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.509710431098938,
+      "learning_rate": 4.522913816689467e-05,
+      "loss": 3.2276,
+      "step": 1116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5213786363601685,
+      "learning_rate": 4.521203830369357e-05,
+      "loss": 2.9433,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.51616632938385,
+      "learning_rate": 4.519493844049248e-05,
+      "loss": 2.9902,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5429805517196655,
+      "learning_rate": 4.517783857729138e-05,
+      "loss": 3.0884,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4714343547821045,
+      "learning_rate": 4.5160738714090286e-05,
+      "loss": 3.0322,
+      "step": 1132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4637647867202759,
+      "learning_rate": 4.51436388508892e-05,
+      "loss": 3.0352,
+      "step": 1136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3424347639083862,
+      "learning_rate": 4.51265389876881e-05,
+      "loss": 3.0866,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.689968228340149,
+      "learning_rate": 4.510943912448701e-05,
+      "loss": 3.0002,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9935293197631836,
+      "learning_rate": 4.509233926128591e-05,
+      "loss": 2.9518,
+      "step": 1148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5461673736572266,
+      "learning_rate": 4.5075239398084816e-05,
+      "loss": 3.0394,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5787737369537354,
+      "learning_rate": 4.505813953488372e-05,
+      "loss": 3.0934,
+      "step": 1156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4527121782302856,
+      "learning_rate": 4.504103967168263e-05,
+      "loss": 2.9734,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6236604452133179,
+      "learning_rate": 4.502393980848154e-05,
+      "loss": 3.021,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5616329908370972,
+      "learning_rate": 4.500683994528044e-05,
+      "loss": 3.0602,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6745967864990234,
+      "learning_rate": 4.4989740082079346e-05,
+      "loss": 3.2712,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5813934803009033,
+      "learning_rate": 4.497264021887825e-05,
+      "loss": 3.2255,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5584092140197754,
+      "learning_rate": 4.4955540355677154e-05,
+      "loss": 3.082,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5441848039627075,
+      "learning_rate": 4.493844049247606e-05,
+      "loss": 2.978,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6514015197753906,
+      "learning_rate": 4.492134062927497e-05,
+      "loss": 3.2604,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4784066677093506,
+      "learning_rate": 4.4904240766073876e-05,
+      "loss": 3.0761,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.493974208831787,
+      "learning_rate": 4.488714090287278e-05,
+      "loss": 3.1323,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4246760606765747,
+      "learning_rate": 4.4870041039671684e-05,
+      "loss": 2.7518,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6689587831497192,
+      "learning_rate": 4.485294117647059e-05,
+      "loss": 3.2381,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.615942120552063,
+      "learning_rate": 4.483584131326949e-05,
+      "loss": 3.0886,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3292858600616455,
+      "learning_rate": 4.48187414500684e-05,
+      "loss": 3.0968,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7597836256027222,
+      "learning_rate": 4.4801641586867306e-05,
+      "loss": 3.0037,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7249919176101685,
+      "learning_rate": 4.4784541723666214e-05,
+      "loss": 2.9228,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.630422830581665,
+      "learning_rate": 4.476744186046512e-05,
+      "loss": 3.0722,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4283276796340942,
+      "learning_rate": 4.475034199726402e-05,
+      "loss": 2.8726,
+      "step": 1228
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.786851167678833,
+      "learning_rate": 4.473324213406293e-05,
+      "loss": 3.1425,
+      "step": 1232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.669625997543335,
+      "learning_rate": 4.471614227086183e-05,
+      "loss": 3.013,
+      "step": 1236
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4979221820831299,
+      "learning_rate": 4.4699042407660744e-05,
+      "loss": 3.106,
+      "step": 1240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.67727792263031,
+      "learning_rate": 4.468194254445965e-05,
+      "loss": 2.9748,
+      "step": 1244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3726154565811157,
+      "learning_rate": 4.466484268125855e-05,
+      "loss": 3.0766,
+      "step": 1248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4735255241394043,
+      "learning_rate": 4.464774281805746e-05,
+      "loss": 2.9678,
+      "step": 1252
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.751224160194397,
+      "learning_rate": 4.463064295485636e-05,
+      "loss": 2.923,
+      "step": 1256
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4682483673095703,
+      "learning_rate": 4.461354309165527e-05,
+      "loss": 2.9822,
+      "step": 1260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7135084867477417,
+      "learning_rate": 4.4596443228454174e-05,
+      "loss": 2.9087,
+      "step": 1264
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6292390823364258,
+      "learning_rate": 4.457934336525308e-05,
+      "loss": 2.9948,
+      "step": 1268
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3494658470153809,
+      "learning_rate": 4.456224350205199e-05,
+      "loss": 2.8939,
+      "step": 1272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.584025263786316,
+      "learning_rate": 4.454514363885089e-05,
+      "loss": 2.9289,
+      "step": 1276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7876780033111572,
+      "learning_rate": 4.45280437756498e-05,
+      "loss": 3.0326,
+      "step": 1280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6452478170394897,
+      "learning_rate": 4.4510943912448704e-05,
+      "loss": 3.2085,
+      "step": 1284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.733777642250061,
+      "learning_rate": 4.4493844049247605e-05,
+      "loss": 2.8625,
+      "step": 1288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5891138315200806,
+      "learning_rate": 4.447674418604651e-05,
+      "loss": 2.8191,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5481853485107422,
+      "learning_rate": 4.445964432284542e-05,
+      "loss": 2.9586,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4756824970245361,
+      "learning_rate": 4.444254445964433e-05,
+      "loss": 2.9225,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4950698614120483,
+      "learning_rate": 4.4425444596443234e-05,
+      "loss": 2.8898,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.938930869102478,
+      "learning_rate": 4.4408344733242135e-05,
+      "loss": 3.0966,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.578041672706604,
+      "learning_rate": 4.439124487004104e-05,
+      "loss": 2.9547,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5932279825210571,
+      "learning_rate": 4.437414500683994e-05,
+      "loss": 3.0289,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5020122528076172,
+      "learning_rate": 4.435704514363886e-05,
+      "loss": 3.0404,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6642231941223145,
+      "learning_rate": 4.4339945280437764e-05,
+      "loss": 3.0883,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6281254291534424,
+      "learning_rate": 4.4322845417236665e-05,
+      "loss": 2.8135,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4702434539794922,
+      "learning_rate": 4.430574555403557e-05,
+      "loss": 2.8406,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.073446750640869,
+      "learning_rate": 4.428864569083447e-05,
+      "loss": 3.0211,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.543094515800476,
+      "learning_rate": 4.427154582763338e-05,
+      "loss": 2.8459,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5158188343048096,
+      "learning_rate": 4.425444596443229e-05,
+      "loss": 2.9384,
+      "step": 1344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6903549432754517,
+      "learning_rate": 4.4237346101231194e-05,
+      "loss": 3.0647,
+      "step": 1348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6733980178833008,
+      "learning_rate": 4.42202462380301e-05,
+      "loss": 2.8504,
+      "step": 1352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7396080493927002,
+      "learning_rate": 4.4203146374829e-05,
+      "loss": 3.1324,
+      "step": 1356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6228291988372803,
+      "learning_rate": 4.418604651162791e-05,
+      "loss": 2.9918,
+      "step": 1360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5741888284683228,
+      "learning_rate": 4.416894664842681e-05,
+      "loss": 3.0242,
+      "step": 1364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4788377285003662,
+      "learning_rate": 4.415184678522572e-05,
+      "loss": 2.7948,
+      "step": 1368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4203468561172485,
+      "learning_rate": 4.4134746922024625e-05,
+      "loss": 3.0096,
+      "step": 1372
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4492088556289673,
+      "learning_rate": 4.411764705882353e-05,
+      "loss": 2.9375,
+      "step": 1376
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4139641523361206,
+      "learning_rate": 4.410054719562244e-05,
+      "loss": 2.7041,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5051127672195435,
+      "learning_rate": 4.408344733242134e-05,
+      "loss": 2.9992,
+      "step": 1384
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6225379705429077,
+      "learning_rate": 4.406634746922025e-05,
+      "loss": 2.9139,
+      "step": 1388
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4177663326263428,
+      "learning_rate": 4.4049247606019155e-05,
+      "loss": 3.1328,
+      "step": 1392
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.342938780784607,
+      "learning_rate": 4.4032147742818055e-05,
+      "loss": 2.7651,
+      "step": 1396
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5941474437713623,
+      "learning_rate": 4.401504787961697e-05,
+      "loss": 3.0623,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9736019372940063,
+      "learning_rate": 4.399794801641587e-05,
+      "loss": 2.9231,
+      "step": 1404
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7911643981933594,
+      "learning_rate": 4.398084815321478e-05,
+      "loss": 2.9161,
+      "step": 1408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.585992693901062,
+      "learning_rate": 4.3963748290013685e-05,
+      "loss": 3.0079,
+      "step": 1412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.537345290184021,
+      "learning_rate": 4.3946648426812585e-05,
+      "loss": 2.9238,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5920599699020386,
+      "learning_rate": 4.392954856361149e-05,
+      "loss": 2.8771,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.606899619102478,
+      "learning_rate": 4.39124487004104e-05,
+      "loss": 2.7501,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5318756103515625,
+      "learning_rate": 4.389534883720931e-05,
+      "loss": 2.8197,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.987911581993103,
+      "learning_rate": 4.3878248974008215e-05,
+      "loss": 2.971,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5010288953781128,
+      "learning_rate": 4.3861149110807115e-05,
+      "loss": 2.9632,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.488486647605896,
+      "learning_rate": 4.384404924760602e-05,
+      "loss": 3.0563,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7354810237884521,
+      "learning_rate": 4.382694938440492e-05,
+      "loss": 2.866,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.509134292602539,
+      "learning_rate": 4.380984952120383e-05,
+      "loss": 2.8723,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.687049150466919,
+      "learning_rate": 4.379274965800274e-05,
+      "loss": 2.9496,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5449968576431274,
+      "learning_rate": 4.3775649794801645e-05,
+      "loss": 2.884,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7197214365005493,
+      "learning_rate": 4.375854993160055e-05,
+      "loss": 2.8686,
+      "step": 1460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3042534589767456,
+      "learning_rate": 4.374145006839945e-05,
+      "loss": 2.739,
+      "step": 1464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4646921157836914,
+      "learning_rate": 4.372435020519836e-05,
+      "loss": 2.8347,
+      "step": 1468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.350310206413269,
+      "learning_rate": 4.370725034199726e-05,
+      "loss": 2.8797,
+      "step": 1472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5870814323425293,
+      "learning_rate": 4.369015047879617e-05,
+      "loss": 3.025,
+      "step": 1476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5572704076766968,
+      "learning_rate": 4.3673050615595076e-05,
+      "loss": 2.8424,
+      "step": 1480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4381687641143799,
+      "learning_rate": 4.365595075239398e-05,
+      "loss": 3.0395,
+      "step": 1484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7877554893493652,
+      "learning_rate": 4.363885088919289e-05,
+      "loss": 2.8723,
+      "step": 1488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7916138172149658,
+      "learning_rate": 4.362175102599179e-05,
+      "loss": 2.8751,
+      "step": 1492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8024954795837402,
+      "learning_rate": 4.36046511627907e-05,
+      "loss": 3.0156,
+      "step": 1496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4376300573349,
+      "learning_rate": 4.3587551299589606e-05,
+      "loss": 2.9152,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5347250699996948,
+      "learning_rate": 4.3570451436388506e-05,
+      "loss": 2.8219,
+      "step": 1504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7805489301681519,
+      "learning_rate": 4.355335157318742e-05,
+      "loss": 2.9276,
+      "step": 1508
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5822547674179077,
+      "learning_rate": 4.353625170998632e-05,
+      "loss": 2.9262,
+      "step": 1512
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5745974779129028,
+      "learning_rate": 4.351915184678523e-05,
+      "loss": 2.8812,
+      "step": 1516
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5027135610580444,
+      "learning_rate": 4.3502051983584136e-05,
+      "loss": 2.7748,
+      "step": 1520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5027369260787964,
+      "learning_rate": 4.3484952120383036e-05,
+      "loss": 2.8619,
+      "step": 1524
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6836864948272705,
+      "learning_rate": 4.3467852257181944e-05,
+      "loss": 2.9688,
+      "step": 1528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.682340145111084,
+      "learning_rate": 4.345075239398085e-05,
+      "loss": 2.9849,
+      "step": 1532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3924282789230347,
+      "learning_rate": 4.343365253077976e-05,
+      "loss": 2.8154,
+      "step": 1536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.518629789352417,
+      "learning_rate": 4.3416552667578666e-05,
+      "loss": 2.8265,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.482877254486084,
+      "learning_rate": 4.3399452804377566e-05,
+      "loss": 2.8364,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.52615487575531,
+      "learning_rate": 4.3382352941176474e-05,
+      "loss": 2.9488,
+      "step": 1548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8858604431152344,
+      "learning_rate": 4.3365253077975374e-05,
+      "loss": 3.0754,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3640977144241333,
+      "learning_rate": 4.334815321477428e-05,
+      "loss": 2.7192,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6419317722320557,
+      "learning_rate": 4.333105335157319e-05,
+      "loss": 2.8016,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4059332609176636,
+      "learning_rate": 4.3313953488372096e-05,
+      "loss": 2.8559,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5837764739990234,
+      "learning_rate": 4.3296853625171004e-05,
+      "loss": 3.0297,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8545840978622437,
+      "learning_rate": 4.3279753761969904e-05,
+      "loss": 2.9717,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2715051174163818,
+      "learning_rate": 4.326265389876881e-05,
+      "loss": 2.7563,
+      "step": 1576
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.742698073387146,
+      "learning_rate": 4.324555403556772e-05,
+      "loss": 2.7826,
+      "step": 1580
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7338377237319946,
+      "learning_rate": 4.322845417236662e-05,
+      "loss": 2.8006,
+      "step": 1584
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5166951417922974,
+      "learning_rate": 4.3211354309165533e-05,
+      "loss": 3.0044,
+      "step": 1588
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5267099142074585,
+      "learning_rate": 4.3194254445964434e-05,
+      "loss": 2.817,
+      "step": 1592
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7478032112121582,
+      "learning_rate": 4.317715458276334e-05,
+      "loss": 3.0036,
+      "step": 1596
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9281939268112183,
+      "learning_rate": 4.316005471956224e-05,
+      "loss": 2.9689,
+      "step": 1600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5747674703598022,
+      "learning_rate": 4.314295485636115e-05,
+      "loss": 2.7691,
+      "step": 1604
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.530669093132019,
+      "learning_rate": 4.312585499316006e-05,
+      "loss": 2.7754,
+      "step": 1608
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4007419347763062,
+      "learning_rate": 4.3108755129958964e-05,
+      "loss": 2.6687,
+      "step": 1612
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7224704027175903,
+      "learning_rate": 4.309165526675787e-05,
+      "loss": 3.0238,
+      "step": 1616
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.688553810119629,
+      "learning_rate": 4.307455540355677e-05,
+      "loss": 2.7985,
+      "step": 1620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3165500164031982,
+      "learning_rate": 4.305745554035568e-05,
+      "loss": 2.8546,
+      "step": 1624
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4722650051116943,
+      "learning_rate": 4.3040355677154587e-05,
+      "loss": 2.8455,
+      "step": 1628
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5830330848693848,
+      "learning_rate": 4.302325581395349e-05,
+      "loss": 2.709,
+      "step": 1632
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5168038606643677,
+      "learning_rate": 4.3006155950752394e-05,
+      "loss": 2.8194,
+      "step": 1636
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.527031660079956,
+      "learning_rate": 4.29890560875513e-05,
+      "loss": 2.7123,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.492450475692749,
+      "learning_rate": 4.297195622435021e-05,
+      "loss": 2.9356,
+      "step": 1644
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4466654062271118,
+      "learning_rate": 4.2954856361149116e-05,
+      "loss": 2.8938,
+      "step": 1648
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6020383834838867,
+      "learning_rate": 4.293775649794802e-05,
+      "loss": 2.8547,
+      "step": 1652
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5876679420471191,
+      "learning_rate": 4.2920656634746924e-05,
+      "loss": 2.7769,
+      "step": 1656
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4787455797195435,
+      "learning_rate": 4.2903556771545825e-05,
+      "loss": 2.6641,
+      "step": 1660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6105437278747559,
+      "learning_rate": 4.288645690834473e-05,
+      "loss": 2.775,
+      "step": 1664
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5835809707641602,
+      "learning_rate": 4.2869357045143646e-05,
+      "loss": 3.019,
+      "step": 1668
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.103267192840576,
+      "learning_rate": 4.285225718194255e-05,
+      "loss": 2.828,
+      "step": 1672
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5641860961914062,
+      "learning_rate": 4.2835157318741454e-05,
+      "loss": 2.8392,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.475889801979065,
+      "learning_rate": 4.2818057455540355e-05,
+      "loss": 2.9558,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3875279426574707,
+      "learning_rate": 4.280095759233926e-05,
+      "loss": 2.7178,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6096158027648926,
+      "learning_rate": 4.278385772913817e-05,
+      "loss": 2.7963,
+      "step": 1688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4312164783477783,
+      "learning_rate": 4.276675786593708e-05,
+      "loss": 2.7887,
+      "step": 1692
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0804941654205322,
+      "learning_rate": 4.2749658002735984e-05,
+      "loss": 2.9118,
+      "step": 1696
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.606184482574463,
+      "learning_rate": 4.2732558139534885e-05,
+      "loss": 2.7951,
+      "step": 1700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4609088897705078,
+      "learning_rate": 4.271545827633379e-05,
+      "loss": 2.8464,
+      "step": 1704
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3830453157424927,
+      "learning_rate": 4.26983584131327e-05,
+      "loss": 2.9156,
+      "step": 1708
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4420268535614014,
+      "learning_rate": 4.26812585499316e-05,
+      "loss": 2.5969,
+      "step": 1712
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4451963901519775,
+      "learning_rate": 4.266415868673051e-05,
+      "loss": 2.9125,
+      "step": 1716
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5134412050247192,
+      "learning_rate": 4.2647058823529415e-05,
+      "loss": 2.6605,
+      "step": 1720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.561641812324524,
+      "learning_rate": 4.262995896032832e-05,
+      "loss": 2.8285,
+      "step": 1724
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4210346937179565,
+      "learning_rate": 4.261285909712722e-05,
+      "loss": 2.7912,
+      "step": 1728
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4329392910003662,
+      "learning_rate": 4.259575923392613e-05,
+      "loss": 2.8862,
+      "step": 1732
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6114985942840576,
+      "learning_rate": 4.257865937072504e-05,
+      "loss": 2.7819,
+      "step": 1736
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6259700059890747,
+      "learning_rate": 4.256155950752394e-05,
+      "loss": 2.9296,
+      "step": 1740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.523130178451538,
+      "learning_rate": 4.2544459644322845e-05,
+      "loss": 2.7139,
+      "step": 1744
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7980687618255615,
+      "learning_rate": 4.252735978112175e-05,
+      "loss": 2.7185,
+      "step": 1748
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.73753821849823,
+      "learning_rate": 4.251025991792066e-05,
+      "loss": 2.7322,
+      "step": 1752
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7766021490097046,
+      "learning_rate": 4.249316005471957e-05,
+      "loss": 2.5836,
+      "step": 1756
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4891935586929321,
+      "learning_rate": 4.247606019151847e-05,
+      "loss": 2.931,
+      "step": 1760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4233758449554443,
+      "learning_rate": 4.2458960328317375e-05,
+      "loss": 2.9084,
+      "step": 1764
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5510305166244507,
+      "learning_rate": 4.2441860465116276e-05,
+      "loss": 2.8017,
+      "step": 1768
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3143435716629028,
+      "learning_rate": 4.242476060191519e-05,
+      "loss": 2.7212,
+      "step": 1772
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5766334533691406,
+      "learning_rate": 4.24076607387141e-05,
+      "loss": 2.7551,
+      "step": 1776
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.551622986793518,
+      "learning_rate": 4.2390560875513e-05,
+      "loss": 2.8498,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2195143699645996,
+      "learning_rate": 4.2373461012311905e-05,
+      "loss": 2.811,
+      "step": 1784
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.530758023262024,
+      "learning_rate": 4.2356361149110806e-05,
+      "loss": 2.772,
+      "step": 1788
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5073307752609253,
+      "learning_rate": 4.233926128590971e-05,
+      "loss": 2.7401,
+      "step": 1792
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6950387954711914,
+      "learning_rate": 4.232216142270862e-05,
+      "loss": 2.7554,
+      "step": 1796
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.389538288116455,
+      "learning_rate": 4.230506155950753e-05,
+      "loss": 2.8187,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.592817783355713,
+      "learning_rate": 4.2287961696306435e-05,
+      "loss": 2.8686,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6047254800796509,
+      "learning_rate": 4.2270861833105336e-05,
+      "loss": 2.8033,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4730092287063599,
+      "learning_rate": 4.225376196990424e-05,
+      "loss": 2.6616,
+      "step": 1812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4784797430038452,
+      "learning_rate": 4.223666210670315e-05,
+      "loss": 2.8108,
+      "step": 1816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6881104707717896,
+      "learning_rate": 4.221956224350205e-05,
+      "loss": 2.8155,
+      "step": 1820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4548227787017822,
+      "learning_rate": 4.220246238030096e-05,
+      "loss": 2.7195,
+      "step": 1824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5578104257583618,
+      "learning_rate": 4.2185362517099866e-05,
+      "loss": 2.8058,
+      "step": 1828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6127967834472656,
+      "learning_rate": 4.216826265389877e-05,
+      "loss": 2.8783,
+      "step": 1832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.883389949798584,
+      "learning_rate": 4.215116279069768e-05,
+      "loss": 2.6328,
+      "step": 1836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6216742992401123,
+      "learning_rate": 4.213406292749658e-05,
+      "loss": 2.8021,
+      "step": 1840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7198604345321655,
+      "learning_rate": 4.211696306429549e-05,
+      "loss": 2.6467,
+      "step": 1844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3806415796279907,
+      "learning_rate": 4.209986320109439e-05,
+      "loss": 2.8337,
+      "step": 1848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9981590509414673,
+      "learning_rate": 4.2082763337893296e-05,
+      "loss": 2.7637,
+      "step": 1852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4233486652374268,
+      "learning_rate": 4.2065663474692204e-05,
+      "loss": 2.6964,
+      "step": 1856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4599206447601318,
+      "learning_rate": 4.204856361149111e-05,
+      "loss": 2.8918,
+      "step": 1860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6856427192687988,
+      "learning_rate": 4.203146374829002e-05,
+      "loss": 2.7184,
+      "step": 1864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5229926109313965,
+      "learning_rate": 4.201436388508892e-05,
+      "loss": 2.7457,
+      "step": 1868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8297462463378906,
+      "learning_rate": 4.1997264021887826e-05,
+      "loss": 2.8097,
+      "step": 1872
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4381957054138184,
+      "learning_rate": 4.198016415868673e-05,
+      "loss": 2.6845,
+      "step": 1876
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8839136362075806,
+      "learning_rate": 4.196306429548564e-05,
+      "loss": 2.7721,
+      "step": 1880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6147316694259644,
+      "learning_rate": 4.194596443228455e-05,
+      "loss": 2.7529,
+      "step": 1884
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4473137855529785,
+      "learning_rate": 4.192886456908345e-05,
+      "loss": 2.7582,
+      "step": 1888
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6229139566421509,
+      "learning_rate": 4.1911764705882356e-05,
+      "loss": 2.7579,
+      "step": 1892
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.669582724571228,
+      "learning_rate": 4.189466484268126e-05,
+      "loss": 2.7924,
+      "step": 1896
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4859235286712646,
+      "learning_rate": 4.1877564979480164e-05,
+      "loss": 2.6415,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4661961793899536,
+      "learning_rate": 4.186046511627907e-05,
+      "loss": 2.5649,
+      "step": 1904
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4079793691635132,
+      "learning_rate": 4.184336525307798e-05,
+      "loss": 2.6718,
+      "step": 1908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7637715339660645,
+      "learning_rate": 4.1826265389876886e-05,
+      "loss": 2.7626,
+      "step": 1912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4910571575164795,
+      "learning_rate": 4.1809165526675787e-05,
+      "loss": 2.683,
+      "step": 1916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.709197759628296,
+      "learning_rate": 4.1792065663474694e-05,
+      "loss": 2.8062,
+      "step": 1920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8040099143981934,
+      "learning_rate": 4.17749658002736e-05,
+      "loss": 2.8158,
+      "step": 1924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4271622896194458,
+      "learning_rate": 4.17578659370725e-05,
+      "loss": 2.8655,
+      "step": 1928
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6832592487335205,
+      "learning_rate": 4.174076607387141e-05,
+      "loss": 2.9054,
+      "step": 1932
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.674769639968872,
+      "learning_rate": 4.1723666210670316e-05,
+      "loss": 2.7739,
+      "step": 1936
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3221181631088257,
+      "learning_rate": 4.1706566347469224e-05,
+      "loss": 2.7352,
+      "step": 1940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6557152271270752,
+      "learning_rate": 4.168946648426813e-05,
+      "loss": 2.7729,
+      "step": 1944
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5999726057052612,
+      "learning_rate": 4.167236662106703e-05,
+      "loss": 2.8399,
+      "step": 1948
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5852841138839722,
+      "learning_rate": 4.165526675786594e-05,
+      "loss": 2.8804,
+      "step": 1952
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.634738802909851,
+      "learning_rate": 4.163816689466484e-05,
+      "loss": 2.6103,
+      "step": 1956
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5783636569976807,
+      "learning_rate": 4.1621067031463754e-05,
+      "loss": 2.4903,
+      "step": 1960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4169994592666626,
+      "learning_rate": 4.160396716826266e-05,
+      "loss": 2.6708,
+      "step": 1964
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4021108150482178,
+      "learning_rate": 4.158686730506156e-05,
+      "loss": 2.6697,
+      "step": 1968
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4128559827804565,
+      "learning_rate": 4.156976744186047e-05,
+      "loss": 2.7368,
+      "step": 1972
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5087906122207642,
+      "learning_rate": 4.155266757865937e-05,
+      "loss": 2.7219,
+      "step": 1976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.4628214836120605,
+      "learning_rate": 4.153556771545828e-05,
+      "loss": 2.7273,
+      "step": 1980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5917600393295288,
+      "learning_rate": 4.1518467852257184e-05,
+      "loss": 2.885,
+      "step": 1984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4118601083755493,
+      "learning_rate": 4.150136798905609e-05,
+      "loss": 2.7849,
+      "step": 1988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4783868789672852,
+      "learning_rate": 4.1484268125855e-05,
+      "loss": 2.8531,
+      "step": 1992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6047073602676392,
+      "learning_rate": 4.14671682626539e-05,
+      "loss": 2.784,
+      "step": 1996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.565877079963684,
+      "learning_rate": 4.145006839945281e-05,
+      "loss": 2.6856,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4095959663391113,
+      "learning_rate": 4.143296853625171e-05,
+      "loss": 2.7933,
+      "step": 2004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3851944208145142,
+      "learning_rate": 4.1415868673050615e-05,
+      "loss": 2.7286,
+      "step": 2008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4448835849761963,
+      "learning_rate": 4.139876880984952e-05,
+      "loss": 2.667,
+      "step": 2012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2041794061660767,
+      "learning_rate": 4.138166894664843e-05,
+      "loss": 2.553,
+      "step": 2016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5906238555908203,
+      "learning_rate": 4.136456908344734e-05,
+      "loss": 2.7194,
+      "step": 2020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.493734359741211,
+      "learning_rate": 4.134746922024624e-05,
+      "loss": 2.6597,
+      "step": 2024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.485025405883789,
+      "learning_rate": 4.1330369357045145e-05,
+      "loss": 2.7412,
+      "step": 2028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6668092012405396,
+      "learning_rate": 4.131326949384405e-05,
+      "loss": 2.7559,
+      "step": 2032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6972142457962036,
+      "learning_rate": 4.129616963064295e-05,
+      "loss": 2.6343,
+      "step": 2036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6248745918273926,
+      "learning_rate": 4.127906976744187e-05,
+      "loss": 2.749,
+      "step": 2040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.486807942390442,
+      "learning_rate": 4.126196990424077e-05,
+      "loss": 2.5962,
+      "step": 2044
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.558280110359192,
+      "learning_rate": 4.1244870041039675e-05,
+      "loss": 2.7067,
+      "step": 2048
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.78428053855896,
+      "learning_rate": 4.122777017783858e-05,
+      "loss": 2.7901,
+      "step": 2052
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5545462369918823,
+      "learning_rate": 4.121067031463748e-05,
+      "loss": 2.7868,
+      "step": 2056
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5399055480957031,
+      "learning_rate": 4.119357045143639e-05,
+      "loss": 2.8405,
+      "step": 2060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.535919189453125,
+      "learning_rate": 4.11764705882353e-05,
+      "loss": 2.6297,
+      "step": 2064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3103563785552979,
+      "learning_rate": 4.1159370725034205e-05,
+      "loss": 2.6334,
+      "step": 2068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4171690940856934,
+      "learning_rate": 4.114227086183311e-05,
+      "loss": 2.5261,
+      "step": 2072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5107210874557495,
+      "learning_rate": 4.112517099863201e-05,
+      "loss": 2.5669,
+      "step": 2076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4645127058029175,
+      "learning_rate": 4.110807113543092e-05,
+      "loss": 2.6166,
+      "step": 2080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8557233810424805,
+      "learning_rate": 4.109097127222982e-05,
+      "loss": 2.6651,
+      "step": 2084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4747710227966309,
+      "learning_rate": 4.107387140902873e-05,
+      "loss": 2.9066,
+      "step": 2088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4575109481811523,
+      "learning_rate": 4.1056771545827635e-05,
+      "loss": 2.7079,
+      "step": 2092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5356212854385376,
+      "learning_rate": 4.103967168262654e-05,
+      "loss": 2.711,
+      "step": 2096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4428212642669678,
+      "learning_rate": 4.102257181942545e-05,
+      "loss": 2.652,
+      "step": 2100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6778663396835327,
+      "learning_rate": 4.100547195622435e-05,
+      "loss": 2.7872,
+      "step": 2104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4642986059188843,
+      "learning_rate": 4.098837209302326e-05,
+      "loss": 2.587,
+      "step": 2108
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5550886392593384,
+      "learning_rate": 4.097127222982216e-05,
+      "loss": 2.6646,
+      "step": 2112
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6808562278747559,
+      "learning_rate": 4.0954172366621066e-05,
+      "loss": 2.588,
+      "step": 2116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7080578804016113,
+      "learning_rate": 4.093707250341998e-05,
+      "loss": 2.6144,
+      "step": 2120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8413456678390503,
+      "learning_rate": 4.091997264021888e-05,
+      "loss": 2.8098,
+      "step": 2124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4448609352111816,
+      "learning_rate": 4.090287277701779e-05,
+      "loss": 2.7619,
+      "step": 2128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5994160175323486,
+      "learning_rate": 4.088577291381669e-05,
+      "loss": 2.6444,
+      "step": 2132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5190199613571167,
+      "learning_rate": 4.0868673050615596e-05,
+      "loss": 2.5597,
+      "step": 2136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5521091222763062,
+      "learning_rate": 4.08515731874145e-05,
+      "loss": 2.6333,
+      "step": 2140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6080738306045532,
+      "learning_rate": 4.083447332421341e-05,
+      "loss": 2.5827,
+      "step": 2144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0923423767089844,
+      "learning_rate": 4.081737346101232e-05,
+      "loss": 2.6463,
+      "step": 2148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4243260622024536,
+      "learning_rate": 4.080027359781122e-05,
+      "loss": 2.6702,
+      "step": 2152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5463379621505737,
+      "learning_rate": 4.0783173734610126e-05,
+      "loss": 2.6284,
+      "step": 2156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4061999320983887,
+      "learning_rate": 4.076607387140903e-05,
+      "loss": 2.4983,
+      "step": 2160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3713377714157104,
+      "learning_rate": 4.0748974008207933e-05,
+      "loss": 2.5406,
+      "step": 2164
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6995548009872437,
+      "learning_rate": 4.073187414500684e-05,
+      "loss": 2.5039,
+      "step": 2168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.644362449645996,
+      "learning_rate": 4.071477428180575e-05,
+      "loss": 2.7398,
+      "step": 2172
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5956717729568481,
+      "learning_rate": 4.0697674418604655e-05,
+      "loss": 2.7333,
+      "step": 2176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.567484736442566,
+      "learning_rate": 4.068057455540356e-05,
+      "loss": 2.727,
+      "step": 2180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.573633074760437,
+      "learning_rate": 4.066347469220246e-05,
+      "loss": 2.639,
+      "step": 2184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5799304246902466,
+      "learning_rate": 4.064637482900137e-05,
+      "loss": 2.735,
+      "step": 2188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5772294998168945,
+      "learning_rate": 4.062927496580027e-05,
+      "loss": 2.5753,
+      "step": 2192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4983608722686768,
+      "learning_rate": 4.061217510259918e-05,
+      "loss": 2.5989,
+      "step": 2196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.625112533569336,
+      "learning_rate": 4.0595075239398086e-05,
+      "loss": 2.6539,
+      "step": 2200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4899359941482544,
+      "learning_rate": 4.057797537619699e-05,
+      "loss": 2.7391,
+      "step": 2204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4032224416732788,
+      "learning_rate": 4.05608755129959e-05,
+      "loss": 2.4294,
+      "step": 2208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.379164218902588,
+      "learning_rate": 4.05437756497948e-05,
+      "loss": 2.8045,
+      "step": 2212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4953522682189941,
+      "learning_rate": 4.052667578659371e-05,
+      "loss": 2.6627,
+      "step": 2216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3828809261322021,
+      "learning_rate": 4.0509575923392616e-05,
+      "loss": 2.5951,
+      "step": 2220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5651103258132935,
+      "learning_rate": 4.0492476060191516e-05,
+      "loss": 2.7003,
+      "step": 2224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7900563478469849,
+      "learning_rate": 4.047537619699043e-05,
+      "loss": 2.5935,
+      "step": 2228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9265758991241455,
+      "learning_rate": 4.045827633378933e-05,
+      "loss": 2.6538,
+      "step": 2232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4881079196929932,
+      "learning_rate": 4.044117647058824e-05,
+      "loss": 2.4857,
+      "step": 2236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4665502309799194,
+      "learning_rate": 4.042407660738714e-05,
+      "loss": 2.5874,
+      "step": 2240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5124859809875488,
+      "learning_rate": 4.0406976744186046e-05,
+      "loss": 2.8067,
+      "step": 2244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4011058807373047,
+      "learning_rate": 4.0389876880984954e-05,
+      "loss": 2.5625,
+      "step": 2248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.469796895980835,
+      "learning_rate": 4.037277701778386e-05,
+      "loss": 2.7695,
+      "step": 2252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.610304355621338,
+      "learning_rate": 4.035567715458277e-05,
+      "loss": 2.5637,
+      "step": 2256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3841075897216797,
+      "learning_rate": 4.033857729138167e-05,
+      "loss": 2.7228,
+      "step": 2260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6841098070144653,
+      "learning_rate": 4.0321477428180576e-05,
+      "loss": 2.6203,
+      "step": 2264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5729596614837646,
+      "learning_rate": 4.0304377564979484e-05,
+      "loss": 2.5021,
+      "step": 2268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6014264822006226,
+      "learning_rate": 4.0287277701778384e-05,
+      "loss": 2.6825,
+      "step": 2272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3496001958847046,
+      "learning_rate": 4.027017783857729e-05,
+      "loss": 2.5271,
+      "step": 2276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5231562852859497,
+      "learning_rate": 4.02530779753762e-05,
+      "loss": 2.6128,
+      "step": 2280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7021198272705078,
+      "learning_rate": 4.0235978112175106e-05,
+      "loss": 2.7152,
+      "step": 2284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4971519708633423,
+      "learning_rate": 4.0218878248974014e-05,
+      "loss": 2.6637,
+      "step": 2288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6225758790969849,
+      "learning_rate": 4.0201778385772914e-05,
+      "loss": 2.6871,
+      "step": 2292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5330243110656738,
+      "learning_rate": 4.018467852257182e-05,
+      "loss": 2.5721,
+      "step": 2296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4374375343322754,
+      "learning_rate": 4.016757865937072e-05,
+      "loss": 2.6393,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5932408571243286,
+      "learning_rate": 4.015047879616963e-05,
+      "loss": 2.5341,
+      "step": 2304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7230651378631592,
+      "learning_rate": 4.0133378932968544e-05,
+      "loss": 2.6912,
+      "step": 2308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5877001285552979,
+      "learning_rate": 4.0116279069767444e-05,
+      "loss": 2.6632,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6193994283676147,
+      "learning_rate": 4.009917920656635e-05,
+      "loss": 2.3792,
+      "step": 2316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5001564025878906,
+      "learning_rate": 4.008207934336525e-05,
+      "loss": 2.538,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5603808164596558,
+      "learning_rate": 4.006497948016416e-05,
+      "loss": 2.5206,
+      "step": 2324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4274383783340454,
+      "learning_rate": 4.004787961696307e-05,
+      "loss": 2.6341,
+      "step": 2328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5026893615722656,
+      "learning_rate": 4.0030779753761974e-05,
+      "loss": 2.5531,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.415287733078003,
+      "learning_rate": 4.001367989056088e-05,
+      "loss": 2.6344,
+      "step": 2336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.61545991897583,
+      "learning_rate": 3.999658002735978e-05,
+      "loss": 2.5476,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4016505479812622,
+      "learning_rate": 3.997948016415869e-05,
+      "loss": 2.562,
+      "step": 2344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.665184497833252,
+      "learning_rate": 3.99623803009576e-05,
+      "loss": 2.6033,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.562911868095398,
+      "learning_rate": 3.99452804377565e-05,
+      "loss": 2.6955,
+      "step": 2352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.477481484413147,
+      "learning_rate": 3.9928180574555405e-05,
+      "loss": 2.66,
+      "step": 2356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4621987342834473,
+      "learning_rate": 3.991108071135431e-05,
+      "loss": 2.4728,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3315041065216064,
+      "learning_rate": 3.989398084815322e-05,
+      "loss": 2.5248,
+      "step": 2364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.876422643661499,
+      "learning_rate": 3.987688098495212e-05,
+      "loss": 2.6436,
+      "step": 2368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.47256600856781,
+      "learning_rate": 3.985978112175103e-05,
+      "loss": 2.5254,
+      "step": 2372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6386239528656006,
+      "learning_rate": 3.9842681258549935e-05,
+      "loss": 2.6176,
+      "step": 2376
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3476667404174805,
+      "learning_rate": 3.9825581395348835e-05,
+      "loss": 2.4059,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.436629056930542,
+      "learning_rate": 3.980848153214774e-05,
+      "loss": 2.6304,
+      "step": 2384
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3741427659988403,
+      "learning_rate": 3.979138166894665e-05,
+      "loss": 2.5209,
+      "step": 2388
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5190575122833252,
+      "learning_rate": 3.977428180574556e-05,
+      "loss": 2.4463,
+      "step": 2392
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.389143943786621,
+      "learning_rate": 3.9757181942544465e-05,
+      "loss": 2.6484,
+      "step": 2396
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3748509883880615,
+      "learning_rate": 3.9740082079343365e-05,
+      "loss": 2.4275,
+      "step": 2400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7997660636901855,
+      "learning_rate": 3.972298221614227e-05,
+      "loss": 2.6384,
+      "step": 2404
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4528918266296387,
+      "learning_rate": 3.970588235294117e-05,
+      "loss": 2.6204,
+      "step": 2408
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.421633243560791,
+      "learning_rate": 3.968878248974009e-05,
+      "loss": 2.5918,
+      "step": 2412
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4433393478393555,
+      "learning_rate": 3.9671682626538994e-05,
+      "loss": 2.6482,
+      "step": 2416
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2582066059112549,
+      "learning_rate": 3.9654582763337895e-05,
+      "loss": 2.413,
+      "step": 2420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.394639253616333,
+      "learning_rate": 3.96374829001368e-05,
+      "loss": 2.4778,
+      "step": 2424
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.424068570137024,
+      "learning_rate": 3.96203830369357e-05,
+      "loss": 2.6462,
+      "step": 2428
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5364696979522705,
+      "learning_rate": 3.960328317373461e-05,
+      "loss": 2.5108,
+      "step": 2432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5929545164108276,
+      "learning_rate": 3.958618331053352e-05,
+      "loss": 2.5638,
+      "step": 2436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5119692087173462,
+      "learning_rate": 3.9569083447332425e-05,
+      "loss": 2.6631,
+      "step": 2440
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.573603630065918,
+      "learning_rate": 3.955198358413133e-05,
+      "loss": 2.7071,
+      "step": 2444
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.563513159751892,
+      "learning_rate": 3.953488372093023e-05,
+      "loss": 2.4158,
+      "step": 2448
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.572964072227478,
+      "learning_rate": 3.951778385772914e-05,
+      "loss": 2.6124,
+      "step": 2452
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4729305505752563,
+      "learning_rate": 3.950068399452805e-05,
+      "loss": 2.4546,
+      "step": 2456
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3578672409057617,
+      "learning_rate": 3.948358413132695e-05,
+      "loss": 2.5572,
+      "step": 2460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.456809401512146,
+      "learning_rate": 3.9466484268125855e-05,
+      "loss": 2.6155,
+      "step": 2464
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.417240023612976,
+      "learning_rate": 3.944938440492476e-05,
+      "loss": 2.6249,
+      "step": 2468
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6173032522201538,
+      "learning_rate": 3.943228454172367e-05,
+      "loss": 2.3838,
