diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.2076465855114595,
+  "eval_steps": 500,
+  "global_step": 8000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.76329803466797,
+      "learning_rate": 5e-06,
+      "loss": 13.6429,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.773799896240234,
+      "learning_rate": 1e-05,
+      "loss": 13.2108,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.7246150970459,
+      "learning_rate": 1.5e-05,
+      "loss": 12.1219,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.235396385192871,
+      "learning_rate": 2e-05,
+      "loss": 10.7281,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.372308731079102,
+      "learning_rate": 2.5e-05,
+      "loss": 9.5067,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.502007961273193,
+      "learning_rate": 3e-05,
+      "loss": 8.8494,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3103082180023193,
+      "learning_rate": 3.5e-05,
+      "loss": 8.3923,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.845451831817627,
+      "learning_rate": 4e-05,
+      "loss": 8.1076,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3222850561141968,
+      "learning_rate": 4.5e-05,
+      "loss": 7.9616,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2554008960723877,
+      "learning_rate": 5e-05,
+      "loss": 7.8755,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.062578558921814,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 7.8111,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.251187801361084,
+      "learning_rate": 6e-05,
+      "loss": 7.7546,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9874353408813477,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 7.6705,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3269367218017578,
+      "learning_rate": 7e-05,
+      "loss": 7.5703,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2412292957305908,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 7.3716,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.549961805343628,
+      "learning_rate": 8e-05,
+      "loss": 7.2603,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9291224479675293,
+      "learning_rate": 8.5e-05,
+      "loss": 7.0493,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.732795238494873,
+      "learning_rate": 9e-05,
+      "loss": 6.9551,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0821665525436401,
+      "learning_rate": 9.5e-05,
+      "loss": 6.8147,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1155742406845093,
+      "learning_rate": 0.0001,
+      "loss": 6.5402,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.240027904510498,
+      "learning_rate": 9.999999582259295e-05,
+      "loss": 6.3528,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.420423150062561,
+      "learning_rate": 9.999998329037248e-05,
+      "loss": 6.2188,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.957785129547119,
+      "learning_rate": 9.999996240334068e-05,
+      "loss": 6.1174,
+      "step": 115
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7262574434280396,
+      "learning_rate": 9.999993316150106e-05,
+      "loss": 6.0459,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.235532522201538,
+      "learning_rate": 9.999989556485852e-05,
+      "loss": 5.8964,
+      "step": 125
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.064032793045044,
+      "learning_rate": 9.999984961341928e-05,
+      "loss": 5.8453,
+      "step": 130
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2455618381500244,
+      "learning_rate": 9.999979530719108e-05,
+      "loss": 5.8238,
+      "step": 135
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5943653583526611,
+      "learning_rate": 9.999973264618297e-05,
+      "loss": 5.7566,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4457610845565796,
+      "learning_rate": 9.999966163040541e-05,
+      "loss": 5.6157,
+      "step": 145
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8074263334274292,
+      "learning_rate": 9.99995822598703e-05,
+      "loss": 5.6586,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4530527591705322,
+      "learning_rate": 9.999949453459088e-05,
+      "loss": 5.493,
+      "step": 155
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2159135341644287,
+      "learning_rate": 9.999939845458179e-05,
+      "loss": 5.4517,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8662954568862915,
+      "learning_rate": 9.999929401985911e-05,
+      "loss": 5.412,
+      "step": 165
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5726619958877563,
+      "learning_rate": 9.99991812304403e-05,
+      "loss": 5.3791,
+      "step": 170
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0798999071121216,
+      "learning_rate": 9.999906008634416e-05,
+      "loss": 5.4143,
+      "step": 175
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6859450340270996,
+      "learning_rate": 9.9998930587591e-05,
+      "loss": 5.2788,
+      "step": 180
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1738460063934326,
+      "learning_rate": 9.99987927342024e-05,
+      "loss": 5.2636,
+      "step": 185
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.013193130493164,
+      "learning_rate": 9.999864652620143e-05,
+      "loss": 5.3485,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.269242525100708,
+      "learning_rate": 9.999849196361251e-05,
+      "loss": 5.0791,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2379977703094482,
+      "learning_rate": 9.999832904646147e-05,