+      "step": 2472
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5841619968414307,
+      "learning_rate": 3.941518467852258e-05,
+      "loss": 2.5878,
+      "step": 2476
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.526047945022583,
+      "learning_rate": 3.939808481532148e-05,
+      "loss": 2.6055,
+      "step": 2480
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.375135064125061,
+      "learning_rate": 3.9380984952120385e-05,
+      "loss": 2.4982,
+      "step": 2484
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0735349655151367,
+      "learning_rate": 3.9363885088919286e-05,
+      "loss": 2.5964,
+      "step": 2488
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4464576244354248,
+      "learning_rate": 3.93467852257182e-05,
+      "loss": 2.5043,
+      "step": 2492
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4004943370819092,
+      "learning_rate": 3.93296853625171e-05,
+      "loss": 2.3337,
+      "step": 2496
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5199224948883057,
+      "learning_rate": 3.931258549931601e-05,
+      "loss": 2.5759,
+      "step": 2500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1466448307037354,
+      "learning_rate": 3.9295485636114915e-05,
+      "loss": 2.6014,
+      "step": 2504
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3880302906036377,
+      "learning_rate": 3.9278385772913816e-05,
+      "loss": 2.4023,
+      "step": 2508
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.501216173171997,
+      "learning_rate": 3.926128590971272e-05,
+      "loss": 2.6306,
+      "step": 2512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5279217958450317,
+      "learning_rate": 3.924418604651163e-05,
+      "loss": 2.5785,
+      "step": 2516
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4741731882095337,
+      "learning_rate": 3.922708618331054e-05,
+      "loss": 2.5174,
+      "step": 2520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6046087741851807,
+      "learning_rate": 3.9209986320109445e-05,
+      "loss": 2.6823,
+      "step": 2524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4944065809249878,
+      "learning_rate": 3.9192886456908346e-05,
+      "loss": 2.7186,
+      "step": 2528
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3260518312454224,
+      "learning_rate": 3.917578659370725e-05,
+      "loss": 2.3741,
+      "step": 2532
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.591991901397705,
+      "learning_rate": 3.9158686730506154e-05,
+      "loss": 2.5534,
+      "step": 2536
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5477372407913208,
+      "learning_rate": 3.914158686730506e-05,
+      "loss": 2.568,
+      "step": 2540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5402345657348633,
+      "learning_rate": 3.912448700410397e-05,
+      "loss": 2.5795,
+      "step": 2544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2363271713256836,
+      "learning_rate": 3.9107387140902876e-05,
+      "loss": 2.4248,
+      "step": 2548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2395968437194824,
+      "learning_rate": 3.909028727770178e-05,
+      "loss": 2.6678,
+      "step": 2552
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5662035942077637,
+      "learning_rate": 3.9073187414500684e-05,
+      "loss": 2.6065,
+      "step": 2556
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3480955362319946,
+      "learning_rate": 3.905608755129959e-05,
+      "loss": 2.5286,
+      "step": 2560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5644841194152832,
+      "learning_rate": 3.90389876880985e-05,
+      "loss": 2.439,
+      "step": 2564
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.556207537651062,
+      "learning_rate": 3.90218878248974e-05,
+      "loss": 2.5694,
+      "step": 2568
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4543012380599976,
+      "learning_rate": 3.9004787961696306e-05,
+      "loss": 2.5277,
+      "step": 2572
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2177674770355225,
+      "learning_rate": 3.8987688098495214e-05,
+      "loss": 2.5114,
+      "step": 2576
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5734175443649292,
+      "learning_rate": 3.897058823529412e-05,
+      "loss": 2.469,
+      "step": 2580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5571191310882568,
+      "learning_rate": 3.895348837209303e-05,
+      "loss": 2.5852,
+      "step": 2584
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5742443799972534,
+      "learning_rate": 3.893638850889193e-05,
+      "loss": 2.429,
+      "step": 2588
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4786051511764526,
+      "learning_rate": 3.8919288645690836e-05,
+      "loss": 2.6498,
+      "step": 2592
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5065261125564575,
+      "learning_rate": 3.890218878248974e-05,
+      "loss": 2.5307,
+      "step": 2596
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9272600412368774,
+      "learning_rate": 3.888508891928865e-05,
+      "loss": 2.5898,
+      "step": 2600
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5606516599655151,
+      "learning_rate": 3.886798905608756e-05,
+      "loss": 2.4504,
+      "step": 2604
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4898043870925903,
+      "learning_rate": 3.885088919288646e-05,
+      "loss": 2.4989,
+      "step": 2608
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.878469467163086,
+      "learning_rate": 3.8833789329685366e-05,
+      "loss": 2.6865,
+      "step": 2612
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8781071901321411,
+      "learning_rate": 3.881668946648427e-05,
+      "loss": 2.5997,
+      "step": 2616
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6376733779907227,
+      "learning_rate": 3.8799589603283174e-05,
+      "loss": 2.5791,
+      "step": 2620
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4639188051223755,
+      "learning_rate": 3.878248974008208e-05,
+      "loss": 2.3323,
+      "step": 2624
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5879504680633545,
+      "learning_rate": 3.876538987688099e-05,
+      "loss": 2.4866,
+      "step": 2628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5608062744140625,
+      "learning_rate": 3.8748290013679896e-05,
+      "loss": 2.3779,
+      "step": 2632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.390249252319336,
+      "learning_rate": 3.87311901504788e-05,
+      "loss": 2.3288,
+      "step": 2636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.357127070426941,
+      "learning_rate": 3.8714090287277704e-05,
+      "loss": 2.5016,
+      "step": 2640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4445414543151855,
+      "learning_rate": 3.8696990424076605e-05,
+      "loss": 2.5401,
+      "step": 2644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4636071920394897,
+      "learning_rate": 3.867989056087551e-05,
+      "loss": 2.3385,
+      "step": 2648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4558255672454834,
+      "learning_rate": 3.866279069767442e-05,
+      "loss": 2.3848,
+      "step": 2652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6711084842681885,
+      "learning_rate": 3.864569083447333e-05,
+      "loss": 2.4541,
+      "step": 2656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5824813842773438,
+      "learning_rate": 3.8628590971272234e-05,
+      "loss": 2.4533,
+      "step": 2660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5478894710540771,
+      "learning_rate": 3.8611491108071135e-05,
+      "loss": 2.4698,
+      "step": 2664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2620136737823486,
+      "learning_rate": 3.859439124487004e-05,
+      "loss": 2.746,
+      "step": 2668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4922306537628174,
+      "learning_rate": 3.857729138166895e-05,
+      "loss": 2.6415,
+      "step": 2672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.476605772972107,
+      "learning_rate": 3.856019151846785e-05,
+      "loss": 2.4816,
+      "step": 2676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7977741956710815,
+      "learning_rate": 3.8543091655266764e-05,
+      "loss": 2.4403,
+      "step": 2680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3775659799575806,
+      "learning_rate": 3.8525991792065665e-05,
+      "loss": 2.425,
+      "step": 2684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.645110845565796,
+      "learning_rate": 3.850889192886457e-05,
+      "loss": 2.4656,
+      "step": 2688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3301700353622437,
+      "learning_rate": 3.849179206566348e-05,
+      "loss": 2.4745,
+      "step": 2692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7708193063735962,
+      "learning_rate": 3.847469220246238e-05,
+      "loss": 2.6024,
+      "step": 2696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4959282875061035,
+      "learning_rate": 3.845759233926129e-05,
+      "loss": 2.5686,
+      "step": 2700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6098191738128662,
+      "learning_rate": 3.8440492476060194e-05,
+      "loss": 2.3471,
+      "step": 2704
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3885496854782104,
+      "learning_rate": 3.84233926128591e-05,
+      "loss": 2.3912,
+      "step": 2708
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4107521772384644,
+      "learning_rate": 3.840629274965801e-05,
+      "loss": 2.397,
+      "step": 2712
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6801328659057617,
+      "learning_rate": 3.838919288645691e-05,
+      "loss": 2.5848,
+      "step": 2716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3853564262390137,
+      "learning_rate": 3.837209302325582e-05,
+      "loss": 2.2871,
+      "step": 2720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3792238235473633,
+      "learning_rate": 3.835499316005472e-05,
+      "loss": 2.45,
+      "step": 2724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6318219900131226,
+      "learning_rate": 3.8337893296853625e-05,
+      "loss": 2.6403,
+      "step": 2728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7002670764923096,
+      "learning_rate": 3.832079343365253e-05,
+      "loss": 2.5752,
+      "step": 2732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3887910842895508,
+      "learning_rate": 3.830369357045144e-05,
+      "loss": 2.4802,
+      "step": 2736
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4336968660354614,
+      "learning_rate": 3.828659370725035e-05,
+      "loss": 2.4769,
+      "step": 2740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5607534646987915,
+      "learning_rate": 3.826949384404925e-05,
+      "loss": 2.4018,
+      "step": 2744
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4709196090698242,
+      "learning_rate": 3.8252393980848155e-05,
+      "loss": 2.4358,
+      "step": 2748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3464138507843018,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 2.5305,
+      "step": 2752
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4054791927337646,
+      "learning_rate": 3.821819425444596e-05,
+      "loss": 2.4714,
+      "step": 2756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.699159026145935,
+      "learning_rate": 3.820109439124488e-05,
+      "loss": 2.4277,
+      "step": 2760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3283201456069946,
+      "learning_rate": 3.818399452804378e-05,
+      "loss": 2.405,
+      "step": 2764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6334400177001953,
+      "learning_rate": 3.8166894664842685e-05,
+      "loss": 2.6551,
+      "step": 2768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4785411357879639,
+      "learning_rate": 3.8149794801641585e-05,
+      "loss": 2.5849,
+      "step": 2772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5953280925750732,
+      "learning_rate": 3.813269493844049e-05,
+      "loss": 2.4022,
+      "step": 2776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4671732187271118,
+      "learning_rate": 3.81155950752394e-05,
+      "loss": 2.3134,
+      "step": 2780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.461626410484314,
+      "learning_rate": 3.809849521203831e-05,
+      "loss": 2.4263,
+      "step": 2784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5867888927459717,
+      "learning_rate": 3.8081395348837215e-05,
+      "loss": 2.4209,
+      "step": 2788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4046286344528198,
+      "learning_rate": 3.8064295485636115e-05,
+      "loss": 2.3673,
+      "step": 2792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6984418630599976,
+      "learning_rate": 3.804719562243502e-05,
+      "loss": 2.4858,
+      "step": 2796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6407222747802734,
+      "learning_rate": 3.803009575923393e-05,
+      "loss": 2.4795,
+      "step": 2800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4830973148345947,
+      "learning_rate": 3.801299589603283e-05,
+      "loss": 2.5164,
+      "step": 2804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2736021280288696,
+      "learning_rate": 3.799589603283174e-05,
+      "loss": 2.3541,
+      "step": 2808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.564693570137024,
+      "learning_rate": 3.7978796169630645e-05,
+      "loss": 2.3652,
+      "step": 2812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5286824703216553,
+      "learning_rate": 3.796169630642955e-05,
+      "loss": 2.5063,
+      "step": 2816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.466086745262146,
+      "learning_rate": 3.794459644322846e-05,
+      "loss": 2.5876,
+      "step": 2820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5650023221969604,
+      "learning_rate": 3.792749658002736e-05,
+      "loss": 2.4316,
+      "step": 2824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4438190460205078,
+      "learning_rate": 3.791039671682627e-05,
+      "loss": 2.3659,
+      "step": 2828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8183629512786865,
+      "learning_rate": 3.789329685362517e-05,
+      "loss": 2.4262,
+      "step": 2832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.856266736984253,
+      "learning_rate": 3.7876196990424076e-05,
+      "loss": 2.457,
+      "step": 2836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6876139640808105,
+      "learning_rate": 3.785909712722299e-05,
+      "loss": 2.4234,
+      "step": 2840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3969238996505737,
+      "learning_rate": 3.784199726402189e-05,
+      "loss": 2.3044,
+      "step": 2844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3718289136886597,
+      "learning_rate": 3.78248974008208e-05,
+      "loss": 2.4889,
+      "step": 2848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5117738246917725,
+      "learning_rate": 3.78077975376197e-05,
+      "loss": 2.3139,
+      "step": 2852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.564997673034668,
+      "learning_rate": 3.7790697674418606e-05,
+      "loss": 2.6801,
+      "step": 2856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5139191150665283,
+      "learning_rate": 3.777359781121751e-05,
+      "loss": 2.3891,
+      "step": 2860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5658632516860962,
+      "learning_rate": 3.775649794801642e-05,
+      "loss": 2.4947,
+      "step": 2864
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5268386602401733,
+      "learning_rate": 3.773939808481533e-05,
+      "loss": 2.4591,
+      "step": 2868
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.977797508239746,
+      "learning_rate": 3.772229822161423e-05,
+      "loss": 2.5294,
+      "step": 2872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.368944764137268,
+      "learning_rate": 3.7705198358413136e-05,
+      "loss": 2.2917,
+      "step": 2876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.502429723739624,
+      "learning_rate": 3.7688098495212036e-05,
+      "loss": 2.5181,
+      "step": 2880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4561350345611572,
+      "learning_rate": 3.7670998632010944e-05,
+      "loss": 2.4059,
+      "step": 2884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9906591176986694,
+      "learning_rate": 3.765389876880985e-05,
+      "loss": 2.475,
+      "step": 2888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.607998251914978,
+      "learning_rate": 3.763679890560876e-05,
+      "loss": 2.668,
+      "step": 2892
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4631425142288208,
+      "learning_rate": 3.7619699042407666e-05,
+      "loss": 2.4278,
+      "step": 2896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3773366212844849,
+      "learning_rate": 3.7602599179206566e-05,
+      "loss": 2.5251,
+      "step": 2900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8001102209091187,
+      "learning_rate": 3.7585499316005474e-05,
+      "loss": 2.4981,
+      "step": 2904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4368932247161865,
+      "learning_rate": 3.756839945280438e-05,
+      "loss": 2.3902,
+      "step": 2908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5067082643508911,
+      "learning_rate": 3.755129958960328e-05,
+      "loss": 2.4835,
+      "step": 2912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6977261304855347,
+      "learning_rate": 3.753419972640219e-05,
+      "loss": 2.4424,
+      "step": 2916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5715231895446777,
+      "learning_rate": 3.7517099863201096e-05,
+      "loss": 2.376,
+      "step": 2920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8464969396591187,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 2.3253,
+      "step": 2924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4582910537719727,
+      "learning_rate": 3.748290013679891e-05,
+      "loss": 2.2077,
+      "step": 2928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3920525312423706,
+      "learning_rate": 3.746580027359781e-05,
+      "loss": 2.2856,
+      "step": 2932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.594838261604309,
+      "learning_rate": 3.744870041039672e-05,
+      "loss": 2.441,
+      "step": 2936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.608803629875183,
+      "learning_rate": 3.743160054719562e-05,
+      "loss": 2.5194,
+      "step": 2940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.492957592010498,
+      "learning_rate": 3.741450068399453e-05,
+      "loss": 2.3489,
+      "step": 2944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.186474323272705,
+      "learning_rate": 3.739740082079344e-05,
+      "loss": 2.5155,
+      "step": 2948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3143714666366577,
+      "learning_rate": 3.738030095759234e-05,
+      "loss": 2.5857,
+      "step": 2952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.287429094314575,
+      "learning_rate": 3.736320109439125e-05,
+      "loss": 2.4715,
+      "step": 2956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3473930358886719,
+      "learning_rate": 3.734610123119015e-05,
+      "loss": 2.4606,
+      "step": 2960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6107207536697388,
+      "learning_rate": 3.7329001367989057e-05,
+      "loss": 2.4097,
+      "step": 2964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.959020733833313,
+      "learning_rate": 3.7311901504787964e-05,
+      "loss": 2.5748,
+      "step": 2968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.483560562133789,
+      "learning_rate": 3.729480164158687e-05,
+      "loss": 2.4517,
+      "step": 2972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4312429428100586,
+      "learning_rate": 3.727770177838578e-05,
+      "loss": 2.6805,
+      "step": 2976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3787798881530762,
+      "learning_rate": 3.726060191518468e-05,
+      "loss": 2.3881,
+      "step": 2980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6351797580718994,
+      "learning_rate": 3.7243502051983587e-05,
+      "loss": 2.4977,
+      "step": 2984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.615729808807373,
+      "learning_rate": 3.722640218878249e-05,
+      "loss": 2.4574,
+      "step": 2988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4360637664794922,
+      "learning_rate": 3.7209302325581394e-05,
+      "loss": 2.4282,
+      "step": 2992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5573183298110962,
+      "learning_rate": 3.71922024623803e-05,
+      "loss": 2.3941,
+      "step": 2996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.398614764213562,
+      "learning_rate": 3.717510259917921e-05,
+      "loss": 2.4977,
+      "step": 3000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7220714092254639,
+      "learning_rate": 3.7158002735978116e-05,
+      "loss": 2.3985,
+      "step": 3004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.470420479774475,
+      "learning_rate": 3.714090287277702e-05,
+      "loss": 2.3886,
+      "step": 3008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.478071689605713,
+      "learning_rate": 3.7123803009575924e-05,
+      "loss": 2.3968,
+      "step": 3012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4263502359390259,
+      "learning_rate": 3.710670314637483e-05,
+      "loss": 2.5225,
+      "step": 3016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7249011993408203,
+      "learning_rate": 3.708960328317373e-05,
+      "loss": 2.538,
+      "step": 3020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5792475938796997,
+      "learning_rate": 3.707250341997264e-05,
+      "loss": 2.6076,
+      "step": 3024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.558214783668518,
+      "learning_rate": 3.705540355677155e-05,
+      "loss": 2.4636,
+      "step": 3028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.395617961883545,
+      "learning_rate": 3.7038303693570454e-05,
+      "loss": 2.4319,
+      "step": 3032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6385856866836548,
+      "learning_rate": 3.702120383036936e-05,
+      "loss": 2.7062,
+      "step": 3036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4634343385696411,
+      "learning_rate": 3.700410396716826e-05,
+      "loss": 2.5087,
+      "step": 3040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6941776275634766,
+      "learning_rate": 3.698700410396717e-05,
+      "loss": 2.3967,
+      "step": 3044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5948282480239868,
+      "learning_rate": 3.696990424076607e-05,
+      "loss": 2.371,
+      "step": 3048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6816450357437134,
+      "learning_rate": 3.6952804377564984e-05,
+      "loss": 2.3695,
+      "step": 3052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4531540870666504,
+      "learning_rate": 3.693570451436389e-05,
+      "loss": 2.375,
+      "step": 3056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.337263584136963,
+      "learning_rate": 3.691860465116279e-05,
+      "loss": 2.3965,
+      "step": 3060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5273529291152954,
+      "learning_rate": 3.69015047879617e-05,
+      "loss": 2.3984,
+      "step": 3064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4230504035949707,
+      "learning_rate": 3.68844049247606e-05,
+      "loss": 2.4717,
+      "step": 3068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6386831998825073,
+      "learning_rate": 3.686730506155951e-05,
+      "loss": 2.5523,
+      "step": 3072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5074329376220703,
+      "learning_rate": 3.6850205198358415e-05,
+      "loss": 2.4326,
+      "step": 3076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5377895832061768,
+      "learning_rate": 3.683310533515732e-05,
+      "loss": 2.3838,
+      "step": 3080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5310460329055786,
+      "learning_rate": 3.681600547195623e-05,
+      "loss": 2.5792,
+      "step": 3084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.486627221107483,
+      "learning_rate": 3.679890560875513e-05,
+      "loss": 2.3772,
+      "step": 3088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9239519834518433,
+      "learning_rate": 3.678180574555404e-05,
+      "loss": 2.4844,
+      "step": 3092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4352439641952515,
+      "learning_rate": 3.6764705882352945e-05,
+      "loss": 2.3233,
+      "step": 3096
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.540075421333313,
+      "learning_rate": 3.6747606019151845e-05,
+      "loss": 2.3849,
+      "step": 3100
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5862653255462646,
+      "learning_rate": 3.673050615595075e-05,
+      "loss": 2.3689,
+      "step": 3104
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4095063209533691,
+      "learning_rate": 3.671340629274966e-05,
+      "loss": 2.3854,
+      "step": 3108
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3623970746994019,
+      "learning_rate": 3.669630642954857e-05,
+      "loss": 2.2868,
+      "step": 3112
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4835405349731445,
+      "learning_rate": 3.667920656634747e-05,
+      "loss": 2.3524,
+      "step": 3116
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.790026068687439,
+      "learning_rate": 3.6662106703146375e-05,
+      "loss": 2.379,
+      "step": 3120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.452793836593628,
+      "learning_rate": 3.664500683994528e-05,
+      "loss": 2.2826,
+      "step": 3124
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.438991665840149,
+      "learning_rate": 3.662790697674418e-05,
+      "loss": 2.4365,
+      "step": 3128
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6045351028442383,
+      "learning_rate": 3.66108071135431e-05,
+      "loss": 2.2628,
+      "step": 3132
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.568385362625122,
+      "learning_rate": 3.6593707250342e-05,
+      "loss": 2.465,
+      "step": 3136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.451302170753479,
+      "learning_rate": 3.6576607387140905e-05,
+      "loss": 2.3324,
+      "step": 3140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4595221281051636,
+      "learning_rate": 3.655950752393981e-05,
+      "loss": 2.4677,
+      "step": 3144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5326175689697266,
+      "learning_rate": 3.654240766073871e-05,
+      "loss": 2.458,
+      "step": 3148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.651713490486145,
+      "learning_rate": 3.652530779753762e-05,
+      "loss": 2.3911,
+      "step": 3152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.475189208984375,
+      "learning_rate": 3.650820793433653e-05,
+      "loss": 2.4598,
+      "step": 3156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7967069149017334,
+      "learning_rate": 3.6491108071135435e-05,
+      "loss": 2.41,
+      "step": 3160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.380880355834961,
+      "learning_rate": 3.647400820793434e-05,
+      "loss": 2.3952,
+      "step": 3164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.392062783241272,
+      "learning_rate": 3.645690834473324e-05,
+      "loss": 2.5062,
+      "step": 3168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5198041200637817,
+      "learning_rate": 3.643980848153215e-05,
+      "loss": 2.3908,
+      "step": 3172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3749966621398926,
+      "learning_rate": 3.642270861833105e-05,
+      "loss": 2.3101,
+      "step": 3176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.559767246246338,
+      "learning_rate": 3.640560875512996e-05,
+      "loss": 2.3929,
+      "step": 3180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3676600456237793,
+      "learning_rate": 3.6388508891928866e-05,
+      "loss": 2.4963,
+      "step": 3184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4563826322555542,
+      "learning_rate": 3.637140902872777e-05,
+      "loss": 2.4377,
+      "step": 3188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5789074897766113,
+      "learning_rate": 3.635430916552668e-05,
+      "loss": 2.484,
+      "step": 3192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.633427381515503,
+      "learning_rate": 3.633720930232558e-05,
+      "loss": 2.5337,
+      "step": 3196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6052011251449585,
+      "learning_rate": 3.632010943912449e-05,
+      "loss": 2.2981,
+      "step": 3200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4831585884094238,
+      "learning_rate": 3.6303009575923396e-05,
+      "loss": 2.2886,
+      "step": 3204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.715384840965271,
+      "learning_rate": 3.6285909712722296e-05,
+      "loss": 2.2484,
+      "step": 3208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5094373226165771,
+      "learning_rate": 3.626880984952121e-05,
+      "loss": 2.3951,
+      "step": 3212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4048563241958618,
+      "learning_rate": 3.625170998632011e-05,
+      "loss": 2.3434,
+      "step": 3216
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.033588171005249,
+      "learning_rate": 3.623461012311902e-05,
+      "loss": 2.4892,
+      "step": 3220
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5815871953964233,
+      "learning_rate": 3.6217510259917926e-05,
+      "loss": 2.4064,
+      "step": 3224
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5218071937561035,
+      "learning_rate": 3.6200410396716826e-05,
+      "loss": 2.4017,
+      "step": 3228
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5122883319854736,
+      "learning_rate": 3.6183310533515733e-05,
+      "loss": 2.2099,
+      "step": 3232
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6052781343460083,
+      "learning_rate": 3.616621067031464e-05,
+      "loss": 2.4246,
+      "step": 3236
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.629193663597107,
+      "learning_rate": 3.614911080711355e-05,
+      "loss": 2.5361,
+      "step": 3240
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.478102445602417,
+      "learning_rate": 3.613201094391245e-05,
+      "loss": 2.2904,
+      "step": 3244
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5051685571670532,
+      "learning_rate": 3.6114911080711356e-05,
+      "loss": 2.5879,
+      "step": 3248
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.39858078956604,
+      "learning_rate": 3.609781121751026e-05,
+      "loss": 2.3035,
+      "step": 3252
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5903834104537964,
+      "learning_rate": 3.6080711354309164e-05,
+      "loss": 2.3173,
+      "step": 3256
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5299763679504395,
+      "learning_rate": 3.606361149110807e-05,
+      "loss": 2.3344,
+      "step": 3260
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5950665473937988,
+      "learning_rate": 3.604651162790698e-05,
+      "loss": 2.2693,
+      "step": 3264
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4136861562728882,
+      "learning_rate": 3.6029411764705886e-05,
+      "loss": 2.3145,
+      "step": 3268
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4014285802841187,
+      "learning_rate": 3.601231190150479e-05,
+      "loss": 2.4621,
+      "step": 3272
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.419997215270996,
+      "learning_rate": 3.5995212038303694e-05,
+      "loss": 2.3412,
+      "step": 3276
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4487018585205078,
+      "learning_rate": 3.59781121751026e-05,
+      "loss": 2.3379,
+      "step": 3280
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4295108318328857,
+      "learning_rate": 3.59610123119015e-05,
+      "loss": 2.3464,
+      "step": 3284
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5380290746688843,
+      "learning_rate": 3.594391244870041e-05,
+      "loss": 2.1826,
+      "step": 3288
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3480552434921265,
+      "learning_rate": 3.5926812585499316e-05,
+      "loss": 2.2881,
+      "step": 3292
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5125305652618408,
+      "learning_rate": 3.5909712722298224e-05,
+      "loss": 2.4374,
+      "step": 3296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5424193143844604,
+      "learning_rate": 3.589261285909713e-05,
+      "loss": 2.3602,
+      "step": 3300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9561680555343628,
+      "learning_rate": 3.587551299589603e-05,
+      "loss": 2.598,
+      "step": 3304
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4579107761383057,
+      "learning_rate": 3.585841313269494e-05,
+      "loss": 2.4397,
+      "step": 3308
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3612507581710815,
+      "learning_rate": 3.5841313269493846e-05,
+      "loss": 2.3842,
+      "step": 3312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6156667470932007,
+      "learning_rate": 3.582421340629275e-05,
+      "loss": 2.4143,
+      "step": 3316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.457672119140625,
+      "learning_rate": 3.580711354309166e-05,
+      "loss": 2.327,
+      "step": 3320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.474558711051941,
+      "learning_rate": 3.579001367989056e-05,
+      "loss": 2.3957,
+      "step": 3324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4732122421264648,
+      "learning_rate": 3.577291381668947e-05,
+      "loss": 2.4016,
+      "step": 3328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4019269943237305,
+      "learning_rate": 3.5755813953488376e-05,
+      "loss": 2.4573,
+      "step": 3332
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6624845266342163,
+      "learning_rate": 3.573871409028728e-05,
+      "loss": 2.2681,
+      "step": 3336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4161522388458252,
+      "learning_rate": 3.5721614227086184e-05,
+      "loss": 2.2106,
+      "step": 3340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3711514472961426,
+      "learning_rate": 3.570451436388509e-05,
+      "loss": 2.1563,
+      "step": 3344
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.575616478919983,
+      "learning_rate": 3.5687414500684e-05,
+      "loss": 2.3763,
+      "step": 3348
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6165823936462402,
+      "learning_rate": 3.5670314637482906e-05,
+      "loss": 2.3312,
+      "step": 3352
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3387306928634644,
+      "learning_rate": 3.565321477428181e-05,
+      "loss": 2.2245,
+      "step": 3356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.541719675064087,
+      "learning_rate": 3.5636114911080714e-05,
+      "loss": 2.3347,
+      "step": 3360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4336274862289429,
+      "learning_rate": 3.5619015047879615e-05,
+      "loss": 2.3484,
+      "step": 3364
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.586674690246582,
+      "learning_rate": 3.560191518467852e-05,
+      "loss": 2.5278,
+      "step": 3368
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.429751992225647,
+      "learning_rate": 3.558481532147743e-05,
+      "loss": 2.2046,
+      "step": 3372
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6239968538284302,
+      "learning_rate": 3.556771545827634e-05,
+      "loss": 2.4159,
+      "step": 3376
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.665915012359619,
+      "learning_rate": 3.5550615595075244e-05,
+      "loss": 2.3933,
+      "step": 3380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.532535195350647,
+      "learning_rate": 3.5533515731874145e-05,
+      "loss": 2.514,
+      "step": 3384
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8504074811935425,
+      "learning_rate": 3.551641586867305e-05,
+      "loss": 2.4337,
+      "step": 3388
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6196644306182861,
+      "learning_rate": 3.549931600547195e-05,
+      "loss": 2.297,
+      "step": 3392
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5914441347122192,
+      "learning_rate": 3.548221614227086e-05,
+      "loss": 2.3395,
+      "step": 3396
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5233490467071533,
+      "learning_rate": 3.5465116279069774e-05,
+      "loss": 2.253,
+      "step": 3400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4033238887786865,
+      "learning_rate": 3.5448016415868675e-05,
+      "loss": 2.3872,
+      "step": 3404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.485261082649231,
+      "learning_rate": 3.543091655266758e-05,
+      "loss": 2.367,
+      "step": 3408
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3384130001068115,
+      "learning_rate": 3.541381668946648e-05,
+      "loss": 2.2609,
+      "step": 3412
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3713083267211914,
+      "learning_rate": 3.539671682626539e-05,
+      "loss": 2.311,
+      "step": 3416
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6065647602081299,
+      "learning_rate": 3.53796169630643e-05,
+      "loss": 2.1545,
+      "step": 3420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.455736517906189,
+      "learning_rate": 3.5362517099863205e-05,
+      "loss": 2.3288,
+      "step": 3424
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6390444040298462,
+      "learning_rate": 3.534541723666211e-05,
+      "loss": 2.3899,
+      "step": 3428
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3747378587722778,
+      "learning_rate": 3.532831737346101e-05,
+      "loss": 2.3622,
+      "step": 3432
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4323629140853882,
+      "learning_rate": 3.531121751025992e-05,
+      "loss": 2.2561,
+      "step": 3436
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4392497539520264,
+      "learning_rate": 3.529411764705883e-05,
+      "loss": 2.4516,
+      "step": 3440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.507541298866272,
+      "learning_rate": 3.527701778385773e-05,
+      "loss": 2.3012,
+      "step": 3444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6980098485946655,
+      "learning_rate": 3.5259917920656635e-05,
+      "loss": 2.3305,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.378029704093933,
+      "learning_rate": 3.524281805745554e-05,
+      "loss": 2.3159,
+      "step": 3452
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4904416799545288,
+      "learning_rate": 3.522571819425445e-05,
+      "loss": 2.2613,
+      "step": 3456
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5655815601348877,
+      "learning_rate": 3.520861833105336e-05,
+      "loss": 2.3091,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4613752365112305,
+      "learning_rate": 3.519151846785226e-05,
+      "loss": 2.3852,
+      "step": 3464
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.735697627067566,
+      "learning_rate": 3.5174418604651165e-05,
+      "loss": 2.3746,
+      "step": 3468
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.56474769115448,
+      "learning_rate": 3.5157318741450066e-05,
+      "loss": 2.4085,
+      "step": 3472
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6144722700119019,
+      "learning_rate": 3.514021887824897e-05,
+      "loss": 2.2048,
+      "step": 3476
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4500874280929565,
+      "learning_rate": 3.512311901504789e-05,
+      "loss": 2.4197,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5875989198684692,
+      "learning_rate": 3.510601915184679e-05,
+      "loss": 2.3395,
+      "step": 3484
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5678739547729492,
+      "learning_rate": 3.5088919288645695e-05,
+      "loss": 2.4267,
+      "step": 3488
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4578819274902344,
+      "learning_rate": 3.5071819425444596e-05,
+      "loss": 2.3653,
+      "step": 3492
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8042303323745728,
+      "learning_rate": 3.50547195622435e-05,
+      "loss": 2.3785,
+      "step": 3496
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4263440370559692,
+      "learning_rate": 3.5037619699042403e-05,
+      "loss": 2.3203,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5346360206604004,
+      "learning_rate": 3.502051983584132e-05,
+      "loss": 2.3125,
+      "step": 3504
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5747292041778564,
+      "learning_rate": 3.5003419972640225e-05,
+      "loss": 2.4275,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7760605812072754,
+      "learning_rate": 3.4986320109439126e-05,
+      "loss": 2.6035,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5713919401168823,
+      "learning_rate": 3.496922024623803e-05,
+      "loss": 2.3099,
+      "step": 3516
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6528106927871704,
+      "learning_rate": 3.4952120383036933e-05,
+      "loss": 2.4168,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4009822607040405,
+      "learning_rate": 3.493502051983584e-05,
+      "loss": 2.3955,
+      "step": 3524
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7046074867248535,
+      "learning_rate": 3.491792065663475e-05,
+      "loss": 2.3926,
+      "step": 3528
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3500627279281616,
+      "learning_rate": 3.4900820793433655e-05,
+      "loss": 2.3855,
+      "step": 3532
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4953322410583496,
+      "learning_rate": 3.488372093023256e-05,
+      "loss": 2.3912,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3733028173446655,
+      "learning_rate": 3.486662106703146e-05,
+      "loss": 2.2607,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.427222490310669,
+      "learning_rate": 3.484952120383037e-05,
+      "loss": 2.3234,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4577361345291138,
+      "learning_rate": 3.483242134062928e-05,
+      "loss": 2.3509,
+      "step": 3548
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3897165060043335,
+      "learning_rate": 3.481532147742818e-05,
+      "loss": 2.154,
+      "step": 3552
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4338810443878174,
+      "learning_rate": 3.4798221614227086e-05,
+      "loss": 2.2335,
+      "step": 3556
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6268779039382935,
+      "learning_rate": 3.478112175102599e-05,
+      "loss": 2.2263,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4855129718780518,
+      "learning_rate": 3.47640218878249e-05,
+      "loss": 2.1839,
+      "step": 3564
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.525172233581543,
+      "learning_rate": 3.474692202462381e-05,
+      "loss": 2.3011,
+      "step": 3568
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.543357253074646,
+      "learning_rate": 3.472982216142271e-05,
+      "loss": 2.34,
+      "step": 3572
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4775084257125854,
+      "learning_rate": 3.4712722298221616e-05,
+      "loss": 2.2593,
+      "step": 3576
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3302327394485474,
+      "learning_rate": 3.4695622435020516e-05,
+      "loss": 2.1703,
+      "step": 3580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1734111309051514,
+      "learning_rate": 3.467852257181943e-05,
+      "loss": 2.3284,
+      "step": 3584
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5368781089782715,
+      "learning_rate": 3.466142270861834e-05,
+      "loss": 2.3187,
+      "step": 3588
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.526089072227478,
+      "learning_rate": 3.464432284541724e-05,
+      "loss": 2.3247,
+      "step": 3592
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.460354208946228,
+      "learning_rate": 3.4627222982216146e-05,
+      "loss": 2.2871,
+      "step": 3596
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5204929113388062,
+      "learning_rate": 3.4610123119015046e-05,
+      "loss": 2.2317,
+      "step": 3600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.355778455734253,
+      "learning_rate": 3.4593023255813954e-05,
+      "loss": 2.2038,
+      "step": 3604
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4250309467315674,
+      "learning_rate": 3.457592339261286e-05,
+      "loss": 2.288,
+      "step": 3608
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.529437780380249,
+      "learning_rate": 3.455882352941177e-05,
+      "loss": 2.2193,
+      "step": 3612
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6923147439956665,
+      "learning_rate": 3.4541723666210676e-05,
+      "loss": 2.2283,
+      "step": 3616
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5242345333099365,
+      "learning_rate": 3.4524623803009576e-05,
+      "loss": 2.3881,
+      "step": 3620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4843759536743164,
+      "learning_rate": 3.4507523939808484e-05,
+      "loss": 2.4458,
+      "step": 3624
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.361167550086975,
+      "learning_rate": 3.4490424076607384e-05,
+      "loss": 2.264,
+      "step": 3628
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4210281372070312,
+      "learning_rate": 3.447332421340629e-05,
+      "loss": 2.2471,
+      "step": 3632
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5141572952270508,
+      "learning_rate": 3.44562243502052e-05,
+      "loss": 2.3368,
+      "step": 3636
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3176366090774536,
+      "learning_rate": 3.4439124487004106e-05,
+      "loss": 2.1529,
+      "step": 3640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4620164632797241,
+      "learning_rate": 3.4422024623803014e-05,
+      "loss": 2.3995,
+      "step": 3644
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6192833185195923,
+      "learning_rate": 3.4404924760601914e-05,
+      "loss": 2.3621,
+      "step": 3648
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.459579348564148,
+      "learning_rate": 3.438782489740082e-05,
+      "loss": 2.3437,
+      "step": 3652
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3151201009750366,
+      "learning_rate": 3.437072503419973e-05,
+      "loss": 2.318,
+      "step": 3656
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4882017374038696,
+      "learning_rate": 3.435362517099863e-05,
+      "loss": 2.2267,
+      "step": 3660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3695846796035767,
+      "learning_rate": 3.433652530779754e-05,
+      "loss": 2.2106,
+      "step": 3664
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8380380868911743,
+      "learning_rate": 3.4319425444596444e-05,
+      "loss": 2.3097,
+      "step": 3668
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.560056447982788,
+      "learning_rate": 3.430232558139535e-05,
+      "loss": 2.4494,
+      "step": 3672
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4692696332931519,
+      "learning_rate": 3.428522571819426e-05,
+      "loss": 2.1774,
+      "step": 3676
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.585492730140686,
+      "learning_rate": 3.426812585499316e-05,
+      "loss": 2.2936,
+      "step": 3680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5503042936325073,
+      "learning_rate": 3.425102599179207e-05,
+      "loss": 2.2004,
+      "step": 3684
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.640663504600525,
+      "learning_rate": 3.423392612859097e-05,
+      "loss": 2.2647,
+      "step": 3688
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4468867778778076,
+      "learning_rate": 3.421682626538988e-05,
+      "loss": 2.3467,
+      "step": 3692
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5558197498321533,
+      "learning_rate": 3.419972640218879e-05,
+      "loss": 2.4436,
+      "step": 3696
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.489776372909546,
+      "learning_rate": 3.418262653898769e-05,
+      "loss": 2.3,
+      "step": 3700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4555796384811401,
+      "learning_rate": 3.41655266757866e-05,
+      "loss": 2.2145,
+      "step": 3704
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3881696462631226,
+      "learning_rate": 3.41484268125855e-05,
+      "loss": 2.3039,
+      "step": 3708
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.412907361984253,
+      "learning_rate": 3.4131326949384405e-05,
+      "loss": 2.3594,
+      "step": 3712
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.746773362159729,
+      "learning_rate": 3.411422708618331e-05,
+      "loss": 2.2989,
+      "step": 3716
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4776164293289185,
+      "learning_rate": 3.409712722298222e-05,
+      "loss": 2.3647,
+      "step": 3720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4442262649536133,
+      "learning_rate": 3.408002735978113e-05,
+      "loss": 2.1363,
+      "step": 3724
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.49587082862854,
+      "learning_rate": 3.406292749658003e-05,
+      "loss": 2.3467,
+      "step": 3728
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8401976823806763,
+      "learning_rate": 3.4045827633378935e-05,
+      "loss": 2.4508,
+      "step": 3732
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.379357099533081,
+      "learning_rate": 3.402872777017784e-05,
+      "loss": 2.2629,
+      "step": 3736
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.590693712234497,
+      "learning_rate": 3.401162790697674e-05,
+      "loss": 2.3216,
+      "step": 3740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4456241130828857,
+      "learning_rate": 3.399452804377565e-05,
+      "loss": 2.3142,
+      "step": 3744
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5368770360946655,
+      "learning_rate": 3.397742818057456e-05,
+      "loss": 2.2955,
+      "step": 3748
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6656548976898193,
+      "learning_rate": 3.3960328317373464e-05,
+      "loss": 2.4015,
+      "step": 3752
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.519033432006836,
+      "learning_rate": 3.3943228454172365e-05,
+      "loss": 2.1557,
+      "step": 3756
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.56667959690094,
+      "learning_rate": 3.392612859097127e-05,
+      "loss": 2.5277,
+      "step": 3760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5404236316680908,
+      "learning_rate": 3.390902872777018e-05,
+      "loss": 2.2924,
+      "step": 3764
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.619696021080017,
+      "learning_rate": 3.389192886456908e-05,
+      "loss": 2.2094,
+      "step": 3768
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3681530952453613,
+      "learning_rate": 3.3874829001367994e-05,
+      "loss": 2.231,
+      "step": 3772
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7254544496536255,
+      "learning_rate": 3.3857729138166895e-05,
+      "loss": 2.2382,
+      "step": 3776
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6230955123901367,
+      "learning_rate": 3.38406292749658e-05,
+      "loss": 2.1105,
+      "step": 3780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5223900079727173,
+      "learning_rate": 3.382352941176471e-05,
+      "loss": 2.1982,
+      "step": 3784
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3313345909118652,
+      "learning_rate": 3.380642954856361e-05,
+      "loss": 2.1896,
+      "step": 3788
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5461819171905518,
+      "learning_rate": 3.378932968536252e-05,
+      "loss": 2.2544,
+      "step": 3792
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4953100681304932,
+      "learning_rate": 3.3772229822161425e-05,
+      "loss": 2.3158,
+      "step": 3796
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4358595609664917,
+      "learning_rate": 3.375512995896033e-05,
+      "loss": 2.1203,
+      "step": 3800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6208328008651733,
+      "learning_rate": 3.373803009575924e-05,
+      "loss": 2.3804,
+      "step": 3804
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.58054518699646,
+      "learning_rate": 3.372093023255814e-05,
+      "loss": 2.2398,
+      "step": 3808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6291831731796265,
+      "learning_rate": 3.370383036935705e-05,
+      "loss": 2.2565,
+      "step": 3812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.500659465789795,
+      "learning_rate": 3.368673050615595e-05,
+      "loss": 2.2771,
+      "step": 3816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3757811784744263,
+      "learning_rate": 3.3669630642954855e-05,
+      "loss": 2.2549,
+      "step": 3820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5131489038467407,
+      "learning_rate": 3.365253077975376e-05,
+      "loss": 2.1822,
+      "step": 3824
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.511177897453308,
+      "learning_rate": 3.363543091655267e-05,
+      "loss": 2.3104,
+      "step": 3828
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4455082416534424,
+      "learning_rate": 3.361833105335158e-05,
+      "loss": 2.5066,
+      "step": 3832
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3408520221710205,
+      "learning_rate": 3.360123119015048e-05,
+      "loss": 2.2685,
+      "step": 3836
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3482859134674072,
+      "learning_rate": 3.3584131326949385e-05,
+      "loss": 2.4401,
+      "step": 3840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4088398218154907,
+      "learning_rate": 3.356703146374829e-05,
+      "loss": 2.3414,
+      "step": 3844
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.483805775642395,
+      "learning_rate": 3.354993160054719e-05,
+      "loss": 2.1907,
+      "step": 3848
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4749683141708374,
+      "learning_rate": 3.353283173734611e-05,
+      "loss": 2.4228,
+      "step": 3852
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5431963205337524,
+      "learning_rate": 3.351573187414501e-05,
+      "loss": 2.2988,
+      "step": 3856
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7580955028533936,
+      "learning_rate": 3.3498632010943915e-05,
+      "loss": 2.2444,
+      "step": 3860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3347162008285522,
+      "learning_rate": 3.348153214774282e-05,
+      "loss": 2.3015,
+      "step": 3864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4706435203552246,
+      "learning_rate": 3.346443228454172e-05,
+      "loss": 2.3175,
+      "step": 3868
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.63969886302948,
+      "learning_rate": 3.344733242134063e-05,
+      "loss": 2.1465,
+      "step": 3872
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.056724786758423,
+      "learning_rate": 3.343023255813954e-05,
+      "loss": 2.3949,
+      "step": 3876
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.551066279411316,
+      "learning_rate": 3.3413132694938445e-05,
+      "loss": 2.2891,
+      "step": 3880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6610709428787231,
+      "learning_rate": 3.3396032831737346e-05,
+      "loss": 2.5973,
+      "step": 3884
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5867125988006592,
+      "learning_rate": 3.337893296853625e-05,
+      "loss": 2.2614,
+      "step": 3888
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8439687490463257,
+      "learning_rate": 3.336183310533516e-05,
+      "loss": 2.1923,
+      "step": 3892
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4217692613601685,
+      "learning_rate": 3.334473324213406e-05,
+      "loss": 2.2826,
+      "step": 3896
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2552785873413086,
+      "learning_rate": 3.332763337893297e-05,
+      "loss": 2.3539,
+      "step": 3900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5517688989639282,
+      "learning_rate": 3.3310533515731876e-05,
+      "loss": 2.2412,
+      "step": 3904
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6065213680267334,
+      "learning_rate": 3.329343365253078e-05,
+      "loss": 2.2576,
+      "step": 3908
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4800779819488525,
+      "learning_rate": 3.327633378932969e-05,
+      "loss": 2.324,
+      "step": 3912
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3546322584152222,
+      "learning_rate": 3.325923392612859e-05,
+      "loss": 2.323,
+      "step": 3916
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6678704023361206,
+      "learning_rate": 3.32421340629275e-05,
+      "loss": 2.3459,
+      "step": 3920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4288249015808105,
+      "learning_rate": 3.32250341997264e-05,
+      "loss": 2.2059,
+      "step": 3924
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4324655532836914,
+      "learning_rate": 3.3207934336525306e-05,
+      "loss": 2.1479,
+      "step": 3928
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5552873611450195,
+      "learning_rate": 3.319083447332422e-05,
+      "loss": 2.2897,
+      "step": 3932
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4077160358428955,
+      "learning_rate": 3.317373461012312e-05,
+      "loss": 2.3246,
+      "step": 3936
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4134106636047363,
+      "learning_rate": 3.315663474692203e-05,
+      "loss": 2.2485,
+      "step": 3940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.670109510421753,
+      "learning_rate": 3.313953488372093e-05,
+      "loss": 2.2681,
+      "step": 3944
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.706309199333191,
+      "learning_rate": 3.3122435020519836e-05,
+      "loss": 2.2443,
+      "step": 3948
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5278905630111694,
+      "learning_rate": 3.3105335157318744e-05,
+      "loss": 2.1895,
+      "step": 3952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3573322296142578,
+      "learning_rate": 3.308823529411765e-05,
+      "loss": 2.4002,
+      "step": 3956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.383784532546997,
+      "learning_rate": 3.307113543091656e-05,
+      "loss": 2.3252,
+      "step": 3960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5072522163391113,
+      "learning_rate": 3.305403556771546e-05,
+      "loss": 2.3,
+      "step": 3964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5655957460403442,
+      "learning_rate": 3.3036935704514366e-05,
+      "loss": 2.1498,
+      "step": 3968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4027670621871948,
+      "learning_rate": 3.3019835841313274e-05,
+      "loss": 2.2169,
+      "step": 3972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5429667234420776,
+      "learning_rate": 3.3002735978112174e-05,
+      "loss": 2.3642,
+      "step": 3976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4422931671142578,
+      "learning_rate": 3.298563611491108e-05,
+      "loss": 2.1742,
+      "step": 3980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4613866806030273,
+      "learning_rate": 3.296853625170999e-05,
+      "loss": 2.4487,
+      "step": 3984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5176717042922974,
+      "learning_rate": 3.2951436388508896e-05,
+      "loss": 2.3083,
+      "step": 3988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7266799211502075,
+      "learning_rate": 3.2934336525307803e-05,
+      "loss": 2.2963,
+      "step": 3992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5054651498794556,
+      "learning_rate": 3.2917236662106704e-05,
+      "loss": 2.4407,
+      "step": 3996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4415006637573242,
+      "learning_rate": 3.290013679890561e-05,
+      "loss": 2.25,
+      "step": 4000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5419758558273315,
+      "learning_rate": 3.288303693570451e-05,
+      "loss": 2.3623,
+      "step": 4004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7471479177474976,
+      "learning_rate": 3.286593707250342e-05,
+      "loss": 2.4276,
+      "step": 4008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7161086797714233,
+      "learning_rate": 3.284883720930233e-05,
+      "loss": 2.3613,
+      "step": 4012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.622846007347107,
+      "learning_rate": 3.2831737346101234e-05,
+      "loss": 2.2338,
+      "step": 4016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.52423095703125,
+      "learning_rate": 3.281463748290014e-05,
+      "loss": 2.3147,
+      "step": 4020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4126418828964233,
+      "learning_rate": 3.279753761969904e-05,
+      "loss": 2.2087,
+      "step": 4024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.316510796546936,
+      "learning_rate": 3.278043775649795e-05,
+      "loss": 2.2318,
+      "step": 4028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7000290155410767,
+      "learning_rate": 3.276333789329685e-05,
+      "loss": 2.187,
+      "step": 4032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4855263233184814,
+      "learning_rate": 3.274623803009576e-05,
+      "loss": 2.1994,
+      "step": 4036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4632859230041504,
+      "learning_rate": 3.272913816689467e-05,
+      "loss": 2.2279,
+      "step": 4040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.456778645515442,
+      "learning_rate": 3.271203830369357e-05,
+      "loss": 2.1008,
+      "step": 4044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4377546310424805,
+      "learning_rate": 3.269493844049248e-05,
+      "loss": 2.2231,
+      "step": 4048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5278170108795166,
+      "learning_rate": 3.267783857729138e-05,
+      "loss": 2.1359,
+      "step": 4052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.558070182800293,
+      "learning_rate": 3.266073871409029e-05,
+      "loss": 2.383,
+      "step": 4056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5845755338668823,
+      "learning_rate": 3.2643638850889194e-05,
+      "loss": 2.2107,
+      "step": 4060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6732598543167114,
+      "learning_rate": 3.26265389876881e-05,
+      "loss": 2.0189,
+      "step": 4064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5206589698791504,
+      "learning_rate": 3.260943912448701e-05,
+      "loss": 2.2918,
+      "step": 4068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6750423908233643,
+      "learning_rate": 3.259233926128591e-05,
+      "loss": 2.2082,
+      "step": 4072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6563453674316406,
+      "learning_rate": 3.257523939808482e-05,
+      "loss": 2.3358,
+      "step": 4076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.575918197631836,
+      "learning_rate": 3.2558139534883724e-05,
+      "loss": 2.355,
+      "step": 4080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5032954216003418,
+      "learning_rate": 3.2541039671682625e-05,
+      "loss": 2.259,
+      "step": 4084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4477620124816895,
+      "learning_rate": 3.252393980848153e-05,
+      "loss": 2.2821,
+      "step": 4088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.342366337776184,
+      "learning_rate": 3.250683994528044e-05,
+      "loss": 2.1086,
+      "step": 4092
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4458723068237305,
+      "learning_rate": 3.248974008207935e-05,
+      "loss": 2.2197,
+      "step": 4096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4470369815826416,
+      "learning_rate": 3.2472640218878254e-05,
+      "loss": 2.3436,
+      "step": 4100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6926721334457397,
+      "learning_rate": 3.2455540355677155e-05,
+      "loss": 2.3198,
+      "step": 4104
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.483053207397461,
+      "learning_rate": 3.243844049247606e-05,
+      "loss": 2.2944,
+      "step": 4108
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5896413326263428,
+      "learning_rate": 3.242134062927496e-05,
+      "loss": 2.1459,
+      "step": 4112
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.915610432624817,
+      "learning_rate": 3.240424076607387e-05,
+      "loss": 2.2768,
+      "step": 4116
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4022173881530762,
+      "learning_rate": 3.2387140902872784e-05,
+      "loss": 2.1106,
+      "step": 4120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3702282905578613,
+      "learning_rate": 3.2370041039671685e-05,
+      "loss": 2.0183,
+      "step": 4124
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3947548866271973,
+      "learning_rate": 3.235294117647059e-05,
+      "loss": 2.1075,
+      "step": 4128
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5231854915618896,
+      "learning_rate": 3.233584131326949e-05,
+      "loss": 2.3221,
+      "step": 4132
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5928566455841064,
+      "learning_rate": 3.23187414500684e-05,
+      "loss": 2.1517,
+      "step": 4136
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.518593192100525,
+      "learning_rate": 3.23016415868673e-05,
+      "loss": 2.4391,
+      "step": 4140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4657520055770874,
+      "learning_rate": 3.2284541723666215e-05,
+      "loss": 2.2852,
+      "step": 4144
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4905215501785278,
+      "learning_rate": 3.226744186046512e-05,
+      "loss": 2.1462,
+      "step": 4148
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6270649433135986,
+      "learning_rate": 3.225034199726402e-05,
+      "loss": 2.4138,
+      "step": 4152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7110861539840698,
+      "learning_rate": 3.223324213406293e-05,
+      "loss": 2.2067,
+      "step": 4156
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.771278977394104,
+      "learning_rate": 3.221614227086183e-05,
+      "loss": 2.3056,
+      "step": 4160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7753063440322876,
+      "learning_rate": 3.219904240766074e-05,
+      "loss": 2.1597,
+      "step": 4164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4673937559127808,
+      "learning_rate": 3.2181942544459645e-05,
+      "loss": 2.2006,
+      "step": 4168
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4760159254074097,
+      "learning_rate": 3.216484268125855e-05,
+      "loss": 2.213,
+      "step": 4172
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.435507893562317,
+      "learning_rate": 3.214774281805746e-05,
+      "loss": 2.2281,
+      "step": 4176
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3926265239715576,
+      "learning_rate": 3.213064295485636e-05,
+      "loss": 2.1146,
+      "step": 4180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4813460111618042,
+      "learning_rate": 3.211354309165527e-05,
+      "loss": 2.2133,
+      "step": 4184
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.576973795890808,
+      "learning_rate": 3.2096443228454175e-05,
+      "loss": 2.1075,
+      "step": 4188
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.532358169555664,
+      "learning_rate": 3.2079343365253076e-05,
+      "loss": 2.2588,
+      "step": 4192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4591597318649292,
+      "learning_rate": 3.206224350205198e-05,
+      "loss": 2.2587,
+      "step": 4196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5652563571929932,
+      "learning_rate": 3.204514363885089e-05,
+      "loss": 2.217,
+      "step": 4200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4030766487121582,
+      "learning_rate": 3.20280437756498e-05,
+      "loss": 2.2451,
+      "step": 4204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5012942552566528,
+      "learning_rate": 3.2010943912448705e-05,
+      "loss": 2.2296,
+      "step": 4208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3982198238372803,
+      "learning_rate": 3.1993844049247606e-05,
+      "loss": 2.3254,
+      "step": 4212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.359818458557129,
+      "learning_rate": 3.197674418604651e-05,
+      "loss": 2.3709,
+      "step": 4216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.687652349472046,
+      "learning_rate": 3.1959644322845414e-05,
+      "loss": 2.2442,
+      "step": 4220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.577440857887268,
+      "learning_rate": 3.194254445964433e-05,
+      "loss": 2.1392,
+      "step": 4224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4704028367996216,
+      "learning_rate": 3.1925444596443235e-05,
+      "loss": 2.3586,
+      "step": 4228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5463335514068604,
+      "learning_rate": 3.1908344733242136e-05,
+      "loss": 2.3475,
+      "step": 4232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4893290996551514,
+      "learning_rate": 3.189124487004104e-05,
+      "loss": 2.1804,
+      "step": 4236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6646398305892944,
+      "learning_rate": 3.1874145006839944e-05,
+      "loss": 2.2972,
+      "step": 4240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4279797077178955,
+      "learning_rate": 3.185704514363885e-05,
+      "loss": 2.235,
+      "step": 4244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5145081281661987,
+      "learning_rate": 3.183994528043776e-05,
+      "loss": 2.231,
+      "step": 4248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4431096315383911,
+      "learning_rate": 3.1822845417236666e-05,
+      "loss": 2.1942,
+      "step": 4252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.431936264038086,
+      "learning_rate": 3.180574555403557e-05,
+      "loss": 2.2726,
+      "step": 4256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4116493463516235,
+      "learning_rate": 3.1788645690834474e-05,
+      "loss": 2.346,
+      "step": 4260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.480595588684082,
+      "learning_rate": 3.177154582763338e-05,
+      "loss": 2.1486,
+      "step": 4264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4615095853805542,
+      "learning_rate": 3.175444596443228e-05,
+      "loss": 2.2935,
+      "step": 4268
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.414937973022461,
+      "learning_rate": 3.173734610123119e-05,
+      "loss": 2.0618,
+      "step": 4272
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.260504961013794,
+      "learning_rate": 3.1720246238030096e-05,
+      "loss": 2.2604,
+      "step": 4276
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.573109745979309,
+      "learning_rate": 3.1703146374829003e-05,
+      "loss": 2.1983,
+      "step": 4280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4415810108184814,
+      "learning_rate": 3.168604651162791e-05,
+      "loss": 2.1983,
+      "step": 4284
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4468250274658203,
+      "learning_rate": 3.166894664842681e-05,
+      "loss": 2.3544,
+      "step": 4288
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4121309518814087,
+      "learning_rate": 3.165184678522572e-05,
+      "loss": 2.3186,
+      "step": 4292
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3506333827972412,
+      "learning_rate": 3.1634746922024626e-05,
+      "loss": 2.2057,
+      "step": 4296
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5312514305114746,
+      "learning_rate": 3.161764705882353e-05,
+      "loss": 2.1451,
+      "step": 4300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8453575372695923,
+      "learning_rate": 3.160054719562244e-05,
+      "loss": 2.2856,
+      "step": 4304
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5432989597320557,
+      "learning_rate": 3.158344733242134e-05,
+      "loss": 2.2947,
+      "step": 4308
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.575966238975525,
+      "learning_rate": 3.156634746922025e-05,
+      "loss": 2.1444,
+      "step": 4312
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6484041213989258,
+      "learning_rate": 3.1549247606019156e-05,
+      "loss": 2.1952,
+      "step": 4316
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4162933826446533,
+      "learning_rate": 3.1532147742818057e-05,
+      "loss": 2.0885,
+      "step": 4320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.570765495300293,
+      "learning_rate": 3.1515047879616964e-05,
+      "loss": 2.1415,
+      "step": 4324
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.603226661682129,
+      "learning_rate": 3.149794801641587e-05,
+      "loss": 2.2982,
+      "step": 4328
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4410016536712646,
+      "learning_rate": 3.148084815321478e-05,
+      "loss": 2.2572,
+      "step": 4332
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.392094373703003,
+      "learning_rate": 3.1463748290013686e-05,
+      "loss": 2.4107,
+      "step": 4336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4003255367279053,
+      "learning_rate": 3.1446648426812587e-05,
+      "loss": 2.2181,
+      "step": 4340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.496429204940796,
+      "learning_rate": 3.1429548563611494e-05,
+      "loss": 2.241,
+      "step": 4344
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5706188678741455,
+      "learning_rate": 3.1412448700410394e-05,
+      "loss": 2.1635,
+      "step": 4348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4904406070709229,
+      "learning_rate": 3.13953488372093e-05,
+      "loss": 2.2688,
+      "step": 4352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4973410367965698,
+      "learning_rate": 3.137824897400821e-05,
+      "loss": 2.2812,
+      "step": 4356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6343064308166504,
+      "learning_rate": 3.1361149110807116e-05,
+      "loss": 2.3717,
+      "step": 4360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.327519178390503,
+      "learning_rate": 3.1344049247606024e-05,
+      "loss": 2.1028,
+      "step": 4364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.46675705909729,
+      "learning_rate": 3.1326949384404924e-05,
+      "loss": 2.2575,
+      "step": 4368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5625230073928833,
+      "learning_rate": 3.130984952120383e-05,
+      "loss": 2.1085,
+      "step": 4372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4606382846832275,
+      "learning_rate": 3.129274965800274e-05,
+      "loss": 2.1484,
+      "step": 4376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3543037176132202,
+      "learning_rate": 3.127564979480164e-05,
+      "loss": 2.3504,
+      "step": 4380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3623141050338745,
+      "learning_rate": 3.125854993160055e-05,
+      "loss": 2.1932,
+      "step": 4384
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2342579364776611,
+      "learning_rate": 3.1241450068399454e-05,
+      "loss": 2.229,
+      "step": 4388
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5705771446228027,
+      "learning_rate": 3.122435020519836e-05,
+      "loss": 2.3358,
+      "step": 4392
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5220746994018555,
+      "learning_rate": 3.120725034199726e-05,
+      "loss": 2.0569,
+      "step": 4396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5559509992599487,
+      "learning_rate": 3.119015047879617e-05,
+      "loss": 2.348,
+      "step": 4400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4289395809173584,
+      "learning_rate": 3.117305061559508e-05,
+      "loss": 2.1036,
+      "step": 4404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.69617760181427,
+      "learning_rate": 3.115595075239398e-05,
+      "loss": 2.1352,
+      "step": 4408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6432487964630127,
+      "learning_rate": 3.113885088919289e-05,
+      "loss": 2.2481,
+      "step": 4412
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.986351728439331,
+      "learning_rate": 3.112175102599179e-05,
+      "loss": 2.2431,
+      "step": 4416
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.541490912437439,
+      "learning_rate": 3.11046511627907e-05,
+      "loss": 2.1826,
+      "step": 4420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5522269010543823,
+      "learning_rate": 3.108755129958961e-05,
+      "loss": 2.1923,
+      "step": 4424
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4916808605194092,
+      "learning_rate": 3.107045143638851e-05,
+      "loss": 2.1626,
+      "step": 4428
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3527686595916748,
+      "learning_rate": 3.1053351573187415e-05,
+      "loss": 2.0595,
+      "step": 4432
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.568322777748108,
+      "learning_rate": 3.103625170998632e-05,
+      "loss": 2.2729,
+      "step": 4436
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5389866828918457,
+      "learning_rate": 3.101915184678523e-05,
+      "loss": 2.0978,
+      "step": 4440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4940327405929565,
+      "learning_rate": 3.100205198358414e-05,
+      "loss": 2.2146,
+      "step": 4444
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5102323293685913,
+      "learning_rate": 3.098495212038304e-05,
+      "loss": 2.1204,
+      "step": 4448
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4519439935684204,
+      "learning_rate": 3.0967852257181945e-05,
+      "loss": 2.2592,
+      "step": 4452
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3934937715530396,
+      "learning_rate": 3.0950752393980845e-05,
+      "loss": 2.1279,
+      "step": 4456
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3405640125274658,
+      "learning_rate": 3.093365253077975e-05,
+      "loss": 2.1896,
+      "step": 4460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6696523427963257,
+      "learning_rate": 3.091655266757866e-05,
+      "loss": 2.0079,
+      "step": 4464
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.470809817314148,
+      "learning_rate": 3.089945280437757e-05,
+      "loss": 2.2609,
+      "step": 4468
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.54929518699646,
+      "learning_rate": 3.0882352941176475e-05,
+      "loss": 2.18,
+      "step": 4472
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0508859157562256,
+      "learning_rate": 3.0865253077975375e-05,
+      "loss": 2.1995,
+      "step": 4476
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4847677946090698,
+      "learning_rate": 3.084815321477428e-05,
+      "loss": 2.1576,
+      "step": 4480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.490350604057312,
+      "learning_rate": 3.083105335157319e-05,
+      "loss": 2.188,
+      "step": 4484
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5087822675704956,
+      "learning_rate": 3.081395348837209e-05,
+      "loss": 2.2151,
+      "step": 4488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8956283330917358,
+      "learning_rate": 3.0796853625171005e-05,
+      "loss": 2.1073,
+      "step": 4492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.575896978378296,
+      "learning_rate": 3.0779753761969905e-05,
+      "loss": 2.2574,
+      "step": 4496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5014541149139404,
+      "learning_rate": 3.076265389876881e-05,
+      "loss": 2.1651,
+      "step": 4500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8757340908050537,
+      "learning_rate": 3.074555403556772e-05,
+      "loss": 2.0871,
+      "step": 4504
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5806230306625366,
+      "learning_rate": 3.072845417236662e-05,
+      "loss": 2.1398,
+      "step": 4508
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7795255184173584,
+      "learning_rate": 3.071135430916553e-05,
+      "loss": 2.2246,
+      "step": 4512
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6598823070526123,
+      "learning_rate": 3.0694254445964435e-05,
+      "loss": 2.2928,
+      "step": 4516
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5600835084915161,
+      "learning_rate": 3.067715458276334e-05,
+      "loss": 2.3343,
+      "step": 4520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5268632173538208,
+      "learning_rate": 3.066005471956224e-05,
+      "loss": 2.3551,
+      "step": 4524
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8789829015731812,
+      "learning_rate": 3.064295485636115e-05,
+      "loss": 2.3127,
+      "step": 4528
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.607021450996399,
+      "learning_rate": 3.062585499316006e-05,
+      "loss": 2.142,
+      "step": 4532
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4491835832595825,
+      "learning_rate": 3.060875512995896e-05,
+      "loss": 2.2,
+      "step": 4536
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5633890628814697,
+      "learning_rate": 3.0591655266757866e-05,
+      "loss": 2.0443,
+      "step": 4540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6567158699035645,
+      "learning_rate": 3.057455540355677e-05,
+      "loss": 2.1271,
+      "step": 4544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5010050535202026,
+      "learning_rate": 3.055745554035568e-05,
+      "loss": 2.2272,
+      "step": 4548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5739331245422363,
+      "learning_rate": 3.054035567715459e-05,
+      "loss": 2.1923,
+      "step": 4552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4975517988204956,
+      "learning_rate": 3.052325581395349e-05,
+      "loss": 2.2189,
+      "step": 4556
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.571330189704895,
+      "learning_rate": 3.0506155950752396e-05,
+      "loss": 2.1658,
+      "step": 4560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.296584129333496,
+      "learning_rate": 3.04890560875513e-05,
+      "loss": 2.1641,
+      "step": 4564
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.920922875404358,
+      "learning_rate": 3.0471956224350207e-05,
+      "loss": 2.3731,
+      "step": 4568
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5119904279708862,
+      "learning_rate": 3.0454856361149114e-05,
+      "loss": 2.0966,
+      "step": 4572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3779191970825195,
+      "learning_rate": 3.0437756497948018e-05,
+      "loss": 2.0729,
+      "step": 4576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3981715440750122,
+      "learning_rate": 3.0420656634746926e-05,
+      "loss": 2.1438,
+      "step": 4580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4112988710403442,
+      "learning_rate": 3.0403556771545826e-05,
+      "loss": 2.3213,
+      "step": 4584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4584025144577026,
+      "learning_rate": 3.0386456908344733e-05,
+      "loss": 2.2532,
+      "step": 4588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.500198245048523,
+      "learning_rate": 3.0369357045143644e-05,
+      "loss": 2.029,
+      "step": 4592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4857838153839111,
+      "learning_rate": 3.0352257181942545e-05,
+      "loss": 2.0186,
+      "step": 4596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.396033525466919,
+      "learning_rate": 3.0335157318741452e-05,
+      "loss": 2.2293,
+      "step": 4600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5105597972869873,
+      "learning_rate": 3.0318057455540356e-05,
+      "loss": 2.1206,
+      "step": 4604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3721805810928345,
+      "learning_rate": 3.0300957592339263e-05,
+      "loss": 2.1646,
+      "step": 4608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4651081562042236,
+      "learning_rate": 3.028385772913817e-05,
+      "loss": 2.2184,
+      "step": 4612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.468990445137024,
+      "learning_rate": 3.0266757865937075e-05,
+      "loss": 2.1083,
+      "step": 4616
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.047722578048706,
+      "learning_rate": 3.0249658002735982e-05,
+      "loss": 2.1685,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5102801322937012,
+      "learning_rate": 3.0232558139534883e-05,
+      "loss": 2.2332,
+      "step": 4624
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3657864332199097,
+      "learning_rate": 3.021545827633379e-05,
+      "loss": 2.3045,
+      "step": 4628
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.468764305114746,
+      "learning_rate": 3.01983584131327e-05,
+      "loss": 2.2094,
+      "step": 4632
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.451079249382019,
+      "learning_rate": 3.01812585499316e-05,
+      "loss": 2.0958,
+      "step": 4636
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5094703435897827,
+      "learning_rate": 3.016415868673051e-05,
+      "loss": 2.186,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5918653011322021,
+      "learning_rate": 3.0147058823529413e-05,
+      "loss": 2.2616,
+      "step": 4644
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5116558074951172,
+      "learning_rate": 3.012995896032832e-05,
+      "loss": 2.2503,
+      "step": 4648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4294779300689697,
+      "learning_rate": 3.011285909712722e-05,
+      "loss": 2.3022,
+      "step": 4652
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.080925464630127,
+      "learning_rate": 3.009575923392613e-05,
+      "loss": 2.0627,
+      "step": 4656
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4744940996170044,
+      "learning_rate": 3.007865937072504e-05,
+      "loss": 2.1063,
+      "step": 4660
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6344597339630127,
+      "learning_rate": 3.006155950752394e-05,
+      "loss": 2.1501,
+      "step": 4664
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5882827043533325,
+      "learning_rate": 3.0044459644322846e-05,
+      "loss": 2.1208,
+      "step": 4668
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4376238584518433,
+      "learning_rate": 3.002735978112175e-05,
+      "loss": 2.0962,
+      "step": 4672
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6142209768295288,
+      "learning_rate": 3.0010259917920658e-05,
+      "loss": 2.0721,
+      "step": 4676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4628876447677612,
+      "learning_rate": 2.9993160054719565e-05,
+      "loss": 1.959,
+      "step": 4680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3108237981796265,
+      "learning_rate": 2.997606019151847e-05,
+      "loss": 2.1634,
+      "step": 4684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4312976598739624,
+      "learning_rate": 2.9958960328317376e-05,
+      "loss": 2.1643,
+      "step": 4688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.449655294418335,
+      "learning_rate": 2.9941860465116277e-05,
+      "loss": 2.2098,
+      "step": 4692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5844769477844238,
+      "learning_rate": 2.9924760601915188e-05,
+      "loss": 2.1497,
+      "step": 4696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4893046617507935,
+      "learning_rate": 2.9907660738714095e-05,
+      "loss": 2.2939,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3878313302993774,
+      "learning_rate": 2.9890560875512996e-05,
+      "loss": 2.1361,
+      "step": 4704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9359514713287354,
+      "learning_rate": 2.9873461012311903e-05,
+      "loss": 2.178,
+      "step": 4708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.581032633781433,
+      "learning_rate": 2.9856361149110807e-05,
+      "loss": 2.0514,
+      "step": 4712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3860806226730347,
+      "learning_rate": 2.9839261285909714e-05,
+      "loss": 2.2888,
+      "step": 4716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4699238538742065,
+      "learning_rate": 2.982216142270862e-05,
+      "loss": 2.167,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.459145188331604,
+      "learning_rate": 2.9805061559507526e-05,
+      "loss": 2.1719,
+      "step": 4724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7492071390151978,
+      "learning_rate": 2.9787961696306433e-05,
+      "loss": 2.314,
+      "step": 4728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9393258094787598,
+      "learning_rate": 2.9770861833105333e-05,
+      "loss": 2.3491,
+      "step": 4732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8979884386062622,
+      "learning_rate": 2.9753761969904244e-05,
+      "loss": 2.1962,
+      "step": 4736
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5597001314163208,
+      "learning_rate": 2.973666210670315e-05,
+      "loss": 2.1819,
+      "step": 4740
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6771557331085205,
+      "learning_rate": 2.9719562243502052e-05,
+      "loss": 2.1407,
+      "step": 4744
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5106451511383057,
+      "learning_rate": 2.970246238030096e-05,
+      "loss": 2.1655,
+      "step": 4748
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.531887173652649,
+      "learning_rate": 2.9685362517099863e-05,
+      "loss": 2.1351,
+      "step": 4752
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3597395420074463,
+      "learning_rate": 2.966826265389877e-05,
+      "loss": 2.0589,
+      "step": 4756
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.761173129081726,
+      "learning_rate": 2.9651162790697678e-05,
+      "loss": 2.0529,
+      "step": 4760
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.44886314868927,
+      "learning_rate": 2.9634062927496582e-05,
+      "loss": 1.9243,
+      "step": 4764
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.516022801399231,
+      "learning_rate": 2.961696306429549e-05,
+      "loss": 2.0999,
+      "step": 4768
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3773995637893677,
+      "learning_rate": 2.959986320109439e-05,
+      "loss": 2.2363,
+      "step": 4772
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5385714769363403,
+      "learning_rate": 2.9582763337893297e-05,
+      "loss": 2.1583,
+      "step": 4776
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.746711015701294,
+      "learning_rate": 2.95656634746922e-05,
+      "loss": 2.1572,
+      "step": 4780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6620087623596191,
+      "learning_rate": 2.954856361149111e-05,
+      "loss": 2.102,
+      "step": 4784
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4622972011566162,
+      "learning_rate": 2.9531463748290016e-05,
+      "loss": 2.2325,
+      "step": 4788
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3566945791244507,
+      "learning_rate": 2.951436388508892e-05,
+      "loss": 2.2317,
+      "step": 4792
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4192919731140137,
+      "learning_rate": 2.9497264021887827e-05,
+      "loss": 2.1342,
+      "step": 4796
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7992663383483887,
+      "learning_rate": 2.9480164158686728e-05,
+      "loss": 1.9621,
+      "step": 4800
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.591426968574524,
+      "learning_rate": 2.946306429548564e-05,
+      "loss": 2.0418,
+      "step": 4804
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5884109735488892,
+      "learning_rate": 2.9445964432284546e-05,
+      "loss": 2.2499,
+      "step": 4808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4876173734664917,
+      "learning_rate": 2.9428864569083446e-05,
+      "loss": 2.172,
+      "step": 4812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9083555936813354,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 2.1113,
+      "step": 4816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6316877603530884,
+      "learning_rate": 2.9394664842681258e-05,
+      "loss": 2.2434,
+      "step": 4820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.578415036201477,
+      "learning_rate": 2.9377564979480165e-05,
+      "loss": 2.1634,
+      "step": 4824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7644866704940796,
+      "learning_rate": 2.9360465116279072e-05,
+      "loss": 2.0934,
+      "step": 4828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3768229484558105,
+      "learning_rate": 2.9343365253077976e-05,
+      "loss": 1.9734,
+      "step": 4832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6598727703094482,
+      "learning_rate": 2.9326265389876884e-05,
+      "loss": 2.2622,
+      "step": 4836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.346545934677124,
+      "learning_rate": 2.9309165526675784e-05,
+      "loss": 2.1191,
+      "step": 4840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5258831977844238,
+      "learning_rate": 2.9292065663474695e-05,
+      "loss": 1.9744,
+      "step": 4844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.573669195175171,
+      "learning_rate": 2.9274965800273602e-05,
+      "loss": 2.1529,
+      "step": 4848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3979613780975342,
+      "learning_rate": 2.9257865937072503e-05,
+      "loss": 2.1268,
+      "step": 4852
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5907671451568604,
+      "learning_rate": 2.924076607387141e-05,
+      "loss": 2.0754,
+      "step": 4856
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4485595226287842,
+      "learning_rate": 2.9223666210670314e-05,
+      "loss": 2.2218,
+      "step": 4860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4896790981292725,
+      "learning_rate": 2.920656634746922e-05,
+      "loss": 2.2253,
+      "step": 4864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5012835264205933,
+      "learning_rate": 2.918946648426813e-05,
+      "loss": 2.3698,
+      "step": 4868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4697033166885376,
+      "learning_rate": 2.9172366621067033e-05,
+      "loss": 2.2947,
+      "step": 4872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.693452000617981,
+      "learning_rate": 2.915526675786594e-05,
+      "loss": 2.1249,
+      "step": 4876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5979125499725342,
+      "learning_rate": 2.913816689466484e-05,
+      "loss": 2.3482,
+      "step": 4880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.587105631828308,
+      "learning_rate": 2.912106703146375e-05,
+      "loss": 2.1287,
+      "step": 4884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5790472030639648,
+      "learning_rate": 2.910396716826266e-05,
+      "loss": 2.1099,
+      "step": 4888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4962116479873657,
+      "learning_rate": 2.908686730506156e-05,
+      "loss": 2.1216,
+      "step": 4892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.657333254814148,
+      "learning_rate": 2.9069767441860467e-05,
+      "loss": 2.1813,
+      "step": 4896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.520858645439148,
+      "learning_rate": 2.905266757865937e-05,
+      "loss": 2.1585,
+      "step": 4900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4019360542297363,
+      "learning_rate": 2.9035567715458278e-05,
+      "loss": 2.1757,
+      "step": 4904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4157730340957642,
+      "learning_rate": 2.9018467852257182e-05,
+      "loss": 2.313,
+      "step": 4908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4938081502914429,
+      "learning_rate": 2.900136798905609e-05,
+      "loss": 2.2372,
+      "step": 4912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5117088556289673,
+      "learning_rate": 2.8984268125854997e-05,
+      "loss": 2.3339,
+      "step": 4916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4263554811477661,
+      "learning_rate": 2.8967168262653897e-05,
+      "loss": 2.1498,
+      "step": 4920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3601114749908447,
+      "learning_rate": 2.8950068399452808e-05,
+      "loss": 2.1967,
+      "step": 4924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5118712186813354,
+      "learning_rate": 2.893296853625171e-05,
+      "loss": 2.2998,
+      "step": 4928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.63152015209198,
+      "learning_rate": 2.8915868673050616e-05,
+      "loss": 2.085,
+      "step": 4932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4992414712905884,
+      "learning_rate": 2.8898768809849523e-05,
+      "loss": 2.2337,
+      "step": 4936
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4842474460601807,
+      "learning_rate": 2.8881668946648427e-05,
+      "loss": 2.2246,
+      "step": 4940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.064748525619507,
+      "learning_rate": 2.8864569083447335e-05,
+      "loss": 2.2128,
+      "step": 4944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7368000745773315,
+      "learning_rate": 2.884746922024624e-05,
+      "loss": 2.1803,
+      "step": 4948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6576658487319946,
+      "learning_rate": 2.8830369357045146e-05,
+      "loss": 2.0681,
+      "step": 4952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.886851191520691,
+      "learning_rate": 2.8813269493844053e-05,
+      "loss": 2.3038,
+      "step": 4956
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6933567523956299,
+      "learning_rate": 2.8796169630642954e-05,
+      "loss": 2.1189,
+      "step": 4960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5234380960464478,
+      "learning_rate": 2.8779069767441864e-05,
+      "loss": 2.239,
+      "step": 4964
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4364981651306152,
+      "learning_rate": 2.8761969904240765e-05,
+      "loss": 2.1525,
+      "step": 4968
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5221827030181885,
+      "learning_rate": 2.8744870041039672e-05,
+      "loss": 2.0648,
+      "step": 4972
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4764840602874756,
+      "learning_rate": 2.872777017783858e-05,
+      "loss": 2.1009,
+      "step": 4976
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.361648440361023,
+      "learning_rate": 2.8710670314637484e-05,
+      "loss": 2.0012,
+      "step": 4980
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4987961053848267,
+      "learning_rate": 2.869357045143639e-05,
+      "loss": 2.3403,
+      "step": 4984
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.422187328338623,
+      "learning_rate": 2.8676470588235295e-05,
+      "loss": 2.1681,
+      "step": 4988
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5344418287277222,
+      "learning_rate": 2.8659370725034202e-05,
+      "loss": 2.1772,
+      "step": 4992
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5092509984970093,
+      "learning_rate": 2.864227086183311e-05,
+      "loss": 2.207,
+      "step": 4996
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6343382596969604,
+      "learning_rate": 2.862517099863201e-05,
+      "loss": 2.1994,
+      "step": 5000
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.447045087814331,
+      "learning_rate": 2.860807113543092e-05,
+      "loss": 2.1439,
+      "step": 5004
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4870860576629639,
+      "learning_rate": 2.859097127222982e-05,
+      "loss": 2.1185,
+      "step": 5008
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.538109302520752,
+      "learning_rate": 2.857387140902873e-05,
+      "loss": 2.1966,
+      "step": 5012
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5985993146896362,
+      "learning_rate": 2.8556771545827636e-05,
+      "loss": 2.1939,
+      "step": 5016
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.562982439994812,
+      "learning_rate": 2.853967168262654e-05,
+      "loss": 2.1722,
+      "step": 5020
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4697779417037964,
+      "learning_rate": 2.8522571819425448e-05,
+      "loss": 2.1089,
+      "step": 5024
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3964099884033203,
+      "learning_rate": 2.850547195622435e-05,
+      "loss": 2.1349,
+      "step": 5028
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.446781039237976,
+      "learning_rate": 2.848837209302326e-05,
+      "loss": 2.0189,
+      "step": 5032
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.305580735206604,
+      "learning_rate": 2.847127222982216e-05,
+      "loss": 2.1448,
+      "step": 5036
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5852243900299072,
+      "learning_rate": 2.8454172366621067e-05,
+      "loss": 2.0051,
+      "step": 5040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6528069972991943,
+      "learning_rate": 2.8437072503419977e-05,
+      "loss": 2.0865,
+      "step": 5044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5359212160110474,
+      "learning_rate": 2.8419972640218878e-05,
+      "loss": 2.301,
+      "step": 5048
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2999087572097778,
+      "learning_rate": 2.8402872777017785e-05,
+      "loss": 2.074,
+      "step": 5052
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6305381059646606,
+      "learning_rate": 2.838577291381669e-05,
+      "loss": 2.1127,
+      "step": 5056
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4597679376602173,
+      "learning_rate": 2.8368673050615597e-05,
+      "loss": 2.0304,
+      "step": 5060
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6588799953460693,
+      "learning_rate": 2.8351573187414504e-05,
+      "loss": 2.2835,
+      "step": 5064
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5638916492462158,
+      "learning_rate": 2.8334473324213408e-05,
+      "loss": 2.0453,
+      "step": 5068
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3972184658050537,
+      "learning_rate": 2.8317373461012315e-05,
+      "loss": 2.0929,
+      "step": 5072
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.448038935661316,
+      "learning_rate": 2.8300273597811216e-05,
+      "loss": 2.2133,
+      "step": 5076
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.492927074432373,
+      "learning_rate": 2.8283173734610123e-05,
+      "loss": 2.0079,
+      "step": 5080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.555188775062561,
+      "learning_rate": 2.8266073871409034e-05,
+      "loss": 2.1773,
+      "step": 5084
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4068621397018433,
+      "learning_rate": 2.8248974008207935e-05,
+      "loss": 2.288,
+      "step": 5088
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.592962622642517,
+      "learning_rate": 2.8231874145006842e-05,
+      "loss": 2.2457,
+      "step": 5092
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6329854726791382,
+      "learning_rate": 2.8214774281805746e-05,
+      "loss": 2.1736,
+      "step": 5096
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5315256118774414,
+      "learning_rate": 2.8197674418604653e-05,
+      "loss": 2.0541,
+      "step": 5100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5884466171264648,
+      "learning_rate": 2.818057455540356e-05,
+      "loss": 2.0702,
+      "step": 5104
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4373297691345215,
+      "learning_rate": 2.8163474692202464e-05,
+      "loss": 2.0738,
+      "step": 5108
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7089366912841797,
+      "learning_rate": 2.8146374829001372e-05,
+      "loss": 2.2945,
+      "step": 5112
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4914695024490356,
+      "learning_rate": 2.8129274965800272e-05,
+      "loss": 2.0974,
+      "step": 5116
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7952321767807007,
+      "learning_rate": 2.811217510259918e-05,
+      "loss": 2.0721,
+      "step": 5120
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4391483068466187,
+      "learning_rate": 2.8095075239398087e-05,
+      "loss": 2.1753,
+      "step": 5124
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6410449743270874,
+      "learning_rate": 2.807797537619699e-05,
+      "loss": 2.0375,
+      "step": 5128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5932730436325073,
+      "learning_rate": 2.80608755129959e-05,
+      "loss": 2.0907,
+      "step": 5132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5292744636535645,
+      "learning_rate": 2.8043775649794802e-05,
+      "loss": 2.266,
+      "step": 5136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.472765326499939,
+      "learning_rate": 2.802667578659371e-05,
+      "loss": 2.075,
+      "step": 5140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5142451524734497,
+      "learning_rate": 2.8009575923392617e-05,
+      "loss": 2.0953,
+      "step": 5144
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8267227411270142,
+      "learning_rate": 2.7992476060191518e-05,
+      "loss": 2.183,
+      "step": 5148
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5693491697311401,
+      "learning_rate": 2.797537619699043e-05,
+      "loss": 2.1086,
+      "step": 5152
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3335392475128174,
+      "learning_rate": 2.795827633378933e-05,
+      "loss": 2.0814,
+      "step": 5156
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5873955488204956,
+      "learning_rate": 2.7941176470588236e-05,
+      "loss": 2.132,
+      "step": 5160
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5618573427200317,
+      "learning_rate": 2.792407660738714e-05,
+      "loss": 2.1605,
+      "step": 5164
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.355206847190857,
+      "learning_rate": 2.7906976744186048e-05,
+      "loss": 2.0064,
+      "step": 5168
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4535104036331177,
+      "learning_rate": 2.7889876880984955e-05,
+      "loss": 2.092,
+      "step": 5172
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.477333426475525,
+      "learning_rate": 2.787277701778386e-05,
+      "loss": 2.0415,
+      "step": 5176
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4349262714385986,
+      "learning_rate": 2.7855677154582766e-05,
+      "loss": 2.1695,
+      "step": 5180
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4431016445159912,
+      "learning_rate": 2.7838577291381667e-05,
+      "loss": 2.1246,
+      "step": 5184
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.419053554534912,
+      "learning_rate": 2.7821477428180574e-05,
+      "loss": 2.1279,
+      "step": 5188
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5947861671447754,
+      "learning_rate": 2.7804377564979485e-05,
+      "loss": 2.1138,
+      "step": 5192
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8315672874450684,
+      "learning_rate": 2.7787277701778385e-05,
+      "loss": 2.172,
+      "step": 5196
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.532003402709961,
+      "learning_rate": 2.7770177838577293e-05,
+      "loss": 2.0775,
+      "step": 5200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6215981245040894,
+      "learning_rate": 2.7753077975376197e-05,
+      "loss": 2.0522,
+      "step": 5204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8606538772583008,
+      "learning_rate": 2.7735978112175104e-05,
+      "loss": 2.0371,
+      "step": 5208
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9666029214859009,
+      "learning_rate": 2.771887824897401e-05,
+      "loss": 2.256,
+      "step": 5212
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5295203924179077,
+      "learning_rate": 2.7701778385772915e-05,
+      "loss": 2.0297,
+      "step": 5216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.395086407661438,
+      "learning_rate": 2.7684678522571823e-05,
+      "loss": 2.1393,
+      "step": 5220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.396425485610962,
+      "learning_rate": 2.7667578659370723e-05,
+      "loss": 1.9676,
+      "step": 5224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5650209188461304,
+      "learning_rate": 2.765047879616963e-05,
+      "loss": 2.1365,
+      "step": 5228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.447646975517273,
+      "learning_rate": 2.763337893296854e-05,
+      "loss": 1.9595,
+      "step": 5232
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.438038945198059,
+      "learning_rate": 2.7616279069767442e-05,
+      "loss": 1.9664,
+      "step": 5236
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4711318016052246,
+      "learning_rate": 2.759917920656635e-05,
+      "loss": 2.0887,
+      "step": 5240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4893648624420166,
+      "learning_rate": 2.7582079343365253e-05,
+      "loss": 1.9264,
+      "step": 5244
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.579157829284668,
+      "learning_rate": 2.756497948016416e-05,
+      "loss": 2.126,
+      "step": 5248
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.480891466140747,
+      "learning_rate": 2.7547879616963068e-05,
+      "loss": 2.2045,
+      "step": 5252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8084710836410522,
+      "learning_rate": 2.7530779753761972e-05,
+      "loss": 2.2102,
+      "step": 5256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4894654750823975,
+      "learning_rate": 2.751367989056088e-05,
+      "loss": 2.1216,
+      "step": 5260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.478981852531433,
+      "learning_rate": 2.749658002735978e-05,
+      "loss": 2.0361,
+      "step": 5264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5740448236465454,
+      "learning_rate": 2.7479480164158687e-05,
+      "loss": 1.9947,
+      "step": 5268
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4918889999389648,
+      "learning_rate": 2.7462380300957598e-05,
+      "loss": 2.068,
+      "step": 5272
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.586724877357483,
+      "learning_rate": 2.74452804377565e-05,
+      "loss": 1.9932,
+      "step": 5276
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7443528175354004,
+      "learning_rate": 2.7428180574555406e-05,
+      "loss": 2.2906,
+      "step": 5280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.630373239517212,
+      "learning_rate": 2.741108071135431e-05,
+      "loss": 2.161,
+      "step": 5284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3437303304672241,
+      "learning_rate": 2.7393980848153217e-05,
+      "loss": 1.9711,
+      "step": 5288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6629517078399658,
+      "learning_rate": 2.7376880984952118e-05,
+      "loss": 2.1511,
+      "step": 5292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.507279872894287,
+      "learning_rate": 2.735978112175103e-05,
+      "loss": 2.0911,
+      "step": 5296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5633537769317627,
+      "learning_rate": 2.7342681258549936e-05,
+      "loss": 2.0376,
+      "step": 5300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.524722933769226,
+      "learning_rate": 2.7325581395348836e-05,
+      "loss": 2.131,
+      "step": 5304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4950675964355469,
+      "learning_rate": 2.7308481532147744e-05,
+      "loss": 2.1594,
+      "step": 5308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6241108179092407,
+      "learning_rate": 2.7291381668946648e-05,
+      "loss": 2.1674,
+      "step": 5312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6131986379623413,
+      "learning_rate": 2.7274281805745555e-05,
+      "loss": 2.0742,
+      "step": 5316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5340795516967773,
+      "learning_rate": 2.7257181942544462e-05,
+      "loss": 2.0877,
+      "step": 5320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.586997389793396,
+      "learning_rate": 2.7240082079343366e-05,
+      "loss": 2.2434,
+      "step": 5324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.51520836353302,
+      "learning_rate": 2.7222982216142274e-05,
+      "loss": 2.0303,
+      "step": 5328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6436048746109009,
+      "learning_rate": 2.7205882352941174e-05,
+      "loss": 2.0539,
+      "step": 5332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3098795413970947,
+      "learning_rate": 2.7188782489740085e-05,
+      "loss": 2.2811,
+      "step": 5336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6447290182113647,
+      "learning_rate": 2.7171682626538992e-05,
+      "loss": 2.1018,
+      "step": 5340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8019658327102661,
+      "learning_rate": 2.7154582763337893e-05,
+      "loss": 2.0778,
+      "step": 5344
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7310752868652344,
+      "learning_rate": 2.71374829001368e-05,
+      "loss": 2.1467,
+      "step": 5348
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4999805688858032,
+      "learning_rate": 2.7120383036935704e-05,
+      "loss": 2.134,
+      "step": 5352
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4711189270019531,
+      "learning_rate": 2.710328317373461e-05,
+      "loss": 2.1542,
+      "step": 5356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.477725863456726,
+      "learning_rate": 2.708618331053352e-05,
+      "loss": 2.0225,
+      "step": 5360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6154271364212036,
+      "learning_rate": 2.7069083447332423e-05,
+      "loss": 2.1603,
+      "step": 5364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6522784233093262,
+      "learning_rate": 2.705198358413133e-05,
+      "loss": 2.2915,
+      "step": 5368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6237088441848755,
+      "learning_rate": 2.703488372093023e-05,
+      "loss": 1.9629,
+      "step": 5372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8103796243667603,
+      "learning_rate": 2.701778385772914e-05,
+      "loss": 2.0765,
+      "step": 5376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4163451194763184,
+      "learning_rate": 2.700068399452805e-05,
+      "loss": 2.0539,
+      "step": 5380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.560591697692871,
+      "learning_rate": 2.698358413132695e-05,
+      "loss": 2.1445,
+      "step": 5384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5179965496063232,
+      "learning_rate": 2.6966484268125857e-05,
+      "loss": 2.1425,
+      "step": 5388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5749553442001343,
+      "learning_rate": 2.694938440492476e-05,
+      "loss": 1.8312,
+      "step": 5392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3497620820999146,
+      "learning_rate": 2.6932284541723668e-05,
+      "loss": 1.9427,
+      "step": 5396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3808335065841675,
+      "learning_rate": 2.6915184678522575e-05,
+      "loss": 2.2364,
+      "step": 5400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8348784446716309,
+      "learning_rate": 2.689808481532148e-05,
+      "loss": 1.9962,
+      "step": 5404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3197436332702637,
+      "learning_rate": 2.6880984952120387e-05,
+      "loss": 2.1261,
+      "step": 5408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.622539758682251,
+      "learning_rate": 2.6863885088919287e-05,
+      "loss": 2.2002,
+      "step": 5412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4938942193984985,
+      "learning_rate": 2.6846785225718198e-05,
+      "loss": 2.0079,
+      "step": 5416
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4376903772354126,
+      "learning_rate": 2.68296853625171e-05,
+      "loss": 1.9435,
+      "step": 5420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7015161514282227,
+      "learning_rate": 2.6812585499316006e-05,
+      "loss": 2.1806,
+      "step": 5424
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7079006433486938,
+      "learning_rate": 2.6795485636114913e-05,
+      "loss": 2.1705,
+      "step": 5428
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5330369472503662,
+      "learning_rate": 2.6778385772913817e-05,
+      "loss": 2.109,
+      "step": 5432
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.606583595275879,
+      "learning_rate": 2.6761285909712724e-05,
+      "loss": 2.1793,
+      "step": 5436
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5606343746185303,
+      "learning_rate": 2.674418604651163e-05,
+      "loss": 2.1413,
+      "step": 5440
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5075594186782837,
+      "learning_rate": 2.6727086183310536e-05,
+      "loss": 2.1066,
+      "step": 5444
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3254252672195435,
+      "learning_rate": 2.6709986320109443e-05,
+      "loss": 1.9901,
+      "step": 5448
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5555039644241333,
+      "learning_rate": 2.6692886456908344e-05,
+      "loss": 2.1421,
+      "step": 5452
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5434138774871826,
+      "learning_rate": 2.6675786593707254e-05,
+      "loss": 2.1533,
+      "step": 5456
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5345412492752075,
+      "learning_rate": 2.6658686730506155e-05,
+      "loss": 2.1013,
+      "step": 5460
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5029314756393433,
+      "learning_rate": 2.6641586867305062e-05,
+      "loss": 2.0473,
+      "step": 5464
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.668525218963623,
+      "learning_rate": 2.662448700410397e-05,
+      "loss": 2.0636,
+      "step": 5468
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5454894304275513,
+      "learning_rate": 2.6607387140902874e-05,
+      "loss": 2.1219,
+      "step": 5472
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3577351570129395,
+      "learning_rate": 2.659028727770178e-05,
+      "loss": 1.9395,
+      "step": 5476
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7034279108047485,
+      "learning_rate": 2.6573187414500685e-05,
+      "loss": 2.1734,
+      "step": 5480
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.519974708557129,
+      "learning_rate": 2.6556087551299592e-05,
+      "loss": 2.0007,
+      "step": 5484
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6780128479003906,
+      "learning_rate": 2.65389876880985e-05,
+      "loss": 2.2113,
+      "step": 5488
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5499041080474854,
+      "learning_rate": 2.65218878248974e-05,
+      "loss": 2.0985,
+      "step": 5492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5147311687469482,
+      "learning_rate": 2.6504787961696307e-05,
+      "loss": 2.0704,
+      "step": 5496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4439009428024292,
+      "learning_rate": 2.648768809849521e-05,
+      "loss": 1.9784,
+      "step": 5500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5315637588500977,
+      "learning_rate": 2.647058823529412e-05,
+      "loss": 2.0466,
+      "step": 5504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.514778971672058,
+      "learning_rate": 2.6453488372093026e-05,
+      "loss": 2.1113,
+      "step": 5508
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.898563027381897,
+      "learning_rate": 2.643638850889193e-05,
+      "loss": 2.129,
+      "step": 5512
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4862741231918335,
+      "learning_rate": 2.6419288645690837e-05,
+      "loss": 2.1319,
+      "step": 5516
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5534049272537231,
+      "learning_rate": 2.6402188782489738e-05,
+      "loss": 2.1904,
+      "step": 5520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6791940927505493,
+      "learning_rate": 2.638508891928865e-05,
+      "loss": 2.0653,
+      "step": 5524
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4288556575775146,
+      "learning_rate": 2.6367989056087556e-05,
+      "loss": 2.1007,
+      "step": 5528
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4318090677261353,
+      "learning_rate": 2.6350889192886457e-05,
+      "loss": 2.0791,
+      "step": 5532
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5377432107925415,
+      "learning_rate": 2.6333789329685364e-05,
+      "loss": 2.2141,
+      "step": 5536
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6055461168289185,
+      "learning_rate": 2.6316689466484268e-05,
+      "loss": 2.0733,
+      "step": 5540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4851833581924438,
+      "learning_rate": 2.6299589603283175e-05,
+      "loss": 2.1332,
+      "step": 5544
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4060307741165161,
+      "learning_rate": 2.628248974008208e-05,
+      "loss": 2.0943,
+      "step": 5548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5619680881500244,
+      "learning_rate": 2.6265389876880987e-05,
+      "loss": 2.1601,
+      "step": 5552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6302011013031006,
+      "learning_rate": 2.6248290013679894e-05,
+      "loss": 2.1288,
+      "step": 5556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.514349102973938,
+      "learning_rate": 2.6231190150478794e-05,
+      "loss": 1.9701,
+      "step": 5560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5871026515960693,
+      "learning_rate": 2.6214090287277705e-05,
+      "loss": 2.0344,
+      "step": 5564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5523772239685059,
+      "learning_rate": 2.6196990424076606e-05,
+      "loss": 2.0973,
+      "step": 5568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4616386890411377,
+      "learning_rate": 2.6179890560875513e-05,
+      "loss": 2.1624,
+      "step": 5572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.53379487991333,
+      "learning_rate": 2.616279069767442e-05,
+      "loss": 2.0887,
+      "step": 5576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6133915185928345,
+      "learning_rate": 2.6145690834473324e-05,
+      "loss": 2.0548,
+      "step": 5580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7219460010528564,
+      "learning_rate": 2.6128590971272232e-05,
+      "loss": 2.0882,
+      "step": 5584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4539477825164795,
+      "learning_rate": 2.6111491108071136e-05,
+      "loss": 1.9867,
+      "step": 5588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6804320812225342,
+      "learning_rate": 2.6094391244870043e-05,
+      "loss": 2.2284,
+      "step": 5592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6927332878112793,
+      "learning_rate": 2.607729138166895e-05,
+      "loss": 2.0798,
+      "step": 5596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5225088596343994,
+      "learning_rate": 2.606019151846785e-05,
+      "loss": 2.1072,
+      "step": 5600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4824042320251465,
+      "learning_rate": 2.604309165526676e-05,
+      "loss": 2.0789,
+      "step": 5604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3722364902496338,
+      "learning_rate": 2.6025991792065662e-05,
+      "loss": 2.1098,
+      "step": 5608
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4595104455947876,
+      "learning_rate": 2.600889192886457e-05,
+      "loss": 2.119,
+      "step": 5612
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.418157696723938,
+      "learning_rate": 2.5991792065663477e-05,
+      "loss": 2.1059,
+      "step": 5616
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4522316455841064,
+      "learning_rate": 2.597469220246238e-05,
+      "loss": 2.2825,
+      "step": 5620
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.685712456703186,
+      "learning_rate": 2.5957592339261288e-05,
+      "loss": 2.0914,
+      "step": 5624
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5950579643249512,
+      "learning_rate": 2.5940492476060192e-05,
+      "loss": 2.0957,
+      "step": 5628
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5742863416671753,
+      "learning_rate": 2.59233926128591e-05,
+      "loss": 2.2025,
+      "step": 5632
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5395307540893555,
+      "learning_rate": 2.5906292749658007e-05,
+      "loss": 2.0688,
+      "step": 5636
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5343267917633057,
+      "learning_rate": 2.5889192886456907e-05,
+      "loss": 2.1783,
+      "step": 5640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4409688711166382,
+      "learning_rate": 2.5872093023255818e-05,
+      "loss": 1.9016,
+      "step": 5644
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5827747583389282,
+      "learning_rate": 2.585499316005472e-05,
+      "loss": 2.1501,
+      "step": 5648
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6408090591430664,
+      "learning_rate": 2.5837893296853626e-05,
+      "loss": 2.0266,
+      "step": 5652
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.454546332359314,
+      "learning_rate": 2.5820793433652533e-05,
+      "loss": 1.9665,
+      "step": 5656
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4659438133239746,
+      "learning_rate": 2.5803693570451437e-05,
+      "loss": 2.0251,
+      "step": 5660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6807434558868408,
+      "learning_rate": 2.5786593707250345e-05,
+      "loss": 2.0587,
+      "step": 5664
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4312716722488403,
+      "learning_rate": 2.576949384404925e-05,
+      "loss": 1.9798,
+      "step": 5668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5319584608078003,
+      "learning_rate": 2.5752393980848156e-05,
+      "loss": 2.0179,
+      "step": 5672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5227807760238647,
+      "learning_rate": 2.5735294117647057e-05,
+      "loss": 2.079,
+      "step": 5676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5442856550216675,
+      "learning_rate": 2.5718194254445964e-05,
+      "loss": 2.0757,
+      "step": 5680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4400264024734497,
+      "learning_rate": 2.5701094391244875e-05,
+      "loss": 2.0712,
+      "step": 5684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5786032676696777,
+      "learning_rate": 2.5683994528043775e-05,
+      "loss": 2.1628,
+      "step": 5688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5079550743103027,
+      "learning_rate": 2.5666894664842683e-05,
+      "loss": 2.0104,
+      "step": 5692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6693617105484009,
+      "learning_rate": 2.5649794801641587e-05,
+      "loss": 1.9668,
+      "step": 5696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6841495037078857,
+      "learning_rate": 2.5632694938440494e-05,
+      "loss": 2.2648,
+      "step": 5700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4236379861831665,
+      "learning_rate": 2.56155950752394e-05,
+      "loss": 2.1775,
+      "step": 5704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4873919486999512,
+      "learning_rate": 2.5598495212038305e-05,
+      "loss": 1.8591,
+      "step": 5708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5848278999328613,
+      "learning_rate": 2.5581395348837212e-05,
+      "loss": 2.0736,
+      "step": 5712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9348700046539307,
+      "learning_rate": 2.5564295485636113e-05,
+      "loss": 2.1204,
+      "step": 5716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5655988454818726,
+      "learning_rate": 2.554719562243502e-05,
+      "loss": 2.0318,
+      "step": 5720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6480430364608765,
+      "learning_rate": 2.553009575923393e-05,
+      "loss": 2.0188,
+      "step": 5724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4328726530075073,
+      "learning_rate": 2.5512995896032832e-05,
+      "loss": 2.1973,
+      "step": 5728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6577439308166504,
+      "learning_rate": 2.549589603283174e-05,
+      "loss": 2.0672,
+      "step": 5732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3879199028015137,
+      "learning_rate": 2.5478796169630643e-05,
+      "loss": 1.9124,
+      "step": 5736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4111934900283813,
+      "learning_rate": 2.546169630642955e-05,
+      "loss": 1.9643,
+      "step": 5740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6698105335235596,
+      "learning_rate": 2.5444596443228458e-05,
+      "loss": 2.167,
+      "step": 5744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.643890380859375,
+      "learning_rate": 2.542749658002736e-05,
+      "loss": 2.1449,
+      "step": 5748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4590214490890503,
+      "learning_rate": 2.541039671682627e-05,
+      "loss": 2.2577,
+      "step": 5752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5629549026489258,
+      "learning_rate": 2.539329685362517e-05,
+      "loss": 2.2183,
+      "step": 5756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4380923509597778,
+      "learning_rate": 2.5376196990424077e-05,
+      "loss": 1.9198,
+      "step": 5760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4581702947616577,
+      "learning_rate": 2.5359097127222988e-05,
+      "loss": 2.0363,
+      "step": 5764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3352808952331543,
+      "learning_rate": 2.5341997264021888e-05,
+      "loss": 2.1038,
+      "step": 5768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.604493498802185,
+      "learning_rate": 2.5324897400820796e-05,
+      "loss": 2.1413,
+      "step": 5772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.360095739364624,
+      "learning_rate": 2.53077975376197e-05,
+      "loss": 2.0967,
+      "step": 5776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.400731086730957,
+      "learning_rate": 2.5290697674418607e-05,
+      "loss": 2.0638,
+      "step": 5780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.611741304397583,
+      "learning_rate": 2.5273597811217514e-05,
+      "loss": 1.9994,
+      "step": 5784
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.614253044128418,
+      "learning_rate": 2.5256497948016418e-05,
+      "loss": 2.0905,
+      "step": 5788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5998769998550415,
+      "learning_rate": 2.5239398084815325e-05,
+      "loss": 2.2255,
+      "step": 5792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5840561389923096,
+      "learning_rate": 2.5222298221614226e-05,
+      "loss": 2.0949,
+      "step": 5796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5957179069519043,
+      "learning_rate": 2.5205198358413133e-05,
+      "loss": 1.9565,
+      "step": 5800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5299445390701294,
+      "learning_rate": 2.5188098495212037e-05,
+      "loss": 2.0896,
+      "step": 5804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5328844785690308,
+      "learning_rate": 2.5170998632010945e-05,
+      "loss": 2.0734,
+      "step": 5808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.43681800365448,
+      "learning_rate": 2.5153898768809852e-05,
+      "loss": 1.9839,
+      "step": 5812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5096018314361572,
+      "learning_rate": 2.5136798905608756e-05,
+      "loss": 2.1289,
+      "step": 5816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5648272037506104,
+      "learning_rate": 2.5119699042407663e-05,
+      "loss": 1.9608,
+      "step": 5820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7235368490219116,
+      "learning_rate": 2.5102599179206564e-05,
+      "loss": 2.0752,
+      "step": 5824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6707707643508911,
+      "learning_rate": 2.5085499316005475e-05,
+      "loss": 2.2227,
+      "step": 5828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5506484508514404,
+      "learning_rate": 2.5068399452804382e-05,
+      "loss": 1.9812,
+      "step": 5832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.50376296043396,
+      "learning_rate": 2.5051299589603283e-05,
+      "loss": 2.0457,
+      "step": 5836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5945008993148804,
+      "learning_rate": 2.503419972640219e-05,
+      "loss": 2.0204,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6294224262237549,
+      "learning_rate": 2.5017099863201094e-05,
+      "loss": 2.0159,
+      "step": 5844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4806592464447021,
+      "learning_rate": 2.5e-05,
+      "loss": 2.0233,
+      "step": 5848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4078065156936646,
+      "learning_rate": 2.4982900136798905e-05,
+      "loss": 1.9797,
+      "step": 5852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5420806407928467,
+      "learning_rate": 2.4965800273597812e-05,
+      "loss": 2.0582,
+      "step": 5856
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4052869081497192,
+      "learning_rate": 2.494870041039672e-05,
+      "loss": 2.0785,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4887535572052002,
+      "learning_rate": 2.4931600547195624e-05,
+      "loss": 2.1067,
+      "step": 5864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.723113775253296,
+      "learning_rate": 2.4914500683994528e-05,
+      "loss": 1.973,
+      "step": 5868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.109470367431641,
+      "learning_rate": 2.4897400820793435e-05,
+      "loss": 2.2463,
+      "step": 5872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.645058274269104,
+      "learning_rate": 2.488030095759234e-05,
+      "loss": 2.1643,
+      "step": 5876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4571024179458618,
+      "learning_rate": 2.4863201094391243e-05,
+      "loss": 2.0342,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.738869309425354,
+      "learning_rate": 2.4846101231190154e-05,
+      "loss": 2.106,
+      "step": 5884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.58241868019104,
+      "learning_rate": 2.4829001367989058e-05,
+      "loss": 2.1393,
+      "step": 5888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5051425695419312,
+      "learning_rate": 2.481190150478796e-05,
+      "loss": 2.1032,
+      "step": 5892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5358223915100098,
+      "learning_rate": 2.479480164158687e-05,
+      "loss": 2.065,
+      "step": 5896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3522099256515503,
+      "learning_rate": 2.4777701778385773e-05,
+      "loss": 1.9355,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4918060302734375,
+      "learning_rate": 2.476060191518468e-05,
+      "loss": 2.0173,
+      "step": 5904
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5328975915908813,
+      "learning_rate": 2.4743502051983584e-05,
+      "loss": 2.1863,
+      "step": 5908
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4947617053985596,
+      "learning_rate": 2.472640218878249e-05,
+      "loss": 2.1405,
+      "step": 5912
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.535042643547058,
+      "learning_rate": 2.4709302325581396e-05,
+      "loss": 1.9693,
+      "step": 5916
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.497016191482544,
+      "learning_rate": 2.46922024623803e-05,
+      "loss": 1.9618,
+      "step": 5920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.490457534790039,
+      "learning_rate": 2.467510259917921e-05,
+      "loss": 2.0978,
+      "step": 5924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4083185195922852,
+      "learning_rate": 2.4658002735978114e-05,
+      "loss": 1.9533,
+      "step": 5928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.638238549232483,
+      "learning_rate": 2.4640902872777018e-05,
+      "loss": 2.0625,
+      "step": 5932
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6385629177093506,
+      "learning_rate": 2.4623803009575925e-05,
+      "loss": 1.9919,
+      "step": 5936
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3852545022964478,
+      "learning_rate": 2.460670314637483e-05,
+      "loss": 2.0971,
+      "step": 5940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4530996084213257,
+      "learning_rate": 2.4589603283173733e-05,
+      "loss": 2.0959,
+      "step": 5944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.573773980140686,
+      "learning_rate": 2.457250341997264e-05,
+      "loss": 2.1314,
+      "step": 5948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4911787509918213,
+      "learning_rate": 2.4555403556771548e-05,
+      "loss": 1.9789,
+      "step": 5952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5948853492736816,
+      "learning_rate": 2.4538303693570452e-05,
+      "loss": 2.1686,
+      "step": 5956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3944917917251587,
+      "learning_rate": 2.4521203830369356e-05,
+      "loss": 1.9249,
+      "step": 5960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5396506786346436,
+      "learning_rate": 2.4504103967168263e-05,
+      "loss": 2.0486,
+      "step": 5964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5431126356124878,
+      "learning_rate": 2.448700410396717e-05,
+      "loss": 1.9906,
+      "step": 5968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4996250867843628,
+      "learning_rate": 2.4469904240766075e-05,
+      "loss": 2.0529,
+      "step": 5972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.653665542602539,
+      "learning_rate": 2.4452804377564982e-05,
+      "loss": 2.0667,
+      "step": 5976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.543585181236267,
+      "learning_rate": 2.4435704514363886e-05,
+      "loss": 2.2203,
+      "step": 5980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4741888046264648,
+      "learning_rate": 2.441860465116279e-05,
+      "loss": 2.0958,
+      "step": 5984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7316616773605347,
+      "learning_rate": 2.4401504787961697e-05,
+      "loss": 2.0456,
+      "step": 5988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6189284324645996,
+      "learning_rate": 2.4384404924760605e-05,
+      "loss": 2.0137,
+      "step": 5992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3436368703842163,
+      "learning_rate": 2.436730506155951e-05,
+      "loss": 2.0219,
+      "step": 5996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.45026433467865,
+      "learning_rate": 2.4350205198358412e-05,
+      "loss": 1.938,
+      "step": 6000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4618995189666748,
+      "learning_rate": 2.433310533515732e-05,
+      "loss": 2.1091,
+      "step": 6004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3523361682891846,
+      "learning_rate": 2.4316005471956224e-05,
+      "loss": 2.0435,
+      "step": 6008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.414756178855896,
+      "learning_rate": 2.429890560875513e-05,
+      "loss": 2.0553,
+      "step": 6012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8548967838287354,
+      "learning_rate": 2.428180574555404e-05,
+      "loss": 2.1002,
+      "step": 6016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.431153416633606,
+      "learning_rate": 2.4264705882352942e-05,
+      "loss": 1.9441,
+      "step": 6020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4049307107925415,
+      "learning_rate": 2.4247606019151846e-05,
+      "loss": 2.1948,
+      "step": 6024
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4106427431106567,
+      "learning_rate": 2.4230506155950754e-05,
+      "loss": 1.9183,
+      "step": 6028
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8905525207519531,
+      "learning_rate": 2.421340629274966e-05,
+      "loss": 1.961,
+      "step": 6032
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5894415378570557,
+      "learning_rate": 2.4196306429548565e-05,
+      "loss": 2.1565,
+      "step": 6036
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6033257246017456,
+      "learning_rate": 2.417920656634747e-05,
+      "loss": 1.8504,
+      "step": 6040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1501522064208984,
+      "learning_rate": 2.4162106703146376e-05,
+      "loss": 2.0988,
+      "step": 6044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.688828945159912,
+      "learning_rate": 2.414500683994528e-05,
+      "loss": 2.0132,
+      "step": 6048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.479693055152893,
+      "learning_rate": 2.4127906976744188e-05,
+      "loss": 2.0963,
+      "step": 6052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6395564079284668,
+      "learning_rate": 2.4110807113543095e-05,
+      "loss": 2.1367,
+      "step": 6056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4523251056671143,
+      "learning_rate": 2.4093707250342e-05,
+      "loss": 2.0349,
+      "step": 6060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.580451488494873,
+      "learning_rate": 2.4076607387140903e-05,
+      "loss": 1.9535,
+      "step": 6064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5951658487319946,
+      "learning_rate": 2.405950752393981e-05,
+      "loss": 2.1379,
+      "step": 6068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5413308143615723,
+      "learning_rate": 2.4042407660738714e-05,
+      "loss": 2.0473,
+      "step": 6072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5242167711257935,
+      "learning_rate": 2.402530779753762e-05,
+      "loss": 2.026,
+      "step": 6076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.484012484550476,
+      "learning_rate": 2.4008207934336525e-05,
+      "loss": 2.0974,
+      "step": 6080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6217378377914429,
+      "learning_rate": 2.3991108071135433e-05,
+      "loss": 2.0572,
+      "step": 6084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.622316837310791,
+      "learning_rate": 2.3974008207934337e-05,
+      "loss": 2.111,
+      "step": 6088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4300661087036133,
+      "learning_rate": 2.395690834473324e-05,
+      "loss": 2.0311,
+      "step": 6092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4880815744400024,
+      "learning_rate": 2.393980848153215e-05,
+      "loss": 1.9217,
+      "step": 6096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5886203050613403,
+      "learning_rate": 2.3922708618331055e-05,
+      "loss": 2.1043,
+      "step": 6100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4203860759735107,
+      "learning_rate": 2.390560875512996e-05,
+      "loss": 2.1326,
+      "step": 6104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.777569055557251,
+      "learning_rate": 2.3888508891928867e-05,
+      "loss": 2.0363,
+      "step": 6108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4837594032287598,
+      "learning_rate": 2.387140902872777e-05,
+      "loss": 1.8943,
+      "step": 6112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6649631261825562,
+      "learning_rate": 2.3854309165526678e-05,
+      "loss": 2.1792,
+      "step": 6116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4544529914855957,
+      "learning_rate": 2.3837209302325582e-05,
+      "loss": 1.9499,
+      "step": 6120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4706698656082153,
+      "learning_rate": 2.382010943912449e-05,
+      "loss": 1.9453,
+      "step": 6124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8051691055297852,
+      "learning_rate": 2.3803009575923393e-05,
+      "loss": 2.124,
+      "step": 6128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5130858421325684,
+      "learning_rate": 2.3785909712722297e-05,
+      "loss": 2.0967,
+      "step": 6132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6385159492492676,
+      "learning_rate": 2.3768809849521205e-05,
+      "loss": 2.155,
+      "step": 6136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.579830527305603,
+      "learning_rate": 2.3751709986320112e-05,
+      "loss": 2.0291,
+      "step": 6140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6579816341400146,
+      "learning_rate": 2.3734610123119016e-05,
+      "loss": 1.9364,
+      "step": 6144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4621435403823853,
+      "learning_rate": 2.3717510259917923e-05,
+      "loss": 2.0041,
+      "step": 6148
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4901877641677856,
+      "learning_rate": 2.3700410396716827e-05,
+      "loss": 2.0768,
+      "step": 6152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5023964643478394,
+      "learning_rate": 2.368331053351573e-05,
+      "loss": 1.8666,
+      "step": 6156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.394523024559021,
+      "learning_rate": 2.366621067031464e-05,
+      "loss": 2.0315,
+      "step": 6160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3989911079406738,
+      "learning_rate": 2.3649110807113546e-05,
+      "loss": 2.0719,
+      "step": 6164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6071873903274536,
+      "learning_rate": 2.363201094391245e-05,
+      "loss": 2.07,
+      "step": 6168
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4817439317703247,
+      "learning_rate": 2.3614911080711354e-05,
+      "loss": 2.0119,
+      "step": 6172
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3561433553695679,
+      "learning_rate": 2.359781121751026e-05,
+      "loss": 1.9405,
+      "step": 6176
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5416829586029053,
+      "learning_rate": 2.358071135430917e-05,
+      "loss": 2.0001,
+      "step": 6180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4573452472686768,
+      "learning_rate": 2.3563611491108072e-05,
+      "loss": 1.9938,
+      "step": 6184
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6019344329833984,
+      "learning_rate": 2.354651162790698e-05,
+      "loss": 1.9446,
+      "step": 6188
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7344564199447632,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 1.862,
+      "step": 6192
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6940727233886719,
+      "learning_rate": 2.3512311901504788e-05,
+      "loss": 1.9843,
+      "step": 6196
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5194576978683472,
+      "learning_rate": 2.3495212038303695e-05,
+      "loss": 2.0475,
+      "step": 6200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7754817008972168,
+      "learning_rate": 2.3478112175102602e-05,
+      "loss": 2.0975,
+      "step": 6204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.676499605178833,
+      "learning_rate": 2.3461012311901506e-05,
+      "loss": 2.1042,
+      "step": 6208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5539844036102295,
+      "learning_rate": 2.344391244870041e-05,
+      "loss": 2.0322,
+      "step": 6212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5558497905731201,
+      "learning_rate": 2.3426812585499318e-05,
+      "loss": 2.1731,
+      "step": 6216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4876151084899902,
+      "learning_rate": 2.340971272229822e-05,
+      "loss": 1.9864,
+      "step": 6220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5103055238723755,
+      "learning_rate": 2.339261285909713e-05,
+      "loss": 2.0973,
+      "step": 6224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.617957353591919,
+      "learning_rate": 2.3375512995896033e-05,
+      "loss": 1.8867,
+      "step": 6228
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.417094111442566,
+      "learning_rate": 2.335841313269494e-05,
+      "loss": 2.0805,
+      "step": 6232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5233298540115356,
+      "learning_rate": 2.3341313269493844e-05,
+      "loss": 1.9551,
+      "step": 6236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5211933851242065,
+      "learning_rate": 2.3324213406292748e-05,
+      "loss": 2.0242,
+      "step": 6240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4824392795562744,
+      "learning_rate": 2.330711354309166e-05,
+      "loss": 1.9477,
+      "step": 6244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4873578548431396,
+      "learning_rate": 2.3290013679890563e-05,
+      "loss": 2.1701,
+      "step": 6248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5688248872756958,
+      "learning_rate": 2.3272913816689467e-05,
+      "loss": 2.0478,
+      "step": 6252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4556810855865479,
+      "learning_rate": 2.3255813953488374e-05,
+      "loss": 1.9112,
+      "step": 6256
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8388420343399048,
+      "learning_rate": 2.3238714090287278e-05,
+      "loss": 2.1023,
+      "step": 6260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.615323543548584,
+      "learning_rate": 2.3221614227086182e-05,
+      "loss": 2.0831,
+      "step": 6264
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.488853096961975,
+      "learning_rate": 2.320451436388509e-05,
+      "loss": 1.9403,
+      "step": 6268
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5217219591140747,
+      "learning_rate": 2.3187414500683997e-05,
+      "loss": 2.1556,
+      "step": 6272
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4916918277740479,
+      "learning_rate": 2.31703146374829e-05,
+      "loss": 2.0442,
+      "step": 6276
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.976369857788086,
+      "learning_rate": 2.3153214774281805e-05,
+      "loss": 2.0054,
+      "step": 6280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6137586832046509,
+      "learning_rate": 2.3136114911080712e-05,
+      "loss": 2.0149,
+      "step": 6284
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3688173294067383,
+      "learning_rate": 2.311901504787962e-05,
+      "loss": 2.0518,
+      "step": 6288
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.489639163017273,
+      "learning_rate": 2.3101915184678523e-05,
+      "loss": 2.186,
+      "step": 6292
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4436581134796143,
+      "learning_rate": 2.308481532147743e-05,
+      "loss": 1.9726,
+      "step": 6296
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.76146125793457,
+      "learning_rate": 2.3067715458276335e-05,
+      "loss": 1.9351,
+      "step": 6300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7009334564208984,
+      "learning_rate": 2.305061559507524e-05,
+      "loss": 1.9967,
+      "step": 6304
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.611024260520935,
+      "learning_rate": 2.3033515731874146e-05,
+      "loss": 2.0911,
+      "step": 6308
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5418471097946167,
+      "learning_rate": 2.3016415868673053e-05,
+      "loss": 1.9721,
+      "step": 6312
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5658395290374756,
+      "learning_rate": 2.2999316005471957e-05,
+      "loss": 2.0573,
+      "step": 6316
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.365360140800476,
+      "learning_rate": 2.298221614227086e-05,
+      "loss": 1.9245,
+      "step": 6320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7127231359481812,
+      "learning_rate": 2.296511627906977e-05,
+      "loss": 1.9568,
+      "step": 6324
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.572302222251892,
+      "learning_rate": 2.2948016415868672e-05,
+      "loss": 2.0137,
+      "step": 6328
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7327531576156616,
+      "learning_rate": 2.293091655266758e-05,
+      "loss": 2.0781,
+      "step": 6332
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5620208978652954,
+      "learning_rate": 2.2913816689466487e-05,
+      "loss": 2.0689,
+      "step": 6336
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4865524768829346,
+      "learning_rate": 2.289671682626539e-05,
+      "loss": 1.9261,
+      "step": 6340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6262495517730713,
+      "learning_rate": 2.2879616963064295e-05,
+      "loss": 2.1261,
+      "step": 6344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.536917805671692,
+      "learning_rate": 2.2862517099863202e-05,
+      "loss": 2.0757,
+      "step": 6348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4335952997207642,
+      "learning_rate": 2.284541723666211e-05,
+      "loss": 1.967,
+      "step": 6352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1377627849578857,
+      "learning_rate": 2.2828317373461014e-05,
+      "loss": 1.996,
+      "step": 6356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.705106496810913,
+      "learning_rate": 2.2811217510259918e-05,
+      "loss": 2.0619,
+      "step": 6360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5044035911560059,
+      "learning_rate": 2.2794117647058825e-05,
+      "loss": 2.0941,
+      "step": 6364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.485723614692688,
+      "learning_rate": 2.277701778385773e-05,
+      "loss": 2.0471,
+      "step": 6368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3270933628082275,
+      "learning_rate": 2.2759917920656636e-05,
+      "loss": 1.9035,
+      "step": 6372
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8408259153366089,
+      "learning_rate": 2.2742818057455544e-05,
+      "loss": 2.1103,
+      "step": 6376
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5372594594955444,
+      "learning_rate": 2.2725718194254448e-05,
+      "loss": 1.9872,
+      "step": 6380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5519907474517822,
+      "learning_rate": 2.270861833105335e-05,
+      "loss": 2.0445,
+      "step": 6384
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5492377281188965,
+      "learning_rate": 2.269151846785226e-05,
+      "loss": 2.1288,
+      "step": 6388
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5490655899047852,
+      "learning_rate": 2.2674418604651163e-05,
+      "loss": 2.0827,
+      "step": 6392
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.575005292892456,
+      "learning_rate": 2.265731874145007e-05,
+      "loss": 2.0462,
+      "step": 6396
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5636953115463257,
+      "learning_rate": 2.2640218878248974e-05,
+      "loss": 2.0272,
+      "step": 6400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4085521697998047,
+      "learning_rate": 2.262311901504788e-05,
+      "loss": 1.9316,
+      "step": 6404
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.375346064567566,
+      "learning_rate": 2.2606019151846785e-05,
+      "loss": 2.1006,
+      "step": 6408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5177054405212402,
+      "learning_rate": 2.258891928864569e-05,
+      "loss": 2.104,
+      "step": 6412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.590615153312683,
+      "learning_rate": 2.25718194254446e-05,
+      "loss": 2.054,
+      "step": 6416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6680008172988892,
+      "learning_rate": 2.2554719562243504e-05,
+      "loss": 2.0611,
+      "step": 6420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7814908027648926,
+      "learning_rate": 2.2537619699042408e-05,
+      "loss": 2.0015,
+      "step": 6424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.419728398323059,
+      "learning_rate": 2.2520519835841315e-05,
+      "loss": 2.0902,
+      "step": 6428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4798617362976074,
+      "learning_rate": 2.250341997264022e-05,
+      "loss": 2.05,
+      "step": 6432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.506953477859497,
+      "learning_rate": 2.2486320109439127e-05,
+      "loss": 2.0153,
+      "step": 6436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6882991790771484,
+      "learning_rate": 2.246922024623803e-05,
+      "loss": 1.9237,
+      "step": 6440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4662657976150513,
+      "learning_rate": 2.2452120383036938e-05,
+      "loss": 2.0524,
+      "step": 6444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5177686214447021,
+      "learning_rate": 2.2435020519835842e-05,
+      "loss": 2.0207,
+      "step": 6448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7615087032318115,
+      "learning_rate": 2.2417920656634746e-05,
+      "loss": 2.07,
+      "step": 6452
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7587693929672241,
+      "learning_rate": 2.2400820793433653e-05,
+      "loss": 2.0928,
+      "step": 6456
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6532100439071655,
+      "learning_rate": 2.238372093023256e-05,
+      "loss": 2.0425,
+      "step": 6460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4801387786865234,
+      "learning_rate": 2.2366621067031464e-05,
+      "loss": 2.1567,
+      "step": 6464
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6623640060424805,
+      "learning_rate": 2.2349521203830372e-05,
+      "loss": 2.0281,
+      "step": 6468
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5329927206039429,
+      "learning_rate": 2.2332421340629276e-05,
+      "loss": 1.9157,
+      "step": 6472
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5152140855789185,
+      "learning_rate": 2.231532147742818e-05,
+      "loss": 2.0218,
+      "step": 6476
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4606438875198364,
+      "learning_rate": 2.2298221614227087e-05,
+      "loss": 1.9652,
+      "step": 6480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4744431972503662,
+      "learning_rate": 2.2281121751025994e-05,
+      "loss": 1.9493,
+      "step": 6484
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.527673363685608,
+      "learning_rate": 2.22640218878249e-05,
+      "loss": 2.1294,
+      "step": 6488
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5893338918685913,
+      "learning_rate": 2.2246922024623802e-05,
+      "loss": 2.0014,
+      "step": 6492
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.53531014919281,
+      "learning_rate": 2.222982216142271e-05,
+      "loss": 2.1177,
+      "step": 6496
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4468746185302734,
+      "learning_rate": 2.2212722298221617e-05,
+      "loss": 1.9305,
+      "step": 6500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5749852657318115,
+      "learning_rate": 2.219562243502052e-05,
+      "loss": 2.0401,
+      "step": 6504
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6165703535079956,
+      "learning_rate": 2.217852257181943e-05,
+      "loss": 2.1496,
+      "step": 6508
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5743823051452637,
+      "learning_rate": 2.2161422708618332e-05,
+      "loss": 2.0051,
+      "step": 6512
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5725358724594116,
+      "learning_rate": 2.2144322845417236e-05,
+      "loss": 2.1795,
+      "step": 6516
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5228146314620972,
+      "learning_rate": 2.2127222982216144e-05,
+      "loss": 2.0832,
+      "step": 6520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.584122657775879,
+      "learning_rate": 2.211012311901505e-05,
+      "loss": 1.9658,
+      "step": 6524
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6192436218261719,
+      "learning_rate": 2.2093023255813955e-05,
+      "loss": 2.047,
+      "step": 6528
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4840277433395386,
+      "learning_rate": 2.207592339261286e-05,
+      "loss": 2.0195,
+      "step": 6532
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5676360130310059,
+      "learning_rate": 2.2058823529411766e-05,
+      "loss": 2.0233,
+      "step": 6536
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5410796403884888,
+      "learning_rate": 2.204172366621067e-05,
+      "loss": 1.9596,
+      "step": 6540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6324695348739624,
+      "learning_rate": 2.2024623803009577e-05,
+      "loss": 1.9401,
+      "step": 6544
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4897451400756836,
+      "learning_rate": 2.2007523939808485e-05,
+      "loss": 2.0069,
+      "step": 6548
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5398979187011719,
+      "learning_rate": 2.199042407660739e-05,
+      "loss": 2.0448,
+      "step": 6552
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3995730876922607,
+      "learning_rate": 2.1973324213406293e-05,
+      "loss": 1.9606,
+      "step": 6556
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6123522520065308,
+      "learning_rate": 2.19562243502052e-05,
+      "loss": 1.9598,
+      "step": 6560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5927785634994507,
+      "learning_rate": 2.1939124487004107e-05,
+      "loss": 1.8914,
+      "step": 6564
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3952325582504272,
+      "learning_rate": 2.192202462380301e-05,
+      "loss": 1.8675,
+      "step": 6568
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.780047059059143,
+      "learning_rate": 2.1904924760601915e-05,
+      "loss": 2.037,
+      "step": 6572
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.719232439994812,
+      "learning_rate": 2.1887824897400823e-05,
+      "loss": 1.9979,
+      "step": 6576
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8734204769134521,
+      "learning_rate": 2.1870725034199727e-05,
+      "loss": 2.0388,
+      "step": 6580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5046261548995972,
+      "learning_rate": 2.185362517099863e-05,
+      "loss": 2.0383,
+      "step": 6584
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5430729389190674,
+      "learning_rate": 2.1836525307797538e-05,
+      "loss": 2.1083,
+      "step": 6588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.647208571434021,
+      "learning_rate": 2.1819425444596445e-05,
+      "loss": 2.0813,
+      "step": 6592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5349608659744263,
+      "learning_rate": 2.180232558139535e-05,
+      "loss": 1.9528,
+      "step": 6596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6204650402069092,
+      "learning_rate": 2.1785225718194253e-05,
+      "loss": 2.1957,
+      "step": 6600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6103018522262573,
+      "learning_rate": 2.176812585499316e-05,
+      "loss": 1.9506,
+      "step": 6604
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1254968643188477,
+      "learning_rate": 2.1751025991792068e-05,
+      "loss": 2.1188,
+      "step": 6608
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5318729877471924,
+      "learning_rate": 2.1733926128590972e-05,
+      "loss": 2.0858,
+      "step": 6612
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4337228536605835,
+      "learning_rate": 2.171682626538988e-05,
+      "loss": 1.9164,
+      "step": 6616
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4926187992095947,
+      "learning_rate": 2.1699726402188783e-05,
+      "loss": 1.9784,
+      "step": 6620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.419592022895813,
+      "learning_rate": 2.1682626538987687e-05,
+      "loss": 1.9949,
+      "step": 6624
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4496034383773804,
+      "learning_rate": 2.1665526675786594e-05,
+      "loss": 2.068,
+      "step": 6628
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4800695180892944,
+      "learning_rate": 2.1648426812585502e-05,
+      "loss": 1.9682,
+      "step": 6632
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7047966718673706,
+      "learning_rate": 2.1631326949384406e-05,
+      "loss": 2.0647,
+      "step": 6636
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4624961614608765,
+      "learning_rate": 2.161422708618331e-05,
+      "loss": 1.9154,
+      "step": 6640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4518662691116333,
+      "learning_rate": 2.1597127222982217e-05,
+      "loss": 2.1044,
+      "step": 6644
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.488329529762268,
+      "learning_rate": 2.158002735978112e-05,
+      "loss": 2.0274,
+      "step": 6648
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4353028535842896,
+      "learning_rate": 2.156292749658003e-05,
+      "loss": 1.9888,
+      "step": 6652
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.470017433166504,
+      "learning_rate": 2.1545827633378936e-05,
+      "loss": 1.9887,
+      "step": 6656
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5084627866744995,
+      "learning_rate": 2.152872777017784e-05,
+      "loss": 1.9513,
+      "step": 6660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5296393632888794,
+      "learning_rate": 2.1511627906976744e-05,
+      "loss": 1.8803,
+      "step": 6664
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.597259283065796,
+      "learning_rate": 2.149452804377565e-05,
+      "loss": 1.9239,
+      "step": 6668
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.476731300354004,
+      "learning_rate": 2.1477428180574558e-05,
+      "loss": 2.0592,
+      "step": 6672
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5087945461273193,
+      "learning_rate": 2.1460328317373462e-05,
+      "loss": 2.0161,
+      "step": 6676
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6463593244552612,
+      "learning_rate": 2.1443228454172366e-05,
+      "loss": 2.0282,
+      "step": 6680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6463061571121216,
+      "learning_rate": 2.1426128590971274e-05,
+      "loss": 2.0664,
+      "step": 6684
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5039496421813965,
+      "learning_rate": 2.1409028727770177e-05,
+      "loss": 2.0983,
+      "step": 6688
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4754881858825684,
+      "learning_rate": 2.1391928864569085e-05,
+      "loss": 2.0178,
+      "step": 6692
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7630116939544678,
+      "learning_rate": 2.1374829001367992e-05,
+      "loss": 2.0397,
+      "step": 6696
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4631503820419312,
+      "learning_rate": 2.1357729138166896e-05,
+      "loss": 1.9825,
+      "step": 6700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5634981393814087,
+      "learning_rate": 2.13406292749658e-05,
+      "loss": 1.8953,
+      "step": 6704
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5709517002105713,
+      "learning_rate": 2.1323529411764707e-05,
+      "loss": 1.9242,
+      "step": 6708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5792012214660645,
+      "learning_rate": 2.130642954856361e-05,
+      "loss": 2.2404,
+      "step": 6712
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5689932107925415,
+      "learning_rate": 2.128932968536252e-05,
+      "loss": 2.1333,
+      "step": 6716
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3986164331436157,
+      "learning_rate": 2.1272229822161423e-05,
+      "loss": 1.8481,
+      "step": 6720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.62111496925354,
+      "learning_rate": 2.125512995896033e-05,
+      "loss": 2.057,
+      "step": 6724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6501810550689697,
+      "learning_rate": 2.1238030095759234e-05,
+      "loss": 2.0681,
+      "step": 6728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5174288749694824,
+      "learning_rate": 2.1220930232558138e-05,
+      "loss": 2.0297,
+      "step": 6732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5836937427520752,
+      "learning_rate": 2.120383036935705e-05,
+      "loss": 2.0201,
+      "step": 6736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6528754234313965,
+      "learning_rate": 2.1186730506155953e-05,
+      "loss": 2.0832,
+      "step": 6740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.435103416442871,
+      "learning_rate": 2.1169630642954857e-05,
+      "loss": 1.9953,
+      "step": 6744
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4534496068954468,
+      "learning_rate": 2.1152530779753764e-05,
+      "loss": 2.0343,
+      "step": 6748
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5175713300704956,
+      "learning_rate": 2.1135430916552668e-05,
+      "loss": 1.9486,
+      "step": 6752
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3781896829605103,
+      "learning_rate": 2.1118331053351575e-05,
+      "loss": 1.8642,
+      "step": 6756
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5847692489624023,
+      "learning_rate": 2.110123119015048e-05,
+      "loss": 1.9693,
+      "step": 6760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9104549884796143,
+      "learning_rate": 2.1084131326949386e-05,
+      "loss": 2.0385,
+      "step": 6764
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6398965120315552,
+      "learning_rate": 2.106703146374829e-05,
+      "loss": 2.0187,
+      "step": 6768
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4324244260787964,
+      "learning_rate": 2.1049931600547194e-05,
+      "loss": 1.8963,
+      "step": 6772
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.57370126247406,
+      "learning_rate": 2.1032831737346102e-05,
+      "loss": 2.0443,
+      "step": 6776
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.39073646068573,
+      "learning_rate": 2.101573187414501e-05,
+      "loss": 1.9976,
+      "step": 6780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6951498985290527,
+      "learning_rate": 2.0998632010943913e-05,
+      "loss": 1.7654,
+      "step": 6784
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5796420574188232,
+      "learning_rate": 2.098153214774282e-05,
+      "loss": 1.9724,
+      "step": 6788
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.536661148071289,
+      "learning_rate": 2.0964432284541724e-05,
+      "loss": 2.0424,
+      "step": 6792
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6385177373886108,
+      "learning_rate": 2.094733242134063e-05,
+      "loss": 2.0934,
+      "step": 6796
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5548115968704224,
+      "learning_rate": 2.0930232558139536e-05,
+      "loss": 1.8169,
+      "step": 6800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4583497047424316,
+      "learning_rate": 2.0913132694938443e-05,
+      "loss": 1.9631,
+      "step": 6804
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6827950477600098,
+      "learning_rate": 2.0896032831737347e-05,
+      "loss": 1.8003,
+      "step": 6808
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.881792664527893,
+      "learning_rate": 2.087893296853625e-05,
+      "loss": 1.9482,
+      "step": 6812
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5031013488769531,
+      "learning_rate": 2.0861833105335158e-05,
+      "loss": 1.888,
+      "step": 6816
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7408212423324585,
+      "learning_rate": 2.0844733242134066e-05,
+      "loss": 2.1306,
+      "step": 6820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5273849964141846,
+      "learning_rate": 2.082763337893297e-05,
+      "loss": 1.9256,
+      "step": 6824
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6354838609695435,
+      "learning_rate": 2.0810533515731877e-05,
+      "loss": 1.9018,
+      "step": 6828
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5808496475219727,
+      "learning_rate": 2.079343365253078e-05,
+      "loss": 1.9239,
+      "step": 6832
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5880577564239502,
+      "learning_rate": 2.0776333789329685e-05,
+      "loss": 1.9876,
+      "step": 6836
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.554107904434204,
+      "learning_rate": 2.0759233926128592e-05,
+      "loss": 2.1325,
+      "step": 6840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7010103464126587,
+      "learning_rate": 2.07421340629275e-05,
+      "loss": 2.1108,
+      "step": 6844
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4548836946487427,
+      "learning_rate": 2.0725034199726403e-05,
+      "loss": 1.854,
+      "step": 6848
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.602513074874878,
+      "learning_rate": 2.0707934336525307e-05,
+      "loss": 1.9275,
+      "step": 6852
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.379732608795166,
+      "learning_rate": 2.0690834473324215e-05,
+      "loss": 1.8298,
+      "step": 6856
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5978758335113525,
+      "learning_rate": 2.067373461012312e-05,
+      "loss": 2.0292,
+      "step": 6860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5919899940490723,
+      "learning_rate": 2.0656634746922026e-05,
+      "loss": 1.9795,
+      "step": 6864
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4929299354553223,
+      "learning_rate": 2.0639534883720933e-05,
+      "loss": 2.0711,
+      "step": 6868
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5734412670135498,
+      "learning_rate": 2.0622435020519837e-05,
+      "loss": 1.9286,
+      "step": 6872
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3680405616760254,
+      "learning_rate": 2.060533515731874e-05,
+      "loss": 1.9525,
+      "step": 6876
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4015724658966064,
+      "learning_rate": 2.058823529411765e-05,
+      "loss": 1.8651,
+      "step": 6880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4925768375396729,
+      "learning_rate": 2.0571135430916556e-05,
+      "loss": 1.8487,
+      "step": 6884
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5962053537368774,
+      "learning_rate": 2.055403556771546e-05,
+      "loss": 2.0726,
+      "step": 6888
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5657358169555664,
+      "learning_rate": 2.0536935704514364e-05,
+      "loss": 2.0781,
+      "step": 6892
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4741593599319458,
+      "learning_rate": 2.051983584131327e-05,
+      "loss": 1.9721,
+      "step": 6896
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5878019332885742,
+      "learning_rate": 2.0502735978112175e-05,
+      "loss": 1.9813,
+      "step": 6900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4257550239562988,
+      "learning_rate": 2.048563611491108e-05,
+      "loss": 1.9342,
+      "step": 6904
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.603327751159668,
+      "learning_rate": 2.046853625170999e-05,
+      "loss": 2.0587,
+      "step": 6908
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4854941368103027,
+      "learning_rate": 2.0451436388508894e-05,
+      "loss": 1.8446,
+      "step": 6912
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5847572088241577,
+      "learning_rate": 2.0434336525307798e-05,
+      "loss": 2.1653,
+      "step": 6916
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.638177514076233,
+      "learning_rate": 2.0417236662106705e-05,
+      "loss": 2.0012,
+      "step": 6920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.627571702003479,
+      "learning_rate": 2.040013679890561e-05,
+      "loss": 1.9958,
+      "step": 6924
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5089852809906006,
+      "learning_rate": 2.0383036935704516e-05,
+      "loss": 1.9816,
+      "step": 6928
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6622666120529175,
+      "learning_rate": 2.036593707250342e-05,
+      "loss": 2.0482,
+      "step": 6932
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5746067762374878,
+      "learning_rate": 2.0348837209302328e-05,
+      "loss": 2.0877,
+      "step": 6936
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4563333988189697,
+      "learning_rate": 2.033173734610123e-05,
+      "loss": 2.0376,
+      "step": 6940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.417210340499878,
+      "learning_rate": 2.0314637482900136e-05,
+      "loss": 1.8537,
+      "step": 6944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.644472360610962,
+      "learning_rate": 2.0297537619699043e-05,
+      "loss": 2.0294,
+      "step": 6948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.462584137916565,
+      "learning_rate": 2.028043775649795e-05,
+      "loss": 2.0171,
+      "step": 6952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6051154136657715,
+      "learning_rate": 2.0263337893296854e-05,
+      "loss": 2.0095,
+      "step": 6956
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6152054071426392,
+      "learning_rate": 2.0246238030095758e-05,
+      "loss": 1.9219,
+      "step": 6960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3505607843399048,
+      "learning_rate": 2.0229138166894666e-05,
+      "loss": 2.0034,
+      "step": 6964
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.522525429725647,
+      "learning_rate": 2.021203830369357e-05,
+      "loss": 1.8819,
+      "step": 6968
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6121808290481567,
+      "learning_rate": 2.0194938440492477e-05,
+      "loss": 2.2193,
+      "step": 6972
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8233622312545776,
+      "learning_rate": 2.0177838577291384e-05,
+      "loss": 1.8937,
+      "step": 6976
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6225485801696777,
+      "learning_rate": 2.0160738714090288e-05,
+      "loss": 1.9533,
+      "step": 6980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5252420902252197,
+      "learning_rate": 2.0143638850889192e-05,
+      "loss": 1.9956,
+      "step": 6984
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5580767393112183,
+      "learning_rate": 2.01265389876881e-05,
+      "loss": 2.0131,
+      "step": 6988
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5372722148895264,
+      "learning_rate": 2.0109439124487007e-05,
+      "loss": 1.9964,
+      "step": 6992
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.500069499015808,
+      "learning_rate": 2.009233926128591e-05,
+      "loss": 2.0251,
+      "step": 6996
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6729307174682617,
+      "learning_rate": 2.0075239398084815e-05,
+      "loss": 1.9491,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8070749044418335,
+      "learning_rate": 2.0058139534883722e-05,
+      "loss": 1.9621,
+      "step": 7004
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5209249258041382,
+      "learning_rate": 2.0041039671682626e-05,
+      "loss": 1.8494,
+      "step": 7008
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4102156162261963,
+      "learning_rate": 2.0023939808481533e-05,
+      "loss": 1.8024,
+      "step": 7012
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6024938821792603,
+      "learning_rate": 2.000683994528044e-05,
+      "loss": 2.0584,
+      "step": 7016
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5164461135864258,
+      "learning_rate": 1.9989740082079345e-05,
+      "loss": 1.9705,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4683334827423096,
+      "learning_rate": 1.997264021887825e-05,
+      "loss": 1.8526,
+      "step": 7024
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5955207347869873,
+      "learning_rate": 1.9955540355677156e-05,
+      "loss": 2.0018,
+      "step": 7028
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6623783111572266,
+      "learning_rate": 1.993844049247606e-05,
+      "loss": 2.1312,
+      "step": 7032
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.660420298576355,
+      "learning_rate": 1.9921340629274967e-05,
+      "loss": 1.8685,
+      "step": 7036
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7066700458526611,
+      "learning_rate": 1.990424076607387e-05,
+      "loss": 1.9847,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5725295543670654,
+      "learning_rate": 1.988714090287278e-05,
+      "loss": 1.9324,
+      "step": 7044
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5613770484924316,
+      "learning_rate": 1.9870041039671683e-05,
+      "loss": 2.0648,
+      "step": 7048
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4569567441940308,
+      "learning_rate": 1.9852941176470586e-05,
+      "loss": 1.8675,
+      "step": 7052
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.610832691192627,
+      "learning_rate": 1.9835841313269497e-05,
+      "loss": 2.0513,
+      "step": 7056
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7159889936447144,
+      "learning_rate": 1.98187414500684e-05,
+      "loss": 1.8884,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6798243522644043,
+      "learning_rate": 1.9801641586867305e-05,
+      "loss": 1.9021,
+      "step": 7064
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.87516450881958,
+      "learning_rate": 1.9784541723666212e-05,
+      "loss": 2.1314,
+      "step": 7068
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.465100646018982,
+      "learning_rate": 1.9767441860465116e-05,
+      "loss": 1.9359,
+      "step": 7072
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7637323141098022,
+      "learning_rate": 1.9750341997264024e-05,
+      "loss": 2.0628,
+      "step": 7076
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4772495031356812,
+      "learning_rate": 1.9733242134062928e-05,
+      "loss": 2.0258,
+      "step": 7080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5895733833312988,
+      "learning_rate": 1.9716142270861835e-05,
+      "loss": 1.8629,
+      "step": 7084
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5556437969207764,
+      "learning_rate": 1.969904240766074e-05,
+      "loss": 1.9945,
+      "step": 7088
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.502622365951538,
+      "learning_rate": 1.9681942544459643e-05,
+      "loss": 2.0398,
+      "step": 7092
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5926116704940796,
+      "learning_rate": 1.966484268125855e-05,
+      "loss": 1.8625,
+      "step": 7096
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5113784074783325,
+      "learning_rate": 1.9647742818057458e-05,
+      "loss": 2.0816,
+      "step": 7100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5072494745254517,
+      "learning_rate": 1.963064295485636e-05,
+      "loss": 1.9681,
+      "step": 7104
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6104905605316162,
+      "learning_rate": 1.961354309165527e-05,
+      "loss": 2.1155,
+      "step": 7108
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6391701698303223,
+      "learning_rate": 1.9596443228454173e-05,
+      "loss": 1.9331,
+      "step": 7112
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5969183444976807,
+      "learning_rate": 1.9579343365253077e-05,
+      "loss": 1.9331,
+      "step": 7116
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6461735963821411,
+      "learning_rate": 1.9562243502051984e-05,
+      "loss": 2.115,
+      "step": 7120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.572561264038086,
+      "learning_rate": 1.954514363885089e-05,
+      "loss": 2.0272,
+      "step": 7124
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5880368947982788,
+      "learning_rate": 1.9528043775649796e-05,
+      "loss": 2.0262,
+      "step": 7128
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.697801947593689,
+      "learning_rate": 1.95109439124487e-05,
+      "loss": 2.1504,
+      "step": 7132
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5090667009353638,
+      "learning_rate": 1.9493844049247607e-05,
+      "loss": 2.0096,
+      "step": 7136
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6617227792739868,
+      "learning_rate": 1.9476744186046514e-05,
+      "loss": 1.9429,
+      "step": 7140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.489587426185608,
+      "learning_rate": 1.9459644322845418e-05,
+      "loss": 1.9735,
+      "step": 7144
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4970664978027344,
+      "learning_rate": 1.9442544459644325e-05,
+      "loss": 2.0434,
+      "step": 7148
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7819324731826782,
+      "learning_rate": 1.942544459644323e-05,
+      "loss": 2.0847,
+      "step": 7152
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.572629690170288,
+      "learning_rate": 1.9408344733242133e-05,
+      "loss": 2.0716,
+      "step": 7156
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6527976989746094,
+      "learning_rate": 1.939124487004104e-05,
+      "loss": 1.9941,
+      "step": 7160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7278069257736206,
+      "learning_rate": 1.9374145006839948e-05,
+      "loss": 1.9261,
+      "step": 7164
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5078670978546143,
+      "learning_rate": 1.9357045143638852e-05,
+      "loss": 1.9513,
+      "step": 7168
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6161733865737915,
+      "learning_rate": 1.9339945280437756e-05,
+      "loss": 1.9143,
+      "step": 7172
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7204173803329468,
+      "learning_rate": 1.9322845417236663e-05,
+      "loss": 1.8862,
+      "step": 7176
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.607851266860962,
+      "learning_rate": 1.9305745554035567e-05,
+      "loss": 1.98,
+      "step": 7180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5279322862625122,
+      "learning_rate": 1.9288645690834475e-05,
+      "loss": 1.8538,
+      "step": 7184
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5147992372512817,
+      "learning_rate": 1.9271545827633382e-05,
+      "loss": 1.9975,
+      "step": 7188
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6696971654891968,
+      "learning_rate": 1.9254445964432286e-05,
+      "loss": 1.8899,
+      "step": 7192
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3951395750045776,
+      "learning_rate": 1.923734610123119e-05,
+      "loss": 1.9392,
+      "step": 7196
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5910269021987915,
+      "learning_rate": 1.9220246238030097e-05,
+      "loss": 2.082,
+      "step": 7200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5352290868759155,
+      "learning_rate": 1.9203146374829005e-05,
+      "loss": 2.0371,
+      "step": 7204
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.521520972251892,
+      "learning_rate": 1.918604651162791e-05,
+      "loss": 1.9586,
+      "step": 7208
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.501772403717041,
+      "learning_rate": 1.9168946648426812e-05,
+      "loss": 1.982,
+      "step": 7212
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.548953890800476,
+      "learning_rate": 1.915184678522572e-05,
+      "loss": 2.0191,
+      "step": 7216
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6001508235931396,
+      "learning_rate": 1.9134746922024624e-05,
+      "loss": 2.0142,
+      "step": 7220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6897025108337402,
+      "learning_rate": 1.9117647058823528e-05,
+      "loss": 1.9539,
+      "step": 7224
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6484875679016113,
+      "learning_rate": 1.910054719562244e-05,
+      "loss": 1.9652,
+      "step": 7228
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.504267692565918,
+      "learning_rate": 1.9083447332421342e-05,
+      "loss": 1.8661,
+      "step": 7232
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.442720890045166,
+      "learning_rate": 1.9066347469220246e-05,
+      "loss": 1.849,
+      "step": 7236
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3908931016921997,
+      "learning_rate": 1.9049247606019154e-05,
+      "loss": 2.0333,
+      "step": 7240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7459537982940674,
+      "learning_rate": 1.9032147742818058e-05,
+      "loss": 1.98,
+      "step": 7244
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5512782335281372,
+      "learning_rate": 1.9015047879616965e-05,
+      "loss": 2.1063,
+      "step": 7248
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5178790092468262,
+      "learning_rate": 1.899794801641587e-05,
+      "loss": 2.0102,
+      "step": 7252
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7210770845413208,
+      "learning_rate": 1.8980848153214776e-05,
+      "loss": 1.9342,
+      "step": 7256
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.472294569015503,
+      "learning_rate": 1.896374829001368e-05,
+      "loss": 2.0667,
+      "step": 7260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5506442785263062,
+      "learning_rate": 1.8946648426812584e-05,
+      "loss": 2.2188,
+      "step": 7264
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6018651723861694,
+      "learning_rate": 1.8929548563611495e-05,
+      "loss": 1.9151,
+      "step": 7268
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6608229875564575,
+      "learning_rate": 1.89124487004104e-05,
+      "loss": 1.8974,
+      "step": 7272
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.580942153930664,
+      "learning_rate": 1.8895348837209303e-05,
+      "loss": 1.9818,
+      "step": 7276
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5641908645629883,
+      "learning_rate": 1.887824897400821e-05,
+      "loss": 1.8878,
+      "step": 7280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6174907684326172,
+      "learning_rate": 1.8861149110807114e-05,
+      "loss": 2.0124,
+      "step": 7284
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6148037910461426,
+      "learning_rate": 1.8844049247606018e-05,
+      "loss": 1.902,
+      "step": 7288
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6356844902038574,
+      "learning_rate": 1.8826949384404925e-05,
+      "loss": 2.0167,
+      "step": 7292
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.441463589668274,
+      "learning_rate": 1.8809849521203833e-05,
+      "loss": 1.8633,
+      "step": 7296
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5125188827514648,
+      "learning_rate": 1.8792749658002737e-05,
+      "loss": 1.7871,
+      "step": 7300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4316954612731934,
+      "learning_rate": 1.877564979480164e-05,
+      "loss": 2.0366,
+      "step": 7304
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7826207876205444,
+      "learning_rate": 1.8758549931600548e-05,
+      "loss": 1.9861,
+      "step": 7308
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7452281713485718,
+      "learning_rate": 1.8741450068399455e-05,
+      "loss": 2.1054,
+      "step": 7312
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5087822675704956,
+      "learning_rate": 1.872435020519836e-05,
+      "loss": 1.9194,
+      "step": 7316
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.560201644897461,
+      "learning_rate": 1.8707250341997263e-05,
+      "loss": 1.983,
+      "step": 7320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5828901529312134,
+      "learning_rate": 1.869015047879617e-05,
+      "loss": 2.0051,
+      "step": 7324
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5497353076934814,
+      "learning_rate": 1.8673050615595075e-05,
+      "loss": 2.0197,
+      "step": 7328
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5281920433044434,
+      "learning_rate": 1.8655950752393982e-05,
+      "loss": 1.9502,
+      "step": 7332
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6356514692306519,
+      "learning_rate": 1.863885088919289e-05,
+      "loss": 2.069,
+      "step": 7336
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4687514305114746,
+      "learning_rate": 1.8621751025991793e-05,
+      "loss": 1.931,
+      "step": 7340
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.604845643043518,
+      "learning_rate": 1.8604651162790697e-05,
+      "loss": 2.0305,
+      "step": 7344
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5669852495193481,
+      "learning_rate": 1.8587551299589605e-05,
+      "loss": 2.0094,
+      "step": 7348
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.586909532546997,
+      "learning_rate": 1.857045143638851e-05,
+      "loss": 2.0053,
+      "step": 7352
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5457959175109863,
+      "learning_rate": 1.8553351573187416e-05,
+      "loss": 1.9575,
+      "step": 7356
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4828368425369263,
+      "learning_rate": 1.853625170998632e-05,
+      "loss": 1.9047,
+      "step": 7360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5736109018325806,
+      "learning_rate": 1.8519151846785227e-05,
+      "loss": 1.93,
+      "step": 7364
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7825120687484741,
+      "learning_rate": 1.850205198358413e-05,
+      "loss": 1.9667,
+      "step": 7368
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.529508113861084,
+      "learning_rate": 1.8484952120383035e-05,
+      "loss": 1.8791,
+      "step": 7372
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5138580799102783,
+      "learning_rate": 1.8467852257181946e-05,
+      "loss": 1.932,
+      "step": 7376
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.619393229484558,
+      "learning_rate": 1.845075239398085e-05,
+      "loss": 1.9461,
+      "step": 7380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4028980731964111,
+      "learning_rate": 1.8433652530779754e-05,
+      "loss": 1.9428,
+      "step": 7384
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.631604552268982,
+      "learning_rate": 1.841655266757866e-05,
+      "loss": 1.9728,
+      "step": 7388
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4649015665054321,
+      "learning_rate": 1.8399452804377565e-05,
+      "loss": 1.8762,
+      "step": 7392
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.48709237575531,
+      "learning_rate": 1.8382352941176472e-05,
+      "loss": 1.9312,
+      "step": 7396
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6089003086090088,
+      "learning_rate": 1.8365253077975376e-05,
+      "loss": 1.8671,
+      "step": 7400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.723889708518982,
+      "learning_rate": 1.8348153214774284e-05,
+      "loss": 1.9107,
+      "step": 7404
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5013092756271362,
+      "learning_rate": 1.8331053351573188e-05,
+      "loss": 1.9322,
+      "step": 7408
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8294460773468018,
+      "learning_rate": 1.831395348837209e-05,
+      "loss": 2.1004,
+      "step": 7412
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4940433502197266,
+      "learning_rate": 1.8296853625171e-05,
+      "loss": 1.882,
+      "step": 7416
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6243213415145874,
+      "learning_rate": 1.8279753761969906e-05,
+      "loss": 2.1548,
+      "step": 7420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.934673547744751,
+      "learning_rate": 1.826265389876881e-05,
+      "loss": 1.9928,
+      "step": 7424
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7047098875045776,
+      "learning_rate": 1.8245554035567718e-05,
+      "loss": 1.9881,
+      "step": 7428
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5831687450408936,
+      "learning_rate": 1.822845417236662e-05,
+      "loss": 1.9889,
+      "step": 7432
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6739925146102905,
+      "learning_rate": 1.8211354309165525e-05,
+      "loss": 1.9835,
+      "step": 7436
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4848778247833252,
+      "learning_rate": 1.8194254445964433e-05,
+      "loss": 1.9711,
+      "step": 7440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.499522089958191,
+      "learning_rate": 1.817715458276334e-05,
+      "loss": 1.9454,
+      "step": 7444
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6421815156936646,
+      "learning_rate": 1.8160054719562244e-05,
+      "loss": 1.9402,
+      "step": 7448
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7279685735702515,
+      "learning_rate": 1.8142954856361148e-05,
+      "loss": 1.7985,
+      "step": 7452
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6793206930160522,
+      "learning_rate": 1.8125854993160055e-05,
+      "loss": 2.0187,
+      "step": 7456
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4948447942733765,
+      "learning_rate": 1.8108755129958963e-05,
+      "loss": 1.9569,
+      "step": 7460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5078566074371338,
+      "learning_rate": 1.8091655266757867e-05,
+      "loss": 1.8296,
+      "step": 7464
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7886958122253418,
+      "learning_rate": 1.8074555403556774e-05,
+      "loss": 2.0134,
+      "step": 7468
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4695848226547241,
+      "learning_rate": 1.8057455540355678e-05,
+      "loss": 1.9271,
+      "step": 7472
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4936541318893433,
+      "learning_rate": 1.8040355677154582e-05,
+      "loss": 1.9308,
+      "step": 7476
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5521645545959473,
+      "learning_rate": 1.802325581395349e-05,
+      "loss": 1.8857,
+      "step": 7480
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5674141645431519,
+      "learning_rate": 1.8006155950752397e-05,
+      "loss": 1.9064,
+      "step": 7484
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6228454113006592,
+      "learning_rate": 1.79890560875513e-05,
+      "loss": 1.8739,
+      "step": 7488
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4367411136627197,
+      "learning_rate": 1.7971956224350205e-05,
+      "loss": 1.9576,
+      "step": 7492
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5149922370910645,
+      "learning_rate": 1.7954856361149112e-05,
+      "loss": 1.9571,
+      "step": 7496
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6180402040481567,
+      "learning_rate": 1.7937756497948016e-05,
+      "loss": 1.8322,
+      "step": 7500
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.410872459411621,
+      "learning_rate": 1.7920656634746923e-05,
+      "loss": 1.8944,
+      "step": 7504
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4997954368591309,
+      "learning_rate": 1.790355677154583e-05,
+      "loss": 2.014,
+      "step": 7508
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4796793460845947,
+      "learning_rate": 1.7886456908344735e-05,
+      "loss": 1.908,
+      "step": 7512
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6083449125289917,
+      "learning_rate": 1.786935704514364e-05,
+      "loss": 1.9159,
+      "step": 7516
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6810011863708496,
+      "learning_rate": 1.7852257181942546e-05,
+      "loss": 1.9426,
+      "step": 7520
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5123205184936523,
+      "learning_rate": 1.7835157318741453e-05,
+      "loss": 1.8627,
+      "step": 7524
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4853988885879517,
+      "learning_rate": 1.7818057455540357e-05,
+      "loss": 1.9814,
+      "step": 7528
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.720802903175354,
+      "learning_rate": 1.780095759233926e-05,
+      "loss": 1.9168,
+      "step": 7532
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6009372472763062,
+      "learning_rate": 1.778385772913817e-05,
+      "loss": 1.9071,
+      "step": 7536
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1640713214874268,
+      "learning_rate": 1.7766757865937072e-05,
+      "loss": 1.9403,
+      "step": 7540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9134138822555542,
+      "learning_rate": 1.7749658002735976e-05,
+      "loss": 2.013,
+      "step": 7544
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6833550930023193,
+      "learning_rate": 1.7732558139534887e-05,
+      "loss": 2.0307,
+      "step": 7548
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.643018364906311,
+      "learning_rate": 1.771545827633379e-05,
+      "loss": 1.8817,
+      "step": 7552
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5238628387451172,
+      "learning_rate": 1.7698358413132695e-05,
+      "loss": 2.0012,
+      "step": 7556
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.612910270690918,
+      "learning_rate": 1.7681258549931602e-05,
+      "loss": 1.9054,
+      "step": 7560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4849001169204712,
+      "learning_rate": 1.7664158686730506e-05,
+      "loss": 1.8482,
+      "step": 7564
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7270807027816772,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 2.0145,
+      "step": 7568
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.676148533821106,
+      "learning_rate": 1.7629958960328318e-05,
+      "loss": 2.0066,
+      "step": 7572
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3805028200149536,
+      "learning_rate": 1.7612859097127225e-05,
+      "loss": 1.813,
+      "step": 7576
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.528660774230957,
+      "learning_rate": 1.759575923392613e-05,
+      "loss": 1.9291,
+      "step": 7580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.623838186264038,
+      "learning_rate": 1.7578659370725033e-05,
+      "loss": 1.9315,
+      "step": 7584
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4332454204559326,
+      "learning_rate": 1.7561559507523944e-05,
+      "loss": 1.9888,
+      "step": 7588
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5631705522537231,
+      "learning_rate": 1.7544459644322848e-05,
+      "loss": 2.1324,
+      "step": 7592
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.10516357421875,
+      "learning_rate": 1.752735978112175e-05,
+      "loss": 1.9798,
+      "step": 7596
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6901365518569946,
+      "learning_rate": 1.751025991792066e-05,
+      "loss": 1.9679,
+      "step": 7600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9404395818710327,
+      "learning_rate": 1.7493160054719563e-05,
+      "loss": 2.0198,
+      "step": 7604
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5954433679580688,
+      "learning_rate": 1.7476060191518467e-05,
+      "loss": 1.887,
+      "step": 7608
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4495494365692139,
+      "learning_rate": 1.7458960328317374e-05,
+      "loss": 2.0086,
+      "step": 7612
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.513266682624817,
+      "learning_rate": 1.744186046511628e-05,
+      "loss": 1.9934,
+      "step": 7616
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6928620338439941,
+      "learning_rate": 1.7424760601915185e-05,
+      "loss": 1.9103,
+      "step": 7620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7382714748382568,
+      "learning_rate": 1.740766073871409e-05,
+      "loss": 2.015,
+      "step": 7624
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4485249519348145,
+      "learning_rate": 1.7390560875512997e-05,
+      "loss": 2.076,
+      "step": 7628
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5897313356399536,
+      "learning_rate": 1.7373461012311904e-05,
+      "loss": 1.8503,
+      "step": 7632
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6106144189834595,
+      "learning_rate": 1.7356361149110808e-05,
+      "loss": 1.7611,
+      "step": 7636
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6251461505889893,
+      "learning_rate": 1.7339261285909715e-05,
+      "loss": 1.8426,
+      "step": 7640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6952441930770874,
+      "learning_rate": 1.732216142270862e-05,
+      "loss": 1.9492,
+      "step": 7644
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6333529949188232,
+      "learning_rate": 1.7305061559507523e-05,
+      "loss": 1.9824,
+      "step": 7648
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5729711055755615,
+      "learning_rate": 1.728796169630643e-05,
+      "loss": 1.9888,
+      "step": 7652
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5947141647338867,
+      "learning_rate": 1.7270861833105338e-05,
+      "loss": 1.8368,
+      "step": 7656
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5854895114898682,
+      "learning_rate": 1.7253761969904242e-05,
+      "loss": 1.9155,
+      "step": 7660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6690934896469116,
+      "learning_rate": 1.7236662106703146e-05,
+      "loss": 1.9139,
+      "step": 7664
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5118485689163208,
+      "learning_rate": 1.7219562243502053e-05,
+      "loss": 1.8872,
+      "step": 7668
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6435084342956543,
+      "learning_rate": 1.7202462380300957e-05,
+      "loss": 2.1278,
+      "step": 7672
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6225100755691528,
+      "learning_rate": 1.7185362517099864e-05,
+      "loss": 1.9057,
+      "step": 7676
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.603485345840454,
+      "learning_rate": 1.716826265389877e-05,
+      "loss": 1.7631,
+      "step": 7680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.588488221168518,
+      "learning_rate": 1.7151162790697676e-05,
+      "loss": 2.0738,
+      "step": 7684
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5990886688232422,
+      "learning_rate": 1.713406292749658e-05,
+      "loss": 1.9249,
+      "step": 7688
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5695997476577759,
+      "learning_rate": 1.7116963064295484e-05,
+      "loss": 1.8807,
+      "step": 7692
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8457928895950317,
+      "learning_rate": 1.7099863201094394e-05,
+      "loss": 2.0667,
+      "step": 7696
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7506129741668701,
+      "learning_rate": 1.70827633378933e-05,
+      "loss": 1.9462,
+      "step": 7700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5219964981079102,
+      "learning_rate": 1.7065663474692202e-05,
+      "loss": 1.8061,
+      "step": 7704
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5406177043914795,
+      "learning_rate": 1.704856361149111e-05,
+      "loss": 1.9219,
+      "step": 7708
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5064541101455688,
+      "learning_rate": 1.7031463748290014e-05,
+      "loss": 1.9149,
+      "step": 7712
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7326499223709106,
+      "learning_rate": 1.701436388508892e-05,
+      "loss": 1.8593,
+      "step": 7716
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6578881740570068,
+      "learning_rate": 1.6997264021887825e-05,
+      "loss": 2.0094,
+      "step": 7720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.724426031112671,
+      "learning_rate": 1.6980164158686732e-05,
+      "loss": 2.0086,
+      "step": 7724
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7272109985351562,
+      "learning_rate": 1.6963064295485636e-05,
+      "loss": 2.0135,
+      "step": 7728
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8231357336044312,
+      "learning_rate": 1.694596443228454e-05,
+      "loss": 1.9537,
+      "step": 7732
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6103249788284302,
+      "learning_rate": 1.6928864569083448e-05,
+      "loss": 2.0054,
+      "step": 7736
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.507234811782837,
+      "learning_rate": 1.6911764705882355e-05,
+      "loss": 2.0489,
+      "step": 7740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.49418306350708,
+      "learning_rate": 1.689466484268126e-05,
+      "loss": 1.8426,
+      "step": 7744
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.828412652015686,
+      "learning_rate": 1.6877564979480166e-05,
+      "loss": 2.0645,
+      "step": 7748
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5563526153564453,
+      "learning_rate": 1.686046511627907e-05,
+      "loss": 1.9228,
+      "step": 7752
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.680085301399231,
+      "learning_rate": 1.6843365253077974e-05,
+      "loss": 2.0503,
+      "step": 7756
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8583743572235107,
+      "learning_rate": 1.682626538987688e-05,
+      "loss": 1.9134,
+      "step": 7760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5803444385528564,
+      "learning_rate": 1.680916552667579e-05,
+      "loss": 1.8787,
+      "step": 7764
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4421805143356323,
+      "learning_rate": 1.6792065663474693e-05,
+      "loss": 1.7833,
+      "step": 7768
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6548757553100586,
+      "learning_rate": 1.6774965800273597e-05,
+      "loss": 1.9397,
+      "step": 7772
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6442538499832153,
+      "learning_rate": 1.6757865937072504e-05,
+      "loss": 2.1046,
+      "step": 7776
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.643537163734436,
+      "learning_rate": 1.674076607387141e-05,
+      "loss": 1.9088,
+      "step": 7780
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5146218538284302,
+      "learning_rate": 1.6723666210670315e-05,
+      "loss": 1.9828,
+      "step": 7784
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.689788579940796,
+      "learning_rate": 1.6706566347469223e-05,
+      "loss": 1.9587,
+      "step": 7788
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.498406171798706,
+      "learning_rate": 1.6689466484268127e-05,
+      "loss": 1.8098,
+      "step": 7792
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6559585332870483,
+      "learning_rate": 1.667236662106703e-05,
+      "loss": 2.0373,
+      "step": 7796
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.675614595413208,
+      "learning_rate": 1.6655266757865938e-05,
+      "loss": 1.9298,
+      "step": 7800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6751221418380737,
+      "learning_rate": 1.6638166894664845e-05,
+      "loss": 1.9871,
+      "step": 7804
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4255478382110596,
+      "learning_rate": 1.662106703146375e-05,
+      "loss": 1.821,
+      "step": 7808
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5127853155136108,
+      "learning_rate": 1.6603967168262653e-05,
+      "loss": 2.0976,
+      "step": 7812
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5522574186325073,
+      "learning_rate": 1.658686730506156e-05,
+      "loss": 2.027,
+      "step": 7816
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.492023229598999,
+      "learning_rate": 1.6569767441860464e-05,
+      "loss": 1.9444,
+      "step": 7820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5379337072372437,
+      "learning_rate": 1.6552667578659372e-05,
+      "loss": 1.9637,
+      "step": 7824
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.615531086921692,
+      "learning_rate": 1.653556771545828e-05,
+      "loss": 1.9079,
+      "step": 7828
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5299206972122192,
+      "learning_rate": 1.6518467852257183e-05,
+      "loss": 1.8934,
+      "step": 7832
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5533653497695923,
+      "learning_rate": 1.6501367989056087e-05,
+      "loss": 1.8824,
+      "step": 7836
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5589085817337036,
+      "learning_rate": 1.6484268125854994e-05,
+      "loss": 1.8289,
+      "step": 7840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5232523679733276,
+      "learning_rate": 1.6467168262653902e-05,
+      "loss": 1.9622,
+      "step": 7844
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4827030897140503,
+      "learning_rate": 1.6450068399452806e-05,
+      "loss": 2.0214,
+      "step": 7848
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5622025728225708,
+      "learning_rate": 1.643296853625171e-05,
+      "loss": 1.9528,
+      "step": 7852
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.578583836555481,
+      "learning_rate": 1.6415868673050617e-05,
+      "loss": 1.9292,
+      "step": 7856
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5761486291885376,
+      "learning_rate": 1.639876880984952e-05,
+      "loss": 1.8721,
+      "step": 7860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.434203028678894,
+      "learning_rate": 1.6381668946648425e-05,
+      "loss": 1.882,
+      "step": 7864
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7951371669769287,
+      "learning_rate": 1.6364569083447336e-05,
+      "loss": 1.8724,
+      "step": 7868
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5747816562652588,
+      "learning_rate": 1.634746922024624e-05,
+      "loss": 1.9838,
+      "step": 7872
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6416630744934082,
+      "learning_rate": 1.6330369357045144e-05,
+      "loss": 2.0448,
+      "step": 7876
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5873196125030518,
+      "learning_rate": 1.631326949384405e-05,
+      "loss": 1.8606,
+      "step": 7880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.708465576171875,
+      "learning_rate": 1.6296169630642955e-05,
+      "loss": 2.0646,
+      "step": 7884
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.764533281326294,
+      "learning_rate": 1.6279069767441862e-05,
+      "loss": 1.9685,
+      "step": 7888
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6329145431518555,
+      "learning_rate": 1.6261969904240766e-05,
+      "loss": 1.9665,
+      "step": 7892
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5862176418304443,
+      "learning_rate": 1.6244870041039673e-05,
+      "loss": 1.9459,
+      "step": 7896
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.66335129737854,
+      "learning_rate": 1.6227770177838577e-05,
+      "loss": 1.9046,
+      "step": 7900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.854282259941101,
+      "learning_rate": 1.621067031463748e-05,
+      "loss": 1.8697,
+      "step": 7904
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5997343063354492,
+      "learning_rate": 1.6193570451436392e-05,
+      "loss": 1.8564,
+      "step": 7908
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5500624179840088,
+      "learning_rate": 1.6176470588235296e-05,
+      "loss": 1.9197,
+      "step": 7912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7503950595855713,
+      "learning_rate": 1.61593707250342e-05,
+      "loss": 1.8817,
+      "step": 7916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5188136100769043,
+      "learning_rate": 1.6142270861833107e-05,
+      "loss": 1.9714,
+      "step": 7920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.632093906402588,
+      "learning_rate": 1.612517099863201e-05,
+      "loss": 1.9189,
+      "step": 7924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.079958915710449,
+      "learning_rate": 1.6108071135430915e-05,
+      "loss": 1.9399,
+      "step": 7928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5415632724761963,
+      "learning_rate": 1.6090971272229823e-05,
+      "loss": 1.977,
+      "step": 7932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6359429359436035,
+      "learning_rate": 1.607387140902873e-05,
+      "loss": 1.9225,
+      "step": 7936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6793197393417358,
+      "learning_rate": 1.6056771545827634e-05,
+      "loss": 1.891,
+      "step": 7940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5686537027359009,
+      "learning_rate": 1.6039671682626538e-05,
+      "loss": 1.8904,
+      "step": 7944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8018839359283447,
+      "learning_rate": 1.6022571819425445e-05,
+      "loss": 1.8233,
+      "step": 7948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.547400951385498,
+      "learning_rate": 1.6005471956224353e-05,
+      "loss": 1.9372,
+      "step": 7952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5246180295944214,
+      "learning_rate": 1.5988372093023257e-05,
+      "loss": 1.9839,
+      "step": 7956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4722802639007568,
+      "learning_rate": 1.5971272229822164e-05,
+      "loss": 1.7585,
+      "step": 7960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5736894607543945,
+      "learning_rate": 1.5954172366621068e-05,
+      "loss": 1.9647,
+      "step": 7964
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6675376892089844,
+      "learning_rate": 1.5937072503419972e-05,
+      "loss": 1.9783,
+      "step": 7968
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5381276607513428,
+      "learning_rate": 1.591997264021888e-05,
+      "loss": 1.8318,
+      "step": 7972
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.672789454460144,
+      "learning_rate": 1.5902872777017786e-05,
+      "loss": 1.9432,
+      "step": 7976
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5092620849609375,
+      "learning_rate": 1.588577291381669e-05,
+      "loss": 1.8645,
+      "step": 7980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7272820472717285,
+      "learning_rate": 1.5868673050615594e-05,
+      "loss": 2.0409,
+      "step": 7984
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5391128063201904,
+      "learning_rate": 1.5851573187414502e-05,
+      "loss": 1.7384,
+      "step": 7988
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4302643537521362,
+      "learning_rate": 1.5834473324213406e-05,
+      "loss": 1.891,
+      "step": 7992
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5676170587539673,
+      "learning_rate": 1.5817373461012313e-05,
+      "loss": 1.7741,
+      "step": 7996
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7254034280776978,
+      "learning_rate": 1.580027359781122e-05,
+      "loss": 1.9008,
+      "step": 8000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5930966138839722,
+      "learning_rate": 1.5783173734610124e-05,
+      "loss": 1.9161,
+      "step": 8004
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.011584758758545,
+      "learning_rate": 1.5766073871409028e-05,
+      "loss": 1.9344,
+      "step": 8008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5066322088241577,
+      "learning_rate": 1.5748974008207936e-05,
+      "loss": 1.8778,
+      "step": 8012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.775755763053894,
+      "learning_rate": 1.5731874145006843e-05,
+      "loss": 1.8985,
+      "step": 8016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5539259910583496,
+      "learning_rate": 1.5714774281805747e-05,
+      "loss": 1.8531,
+      "step": 8020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7370346784591675,
+      "learning_rate": 1.569767441860465e-05,
+      "loss": 1.9098,
+      "step": 8024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4785212278366089,
+      "learning_rate": 1.5680574555403558e-05,
+      "loss": 2.0605,
+      "step": 8028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6083650588989258,
+      "learning_rate": 1.5663474692202462e-05,
+      "loss": 2.0839,
+      "step": 8032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6726301908493042,
+      "learning_rate": 1.564637482900137e-05,
+      "loss": 1.8263,
+      "step": 8036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.202913761138916,
+      "learning_rate": 1.5629274965800273e-05,
+      "loss": 1.8298,
+      "step": 8040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6206969022750854,
+      "learning_rate": 1.561217510259918e-05,
+      "loss": 1.9383,
+      "step": 8044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6992504596710205,
+      "learning_rate": 1.5595075239398085e-05,
+      "loss": 1.9817,
+      "step": 8048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.685895323753357,
+      "learning_rate": 1.557797537619699e-05,
+      "loss": 2.0378,
+      "step": 8052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6659513711929321,
+      "learning_rate": 1.5560875512995896e-05,
+      "loss": 1.923,
+      "step": 8056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6059319972991943,
+      "learning_rate": 1.5543775649794803e-05,
+      "loss": 1.9304,
+      "step": 8060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.725585699081421,
+      "learning_rate": 1.5526675786593707e-05,
+      "loss": 2.0079,
+      "step": 8064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7117788791656494,
+      "learning_rate": 1.5509575923392615e-05,
+      "loss": 2.1715,
+      "step": 8068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5837129354476929,
+      "learning_rate": 1.549247606019152e-05,
+      "loss": 1.9936,
+      "step": 8072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9159271717071533,
+      "learning_rate": 1.5475376196990423e-05,
+      "loss": 1.9387,
+      "step": 8076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7202599048614502,
+      "learning_rate": 1.545827633378933e-05,
+      "loss": 1.9975,
+      "step": 8080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5703368186950684,
+      "learning_rate": 1.5441176470588237e-05,
+      "loss": 1.8077,
+      "step": 8084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.520732045173645,
+      "learning_rate": 1.542407660738714e-05,
+      "loss": 2.0593,
+      "step": 8088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8129254579544067,
+      "learning_rate": 1.5406976744186045e-05,
+      "loss": 1.8897,
+      "step": 8092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4720978736877441,
+      "learning_rate": 1.5389876880984953e-05,
+      "loss": 1.9366,
+      "step": 8096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5979335308074951,
+      "learning_rate": 1.537277701778386e-05,
+      "loss": 1.843,
+      "step": 8100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4189525842666626,
+      "learning_rate": 1.5355677154582764e-05,
+      "loss": 1.7472,
+      "step": 8104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.724085807800293,
+      "learning_rate": 1.533857729138167e-05,
+      "loss": 2.108,
+      "step": 8108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5960993766784668,
+      "learning_rate": 1.5321477428180575e-05,
+      "loss": 1.8707,
+      "step": 8112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.624754786491394,
+      "learning_rate": 1.530437756497948e-05,
+      "loss": 1.9318,
+      "step": 8116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.758505940437317,
+      "learning_rate": 1.5287277701778386e-05,
+      "loss": 1.8996,
+      "step": 8120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.633908987045288,
+      "learning_rate": 1.5270177838577294e-05,
+      "loss": 1.8455,
+      "step": 8124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.578237533569336,
+      "learning_rate": 1.5253077975376198e-05,
+      "loss": 2.071,
+      "step": 8128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5181132555007935,
+      "learning_rate": 1.5235978112175103e-05,
+      "loss": 1.766,
+      "step": 8132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6433082818984985,
+      "learning_rate": 1.5218878248974009e-05,
+      "loss": 2.0446,
+      "step": 8136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6200603246688843,
+      "learning_rate": 1.5201778385772913e-05,
+      "loss": 1.7629,
+      "step": 8140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7067992687225342,
+      "learning_rate": 1.5184678522571822e-05,
+      "loss": 1.8624,
+      "step": 8144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.515853762626648,
+      "learning_rate": 1.5167578659370726e-05,
+      "loss": 1.8187,
+      "step": 8148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7134236097335815,
+      "learning_rate": 1.5150478796169632e-05,
+      "loss": 1.9738,
+      "step": 8152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5560367107391357,
+      "learning_rate": 1.5133378932968537e-05,
+      "loss": 1.9252,
+      "step": 8156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.476830005645752,
+      "learning_rate": 1.5116279069767441e-05,
+      "loss": 1.7564,
+      "step": 8160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6756997108459473,
+      "learning_rate": 1.509917920656635e-05,
+      "loss": 1.8457,
+      "step": 8164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6191574335098267,
+      "learning_rate": 1.5082079343365254e-05,
+      "loss": 1.9078,
+      "step": 8168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7781707048416138,
+      "learning_rate": 1.506497948016416e-05,
+      "loss": 1.9178,
+      "step": 8172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.361154556274414,
+      "learning_rate": 1.5047879616963066e-05,
+      "loss": 1.9068,
+      "step": 8176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6916831731796265,
+      "learning_rate": 1.503077975376197e-05,
+      "loss": 1.9089,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.682424783706665,
+      "learning_rate": 1.5013679890560875e-05,
+      "loss": 2.0016,
+      "step": 8184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4187464714050293,
+      "learning_rate": 1.4996580027359783e-05,
+      "loss": 1.811,
+      "step": 8188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5043885707855225,
+      "learning_rate": 1.4979480164158688e-05,
+      "loss": 1.8448,
+      "step": 8192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6151576042175293,
+      "learning_rate": 1.4962380300957594e-05,
+      "loss": 1.9776,
+      "step": 8196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5769206285476685,
+      "learning_rate": 1.4945280437756498e-05,
+      "loss": 1.937,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5044043064117432,
+      "learning_rate": 1.4928180574555403e-05,
+      "loss": 1.852,
+      "step": 8204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6411508321762085,
+      "learning_rate": 1.491108071135431e-05,
+      "loss": 1.9599,
+      "step": 8208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6469234228134155,
+      "learning_rate": 1.4893980848153216e-05,
+      "loss": 1.9749,
+      "step": 8212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.64458167552948,
+      "learning_rate": 1.4876880984952122e-05,
+      "loss": 1.728,
+      "step": 8216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5000782012939453,
+      "learning_rate": 1.4859781121751026e-05,
+      "loss": 1.8083,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.521461844444275,
+      "learning_rate": 1.4842681258549932e-05,
+      "loss": 1.9935,
+      "step": 8224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5517728328704834,
+      "learning_rate": 1.4825581395348839e-05,
+      "loss": 2.0673,
+      "step": 8228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.566142201423645,
+      "learning_rate": 1.4808481532147745e-05,
+      "loss": 1.8942,
+      "step": 8232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5397378206253052,
+      "learning_rate": 1.4791381668946649e-05,
+      "loss": 1.963,
+      "step": 8236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3787508010864258,
+      "learning_rate": 1.4774281805745554e-05,
+      "loss": 1.9094,
+      "step": 8240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4250317811965942,
+      "learning_rate": 1.475718194254446e-05,
+      "loss": 1.8181,
+      "step": 8244
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.868123173713684,
+      "learning_rate": 1.4740082079343364e-05,
+      "loss": 1.975,
+      "step": 8248
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6380820274353027,
+      "learning_rate": 1.4722982216142273e-05,
+      "loss": 1.8019,
+      "step": 8252
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7641537189483643,
+      "learning_rate": 1.4705882352941177e-05,
+      "loss": 1.8762,
+      "step": 8256
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.704601526260376,
+      "learning_rate": 1.4688782489740083e-05,
+      "loss": 1.938,
+      "step": 8260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5377604961395264,
+      "learning_rate": 1.4671682626538988e-05,
+      "loss": 1.837,
+      "step": 8264
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5826222896575928,
+      "learning_rate": 1.4654582763337892e-05,
+      "loss": 1.9301,
+      "step": 8268
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.691020131111145,
+      "learning_rate": 1.4637482900136801e-05,
+      "loss": 1.9719,
+      "step": 8272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5105916261672974,
+      "learning_rate": 1.4620383036935705e-05,
+      "loss": 1.9823,
+      "step": 8276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5825799703598022,
+      "learning_rate": 1.460328317373461e-05,
+      "loss": 1.8319,
+      "step": 8280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7014378309249878,
+      "learning_rate": 1.4586183310533516e-05,
+      "loss": 1.85,
+      "step": 8284
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5736719369888306,
+      "learning_rate": 1.456908344733242e-05,
+      "loss": 1.7308,
+      "step": 8288
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4500672817230225,
+      "learning_rate": 1.455198358413133e-05,
+      "loss": 1.8872,
+      "step": 8292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5889211893081665,
+      "learning_rate": 1.4534883720930233e-05,
+      "loss": 1.9407,
+      "step": 8296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7505712509155273,
+      "learning_rate": 1.4517783857729139e-05,
+      "loss": 1.9576,
+      "step": 8300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6860861778259277,
+      "learning_rate": 1.4500683994528045e-05,
+      "loss": 2.0681,
+      "step": 8304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6268845796585083,
+      "learning_rate": 1.4483584131326949e-05,
+      "loss": 2.0465,
+      "step": 8308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5005725622177124,
+      "learning_rate": 1.4466484268125854e-05,
+      "loss": 1.8097,
+      "step": 8312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6980187892913818,
+      "learning_rate": 1.4449384404924762e-05,
+      "loss": 1.8862,
+      "step": 8316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7522437572479248,
+      "learning_rate": 1.4432284541723667e-05,
+      "loss": 1.8532,
+      "step": 8320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.590436577796936,
+      "learning_rate": 1.4415184678522573e-05,
+      "loss": 1.8276,
+      "step": 8324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.487074613571167,
+      "learning_rate": 1.4398084815321477e-05,
+      "loss": 1.9804,
+      "step": 8328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9625794887542725,
+      "learning_rate": 1.4380984952120383e-05,
+      "loss": 1.9065,
+      "step": 8332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6127842664718628,
+      "learning_rate": 1.436388508891929e-05,
+      "loss": 1.8961,
+      "step": 8336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7317554950714111,
+      "learning_rate": 1.4346785225718196e-05,
+      "loss": 1.9341,
+      "step": 8340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5197057723999023,
+      "learning_rate": 1.4329685362517101e-05,
+      "loss": 1.9849,
+      "step": 8344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5553900003433228,
+      "learning_rate": 1.4312585499316005e-05,
+      "loss": 1.7892,
+      "step": 8348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.52109956741333,
+      "learning_rate": 1.429548563611491e-05,
+      "loss": 1.8928,
+      "step": 8352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6558705568313599,
+      "learning_rate": 1.4278385772913818e-05,
+      "loss": 1.799,
+      "step": 8356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.612926959991455,
+      "learning_rate": 1.4261285909712724e-05,
+      "loss": 2.0211,
+      "step": 8360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5769792795181274,
+      "learning_rate": 1.424418604651163e-05,
+      "loss": 1.9197,
+      "step": 8364
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.557539939880371,
+      "learning_rate": 1.4227086183310533e-05,
+      "loss": 1.973,
+      "step": 8368
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6096733808517456,
+      "learning_rate": 1.4209986320109439e-05,
+      "loss": 1.9608,
+      "step": 8372
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.759639024734497,
+      "learning_rate": 1.4192886456908345e-05,
+      "loss": 2.0181,
+      "step": 8376
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7144453525543213,
+      "learning_rate": 1.4175786593707252e-05,
+      "loss": 1.9385,
+      "step": 8380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6760494709014893,
+      "learning_rate": 1.4158686730506158e-05,
+      "loss": 2.1005,
+      "step": 8384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7116944789886475,
+      "learning_rate": 1.4141586867305062e-05,
+      "loss": 1.9031,
+      "step": 8388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5736141204833984,
+      "learning_rate": 1.4124487004103967e-05,
+      "loss": 2.0717,
+      "step": 8392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5009993314743042,
+      "learning_rate": 1.4107387140902873e-05,
+      "loss": 1.9062,
+      "step": 8396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.734379529953003,
+      "learning_rate": 1.409028727770178e-05,
+      "loss": 1.9607,
+      "step": 8400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7106738090515137,
+      "learning_rate": 1.4073187414500686e-05,
+      "loss": 1.9597,
+      "step": 8404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.503051996231079,
+      "learning_rate": 1.405608755129959e-05,
+      "loss": 1.8636,
+      "step": 8408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.544773817062378,
+      "learning_rate": 1.4038987688098496e-05,
+      "loss": 1.903,
+      "step": 8412
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6054282188415527,
+      "learning_rate": 1.4021887824897401e-05,
+      "loss": 2.01,
+      "step": 8416
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5877810716629028,
+      "learning_rate": 1.4004787961696309e-05,
+      "loss": 1.8562,
+      "step": 8420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.669568419456482,
+      "learning_rate": 1.3987688098495214e-05,
+      "loss": 1.8486,
+      "step": 8424
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.780972957611084,
+      "learning_rate": 1.3970588235294118e-05,
+      "loss": 1.9747,
+      "step": 8428
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4789724349975586,
+      "learning_rate": 1.3953488372093024e-05,
+      "loss": 1.9874,
+      "step": 8432
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5372753143310547,
+      "learning_rate": 1.393638850889193e-05,
+      "loss": 1.6921,
+      "step": 8436
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5397859811782837,
+      "learning_rate": 1.3919288645690833e-05,
+      "loss": 1.9584,
+      "step": 8440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.774103045463562,
+      "learning_rate": 1.3902188782489742e-05,
+      "loss": 1.9538,
+      "step": 8444
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6088874340057373,
+      "learning_rate": 1.3885088919288646e-05,
+      "loss": 1.9065,
+      "step": 8448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5402274131774902,
+      "learning_rate": 1.3867989056087552e-05,
+      "loss": 1.8515,
+      "step": 8452
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.530392050743103,
+      "learning_rate": 1.3850889192886458e-05,
+      "loss": 1.6816,
+      "step": 8456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4472659826278687,
+      "learning_rate": 1.3833789329685362e-05,
+      "loss": 1.8966,
+      "step": 8460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.587443470954895,
+      "learning_rate": 1.381668946648427e-05,
+      "loss": 1.7889,
+      "step": 8464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7226911783218384,
+      "learning_rate": 1.3799589603283175e-05,
+      "loss": 1.9593,
+      "step": 8468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.587898850440979,
+      "learning_rate": 1.378248974008208e-05,
+      "loss": 1.8951,
+      "step": 8472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9148885011672974,
+      "learning_rate": 1.3765389876880986e-05,
+      "loss": 1.9351,
+      "step": 8476
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8292063474655151,
+      "learning_rate": 1.374829001367989e-05,
+      "loss": 1.8462,
+      "step": 8480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.73773992061615,
+      "learning_rate": 1.3731190150478799e-05,
+      "loss": 1.8633,
+      "step": 8484
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4449963569641113,
+      "learning_rate": 1.3714090287277703e-05,
+      "loss": 1.924,
+      "step": 8488
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5996627807617188,
+      "learning_rate": 1.3696990424076609e-05,
+      "loss": 1.8774,
+      "step": 8492
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6882178783416748,
+      "learning_rate": 1.3679890560875514e-05,
+      "loss": 1.9221,
+      "step": 8496
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6781047582626343,
+      "learning_rate": 1.3662790697674418e-05,
+      "loss": 1.7734,
+      "step": 8500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6286660432815552,
+      "learning_rate": 1.3645690834473324e-05,
+      "loss": 1.8426,
+      "step": 8504
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4449224472045898,
+      "learning_rate": 1.3628590971272231e-05,
+      "loss": 1.8119,
+      "step": 8508
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6157493591308594,
+      "learning_rate": 1.3611491108071137e-05,
+      "loss": 1.7127,
+      "step": 8512
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8456206321716309,
+      "learning_rate": 1.3594391244870042e-05,
+      "loss": 1.8209,
+      "step": 8516
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5280355215072632,
+      "learning_rate": 1.3577291381668946e-05,
+      "loss": 2.0005,
+      "step": 8520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8175022602081299,
+      "learning_rate": 1.3560191518467852e-05,
+      "loss": 1.9212,
+      "step": 8524
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5971537828445435,
+      "learning_rate": 1.354309165526676e-05,
+      "loss": 1.8606,
+      "step": 8528
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7127560377120972,
+      "learning_rate": 1.3525991792065665e-05,
+      "loss": 1.8265,
+      "step": 8532
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.723050832748413,
+      "learning_rate": 1.350889192886457e-05,
+      "loss": 1.9374,
+      "step": 8536
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4820556640625,
+      "learning_rate": 1.3491792065663475e-05,
+      "loss": 1.8634,
+      "step": 8540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.731012225151062,
+      "learning_rate": 1.347469220246238e-05,
+      "loss": 1.9204,
+      "step": 8544
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.586348295211792,
+      "learning_rate": 1.3457592339261288e-05,
+      "loss": 1.8462,
+      "step": 8548
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7355362176895142,
+      "learning_rate": 1.3440492476060193e-05,
+      "loss": 2.0345,
+      "step": 8552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4833712577819824,
+      "learning_rate": 1.3423392612859099e-05,
+      "loss": 1.9132,
+      "step": 8556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.999471664428711,
+      "learning_rate": 1.3406292749658003e-05,
+      "loss": 2.0444,
+      "step": 8560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6343189477920532,
+      "learning_rate": 1.3389192886456909e-05,
+      "loss": 1.8392,
+      "step": 8564
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5069808959960938,
+      "learning_rate": 1.3372093023255814e-05,
+      "loss": 1.8399,
+      "step": 8568
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.521087884902954,
+      "learning_rate": 1.3354993160054722e-05,
+      "loss": 1.9186,
+      "step": 8572
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4739211797714233,
+      "learning_rate": 1.3337893296853627e-05,
+      "loss": 1.9674,
+      "step": 8576
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5574839115142822,
+      "learning_rate": 1.3320793433652531e-05,
+      "loss": 1.8879,
+      "step": 8580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7786009311676025,
+      "learning_rate": 1.3303693570451437e-05,
+      "loss": 1.9308,
+      "step": 8584
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.627382516860962,
+      "learning_rate": 1.3286593707250342e-05,
+      "loss": 1.8588,
+      "step": 8588
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5507442951202393,
+      "learning_rate": 1.326949384404925e-05,
+      "loss": 2.0934,
+      "step": 8592
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.486575722694397,
+      "learning_rate": 1.3252393980848154e-05,
+      "loss": 2.1199,
+      "step": 8596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3391540050506592,
+      "learning_rate": 1.323529411764706e-05,
+      "loss": 1.9324,
+      "step": 8600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.69169020652771,
+      "learning_rate": 1.3218194254445965e-05,
+      "loss": 1.906,
+      "step": 8604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7022899389266968,
+      "learning_rate": 1.3201094391244869e-05,
+      "loss": 1.9422,
+      "step": 8608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4805599451065063,
+      "learning_rate": 1.3183994528043778e-05,
+      "loss": 1.8661,
+      "step": 8612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5892281532287598,
+      "learning_rate": 1.3166894664842682e-05,
+      "loss": 1.8542,
+      "step": 8616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6527740955352783,
+      "learning_rate": 1.3149794801641588e-05,
+      "loss": 1.8107,
+      "step": 8620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5607056617736816,
+      "learning_rate": 1.3132694938440493e-05,
+      "loss": 1.8057,
+      "step": 8624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5519297122955322,
+      "learning_rate": 1.3115595075239397e-05,
+      "loss": 1.8115,
+      "step": 8628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5834949016571045,
+      "learning_rate": 1.3098495212038303e-05,
+      "loss": 1.8907,
+      "step": 8632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6823513507843018,
+      "learning_rate": 1.308139534883721e-05,
+      "loss": 1.7454,
+      "step": 8636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6757451295852661,
+      "learning_rate": 1.3064295485636116e-05,
+      "loss": 1.7607,
+      "step": 8640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6643555164337158,
+      "learning_rate": 1.3047195622435022e-05,
+      "loss": 1.9063,
+      "step": 8644
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7445098161697388,
+      "learning_rate": 1.3030095759233925e-05,
+      "loss": 1.7924,
+      "step": 8648
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6503303050994873,
+      "learning_rate": 1.3012995896032831e-05,
+      "loss": 2.072,
+      "step": 8652
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5941767692565918,
+      "learning_rate": 1.2995896032831738e-05,
+      "loss": 2.0103,
+      "step": 8656
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6478708982467651,
+      "learning_rate": 1.2978796169630644e-05,
+      "loss": 1.7316,
+      "step": 8660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5611132383346558,
+      "learning_rate": 1.296169630642955e-05,
+      "loss": 1.822,
+      "step": 8664
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5451310873031616,
+      "learning_rate": 1.2944596443228454e-05,
+      "loss": 2.0152,
+      "step": 8668
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7043241262435913,
+      "learning_rate": 1.292749658002736e-05,
+      "loss": 1.9034,
+      "step": 8672
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7091752290725708,
+      "learning_rate": 1.2910396716826267e-05,
+      "loss": 1.8423,
+      "step": 8676
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6127060651779175,
+      "learning_rate": 1.2893296853625172e-05,
+      "loss": 1.6937,
+      "step": 8680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7381840944290161,
+      "learning_rate": 1.2876196990424078e-05,
+      "loss": 1.9543,
+      "step": 8684
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5597513914108276,
+      "learning_rate": 1.2859097127222982e-05,
+      "loss": 1.8262,
+      "step": 8688
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6478689908981323,
+      "learning_rate": 1.2841997264021888e-05,
+      "loss": 2.0061,
+      "step": 8692
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6866868734359741,
+      "learning_rate": 1.2824897400820793e-05,
+      "loss": 2.0661,
+      "step": 8696
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6700700521469116,
+      "learning_rate": 1.28077975376197e-05,
+      "loss": 1.8317,
+      "step": 8700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6782305240631104,
+      "learning_rate": 1.2790697674418606e-05,
+      "loss": 1.7571,
+      "step": 8704
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5588642358779907,
+      "learning_rate": 1.277359781121751e-05,
+      "loss": 1.8465,
+      "step": 8708
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5376290082931519,
+      "learning_rate": 1.2756497948016416e-05,
+      "loss": 1.8062,
+      "step": 8712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6274547576904297,
+      "learning_rate": 1.2739398084815322e-05,
+      "loss": 1.8891,
+      "step": 8716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.577243447303772,
+      "learning_rate": 1.2722298221614229e-05,
+      "loss": 2.0104,
+      "step": 8720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.848371148109436,
+      "learning_rate": 1.2705198358413134e-05,
+      "loss": 1.7878,
+      "step": 8724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5511852502822876,
+      "learning_rate": 1.2688098495212038e-05,
+      "loss": 1.765,
+      "step": 8728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.580466389656067,
+      "learning_rate": 1.2670998632010944e-05,
+      "loss": 1.9052,
+      "step": 8732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6786214113235474,
+      "learning_rate": 1.265389876880985e-05,
+      "loss": 2.0108,
+      "step": 8736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.574545979499817,
+      "learning_rate": 1.2636798905608757e-05,
+      "loss": 1.9098,
+      "step": 8740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7882972955703735,
+      "learning_rate": 1.2619699042407663e-05,
+      "loss": 1.9742,
+      "step": 8744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6370630264282227,
+      "learning_rate": 1.2602599179206567e-05,
+      "loss": 1.8861,
+      "step": 8748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.044381618499756,
+      "learning_rate": 1.2585499316005472e-05,
+      "loss": 1.7537,
+      "step": 8752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6843827962875366,
+      "learning_rate": 1.2568399452804378e-05,
+      "loss": 1.9561,
+      "step": 8756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.605976939201355,
+      "learning_rate": 1.2551299589603282e-05,
+      "loss": 1.8912,
+      "step": 8760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6944398880004883,
+      "learning_rate": 1.2534199726402191e-05,
+      "loss": 1.988,
+      "step": 8764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5753448009490967,
+      "learning_rate": 1.2517099863201095e-05,
+      "loss": 1.9168,
+      "step": 8768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6419624090194702,
+      "learning_rate": 1.25e-05,
+      "loss": 1.9642,
+      "step": 8772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6650840044021606,
+      "learning_rate": 1.2482900136798906e-05,
+      "loss": 1.9391,
+      "step": 8776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5701370239257812,
+      "learning_rate": 1.2465800273597812e-05,
+      "loss": 1.7449,
+      "step": 8780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6118509769439697,
+      "learning_rate": 1.2448700410396718e-05,
+      "loss": 1.9024,
+      "step": 8784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4989128112792969,
+      "learning_rate": 1.2431600547195622e-05,
+      "loss": 1.8511,
+      "step": 8788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8597005605697632,
+      "learning_rate": 1.2414500683994529e-05,
+      "loss": 2.0558,
+      "step": 8792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6900529861450195,
+      "learning_rate": 1.2397400820793434e-05,
+      "loss": 1.9513,
+      "step": 8796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8187960386276245,
+      "learning_rate": 1.238030095759234e-05,
+      "loss": 2.057,
+      "step": 8800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5813332796096802,
+      "learning_rate": 1.2363201094391246e-05,
+      "loss": 1.8454,
+      "step": 8804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.655596137046814,
+      "learning_rate": 1.234610123119015e-05,
+      "loss": 1.8531,
+      "step": 8808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.681291937828064,
+      "learning_rate": 1.2329001367989057e-05,
+      "loss": 1.8698,
+      "step": 8812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6871446371078491,
+      "learning_rate": 1.2311901504787963e-05,
+      "loss": 1.7355,
+      "step": 8816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.62662935256958,
+      "learning_rate": 1.2294801641586867e-05,
+      "loss": 2.067,
+      "step": 8820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6432703733444214,
+      "learning_rate": 1.2277701778385774e-05,
+      "loss": 1.8635,
+      "step": 8824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.410051941871643,
+      "learning_rate": 1.2260601915184678e-05,
+      "loss": 1.7336,
+      "step": 8828
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6481599807739258,
+      "learning_rate": 1.2243502051983585e-05,
+      "loss": 2.025,
+      "step": 8832
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7318426370620728,
+      "learning_rate": 1.2226402188782491e-05,
+      "loss": 1.9353,
+      "step": 8836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6890826225280762,
+      "learning_rate": 1.2209302325581395e-05,
+      "loss": 1.8659,
+      "step": 8840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5665086507797241,
+      "learning_rate": 1.2192202462380302e-05,
+      "loss": 1.9448,
+      "step": 8844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7020827531814575,
+      "learning_rate": 1.2175102599179206e-05,
+      "loss": 2.0202,
+      "step": 8848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6117048263549805,
+      "learning_rate": 1.2158002735978112e-05,
+      "loss": 1.9785,
+      "step": 8852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5090599060058594,
+      "learning_rate": 1.214090287277702e-05,
+      "loss": 1.929,
+      "step": 8856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5808215141296387,
+      "learning_rate": 1.2123803009575923e-05,
+      "loss": 1.9534,
+      "step": 8860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5516841411590576,
+      "learning_rate": 1.210670314637483e-05,
+      "loss": 1.9717,
+      "step": 8864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.79678475856781,
+      "learning_rate": 1.2089603283173734e-05,
+      "loss": 1.8398,
+      "step": 8868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5560812950134277,
+      "learning_rate": 1.207250341997264e-05,
+      "loss": 1.9108,
+      "step": 8872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.660593032836914,
+      "learning_rate": 1.2055403556771547e-05,
+      "loss": 1.9138,
+      "step": 8876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6071677207946777,
+      "learning_rate": 1.2038303693570451e-05,
+      "loss": 1.7974,
+      "step": 8880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5675511360168457,
+      "learning_rate": 1.2021203830369357e-05,
+      "loss": 1.7748,
+      "step": 8884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5390551090240479,
+      "learning_rate": 1.2004103967168263e-05,
+      "loss": 1.8618,
+      "step": 8888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4867161512374878,
+      "learning_rate": 1.1987004103967168e-05,
+      "loss": 1.7825,
+      "step": 8892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6912176609039307,
+      "learning_rate": 1.1969904240766076e-05,
+      "loss": 1.9014,
+      "step": 8896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5211007595062256,
+      "learning_rate": 1.195280437756498e-05,
+      "loss": 1.8937,
+      "step": 8900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.582693099975586,
+      "learning_rate": 1.1935704514363885e-05,
+      "loss": 1.8975,
+      "step": 8904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5599603652954102,
+      "learning_rate": 1.1918604651162791e-05,
+      "loss": 1.8864,
+      "step": 8908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5778361558914185,
+      "learning_rate": 1.1901504787961697e-05,
+      "loss": 2.0265,
+      "step": 8912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5679864883422852,
+      "learning_rate": 1.1884404924760602e-05,
+      "loss": 1.8832,
+      "step": 8916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.741749882698059,
+      "learning_rate": 1.1867305061559508e-05,
+      "loss": 1.9069,
+      "step": 8920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6839423179626465,
+      "learning_rate": 1.1850205198358414e-05,
+      "loss": 1.8291,
+      "step": 8924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.638995885848999,
+      "learning_rate": 1.183310533515732e-05,
+      "loss": 1.8998,
+      "step": 8928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6037232875823975,
+      "learning_rate": 1.1816005471956225e-05,
+      "loss": 1.8874,
+      "step": 8932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.105273962020874,
+      "learning_rate": 1.179890560875513e-05,
+      "loss": 1.9178,
+      "step": 8936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7534767389297485,
+      "learning_rate": 1.1781805745554036e-05,
+      "loss": 1.893,
+      "step": 8940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6804143190383911,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 1.9738,
+      "step": 8944
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6752735376358032,
+      "learning_rate": 1.1747606019151847e-05,
+      "loss": 1.774,
+      "step": 8948
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6710683107376099,
+      "learning_rate": 1.1730506155950753e-05,
+      "loss": 1.9189,
+      "step": 8952
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5679121017456055,
+      "learning_rate": 1.1713406292749659e-05,
+      "loss": 1.7889,
+      "step": 8956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6607458591461182,
+      "learning_rate": 1.1696306429548564e-05,
+      "loss": 1.7928,
+      "step": 8960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.458150029182434,
+      "learning_rate": 1.167920656634747e-05,
+      "loss": 1.8047,
+      "step": 8964
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8078303337097168,
+      "learning_rate": 1.1662106703146374e-05,
+      "loss": 2.096,
+      "step": 8968
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7263890504837036,
+      "learning_rate": 1.1645006839945281e-05,
+      "loss": 2.0121,
+      "step": 8972
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5924429893493652,
+      "learning_rate": 1.1627906976744187e-05,
+      "loss": 1.9003,
+      "step": 8976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7311723232269287,
+      "learning_rate": 1.1610807113543091e-05,
+      "loss": 1.8617,
+      "step": 8980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6257747411727905,
+      "learning_rate": 1.1593707250341998e-05,
+      "loss": 1.7866,
+      "step": 8984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6690551042556763,
+      "learning_rate": 1.1576607387140902e-05,
+      "loss": 1.9114,
+      "step": 8988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6894851922988892,
+      "learning_rate": 1.155950752393981e-05,
+      "loss": 1.9979,
+      "step": 8992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7198370695114136,
+      "learning_rate": 1.1542407660738715e-05,
+      "loss": 2.0374,
+      "step": 8996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5919808149337769,
+      "learning_rate": 1.152530779753762e-05,
+      "loss": 1.7399,
+      "step": 9000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6823852062225342,
+      "learning_rate": 1.1508207934336527e-05,
+      "loss": 1.8478,
+      "step": 9004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7140239477157593,
+      "learning_rate": 1.149110807113543e-05,
+      "loss": 1.9611,
+      "step": 9008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5392545461654663,
+      "learning_rate": 1.1474008207934336e-05,
+      "loss": 1.7857,
+      "step": 9012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8829190731048584,
+      "learning_rate": 1.1456908344733244e-05,
+      "loss": 1.8474,
+      "step": 9016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5163602828979492,
+      "learning_rate": 1.1439808481532147e-05,
+      "loss": 1.9006,
+      "step": 9020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5375906229019165,
+      "learning_rate": 1.1422708618331055e-05,
+      "loss": 2.0124,
+      "step": 9024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5933159589767456,
+      "learning_rate": 1.1405608755129959e-05,
+      "loss": 1.8251,
+      "step": 9028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8997576236724854,
+      "learning_rate": 1.1388508891928864e-05,
+      "loss": 1.9688,
+      "step": 9032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5310240983963013,
+      "learning_rate": 1.1371409028727772e-05,
+      "loss": 1.9101,
+      "step": 9036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7153500318527222,
+      "learning_rate": 1.1354309165526676e-05,
+      "loss": 1.7864,
+      "step": 9040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7716693878173828,
+      "learning_rate": 1.1337209302325581e-05,
+      "loss": 1.9119,
+      "step": 9044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8320080041885376,
+      "learning_rate": 1.1320109439124487e-05,
+      "loss": 2.0897,
+      "step": 9048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6159136295318604,
+      "learning_rate": 1.1303009575923393e-05,
+      "loss": 2.0392,
+      "step": 9052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5778326988220215,
+      "learning_rate": 1.12859097127223e-05,
+      "loss": 1.9324,
+      "step": 9056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5392941236495972,
+      "learning_rate": 1.1268809849521204e-05,
+      "loss": 1.8133,
+      "step": 9060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6904296875,
+      "learning_rate": 1.125170998632011e-05,
+      "loss": 1.9265,
+      "step": 9064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7861131429672241,
+      "learning_rate": 1.1234610123119015e-05,
+      "loss": 1.7405,
+      "step": 9068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6475653648376465,
+      "learning_rate": 1.1217510259917921e-05,
+      "loss": 1.8417,
+      "step": 9072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4819046258926392,
+      "learning_rate": 1.1200410396716827e-05,
+      "loss": 1.7927,
+      "step": 9076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5303575992584229,
+      "learning_rate": 1.1183310533515732e-05,
+      "loss": 1.735,
+      "step": 9080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5807466506958008,
+      "learning_rate": 1.1166210670314638e-05,
+      "loss": 1.7733,
+      "step": 9084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6154811382293701,
+      "learning_rate": 1.1149110807113544e-05,
+      "loss": 1.985,
+      "step": 9088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.588836908340454,
+      "learning_rate": 1.113201094391245e-05,
+      "loss": 1.8861,
+      "step": 9092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6571130752563477,
+      "learning_rate": 1.1114911080711355e-05,
+      "loss": 1.9013,
+      "step": 9096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7480363845825195,
+      "learning_rate": 1.109781121751026e-05,
+      "loss": 2.0298,
+      "step": 9100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.53179931640625,
+      "learning_rate": 1.1080711354309166e-05,
+      "loss": 1.8883,
+      "step": 9104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7249871492385864,
+      "learning_rate": 1.1063611491108072e-05,
+      "loss": 1.888,
+      "step": 9108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.771964192390442,
+      "learning_rate": 1.1046511627906977e-05,
+      "loss": 1.7898,
+      "step": 9112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8706704378128052,
+      "learning_rate": 1.1029411764705883e-05,
+      "loss": 1.8936,
+      "step": 9116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6252784729003906,
+      "learning_rate": 1.1012311901504789e-05,
+      "loss": 1.909,
+      "step": 9120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5766993761062622,
+      "learning_rate": 1.0995212038303694e-05,
+      "loss": 1.7883,
+      "step": 9124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5931488275527954,
+      "learning_rate": 1.09781121751026e-05,
+      "loss": 1.8009,
+      "step": 9128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.652861475944519,
+      "learning_rate": 1.0961012311901506e-05,
+      "loss": 1.9441,
+      "step": 9132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.604109764099121,
+      "learning_rate": 1.0943912448700411e-05,
+      "loss": 1.944,
+      "step": 9136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.648501992225647,
+      "learning_rate": 1.0926812585499315e-05,
+      "loss": 1.7164,
+      "step": 9140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.697082757949829,
+      "learning_rate": 1.0909712722298223e-05,
+      "loss": 1.9967,
+      "step": 9144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5939291715621948,
+      "learning_rate": 1.0892612859097127e-05,
+      "loss": 1.8479,
+      "step": 9148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4296135902404785,
+      "learning_rate": 1.0875512995896034e-05,
+      "loss": 1.7912,
+      "step": 9152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5867764949798584,
+      "learning_rate": 1.085841313269494e-05,
+      "loss": 2.0543,
+      "step": 9156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.877381443977356,
+      "learning_rate": 1.0841313269493844e-05,
+      "loss": 1.9605,
+      "step": 9160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5704667568206787,
+      "learning_rate": 1.0824213406292751e-05,
+      "loss": 1.8716,
+      "step": 9164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7016427516937256,
+      "learning_rate": 1.0807113543091655e-05,
+      "loss": 1.8002,
+      "step": 9168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5268316268920898,
+      "learning_rate": 1.079001367989056e-05,
+      "loss": 1.8328,
+      "step": 9172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.773216962814331,
+      "learning_rate": 1.0772913816689468e-05,
+      "loss": 2.0852,
+      "step": 9176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8278284072875977,
+      "learning_rate": 1.0755813953488372e-05,
+      "loss": 1.7678,
+      "step": 9180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9138497114181519,
+      "learning_rate": 1.0738714090287279e-05,
+      "loss": 1.8216,
+      "step": 9184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6149177551269531,
+      "learning_rate": 1.0721614227086183e-05,
+      "loss": 2.0147,
+      "step": 9188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5738332271575928,
+      "learning_rate": 1.0704514363885089e-05,
+      "loss": 1.9592,
+      "step": 9192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.617006778717041,
+      "learning_rate": 1.0687414500683996e-05,
+      "loss": 1.9349,
+      "step": 9196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6684281826019287,
+      "learning_rate": 1.06703146374829e-05,
+      "loss": 1.9102,
+      "step": 9200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5881692171096802,
+      "learning_rate": 1.0653214774281806e-05,
+      "loss": 1.7868,
+      "step": 9204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6087418794631958,
+      "learning_rate": 1.0636114911080711e-05,
+      "loss": 1.8317,
+      "step": 9208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5474470853805542,
+      "learning_rate": 1.0619015047879617e-05,
+      "loss": 1.8203,
+      "step": 9212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6170072555541992,
+      "learning_rate": 1.0601915184678524e-05,
+      "loss": 1.9812,
+      "step": 9216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6150282621383667,
+      "learning_rate": 1.0584815321477428e-05,
+      "loss": 1.831,
+      "step": 9220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.888481855392456,
+      "learning_rate": 1.0567715458276334e-05,
+      "loss": 1.9179,
+      "step": 9224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.499454379081726,
+      "learning_rate": 1.055061559507524e-05,
+      "loss": 1.9117,
+      "step": 9228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8907907009124756,
+      "learning_rate": 1.0533515731874145e-05,
+      "loss": 1.9377,
+      "step": 9232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6337422132492065,
+      "learning_rate": 1.0516415868673051e-05,
+      "loss": 1.946,
+      "step": 9236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.692440390586853,
+      "learning_rate": 1.0499316005471957e-05,
+      "loss": 2.0171,
+      "step": 9240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8192511796951294,
+      "learning_rate": 1.0482216142270862e-05,
+      "loss": 2.0665,
+      "step": 9244
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7105309963226318,
+      "learning_rate": 1.0465116279069768e-05,
+      "loss": 1.8517,
+      "step": 9248
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.570191740989685,
+      "learning_rate": 1.0448016415868673e-05,
+      "loss": 1.879,
+      "step": 9252
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5352414846420288,
+      "learning_rate": 1.0430916552667579e-05,
+      "loss": 1.9844,
+      "step": 9256
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6077239513397217,
+      "learning_rate": 1.0413816689466485e-05,
+      "loss": 1.8066,
+      "step": 9260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6216827630996704,
+      "learning_rate": 1.039671682626539e-05,
+      "loss": 1.7569,
+      "step": 9264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5977295637130737,
+      "learning_rate": 1.0379616963064296e-05,
+      "loss": 1.841,
+      "step": 9268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7415016889572144,
+      "learning_rate": 1.0362517099863202e-05,
+      "loss": 2.0001,
+      "step": 9272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6122945547103882,
+      "learning_rate": 1.0345417236662107e-05,
+      "loss": 1.7887,
+      "step": 9276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7112343311309814,
+      "learning_rate": 1.0328317373461013e-05,
+      "loss": 1.9709,
+      "step": 9280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.583895206451416,
+      "learning_rate": 1.0311217510259919e-05,
+      "loss": 2.0006,
+      "step": 9284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5309836864471436,
+      "learning_rate": 1.0294117647058824e-05,
+      "loss": 1.8978,
+      "step": 9288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6894787549972534,
+      "learning_rate": 1.027701778385773e-05,
+      "loss": 1.856,
+      "step": 9292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5949068069458008,
+      "learning_rate": 1.0259917920656636e-05,
+      "loss": 1.84,
+      "step": 9296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5734741687774658,
+      "learning_rate": 1.024281805745554e-05,
+      "loss": 1.9617,
+      "step": 9300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7839090824127197,
+      "learning_rate": 1.0225718194254447e-05,
+      "loss": 1.7957,
+      "step": 9304
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9882410764694214,
+      "learning_rate": 1.0208618331053353e-05,
+      "loss": 2.0439,
+      "step": 9308
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6469732522964478,
+      "learning_rate": 1.0191518467852258e-05,
+      "loss": 1.82,
+      "step": 9312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.597993016242981,
+      "learning_rate": 1.0174418604651164e-05,
+      "loss": 2.0186,
+      "step": 9316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5456162691116333,
+      "learning_rate": 1.0157318741450068e-05,
+      "loss": 1.9057,
+      "step": 9320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.589836597442627,
+      "learning_rate": 1.0140218878248975e-05,
+      "loss": 1.8665,
+      "step": 9324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5683108568191528,
+      "learning_rate": 1.0123119015047879e-05,
+      "loss": 1.7532,
+      "step": 9328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5445048809051514,
+      "learning_rate": 1.0106019151846785e-05,
+      "loss": 1.9741,
+      "step": 9332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6579010486602783,
+      "learning_rate": 1.0088919288645692e-05,
+      "loss": 1.8469,
+      "step": 9336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.813675045967102,
+      "learning_rate": 1.0071819425444596e-05,
+      "loss": 1.9811,
+      "step": 9340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.69435453414917,
+      "learning_rate": 1.0054719562243503e-05,
+      "loss": 1.6031,
+      "step": 9344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9789328575134277,
+      "learning_rate": 1.0037619699042407e-05,
+      "loss": 2.0603,
+      "step": 9348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7569694519042969,
+      "learning_rate": 1.0020519835841313e-05,
+      "loss": 1.8548,
+      "step": 9352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7058336734771729,
+      "learning_rate": 1.000341997264022e-05,
+      "loss": 1.8937,
+      "step": 9356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7999751567840576,
+      "learning_rate": 9.986320109439124e-06,
+      "loss": 2.0138,
+      "step": 9360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6166166067123413,
+      "learning_rate": 9.96922024623803e-06,
+      "loss": 1.8514,
+      "step": 9364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5393308401107788,
+      "learning_rate": 9.952120383036936e-06,
+      "loss": 1.8538,
+      "step": 9368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7060655355453491,
+      "learning_rate": 9.935020519835841e-06,
+      "loss": 1.8204,
+      "step": 9372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5741777420043945,
+      "learning_rate": 9.917920656634749e-06,
+      "loss": 1.8723,
+      "step": 9376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7236156463623047,
+      "learning_rate": 9.900820793433653e-06,
+      "loss": 2.0826,
+      "step": 9380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6071019172668457,
+      "learning_rate": 9.883720930232558e-06,
+      "loss": 1.8699,
+      "step": 9384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5609468221664429,
+      "learning_rate": 9.866621067031464e-06,
+      "loss": 1.7908,
+      "step": 9388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5828368663787842,
+      "learning_rate": 9.84952120383037e-06,
+      "loss": 1.9035,
+      "step": 9392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6804313659667969,
+      "learning_rate": 9.832421340629275e-06,
+      "loss": 1.873,
+      "step": 9396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7444369792938232,
+      "learning_rate": 9.81532147742818e-06,
+      "loss": 1.9714,
+      "step": 9400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.478379487991333,
+      "learning_rate": 9.798221614227086e-06,
+      "loss": 1.825,
+      "step": 9404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8659464120864868,
+      "learning_rate": 9.781121751025992e-06,
+      "loss": 2.0924,
+      "step": 9408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6995452642440796,
+      "learning_rate": 9.764021887824898e-06,
+      "loss": 1.7909,
+      "step": 9412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7288949489593506,
+      "learning_rate": 9.746922024623803e-06,
+      "loss": 1.9683,
+      "step": 9416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.728184461593628,
+      "learning_rate": 9.729822161422709e-06,
+      "loss": 1.867,
+      "step": 9420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6516950130462646,
+      "learning_rate": 9.712722298221615e-06,
+      "loss": 1.847,
+      "step": 9424
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6084563732147217,
+      "learning_rate": 9.69562243502052e-06,
+      "loss": 1.8446,
+      "step": 9428
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6345360279083252,
+      "learning_rate": 9.678522571819426e-06,
+      "loss": 1.8856,
+      "step": 9432
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6293641328811646,
+      "learning_rate": 9.661422708618332e-06,
+      "loss": 1.8721,
+      "step": 9436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7901579141616821,
+      "learning_rate": 9.644322845417237e-06,
+      "loss": 1.8333,
+      "step": 9440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8444390296936035,
+      "learning_rate": 9.627222982216143e-06,
+      "loss": 1.8012,
+      "step": 9444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9100682735443115,
+      "learning_rate": 9.610123119015049e-06,
+      "loss": 1.8989,
+      "step": 9448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5829551219940186,
+      "learning_rate": 9.593023255813954e-06,
+      "loss": 1.692,
+      "step": 9452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7920702695846558,
+      "learning_rate": 9.57592339261286e-06,
+      "loss": 1.9008,
+      "step": 9456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8126853704452515,
+      "learning_rate": 9.558823529411764e-06,
+      "loss": 1.8993,
+      "step": 9460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6333680152893066,
+      "learning_rate": 9.541723666210671e-06,
+      "loss": 1.8486,
+      "step": 9464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6809686422348022,
+      "learning_rate": 9.524623803009577e-06,
+      "loss": 1.8182,
+      "step": 9468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5512781143188477,
+      "learning_rate": 9.507523939808483e-06,
+      "loss": 1.782,
+      "step": 9472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4863585233688354,
+      "learning_rate": 9.490424076607388e-06,
+      "loss": 1.9,
+      "step": 9476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8519527912139893,
+      "learning_rate": 9.473324213406292e-06,
+      "loss": 1.8792,
+      "step": 9480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.68942391872406,
+      "learning_rate": 9.4562243502052e-06,
+      "loss": 1.81,
+      "step": 9484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6430591344833374,
+      "learning_rate": 9.439124487004105e-06,
+      "loss": 1.7057,
+      "step": 9488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7153677940368652,
+      "learning_rate": 9.422024623803009e-06,
+      "loss": 1.7675,
+      "step": 9492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2399628162384033,
+      "learning_rate": 9.404924760601916e-06,
+      "loss": 1.9985,
+      "step": 9496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6189675331115723,
+      "learning_rate": 9.38782489740082e-06,
+      "loss": 1.8922,
+      "step": 9500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6087392568588257,
+      "learning_rate": 9.370725034199728e-06,
+      "loss": 1.8978,
+      "step": 9504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7213088274002075,
+      "learning_rate": 9.353625170998632e-06,
+      "loss": 1.9047,
+      "step": 9508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6473731994628906,
+      "learning_rate": 9.336525307797537e-06,
+      "loss": 1.922,
+      "step": 9512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6384577751159668,
+      "learning_rate": 9.319425444596445e-06,
+      "loss": 1.9051,
+      "step": 9516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6715625524520874,
+      "learning_rate": 9.302325581395349e-06,
+      "loss": 1.7335,
+      "step": 9520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7468408346176147,
+      "learning_rate": 9.285225718194254e-06,
+      "loss": 1.8317,
+      "step": 9524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6021628379821777,
+      "learning_rate": 9.26812585499316e-06,
+      "loss": 1.7717,
+      "step": 9528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4728972911834717,
+      "learning_rate": 9.251025991792066e-06,
+      "loss": 1.8383,
+      "step": 9532
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.641257643699646,
+      "learning_rate": 9.233926128590973e-06,
+      "loss": 1.9203,
+      "step": 9536
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5959044694900513,
+      "learning_rate": 9.216826265389877e-06,
+      "loss": 1.8186,
+      "step": 9540
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8256573677062988,
+      "learning_rate": 9.199726402188783e-06,
+      "loss": 1.9279,
+      "step": 9544
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6256192922592163,
+      "learning_rate": 9.182626538987688e-06,
+      "loss": 1.8773,
+      "step": 9548
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7054941654205322,
+      "learning_rate": 9.165526675786594e-06,
+      "loss": 1.8826,
+      "step": 9552
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7333673238754272,
+      "learning_rate": 9.1484268125855e-06,
+      "loss": 1.8855,
+      "step": 9556
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9579733610153198,
+      "learning_rate": 9.131326949384405e-06,
+      "loss": 1.7222,
+      "step": 9560
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6406586170196533,
+      "learning_rate": 9.11422708618331e-06,
+      "loss": 1.8063,
+      "step": 9564
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8166844844818115,
+      "learning_rate": 9.097127222982216e-06,
+      "loss": 1.8802,
+      "step": 9568
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.815399169921875,
+      "learning_rate": 9.080027359781122e-06,
+      "loss": 1.9574,
+      "step": 9572
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7717112302780151,
+      "learning_rate": 9.062927496580028e-06,
+      "loss": 1.8687,
+      "step": 9576
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8283933401107788,
+      "learning_rate": 9.045827633378933e-06,
+      "loss": 1.7794,
+      "step": 9580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5563534498214722,
+      "learning_rate": 9.028727770177839e-06,
+      "loss": 1.9697,
+      "step": 9584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7888858318328857,
+      "learning_rate": 9.011627906976745e-06,
+      "loss": 1.8549,
+      "step": 9588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6961920261383057,
+      "learning_rate": 8.99452804377565e-06,
+      "loss": 1.8044,
+      "step": 9592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6775288581848145,
+      "learning_rate": 8.977428180574556e-06,
+      "loss": 1.7884,
+      "step": 9596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5376509428024292,
+      "learning_rate": 8.960328317373462e-06,
+      "loss": 1.9472,
+      "step": 9600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5770347118377686,
+      "learning_rate": 8.943228454172367e-06,
+      "loss": 1.7587,
+      "step": 9604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5645045042037964,
+      "learning_rate": 8.926128590971273e-06,
+      "loss": 1.9036,
+      "step": 9608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6209347248077393,
+      "learning_rate": 8.909028727770179e-06,
+      "loss": 1.9395,
+      "step": 9612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5157575607299805,
+      "learning_rate": 8.891928864569084e-06,
+      "loss": 1.9096,
+      "step": 9616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6890482902526855,
+      "learning_rate": 8.874829001367988e-06,
+      "loss": 1.8541,
+      "step": 9620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6792246103286743,
+      "learning_rate": 8.857729138166896e-06,
+      "loss": 1.9734,
+      "step": 9624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5776066780090332,
+      "learning_rate": 8.840629274965801e-06,
+      "loss": 1.9166,
+      "step": 9628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5487701892852783,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 1.7777,
+      "step": 9632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6531468629837036,
+      "learning_rate": 8.806429548563612e-06,
+      "loss": 1.6488,
+      "step": 9636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6902531385421753,
+      "learning_rate": 8.789329685362516e-06,
+      "loss": 1.8287,
+      "step": 9640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6954971551895142,
+      "learning_rate": 8.772229822161424e-06,
+      "loss": 1.896,
+      "step": 9644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5442079305648804,
+      "learning_rate": 8.75512995896033e-06,
+      "loss": 1.8335,
+      "step": 9648
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.739211916923523,
+      "learning_rate": 8.738030095759233e-06,
+      "loss": 1.8265,
+      "step": 9652
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6437937021255493,
+      "learning_rate": 8.72093023255814e-06,
+      "loss": 1.7487,
+      "step": 9656
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.568172812461853,
+      "learning_rate": 8.703830369357045e-06,
+      "loss": 1.8618,
+      "step": 9660
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.84135103225708,
+      "learning_rate": 8.686730506155952e-06,
+      "loss": 1.8962,
+      "step": 9664
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.599940538406372,
+      "learning_rate": 8.669630642954858e-06,
+      "loss": 1.7502,
+      "step": 9668
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9855149984359741,
+      "learning_rate": 8.652530779753762e-06,
+      "loss": 2.0228,
+      "step": 9672
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.641782283782959,
+      "learning_rate": 8.635430916552669e-06,
+      "loss": 2.0184,
+      "step": 9676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9988453388214111,
+      "learning_rate": 8.618331053351573e-06,
+      "loss": 2.0865,
+      "step": 9680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7934141159057617,
+      "learning_rate": 8.601231190150479e-06,
+      "loss": 1.8933,
+      "step": 9684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5691959857940674,
+      "learning_rate": 8.584131326949384e-06,
+      "loss": 1.8344,
+      "step": 9688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7705130577087402,
+      "learning_rate": 8.56703146374829e-06,
+      "loss": 1.646,
+      "step": 9692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6504237651824951,
+      "learning_rate": 8.549931600547197e-06,
+      "loss": 1.7268,
+      "step": 9696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4613388776779175,
+      "learning_rate": 8.532831737346101e-06,
+      "loss": 1.7871,
+      "step": 9700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7658060789108276,
+      "learning_rate": 8.515731874145007e-06,
+      "loss": 1.8519,
+      "step": 9704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.84966242313385,
+      "learning_rate": 8.498632010943912e-06,
+      "loss": 1.8277,
+      "step": 9708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.674241542816162,
+      "learning_rate": 8.481532147742818e-06,
+      "loss": 1.7632,
+      "step": 9712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7811789512634277,
+      "learning_rate": 8.464432284541724e-06,
+      "loss": 1.899,
+      "step": 9716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7027822732925415,
+      "learning_rate": 8.44733242134063e-06,
+      "loss": 1.9129,
+      "step": 9720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7912241220474243,
+      "learning_rate": 8.430232558139535e-06,
+      "loss": 1.8546,
+      "step": 9724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5889939069747925,
+      "learning_rate": 8.41313269493844e-06,
+      "loss": 2.0574,
+      "step": 9728
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.55410635471344,
+      "learning_rate": 8.396032831737346e-06,
+      "loss": 1.8547,
+      "step": 9732
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6358247995376587,
+      "learning_rate": 8.378932968536252e-06,
+      "loss": 1.7924,
+      "step": 9736
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.698111891746521,
+      "learning_rate": 8.361833105335158e-06,
+      "loss": 1.8366,
+      "step": 9740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5512586832046509,
+      "learning_rate": 8.344733242134063e-06,
+      "loss": 1.7097,
+      "step": 9744
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5253231525421143,
+      "learning_rate": 8.327633378932969e-06,
+      "loss": 1.9076,
+      "step": 9748
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6231822967529297,
+      "learning_rate": 8.310533515731875e-06,
+      "loss": 1.9026,
+      "step": 9752
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.726797103881836,
+      "learning_rate": 8.29343365253078e-06,
+      "loss": 1.9518,
+      "step": 9756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7323880195617676,
+      "learning_rate": 8.276333789329686e-06,
+      "loss": 1.9352,
+      "step": 9760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7740181684494019,
+      "learning_rate": 8.259233926128592e-06,
+      "loss": 1.8697,
+      "step": 9764
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6174265146255493,
+      "learning_rate": 8.242134062927497e-06,
+      "loss": 1.8365,
+      "step": 9768
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7219985723495483,
+      "learning_rate": 8.225034199726403e-06,
+      "loss": 1.8343,
+      "step": 9772
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.618703007698059,
+      "learning_rate": 8.207934336525308e-06,
+      "loss": 1.914,
+      "step": 9776
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5896012783050537,
+      "learning_rate": 8.190834473324212e-06,
+      "loss": 1.8311,
+      "step": 9780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.614791989326477,
+      "learning_rate": 8.17373461012312e-06,
+      "loss": 1.8189,
+      "step": 9784
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6827095746994019,
+      "learning_rate": 8.156634746922025e-06,
+      "loss": 1.7639,
+      "step": 9788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.563234806060791,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 1.8255,
+      "step": 9792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.871492624282837,
+      "learning_rate": 8.122435020519837e-06,
+      "loss": 1.7558,
+      "step": 9796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.827107548713684,
+      "learning_rate": 8.10533515731874e-06,
+      "loss": 1.93,
+      "step": 9800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7132670879364014,
+      "learning_rate": 8.088235294117648e-06,
+      "loss": 1.9702,
+      "step": 9804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.71363365650177,
+      "learning_rate": 8.071135430916554e-06,
+      "loss": 1.9638,
+      "step": 9808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6538909673690796,
+      "learning_rate": 8.054035567715458e-06,
+      "loss": 1.9329,
+      "step": 9812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4397317171096802,
+      "learning_rate": 8.036935704514365e-06,
+      "loss": 1.7031,
+      "step": 9816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7500611543655396,
+      "learning_rate": 8.019835841313269e-06,
+      "loss": 1.7927,
+      "step": 9820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7213584184646606,
+      "learning_rate": 8.002735978112176e-06,
+      "loss": 1.8121,
+      "step": 9824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8555186986923218,
+      "learning_rate": 7.985636114911082e-06,
+      "loss": 1.9388,
+      "step": 9828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5887397527694702,
+      "learning_rate": 7.968536251709986e-06,
+      "loss": 1.8344,
+      "step": 9832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5652546882629395,
+      "learning_rate": 7.951436388508893e-06,
+      "loss": 1.9696,
+      "step": 9836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5642329454421997,
+      "learning_rate": 7.934336525307797e-06,
+      "loss": 1.6744,
+      "step": 9840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.631618618965149,
+      "learning_rate": 7.917236662106703e-06,
+      "loss": 1.8742,
+      "step": 9844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6112003326416016,
+      "learning_rate": 7.90013679890561e-06,
+      "loss": 1.9281,
+      "step": 9848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.612426996231079,
+      "learning_rate": 7.883036935704514e-06,
+      "loss": 1.9576,
+      "step": 9852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7601630687713623,
+      "learning_rate": 7.865937072503421e-06,
+      "loss": 1.7478,
+      "step": 9856
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.782696008682251,
+      "learning_rate": 7.848837209302325e-06,
+      "loss": 1.8667,
+      "step": 9860
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.716011881828308,
+      "learning_rate": 7.831737346101231e-06,
+      "loss": 1.9523,
+      "step": 9864
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5538578033447266,
+      "learning_rate": 7.814637482900137e-06,
+      "loss": 1.9909,
+      "step": 9868
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6781466007232666,
+      "learning_rate": 7.797537619699042e-06,
+      "loss": 1.7684,
+      "step": 9872
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6765227317810059,
+      "learning_rate": 7.780437756497948e-06,
+      "loss": 1.8663,
+      "step": 9876
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5819069147109985,
+      "learning_rate": 7.763337893296854e-06,
+      "loss": 1.8592,
+      "step": 9880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6805880069732666,
+      "learning_rate": 7.74623803009576e-06,
+      "loss": 1.8612,
+      "step": 9884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6755802631378174,
+      "learning_rate": 7.729138166894665e-06,
+      "loss": 1.8602,
+      "step": 9888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7085371017456055,
+      "learning_rate": 7.71203830369357e-06,
+      "loss": 1.6679,
+      "step": 9892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5899639129638672,
+      "learning_rate": 7.694938440492476e-06,
+      "loss": 1.7626,
+      "step": 9896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.740966796875,
+      "learning_rate": 7.677838577291382e-06,
+      "loss": 1.7984,
+      "step": 9900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8925389051437378,
+      "learning_rate": 7.660738714090288e-06,
+      "loss": 1.8242,
+      "step": 9904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7259442806243896,
+      "learning_rate": 7.643638850889193e-06,
+      "loss": 1.7816,
+      "step": 9908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5764368772506714,
+      "learning_rate": 7.626538987688099e-06,
+      "loss": 1.7718,
+      "step": 9912
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7654069662094116,
+      "learning_rate": 7.6094391244870045e-06,
+      "loss": 1.7728,
+      "step": 9916
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5519089698791504,
+      "learning_rate": 7.592339261285911e-06,
+      "loss": 1.9182,
+      "step": 9920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.690704584121704,
+      "learning_rate": 7.575239398084816e-06,
+      "loss": 1.9885,
+      "step": 9924
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5261144638061523,
+      "learning_rate": 7.558139534883721e-06,
+      "loss": 1.7169,
+      "step": 9928
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5968029499053955,
+      "learning_rate": 7.541039671682627e-06,
+      "loss": 1.7766,
+      "step": 9932
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5991588830947876,
+      "learning_rate": 7.523939808481533e-06,
+      "loss": 1.8359,
+      "step": 9936
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.591503620147705,
+      "learning_rate": 7.506839945280438e-06,
+      "loss": 1.7582,
+      "step": 9940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5666191577911377,
+      "learning_rate": 7.489740082079344e-06,
+      "loss": 1.8947,
+      "step": 9944
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7336504459381104,
+      "learning_rate": 7.472640218878249e-06,
+      "loss": 1.8194,
+      "step": 9948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7414416074752808,
+      "learning_rate": 7.455540355677155e-06,
+      "loss": 1.8543,
+      "step": 9952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6664258241653442,
+      "learning_rate": 7.438440492476061e-06,
+      "loss": 1.786,
+      "step": 9956
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.594941258430481,
+      "learning_rate": 7.421340629274966e-06,
+      "loss": 1.773,
+      "step": 9960
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.71213698387146,
+      "learning_rate": 7.404240766073872e-06,
+      "loss": 1.9381,
+      "step": 9964
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6945880651474,
+      "learning_rate": 7.387140902872777e-06,
+      "loss": 1.7854,
+      "step": 9968
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.71058189868927,
+      "learning_rate": 7.370041039671682e-06,
+      "loss": 1.8552,
+      "step": 9972
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8152309656143188,
+      "learning_rate": 7.3529411764705884e-06,
+      "loss": 1.8759,
+      "step": 9976
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6043202877044678,
+      "learning_rate": 7.335841313269494e-06,
+      "loss": 1.7945,
+      "step": 9980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7119457721710205,
+      "learning_rate": 7.318741450068401e-06,
+      "loss": 1.9138,
+      "step": 9984
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9094306230545044,
+      "learning_rate": 7.301641586867305e-06,
+      "loss": 1.8234,
+      "step": 9988
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6789647340774536,
+      "learning_rate": 7.28454172366621e-06,
+      "loss": 1.8883,
+      "step": 9992
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6013123989105225,
+      "learning_rate": 7.267441860465117e-06,
+      "loss": 1.6778,
+      "step": 9996
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7987964153289795,
+      "learning_rate": 7.250341997264022e-06,
+      "loss": 1.6727,
+      "step": 10000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6856164932250977,
+      "learning_rate": 7.233242134062927e-06,
+      "loss": 1.8289,
+      "step": 10004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7940067052841187,
+      "learning_rate": 7.216142270861834e-06,
+      "loss": 1.917,
+      "step": 10008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.806894063949585,
+      "learning_rate": 7.1990424076607384e-06,
+      "loss": 1.9523,
+      "step": 10012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6045538187026978,
+      "learning_rate": 7.181942544459645e-06,
+      "loss": 1.8155,
+      "step": 10016
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6840627193450928,
+      "learning_rate": 7.164842681258551e-06,
+      "loss": 1.87,
+      "step": 10020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5683917999267578,
+      "learning_rate": 7.147742818057455e-06,
+      "loss": 1.9501,
+      "step": 10024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6646571159362793,
+      "learning_rate": 7.130642954856362e-06,
+      "loss": 1.862,
+      "step": 10028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6929519176483154,
+      "learning_rate": 7.113543091655267e-06,
+      "loss": 1.9485,
+      "step": 10032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.600963830947876,
+      "learning_rate": 7.096443228454172e-06,
+      "loss": 1.7745,
+      "step": 10036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6967799663543701,
+      "learning_rate": 7.079343365253079e-06,
+      "loss": 1.9167,
+      "step": 10040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6387646198272705,
+      "learning_rate": 7.062243502051984e-06,
+      "loss": 1.9051,
+      "step": 10044
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5275384187698364,
+      "learning_rate": 7.04514363885089e-06,
+      "loss": 1.9997,
+      "step": 10048
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.677001953125,
+      "learning_rate": 7.028043775649795e-06,
+      "loss": 1.9359,
+      "step": 10052
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5836267471313477,
+      "learning_rate": 7.010943912448701e-06,
+      "loss": 1.7359,
+      "step": 10056
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7847533226013184,
+      "learning_rate": 6.993844049247607e-06,
+      "loss": 1.9568,
+      "step": 10060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6765036582946777,
+      "learning_rate": 6.976744186046512e-06,
+      "loss": 1.8862,
+      "step": 10064
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8817429542541504,
+      "learning_rate": 6.959644322845417e-06,
+      "loss": 1.8935,
+      "step": 10068
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.765724778175354,
+      "learning_rate": 6.942544459644323e-06,
+      "loss": 1.7135,
+      "step": 10072
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5911288261413574,
+      "learning_rate": 6.925444596443229e-06,
+      "loss": 1.7035,
+      "step": 10076
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.738471508026123,
+      "learning_rate": 6.908344733242135e-06,
+      "loss": 1.7544,
+      "step": 10080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.719092845916748,
+      "learning_rate": 6.89124487004104e-06,
+      "loss": 1.8507,
+      "step": 10084
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.786017656326294,
+      "learning_rate": 6.874145006839945e-06,
+      "loss": 1.9009,
+      "step": 10088
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0530200004577637,
+      "learning_rate": 6.8570451436388514e-06,
+      "loss": 1.8839,
+      "step": 10092
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6444154977798462,
+      "learning_rate": 6.839945280437757e-06,
+      "loss": 1.8756,
+      "step": 10096
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6261224746704102,
+      "learning_rate": 6.822845417236662e-06,
+      "loss": 1.6744,
+      "step": 10100
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6390396356582642,
+      "learning_rate": 6.805745554035568e-06,
+      "loss": 1.8752,
+      "step": 10104
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9215726852416992,
+      "learning_rate": 6.788645690834473e-06,
+      "loss": 1.9022,
+      "step": 10108
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.839656949043274,
+      "learning_rate": 6.77154582763338e-06,
+      "loss": 2.1024,
+      "step": 10112
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6141258478164673,
+      "learning_rate": 6.754445964432285e-06,
+      "loss": 1.7609,
+      "step": 10116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6340337991714478,
+      "learning_rate": 6.73734610123119e-06,
+      "loss": 1.8939,
+      "step": 10120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.529555082321167,
+      "learning_rate": 6.720246238030097e-06,
+      "loss": 1.7863,
+      "step": 10124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.670279860496521,
+      "learning_rate": 6.7031463748290014e-06,
+      "loss": 1.8213,
+      "step": 10128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7106720209121704,
+      "learning_rate": 6.686046511627907e-06,
+      "loss": 1.6828,
+      "step": 10132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8096644878387451,
+      "learning_rate": 6.668946648426814e-06,
+      "loss": 1.8081,
+      "step": 10136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.636301040649414,
+      "learning_rate": 6.651846785225718e-06,
+      "loss": 1.8518,
+      "step": 10140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5952695608139038,
+      "learning_rate": 6.634746922024625e-06,
+      "loss": 1.7897,
+      "step": 10144
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5805269479751587,
+      "learning_rate": 6.61764705882353e-06,
+      "loss": 1.9542,
+      "step": 10148
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8776960372924805,
+      "learning_rate": 6.6005471956224345e-06,
+      "loss": 1.8719,
+      "step": 10152
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5122414827346802,
+      "learning_rate": 6.583447332421341e-06,
+      "loss": 1.7259,
+      "step": 10156
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7734562158584595,
+      "learning_rate": 6.566347469220247e-06,
+      "loss": 1.8812,
+      "step": 10160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.511674165725708,
+      "learning_rate": 6.5492476060191514e-06,
+      "loss": 1.841,
+      "step": 10164
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6405954360961914,
+      "learning_rate": 6.532147742818058e-06,
+      "loss": 1.8166,
+      "step": 10168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.822506070137024,
+      "learning_rate": 6.515047879616963e-06,
+      "loss": 1.8154,
+      "step": 10172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6633646488189697,
+      "learning_rate": 6.497948016415869e-06,
+      "loss": 1.8579,
+      "step": 10176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8290987014770508,
+      "learning_rate": 6.480848153214775e-06,
+      "loss": 1.8602,
+      "step": 10180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6676056385040283,
+      "learning_rate": 6.46374829001368e-06,
+      "loss": 1.7908,
+      "step": 10184
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4981609582901,
+      "learning_rate": 6.446648426812586e-06,
+      "loss": 1.7605,
+      "step": 10188
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7770746946334839,
+      "learning_rate": 6.429548563611491e-06,
+      "loss": 2.005,
+      "step": 10192
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6636948585510254,
+      "learning_rate": 6.412448700410397e-06,
+      "loss": 1.9328,
+      "step": 10196
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5431137084960938,
+      "learning_rate": 6.395348837209303e-06,
+      "loss": 1.7199,
+      "step": 10200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6235666275024414,
+      "learning_rate": 6.378248974008208e-06,
+      "loss": 1.8701,
+      "step": 10204
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7597670555114746,
+      "learning_rate": 6.3611491108071144e-06,
+      "loss": 1.7161,
+      "step": 10208
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8666725158691406,
+      "learning_rate": 6.344049247606019e-06,
+      "loss": 1.8283,
+      "step": 10212
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8828492164611816,
+      "learning_rate": 6.326949384404925e-06,
+      "loss": 1.9689,
+      "step": 10216
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.539336919784546,
+      "learning_rate": 6.309849521203831e-06,
+      "loss": 1.731,
+      "step": 10220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6746304035186768,
+      "learning_rate": 6.292749658002736e-06,
+      "loss": 1.7627,
+      "step": 10224
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6880711317062378,
+      "learning_rate": 6.275649794801641e-06,
+      "loss": 1.8081,
+      "step": 10228
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6686969995498657,
+      "learning_rate": 6.2585499316005475e-06,
+      "loss": 1.7955,
+      "step": 10232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6514986753463745,
+      "learning_rate": 6.241450068399453e-06,
+      "loss": 1.7788,
+      "step": 10236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5582215785980225,
+      "learning_rate": 6.224350205198359e-06,
+      "loss": 1.7514,
+      "step": 10240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6330581903457642,
+      "learning_rate": 6.2072503419972644e-06,
+      "loss": 1.9055,
+      "step": 10244
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6368670463562012,
+      "learning_rate": 6.19015047879617e-06,
+      "loss": 1.732,
+      "step": 10248
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7110068798065186,
+      "learning_rate": 6.173050615595075e-06,
+      "loss": 1.9471,
+      "step": 10252
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7540807723999023,
+      "learning_rate": 6.155950752393981e-06,
+      "loss": 1.8072,
+      "step": 10256
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6526390314102173,
+      "learning_rate": 6.138850889192887e-06,
+      "loss": 1.7913,
+      "step": 10260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7240028381347656,
+      "learning_rate": 6.121751025991793e-06,
+      "loss": 1.8526,
+      "step": 10264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.683828592300415,
+      "learning_rate": 6.1046511627906975e-06,
+      "loss": 1.7572,
+      "step": 10268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5554591417312622,
+      "learning_rate": 6.087551299589603e-06,
+      "loss": 1.7571,
+      "step": 10272
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6821385622024536,
+      "learning_rate": 6.07045143638851e-06,
+      "loss": 1.7848,
+      "step": 10276
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6544996500015259,
+      "learning_rate": 6.053351573187415e-06,
+      "loss": 1.8252,
+      "step": 10280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6138795614242554,
+      "learning_rate": 6.03625170998632e-06,
+      "loss": 1.979,
+      "step": 10284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.004995346069336,
+      "learning_rate": 6.019151846785226e-06,
+      "loss": 1.7996,
+      "step": 10288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8927818536758423,
+      "learning_rate": 6.002051983584131e-06,
+      "loss": 1.8892,
+      "step": 10292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9253323078155518,
+      "learning_rate": 5.984952120383038e-06,
+      "loss": 1.9798,
+      "step": 10296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.623855471611023,
+      "learning_rate": 5.967852257181943e-06,
+      "loss": 1.6606,
+      "step": 10300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5401018857955933,
+      "learning_rate": 5.950752393980848e-06,
+      "loss": 1.7557,
+      "step": 10304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.643587350845337,
+      "learning_rate": 5.933652530779754e-06,
+      "loss": 1.9448,
+      "step": 10308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7403227090835571,
+      "learning_rate": 5.91655266757866e-06,
+      "loss": 1.9609,
+      "step": 10312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7716693878173828,
+      "learning_rate": 5.899452804377565e-06,
+      "loss": 1.8861,
+      "step": 10316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7579643726348877,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 1.8461,
+      "step": 10320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.512486219406128,
+      "learning_rate": 5.8652530779753766e-06,
+      "loss": 1.8184,
+      "step": 10324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8807471990585327,
+      "learning_rate": 5.848153214774282e-06,
+      "loss": 1.7981,
+      "step": 10328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8313617706298828,
+      "learning_rate": 5.831053351573187e-06,
+      "loss": 1.7566,
+      "step": 10332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7667449712753296,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 1.8996,
+      "step": 10336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5892364978790283,
+      "learning_rate": 5.796853625170999e-06,
+      "loss": 1.7042,
+      "step": 10340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9081833362579346,
+      "learning_rate": 5.779753761969905e-06,
+      "loss": 2.0525,
+      "step": 10344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.891441822052002,
+      "learning_rate": 5.76265389876881e-06,
+      "loss": 1.7879,
+      "step": 10348
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9260497093200684,
+      "learning_rate": 5.745554035567715e-06,
+      "loss": 1.8796,
+      "step": 10352
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6760286092758179,
+      "learning_rate": 5.728454172366622e-06,
+      "loss": 1.9087,
+      "step": 10356
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6360900402069092,
+      "learning_rate": 5.711354309165527e-06,
+      "loss": 1.7715,
+      "step": 10360
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7029601335525513,
+      "learning_rate": 5.694254445964432e-06,
+      "loss": 1.7859,
+      "step": 10364
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5900228023529053,
+      "learning_rate": 5.677154582763338e-06,
+      "loss": 1.8456,
+      "step": 10368
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6234673261642456,
+      "learning_rate": 5.6600547195622435e-06,
+      "loss": 1.8007,
+      "step": 10372
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5690399408340454,
+      "learning_rate": 5.64295485636115e-06,
+      "loss": 1.8154,
+      "step": 10376
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.622010588645935,
+      "learning_rate": 5.625854993160055e-06,
+      "loss": 1.9555,
+      "step": 10380
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7594480514526367,
+      "learning_rate": 5.6087551299589605e-06,
+      "loss": 1.9146,
+      "step": 10384
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.644951581954956,
+      "learning_rate": 5.591655266757866e-06,
+      "loss": 1.9562,
+      "step": 10388
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6730928421020508,
+      "learning_rate": 5.574555403556772e-06,
+      "loss": 1.927,
+      "step": 10392
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.749422311782837,
+      "learning_rate": 5.557455540355677e-06,
+      "loss": 1.8613,
+      "step": 10396
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6241000890731812,
+      "learning_rate": 5.540355677154583e-06,
+      "loss": 1.6963,
+      "step": 10400
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6478880643844604,
+      "learning_rate": 5.523255813953489e-06,
+      "loss": 1.7823,
+      "step": 10404
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8106017112731934,
+      "learning_rate": 5.506155950752394e-06,
+      "loss": 1.8875,
+      "step": 10408
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.792172908782959,
+      "learning_rate": 5.4890560875513e-06,
+      "loss": 1.9241,
+      "step": 10412
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5553025007247925,
+      "learning_rate": 5.471956224350206e-06,
+      "loss": 1.7159,
+      "step": 10416
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5697897672653198,
+      "learning_rate": 5.454856361149111e-06,
+      "loss": 1.7839,
+      "step": 10420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7735848426818848,
+      "learning_rate": 5.437756497948017e-06,
+      "loss": 1.7892,
+      "step": 10424
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8028829097747803,
+      "learning_rate": 5.420656634746922e-06,
+      "loss": 1.8241,
+      "step": 10428
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7222471237182617,
+      "learning_rate": 5.403556771545827e-06,
+      "loss": 1.7322,
+      "step": 10432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5926076173782349,
+      "learning_rate": 5.386456908344734e-06,
+      "loss": 1.8576,
+      "step": 10436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.642992377281189,
+      "learning_rate": 5.3693570451436396e-06,
+      "loss": 1.83,
+      "step": 10440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.577621579170227,
+      "learning_rate": 5.352257181942544e-06,
+      "loss": 1.7914,
+      "step": 10444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5518428087234497,
+      "learning_rate": 5.33515731874145e-06,
+      "loss": 1.9518,
+      "step": 10448
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6200405359268188,
+      "learning_rate": 5.318057455540356e-06,
+      "loss": 1.7868,
+      "step": 10452
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5001662969589233,
+      "learning_rate": 5.300957592339262e-06,
+      "loss": 1.8093,
+      "step": 10456
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6836313009262085,
+      "learning_rate": 5.283857729138167e-06,
+      "loss": 1.8562,
+      "step": 10460
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7168313264846802,
+      "learning_rate": 5.266757865937073e-06,
+      "loss": 1.8735,
+      "step": 10464
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7337522506713867,
+      "learning_rate": 5.249658002735978e-06,
+      "loss": 1.9054,
+      "step": 10468
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4755985736846924,
+      "learning_rate": 5.232558139534884e-06,
+      "loss": 1.8456,
+      "step": 10472
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7717063426971436,
+      "learning_rate": 5.2154582763337896e-06,
+      "loss": 1.8035,
+      "step": 10476
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8099918365478516,
+      "learning_rate": 5.198358413132695e-06,
+      "loss": 1.9099,
+      "step": 10480
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6142700910568237,
+      "learning_rate": 5.181258549931601e-06,
+      "loss": 1.7275,
+      "step": 10484
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6501797437667847,
+      "learning_rate": 5.1641586867305065e-06,
+      "loss": 1.7888,
+      "step": 10488
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6632269620895386,
+      "learning_rate": 5.147058823529412e-06,
+      "loss": 1.77,
+      "step": 10492
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.698384165763855,
+      "learning_rate": 5.129958960328318e-06,
+      "loss": 1.7493,
+      "step": 10496
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6063936948776245,
+      "learning_rate": 5.1128590971272235e-06,
+      "loss": 1.7966,
+      "step": 10500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9107542037963867,
+      "learning_rate": 5.095759233926129e-06,
+      "loss": 1.7634,
+      "step": 10504
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.839381217956543,
+      "learning_rate": 5.078659370725034e-06,
+      "loss": 1.7991,
+      "step": 10508
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7109549045562744,
+      "learning_rate": 5.0615595075239396e-06,
+      "loss": 1.7911,
+      "step": 10512
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7641953229904175,
+      "learning_rate": 5.044459644322846e-06,
+      "loss": 1.9,
+      "step": 10516
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.981584072113037,
+      "learning_rate": 5.027359781121752e-06,
+      "loss": 1.8533,
+      "step": 10520
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5387113094329834,
+      "learning_rate": 5.0102599179206565e-06,
+      "loss": 1.9877,
+      "step": 10524
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.630399465560913,
+      "learning_rate": 4.993160054719562e-06,
+      "loss": 1.9056,
+      "step": 10528
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8931196928024292,
+      "learning_rate": 4.976060191518468e-06,
+      "loss": 1.7797,
+      "step": 10532
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5811591148376465,
+      "learning_rate": 4.958960328317374e-06,
+      "loss": 2.0371,
+      "step": 10536
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7181917428970337,
+      "learning_rate": 4.941860465116279e-06,
+      "loss": 1.7816,
+      "step": 10540
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6405460834503174,
+      "learning_rate": 4.924760601915185e-06,
+      "loss": 1.7458,
+      "step": 10544
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.758260726928711,
+      "learning_rate": 4.90766073871409e-06,
+      "loss": 1.948,
+      "step": 10548
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.845942497253418,
+      "learning_rate": 4.890560875512996e-06,
+      "loss": 1.7851,
+      "step": 10552
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.79182767868042,
+      "learning_rate": 4.873461012311902e-06,
+      "loss": 1.8356,
+      "step": 10556
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8030176162719727,
+      "learning_rate": 4.856361149110807e-06,
+      "loss": 1.9056,
+      "step": 10560
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7074739933013916,
+      "learning_rate": 4.839261285909713e-06,
+      "loss": 1.8684,
+      "step": 10564
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5636216402053833,
+      "learning_rate": 4.822161422708619e-06,
+      "loss": 1.7258,
+      "step": 10568
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8994554281234741,
+      "learning_rate": 4.805061559507524e-06,
+      "loss": 1.8462,
+      "step": 10572
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5819472074508667,
+      "learning_rate": 4.78796169630643e-06,
+      "loss": 1.7744,
+      "step": 10576
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7214453220367432,
+      "learning_rate": 4.770861833105336e-06,
+      "loss": 1.8286,
+      "step": 10580
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6858314275741577,
+      "learning_rate": 4.753761969904241e-06,
+      "loss": 1.9328,
+      "step": 10584
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6816788911819458,
+      "learning_rate": 4.736662106703146e-06,
+      "loss": 1.9712,
+      "step": 10588
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5183528661727905,
+      "learning_rate": 4.7195622435020526e-06,
+      "loss": 1.8719,
+      "step": 10592
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7054622173309326,
+      "learning_rate": 4.702462380300958e-06,
+      "loss": 1.8244,
+      "step": 10596
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5704643726348877,
+      "learning_rate": 4.685362517099864e-06,
+      "loss": 1.7187,
+      "step": 10600
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.616736888885498,
+      "learning_rate": 4.668262653898769e-06,
+      "loss": 1.9231,
+      "step": 10604
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8305636644363403,
+      "learning_rate": 4.651162790697674e-06,
+      "loss": 2.0724,
+      "step": 10608
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.623712420463562,
+      "learning_rate": 4.63406292749658e-06,
+      "loss": 1.8901,
+      "step": 10612
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7231340408325195,
+      "learning_rate": 4.6169630642954865e-06,
+      "loss": 1.7852,
+      "step": 10616
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.707742691040039,
+      "learning_rate": 4.599863201094391e-06,
+      "loss": 1.861,
+      "step": 10620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6897032260894775,
+      "learning_rate": 4.582763337893297e-06,
+      "loss": 1.7987,
+      "step": 10624
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4974230527877808,
+      "learning_rate": 4.5656634746922026e-06,
+      "loss": 1.7377,
+      "step": 10628
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7821507453918457,
+      "learning_rate": 4.548563611491108e-06,
+      "loss": 1.7803,
+      "step": 10632
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7004159688949585,
+      "learning_rate": 4.531463748290014e-06,
+      "loss": 1.7409,
+      "step": 10636
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.620969295501709,
+      "learning_rate": 4.5143638850889195e-06,
+      "loss": 1.6708,
+      "step": 10640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6645839214324951,
+      "learning_rate": 4.497264021887825e-06,
+      "loss": 1.8713,
+      "step": 10644
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6770342588424683,
+      "learning_rate": 4.480164158686731e-06,
+      "loss": 1.9282,
+      "step": 10648
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7173173427581787,
+      "learning_rate": 4.4630642954856365e-06,
+      "loss": 1.8702,
+      "step": 10652
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7240304946899414,
+      "learning_rate": 4.445964432284542e-06,
+      "loss": 1.8032,
+      "step": 10656
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.618517518043518,
+      "learning_rate": 4.428864569083448e-06,
+      "loss": 1.846,
+      "step": 10660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6240915060043335,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 1.8224,
+      "step": 10664
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.519727110862732,
+      "learning_rate": 4.394664842681258e-06,
+      "loss": 1.7819,
+      "step": 10668
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0412402153015137,
+      "learning_rate": 4.377564979480165e-06,
+      "loss": 1.8308,
+      "step": 10672
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5424792766571045,
+      "learning_rate": 4.36046511627907e-06,
+      "loss": 1.8432,
+      "step": 10676
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9428133964538574,
+      "learning_rate": 4.343365253077976e-06,
+      "loss": 1.6436,
+      "step": 10680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6989853382110596,
+      "learning_rate": 4.326265389876881e-06,
+      "loss": 1.7265,
+      "step": 10684
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5763792991638184,
+      "learning_rate": 4.3091655266757865e-06,
+      "loss": 1.8635,
+      "step": 10688
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7506353855133057,
+      "learning_rate": 4.292065663474692e-06,
+      "loss": 1.83,
+      "step": 10692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6734927892684937,
+      "learning_rate": 4.274965800273599e-06,
+      "loss": 1.8733,
+      "step": 10696
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7979787588119507,
+      "learning_rate": 4.257865937072503e-06,
+      "loss": 1.8805,
+      "step": 10700
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.643833875656128,
+      "learning_rate": 4.240766073871409e-06,
+      "loss": 1.7028,
+      "step": 10704
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.727108120918274,
+      "learning_rate": 4.223666210670315e-06,
+      "loss": 1.7918,
+      "step": 10708
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6539056301116943,
+      "learning_rate": 4.20656634746922e-06,
+      "loss": 1.8238,
+      "step": 10712
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7414677143096924,
+      "learning_rate": 4.189466484268126e-06,
+      "loss": 1.8796,
+      "step": 10716
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6490898132324219,
+      "learning_rate": 4.172366621067032e-06,
+      "loss": 1.7907,
+      "step": 10720
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7541985511779785,
+      "learning_rate": 4.155266757865937e-06,
+      "loss": 1.8669,
+      "step": 10724
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6836141347885132,
+      "learning_rate": 4.138166894664843e-06,
+      "loss": 1.7802,
+      "step": 10728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.008136034011841,
+      "learning_rate": 4.121067031463749e-06,
+      "loss": 1.9157,
+      "step": 10732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7152680158615112,
+      "learning_rate": 4.103967168262654e-06,
+      "loss": 1.8344,
+      "step": 10736
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5120372772216797,
+      "learning_rate": 4.08686730506156e-06,
+      "loss": 1.7443,
+      "step": 10740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5764340162277222,
+      "learning_rate": 4.0697674418604655e-06,
+      "loss": 1.7508,
+      "step": 10744
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6534470319747925,
+      "learning_rate": 4.05266757865937e-06,
+      "loss": 1.8064,
+      "step": 10748
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8456621170043945,
+      "learning_rate": 4.035567715458277e-06,
+      "loss": 1.8589,
+      "step": 10752
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6170421838760376,
+      "learning_rate": 4.0184678522571825e-06,
+      "loss": 1.6253,
+      "step": 10756
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5028194189071655,
+      "learning_rate": 4.001367989056088e-06,
+      "loss": 1.7209,
+      "step": 10760
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9968106746673584,
+      "learning_rate": 3.984268125854993e-06,
+      "loss": 1.9172,
+      "step": 10764
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7104765176773071,
+      "learning_rate": 3.967168262653899e-06,
+      "loss": 1.7745,
+      "step": 10768
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.739295244216919,
+      "learning_rate": 3.950068399452805e-06,
+      "loss": 1.8637,
+      "step": 10772
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.587810754776001,
+      "learning_rate": 3.932968536251711e-06,
+      "loss": 1.6962,
+      "step": 10776
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6938265562057495,
+      "learning_rate": 3.9158686730506155e-06,
+      "loss": 1.8125,
+      "step": 10780
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7163448333740234,
+      "learning_rate": 3.898768809849521e-06,
+      "loss": 1.8493,
+      "step": 10784
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6594278812408447,
+      "learning_rate": 3.881668946648427e-06,
+      "loss": 1.8301,
+      "step": 10788
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.661649227142334,
+      "learning_rate": 3.8645690834473325e-06,
+      "loss": 1.774,
+      "step": 10792
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8990195989608765,
+      "learning_rate": 3.847469220246238e-06,
+      "loss": 1.8284,
+      "step": 10796
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8179618120193481,
+      "learning_rate": 3.830369357045144e-06,
+      "loss": 1.7436,
+      "step": 10800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0388402938842773,
+      "learning_rate": 3.8132694938440494e-06,
+      "loss": 1.9214,
+      "step": 10804
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.578235387802124,
+      "learning_rate": 3.7961696306429555e-06,
+      "loss": 1.8137,
+      "step": 10808
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9262685775756836,
+      "learning_rate": 3.7790697674418603e-06,
+      "loss": 1.8067,
+      "step": 10812
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7524725198745728,
+      "learning_rate": 3.7619699042407664e-06,
+      "loss": 2.018,
+      "step": 10816
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7184381484985352,
+      "learning_rate": 3.744870041039672e-06,
+      "loss": 1.8733,
+      "step": 10820
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.626380443572998,
+      "learning_rate": 3.7277701778385777e-06,
+      "loss": 1.8379,
+      "step": 10824
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.610595941543579,
+      "learning_rate": 3.710670314637483e-06,
+      "loss": 1.8246,
+      "step": 10828
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6897306442260742,
+      "learning_rate": 3.6935704514363886e-06,
+      "loss": 1.7103,
+      "step": 10832
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8274520635604858,
+      "learning_rate": 3.6764705882352942e-06,
+      "loss": 1.9273,
+      "step": 10836
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8843746185302734,
+      "learning_rate": 3.6593707250342003e-06,
+      "loss": 1.792,
+      "step": 10840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.664485216140747,
+      "learning_rate": 3.642270861833105e-06,
+      "loss": 2.0409,
+      "step": 10844
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5859229564666748,
+      "learning_rate": 3.625170998632011e-06,
+      "loss": 1.9006,
+      "step": 10848
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6038724184036255,
+      "learning_rate": 3.608071135430917e-06,
+      "loss": 1.8245,
+      "step": 10852
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6321396827697754,
+      "learning_rate": 3.5909712722298225e-06,
+      "loss": 1.8706,
+      "step": 10856
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7967422008514404,
+      "learning_rate": 3.5738714090287277e-06,
+      "loss": 1.86,
+      "step": 10860
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5631977319717407,
+      "learning_rate": 3.5567715458276333e-06,
+      "loss": 1.7408,
+      "step": 10864
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5402038097381592,
+      "learning_rate": 3.5396716826265394e-06,
+      "loss": 1.8824,
+      "step": 10868
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5307432413101196,
+      "learning_rate": 3.522571819425445e-06,
+      "loss": 1.7778,
+      "step": 10872
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6216157674789429,
+      "learning_rate": 3.5054719562243503e-06,
+      "loss": 1.6656,
+      "step": 10876
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6624993085861206,
+      "learning_rate": 3.488372093023256e-06,
+      "loss": 1.9027,
+      "step": 10880
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6463823318481445,
+      "learning_rate": 3.4712722298221616e-06,
+      "loss": 1.8464,
+      "step": 10884
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.56813645362854,
+      "learning_rate": 3.4541723666210677e-06,
+      "loss": 1.8271,
+      "step": 10888
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8501081466674805,
+      "learning_rate": 3.4370725034199725e-06,
+      "loss": 1.9077,
+      "step": 10892
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9633204936981201,
+      "learning_rate": 3.4199726402188785e-06,
+      "loss": 1.8458,
+      "step": 10896
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8002575635910034,
+      "learning_rate": 3.402872777017784e-06,
+      "loss": 2.0061,
+      "step": 10900
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6839830875396729,
+      "learning_rate": 3.38577291381669e-06,
+      "loss": 1.8378,
+      "step": 10904
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6927244663238525,
+      "learning_rate": 3.368673050615595e-06,
+      "loss": 1.7045,
+      "step": 10908
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5076004266738892,
+      "learning_rate": 3.3515731874145007e-06,
+      "loss": 1.9581,
+      "step": 10912
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7239080667495728,
+      "learning_rate": 3.334473324213407e-06,
+      "loss": 1.9258,
+      "step": 10916
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8428497314453125,
+      "learning_rate": 3.3173734610123124e-06,
+      "loss": 1.8136,
+      "step": 10920
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6915169954299927,
+      "learning_rate": 3.3002735978112172e-06,
+      "loss": 1.9492,
+      "step": 10924
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.641465425491333,
+      "learning_rate": 3.2831737346101233e-06,
+      "loss": 1.7496,
+      "step": 10928
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6878330707550049,
+      "learning_rate": 3.266073871409029e-06,
+      "loss": 1.85,
+      "step": 10932
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8498550653457642,
+      "learning_rate": 3.2489740082079346e-06,
+      "loss": 1.9107,
+      "step": 10936
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.61210036277771,
+      "learning_rate": 3.23187414500684e-06,
+      "loss": 1.9388,
+      "step": 10940
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.551893949508667,
+      "learning_rate": 3.2147742818057455e-06,
+      "loss": 1.8542,
+      "step": 10944
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.587180733680725,
+      "learning_rate": 3.1976744186046516e-06,
+      "loss": 1.8516,
+      "step": 10948
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.42110276222229,
+      "learning_rate": 3.1805745554035572e-06,
+      "loss": 1.7036,
+      "step": 10952
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7156518697738647,
+      "learning_rate": 3.1634746922024624e-06,
+      "loss": 1.8442,
+      "step": 10956
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6081629991531372,
+      "learning_rate": 3.146374829001368e-06,
+      "loss": 1.9073,
+      "step": 10960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0794620513916016,
+      "learning_rate": 3.1292749658002737e-06,
+      "loss": 2.0049,
+      "step": 10964
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6746197938919067,
+      "learning_rate": 3.1121751025991794e-06,
+      "loss": 1.7059,
+      "step": 10968
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5624709129333496,
+      "learning_rate": 3.095075239398085e-06,
+      "loss": 1.8929,
+      "step": 10972
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6810204982757568,
+      "learning_rate": 3.0779753761969907e-06,
+      "loss": 1.8587,
+      "step": 10976
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.586511492729187,
+      "learning_rate": 3.0608755129958963e-06,
+      "loss": 1.7846,
+      "step": 10980
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6255276203155518,
+      "learning_rate": 3.0437756497948016e-06,
+      "loss": 1.8619,
+      "step": 10984
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8912711143493652,
+      "learning_rate": 3.0266757865937076e-06,
+      "loss": 1.9116,
+      "step": 10988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4620989561080933,
+      "learning_rate": 3.009575923392613e-06,
+      "loss": 1.9151,
+      "step": 10992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5181527137756348,
+      "learning_rate": 2.992476060191519e-06,
+      "loss": 1.6923,
+      "step": 10996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7841485738754272,
+      "learning_rate": 2.975376196990424e-06,
+      "loss": 1.7321,
+      "step": 11000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8255186080932617,
+      "learning_rate": 2.95827633378933e-06,
+      "loss": 1.8564,
+      "step": 11004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6331710815429688,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 1.9261,
+      "step": 11008
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5975050926208496,
+      "learning_rate": 2.924076607387141e-06,
+      "loss": 1.7562,
+      "step": 11012
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5330755710601807,
+      "learning_rate": 2.9069767441860468e-06,
+      "loss": 1.7575,
+      "step": 11016
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7199543714523315,
+      "learning_rate": 2.8898768809849524e-06,
+      "loss": 1.8875,
+      "step": 11020
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7265928983688354,
+      "learning_rate": 2.8727770177838576e-06,
+      "loss": 1.9619,
+      "step": 11024
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.955166220664978,
+      "learning_rate": 2.8556771545827637e-06,
+      "loss": 1.9266,
+      "step": 11028
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.708550214767456,
+      "learning_rate": 2.838577291381669e-06,
+      "loss": 1.8194,
+      "step": 11032
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7049528360366821,
+      "learning_rate": 2.821477428180575e-06,
+      "loss": 1.728,
+      "step": 11036
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8736356496810913,
+      "learning_rate": 2.8043775649794802e-06,
+      "loss": 1.7949,
+      "step": 11040
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8903663158416748,
+      "learning_rate": 2.787277701778386e-06,
+      "loss": 1.7765,
+      "step": 11044
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.690417766571045,
+      "learning_rate": 2.7701778385772915e-06,
+      "loss": 1.8932,
+      "step": 11048
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7920684814453125,
+      "learning_rate": 2.753077975376197e-06,
+      "loss": 1.7986,
+      "step": 11052
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.771796703338623,
+      "learning_rate": 2.735978112175103e-06,
+      "loss": 1.7103,
+      "step": 11056
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.51233971118927,
+      "learning_rate": 2.7188782489740085e-06,
+      "loss": 1.6045,
+      "step": 11060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6907927989959717,
+      "learning_rate": 2.7017783857729137e-06,
+      "loss": 1.9416,
+      "step": 11064
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.575386881828308,
+      "learning_rate": 2.6846785225718198e-06,
+      "loss": 1.7111,
+      "step": 11068
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0558063983917236,
+      "learning_rate": 2.667578659370725e-06,
+      "loss": 1.7934,
+      "step": 11072
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.705366611480713,
+      "learning_rate": 2.650478796169631e-06,
+      "loss": 1.8026,
+      "step": 11076
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.625901222229004,
+      "learning_rate": 2.6333789329685363e-06,
+      "loss": 1.9377,
+      "step": 11080
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7622089385986328,
+      "learning_rate": 2.616279069767442e-06,
+      "loss": 1.7597,
+      "step": 11084
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.841213583946228,
+      "learning_rate": 2.5991792065663476e-06,
+      "loss": 1.901,
+      "step": 11088
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6358860731124878,
+      "learning_rate": 2.5820793433652533e-06,
+      "loss": 1.7777,
+      "step": 11092
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6823545694351196,
+      "learning_rate": 2.564979480164159e-06,
+      "loss": 1.9779,
+      "step": 11096
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8069067001342773,
+      "learning_rate": 2.5478796169630646e-06,
+      "loss": 1.9257,
+      "step": 11100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4878418445587158,
+      "learning_rate": 2.5307797537619698e-06,
+      "loss": 1.8447,
+      "step": 11104
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6673344373703003,
+      "learning_rate": 2.513679890560876e-06,
+      "loss": 1.8752,
+      "step": 11108
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6673535108566284,
+      "learning_rate": 2.496580027359781e-06,
+      "loss": 1.8024,
+      "step": 11112
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7625112533569336,
+      "learning_rate": 2.479480164158687e-06,
+      "loss": 1.7451,
+      "step": 11116
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7562973499298096,
+      "learning_rate": 2.4623803009575924e-06,
+      "loss": 1.8505,
+      "step": 11120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.55601966381073,
+      "learning_rate": 2.445280437756498e-06,
+      "loss": 1.7542,
+      "step": 11124
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5536571741104126,
+      "learning_rate": 2.4281805745554037e-06,
+      "loss": 1.7571,
+      "step": 11128
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.797058343887329,
+      "learning_rate": 2.4110807113543093e-06,
+      "loss": 1.8041,
+      "step": 11132
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5327272415161133,
+      "learning_rate": 2.393980848153215e-06,
+      "loss": 1.9662,
+      "step": 11136
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6519979238510132,
+      "learning_rate": 2.3768809849521206e-06,
+      "loss": 1.8325,
+      "step": 11140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6532912254333496,
+      "learning_rate": 2.3597811217510263e-06,
+      "loss": 1.7309,
+      "step": 11144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6513277292251587,
+      "learning_rate": 2.342681258549932e-06,
+      "loss": 1.8967,
+      "step": 11148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.01287579536438,
+      "learning_rate": 2.325581395348837e-06,
+      "loss": 1.8535,
+      "step": 11152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7155896425247192,
+      "learning_rate": 2.3084815321477432e-06,
+      "loss": 1.9063,
+      "step": 11156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.802552342414856,
+      "learning_rate": 2.2913816689466485e-06,
+      "loss": 1.7353,
+      "step": 11160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.850523591041565,
+      "learning_rate": 2.274281805745554e-06,
+      "loss": 2.0235,
+      "step": 11164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6676249504089355,
+      "learning_rate": 2.2571819425444598e-06,
+      "loss": 1.8164,
+      "step": 11168
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.562305212020874,
+      "learning_rate": 2.2400820793433654e-06,
+      "loss": 1.8171,
+      "step": 11172
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9502869844436646,
+      "learning_rate": 2.222982216142271e-06,
+      "loss": 1.8298,
+      "step": 11176
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.739216923713684,
+      "learning_rate": 2.2058823529411767e-06,
+      "loss": 1.7871,
+      "step": 11180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.595462441444397,
+      "learning_rate": 2.1887824897400824e-06,
+      "loss": 1.7876,
+      "step": 11184
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.28205943107605,
+      "learning_rate": 2.171682626538988e-06,
+      "loss": 1.9108,
+      "step": 11188
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8560374975204468,
+      "learning_rate": 2.1545827633378932e-06,
+      "loss": 1.7791,
+      "step": 11192
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6630357503890991,
+      "learning_rate": 2.1374829001367993e-06,
+      "loss": 1.8153,
+      "step": 11196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6874091625213623,
+      "learning_rate": 2.1203830369357045e-06,
+      "loss": 1.813,
+      "step": 11200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5925251245498657,
+      "learning_rate": 2.10328317373461e-06,
+      "loss": 1.7251,
+      "step": 11204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7215338945388794,
+      "learning_rate": 2.086183310533516e-06,
+      "loss": 1.8046,
+      "step": 11208
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6895273923873901,
+      "learning_rate": 2.0690834473324215e-06,
+      "loss": 1.8298,
+      "step": 11212
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.641157865524292,
+      "learning_rate": 2.051983584131327e-06,
+      "loss": 1.7497,
+      "step": 11216
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.756468653678894,
+      "learning_rate": 2.0348837209302328e-06,
+      "loss": 1.791,
+      "step": 11220
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7440533638000488,
+      "learning_rate": 2.0177838577291384e-06,
+      "loss": 1.7596,
+      "step": 11224
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4534015655517578,
+      "learning_rate": 2.000683994528044e-06,
+      "loss": 1.7385,
+      "step": 11228
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6513500213623047,
+      "learning_rate": 1.9835841313269493e-06,
+      "loss": 2.1346,
+      "step": 11232
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7908762693405151,
+      "learning_rate": 1.9664842681258554e-06,
+      "loss": 1.7386,
+      "step": 11236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5934255123138428,
+      "learning_rate": 1.9493844049247606e-06,
+      "loss": 1.8361,
+      "step": 11240
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4890646934509277,
+      "learning_rate": 1.9322845417236662e-06,
+      "loss": 1.7226,
+      "step": 11244
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6311677694320679,
+      "learning_rate": 1.915184678522572e-06,
+      "loss": 1.7406,
+      "step": 11248
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.662655234336853,
+      "learning_rate": 1.8980848153214778e-06,
+      "loss": 1.894,
+      "step": 11252
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6911251544952393,
+      "learning_rate": 1.8809849521203832e-06,
+      "loss": 2.0025,
+      "step": 11256
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5574413537979126,
+      "learning_rate": 1.8638850889192888e-06,
+      "loss": 1.7462,
+      "step": 11260
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.699474573135376,
+      "learning_rate": 1.8467852257181943e-06,
+      "loss": 1.7242,
+      "step": 11264
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.642033338546753,
+      "learning_rate": 1.8296853625171001e-06,
+      "loss": 1.7673,
+      "step": 11268
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6996281147003174,
+      "learning_rate": 1.8125854993160056e-06,
+      "loss": 1.8045,
+      "step": 11272
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6581562757492065,
+      "learning_rate": 1.7954856361149112e-06,
+      "loss": 1.9656,
+      "step": 11276
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6569535732269287,
+      "learning_rate": 1.7783857729138167e-06,
+      "loss": 1.9959,
+      "step": 11280
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7449274063110352,
+      "learning_rate": 1.7612859097127225e-06,
+      "loss": 1.8343,
+      "step": 11284
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6935420036315918,
+      "learning_rate": 1.744186046511628e-06,
+      "loss": 1.8053,
+      "step": 11288
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5281199216842651,
+      "learning_rate": 1.7270861833105338e-06,
+      "loss": 1.9919,
+      "step": 11292
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.633588433265686,
+      "learning_rate": 1.7099863201094393e-06,
+      "loss": 1.8699,
+      "step": 11296
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7310950756072998,
+      "learning_rate": 1.692886456908345e-06,
+      "loss": 1.9074,
+      "step": 11300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5915794372558594,
+      "learning_rate": 1.6757865937072504e-06,
+      "loss": 1.7659,
+      "step": 11304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7064154148101807,
+      "learning_rate": 1.6586867305061562e-06,
+      "loss": 1.8796,
+      "step": 11308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7068238258361816,
+      "learning_rate": 1.6415868673050617e-06,
+      "loss": 1.824,
+      "step": 11312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.686372995376587,
+      "learning_rate": 1.6244870041039673e-06,
+      "loss": 2.0173,
+      "step": 11316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5897817611694336,
+      "learning_rate": 1.6073871409028727e-06,
+      "loss": 1.8182,
+      "step": 11320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6900789737701416,
+      "learning_rate": 1.5902872777017786e-06,
+      "loss": 1.7615,
+      "step": 11324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7033038139343262,
+      "learning_rate": 1.573187414500684e-06,
+      "loss": 1.7663,
+      "step": 11328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.661864161491394,
+      "learning_rate": 1.5560875512995897e-06,
+      "loss": 1.8007,
+      "step": 11332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.663250207901001,
+      "learning_rate": 1.5389876880984953e-06,
+      "loss": 1.7852,
+      "step": 11336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7646623849868774,
+      "learning_rate": 1.5218878248974008e-06,
+      "loss": 1.7738,
+      "step": 11340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6536412239074707,
+      "learning_rate": 1.5047879616963064e-06,
+      "loss": 1.8398,
+      "step": 11344
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.726285457611084,
+      "learning_rate": 1.487688098495212e-06,
+      "loss": 1.7792,
+      "step": 11348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7666223049163818,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 1.9069,
+      "step": 11352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7547391653060913,
+      "learning_rate": 1.4534883720930234e-06,
+      "loss": 1.8827,
+      "step": 11356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8074520826339722,
+      "learning_rate": 1.4363885088919288e-06,
+      "loss": 1.8349,
+      "step": 11360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5819740295410156,
+      "learning_rate": 1.4192886456908345e-06,
+      "loss": 2.0226,
+      "step": 11364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4653937816619873,
+      "learning_rate": 1.4021887824897401e-06,
+      "loss": 1.7796,
+      "step": 11368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7137483358383179,
+      "learning_rate": 1.3850889192886458e-06,
+      "loss": 1.9836,
+      "step": 11372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5901538133621216,
+      "learning_rate": 1.3679890560875514e-06,
+      "loss": 1.7209,
+      "step": 11376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5994963645935059,
+      "learning_rate": 1.3508891928864569e-06,
+      "loss": 1.8554,
+      "step": 11380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.578391194343567,
+      "learning_rate": 1.3337893296853625e-06,
+      "loss": 1.8694,
+      "step": 11384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6012123823165894,
+      "learning_rate": 1.3166894664842682e-06,
+      "loss": 1.7565,
+      "step": 11388
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6539853811264038,
+      "learning_rate": 1.2995896032831738e-06,
+      "loss": 1.8733,
+      "step": 11392
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7771899700164795,
+      "learning_rate": 1.2824897400820795e-06,
+      "loss": 1.8847,
+      "step": 11396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6242918968200684,
+      "learning_rate": 1.2653898768809849e-06,
+      "loss": 1.7995,
+      "step": 11400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7309808731079102,
+      "learning_rate": 1.2482900136798905e-06,
+      "loss": 1.7911,
+      "step": 11404
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.64167320728302,
+      "learning_rate": 1.2311901504787962e-06,
+      "loss": 1.8132,
+      "step": 11408
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7144200801849365,
+      "learning_rate": 1.2140902872777018e-06,
+      "loss": 1.8484,
+      "step": 11412
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.498760461807251,
+      "learning_rate": 1.1969904240766075e-06,
+      "loss": 1.7778,
+      "step": 11416
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.62421452999115,
+      "learning_rate": 1.1798905608755131e-06,
+      "loss": 1.8898,
+      "step": 11420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7838332653045654,
+      "learning_rate": 1.1627906976744186e-06,
+      "loss": 1.8246,
+      "step": 11424
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7368948459625244,
+      "learning_rate": 1.1456908344733242e-06,
+      "loss": 1.9796,
+      "step": 11428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5505352020263672,
+      "learning_rate": 1.1285909712722299e-06,
+      "loss": 1.8111,
+      "step": 11432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7267060279846191,
+      "learning_rate": 1.1114911080711355e-06,
+      "loss": 1.7691,
+      "step": 11436
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6978029012680054,
+      "learning_rate": 1.0943912448700412e-06,
+      "loss": 1.7508,
+      "step": 11440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.665993571281433,
+      "learning_rate": 1.0772913816689466e-06,
+      "loss": 1.7835,
+      "step": 11444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5911554098129272,
+      "learning_rate": 1.0601915184678523e-06,
+      "loss": 1.8659,
+      "step": 11448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.122206449508667,
+      "learning_rate": 1.043091655266758e-06,
+      "loss": 1.8262,
+      "step": 11452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5077828168869019,
+      "learning_rate": 1.0259917920656636e-06,
+      "loss": 1.8085,
+      "step": 11456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.527305245399475,
+      "learning_rate": 1.0088919288645692e-06,
+      "loss": 1.7245,
+      "step": 11460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8870925903320312,
+      "learning_rate": 9.917920656634746e-07,
+      "loss": 1.891,
+      "step": 11464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.66902756690979,
+      "learning_rate": 9.746922024623803e-07,
+      "loss": 1.8576,
+      "step": 11468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6984986066818237,
+      "learning_rate": 9.57592339261286e-07,
+      "loss": 1.711,
+      "step": 11472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8500134944915771,
+      "learning_rate": 9.404924760601916e-07,
+      "loss": 1.83,
+      "step": 11476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5974342823028564,
+      "learning_rate": 9.233926128590971e-07,
+      "loss": 1.7812,
+      "step": 11480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.678200364112854,
+      "learning_rate": 9.062927496580028e-07,
+      "loss": 1.8443,
+      "step": 11484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.766434907913208,
+      "learning_rate": 8.891928864569083e-07,
+      "loss": 1.9573,
+      "step": 11488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7245994806289673,
+      "learning_rate": 8.72093023255814e-07,
+      "loss": 1.8823,
+      "step": 11492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.598664402961731,
+      "learning_rate": 8.549931600547196e-07,
+      "loss": 1.78,
+      "step": 11496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5603172779083252,
+      "learning_rate": 8.378932968536252e-07,
+      "loss": 1.7606,
+      "step": 11500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5893508195877075,
+      "learning_rate": 8.207934336525308e-07,
+      "loss": 1.7072,
+      "step": 11504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6806880235671997,
+      "learning_rate": 8.036935704514364e-07,
+      "loss": 1.9592,
+      "step": 11508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6872777938842773,
+      "learning_rate": 7.86593707250342e-07,
+      "loss": 1.8096,
+      "step": 11512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6055060625076294,
+      "learning_rate": 7.694938440492477e-07,
+      "loss": 1.991,
+      "step": 11516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5616644620895386,
+      "learning_rate": 7.523939808481532e-07,
+      "loss": 1.8996,
+      "step": 11520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7734068632125854,
+      "learning_rate": 7.352941176470589e-07,
+      "loss": 1.8998,
+      "step": 11524
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7730764150619507,
+      "learning_rate": 7.181942544459644e-07,
+      "loss": 1.82,
+      "step": 11528
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6176447868347168,
+      "learning_rate": 7.010943912448701e-07,
+      "loss": 1.6604,
+      "step": 11532
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6673433780670166,
+      "learning_rate": 6.839945280437757e-07,
+      "loss": 1.8491,
+      "step": 11536
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7543388605117798,
+      "learning_rate": 6.668946648426813e-07,
+      "loss": 1.9856,
+      "step": 11540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.687723159790039,
+      "learning_rate": 6.497948016415869e-07,
+      "loss": 1.7666,
+      "step": 11544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6949374675750732,
+      "learning_rate": 6.326949384404924e-07,
+      "loss": 1.7367,
+      "step": 11548
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.88869047164917,
+      "learning_rate": 6.155950752393981e-07,
+      "loss": 1.8341,
+      "step": 11552
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.616167426109314,
+      "learning_rate": 5.984952120383037e-07,
+      "loss": 1.8967,
+      "step": 11556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6819902658462524,
+      "learning_rate": 5.813953488372093e-07,
+      "loss": 1.6976,
+      "step": 11560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.628939151763916,
+      "learning_rate": 5.642954856361149e-07,
+      "loss": 1.7678,
+      "step": 11564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9879276752471924,
+      "learning_rate": 5.471956224350206e-07,
+      "loss": 1.8467,
+      "step": 11568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9371817111968994,
+      "learning_rate": 5.300957592339261e-07,
+      "loss": 1.8366,
+      "step": 11572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7080752849578857,
+      "learning_rate": 5.129958960328318e-07,
+      "loss": 1.8151,
+      "step": 11576
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6665830612182617,
+      "learning_rate": 4.958960328317373e-07,
+      "loss": 1.7895,
+      "step": 11580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5489290952682495,
+      "learning_rate": 4.78796169630643e-07,
+      "loss": 1.7158,
+      "step": 11584
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6452659368515015,
+      "learning_rate": 4.6169630642954857e-07,
+      "loss": 1.9806,
+      "step": 11588
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5039621591567993,
+      "learning_rate": 4.4459644322845417e-07,
+      "loss": 1.6445,
+      "step": 11592
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6051033735275269,
+      "learning_rate": 4.274965800273598e-07,
+      "loss": 1.812,
+      "step": 11596
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5924128293991089,
+      "learning_rate": 4.103967168262654e-07,
+      "loss": 1.7894,
+      "step": 11600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.752646565437317,
+      "learning_rate": 3.93296853625171e-07,
+      "loss": 1.7116,
+      "step": 11604
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6699579954147339,
+      "learning_rate": 3.761969904240766e-07,
+      "loss": 1.7171,
+      "step": 11608
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7405366897583008,
+      "learning_rate": 3.590971272229822e-07,
+      "loss": 1.9161,
+      "step": 11612
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5944126844406128,
+      "learning_rate": 3.4199726402188785e-07,
+      "loss": 1.758,
+      "step": 11616
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7876613140106201,
+      "learning_rate": 3.2489740082079345e-07,
+      "loss": 1.8338,
+      "step": 11620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.640367031097412,
+      "learning_rate": 3.0779753761969905e-07,
+      "loss": 1.8632,
+      "step": 11624
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.646633267402649,
+      "learning_rate": 2.9069767441860464e-07,
+      "loss": 1.8923,
+      "step": 11628
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.551578402519226,
+      "learning_rate": 2.735978112175103e-07,
+      "loss": 1.8043,
+      "step": 11632
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0039708614349365,
+      "learning_rate": 2.564979480164159e-07,
+      "loss": 1.8416,
+      "step": 11636
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6551536321640015,
+      "learning_rate": 2.393980848153215e-07,
+      "loss": 1.8004,
+      "step": 11640
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.774258017539978,
+      "learning_rate": 2.2229822161422708e-07,
+      "loss": 1.6694,
+      "step": 11644
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6978957653045654,
+      "learning_rate": 2.051983584131327e-07,
+      "loss": 1.9608,
+      "step": 11648
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5244742631912231,
+      "learning_rate": 1.880984952120383e-07,
+      "loss": 1.749,
+      "step": 11652
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7631537914276123,
+      "learning_rate": 1.7099863201094393e-07,
+      "loss": 1.9072,
+      "step": 11656
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6298117637634277,
+      "learning_rate": 1.5389876880984952e-07,
+      "loss": 1.8008,
+      "step": 11660
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6735055446624756,
+      "learning_rate": 1.3679890560875515e-07,
+      "loss": 1.7698,
+      "step": 11664
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6959282159805298,
+      "learning_rate": 1.1969904240766074e-07,
+      "loss": 1.7243,
+      "step": 11668
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6977593898773193,
+      "learning_rate": 1.0259917920656635e-07,
+      "loss": 1.7902,
+      "step": 11672
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5963069200515747,
+      "learning_rate": 8.549931600547196e-08,
+      "loss": 1.6623,
+      "step": 11676
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5912084579467773,
+      "learning_rate": 6.839945280437757e-08,
+      "loss": 1.8806,
+      "step": 11680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8129706382751465,
+      "learning_rate": 5.129958960328318e-08,
+      "loss": 1.9337,
+      "step": 11684
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7241120338439941,
+      "learning_rate": 3.419972640218879e-08,
+      "loss": 1.9163,
+      "step": 11688
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5432144403457642,
+      "learning_rate": 1.7099863201094393e-08,
+      "loss": 1.6597,
+      "step": 11692
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5885206460952759,
+      "learning_rate": 0.0,
+      "loss": 1.9262,
+      "step": 11696
+    },
+    {
+      "epoch": 1.0,
+      "step": 11696,
+      "total_flos": 1.5848503128568627e+17,
+      "train_loss": 2.3174790558589957,
+      "train_runtime": 2998.5108,
+      "train_samples_per_second": 31.202,
+      "train_steps_per_second": 3.901
+    }
+  ],
+  "logging_steps": 4,
+  "max_steps": 11696,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1170,
+  "total_flos": 1.5848503128568627e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}