+      "loss": 5.0881,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.084080457687378,
+      "learning_rate": 9.99981577747755e-05,
+      "loss": 5.1453,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3836238384246826,
+      "learning_rate": 9.999797814858327e-05,
+      "loss": 5.1081,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.892317056655884,
+      "learning_rate": 9.999779016791477e-05,
+      "loss": 5.1076,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1019105911254883,
+      "learning_rate": 9.999759383280141e-05,
+      "loss": 4.9419,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6054673194885254,
+      "learning_rate": 9.9997389143276e-05,
+      "loss": 5.0233,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0115082263946533,
+      "learning_rate": 9.999717609937275e-05,
+      "loss": 4.9838,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.321122407913208,
+      "learning_rate": 9.999695470112725e-05,
+      "loss": 4.9806,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5660901069641113,
+      "learning_rate": 9.99967249485765e-05,
+      "loss": 4.9209,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.20086407661438,
+      "learning_rate": 9.999648684175888e-05,
+      "loss": 4.8584,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8749756813049316,
+      "learning_rate": 9.999624038071418e-05,
+      "loss": 4.912,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.378511905670166,
+      "learning_rate": 9.99959855654836e-05,
+      "loss": 4.9607,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5024422407150269,
+      "learning_rate": 9.999572239610968e-05,
+      "loss": 4.9151,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2261245250701904,
+      "learning_rate": 9.999545087263645e-05,
+      "loss": 4.8319,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.797099232673645,
+      "learning_rate": 9.999517099510923e-05,
+      "loss": 4.8008,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4046310186386108,
+      "learning_rate": 9.999488276357481e-05,
+      "loss": 4.8111,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.530915379524231,
+      "learning_rate": 9.999458617808137e-05,
+      "loss": 4.7673,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6640558242797852,
+      "learning_rate": 9.999428123867843e-05,
+      "loss": 4.696,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4346458911895752,
+      "learning_rate": 9.999396794541695e-05,
+      "loss": 4.8013,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.630961537361145,
+      "learning_rate": 9.999364629834931e-05,
+      "loss": 4.7538,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5089904069900513,
+      "learning_rate": 9.999331629752924e-05,
+      "loss": 4.8037,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.620252013206482,
+      "learning_rate": 9.999297794301188e-05,
+      "loss": 4.6888,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3059202432632446,
+      "learning_rate": 9.999263123485376e-05,
+      "loss": 4.6302,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.609636902809143,
+      "learning_rate": 9.999227617311282e-05,
+      "loss": 4.6804,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0964434146881104,
+      "learning_rate": 9.99919127578484e-05,
+      "loss": 4.6313,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8375858068466187,
+      "learning_rate": 9.999154098912122e-05,
+      "loss": 4.6352,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8606252670288086,
+      "learning_rate": 9.999116086699338e-05,
+      "loss": 4.6751,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7570736408233643,
+      "learning_rate": 9.999077239152843e-05,
+      "loss": 4.5551,
+      "step": 335
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5791484117507935,
+      "learning_rate": 9.999037556279127e-05,
+      "loss": 4.6209,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.674574375152588,
+      "learning_rate": 9.998997038084818e-05,
+      "loss": 4.5579,
+      "step": 345
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.608364462852478,
+      "learning_rate": 9.998955684576692e-05,
+      "loss": 4.6303,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7621687650680542,
+      "learning_rate": 9.998913495761654e-05,
+      "loss": 4.5925,
+      "step": 355
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.084482431411743,
+      "learning_rate": 9.998870471646757e-05,
+      "loss": 4.5258,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8192737102508545,
+      "learning_rate": 9.99882661223919e-05,
+      "loss": 4.4985,
+      "step": 365
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.451216697692871,
+      "learning_rate": 9.998781917546278e-05,
+      "loss": 4.6293,
+      "step": 370
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.027977228164673,
+      "learning_rate": 9.998736387575493e-05,
+      "loss": 4.5551,
+      "step": 375
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2394070625305176,
+      "learning_rate": 9.998690022334442e-05,
+      "loss": 4.4364,
+      "step": 380
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1030092239379883,
+      "learning_rate": 9.998642821830873e-05,
+      "loss": 4.529,
+      "step": 385
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6759475469589233,
+      "learning_rate": 9.99859478607267e-05,
+      "loss": 4.4632,
+      "step": 390
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8269052505493164,
+      "learning_rate": 9.998545915067864e-05,
+      "loss": 4.5811,
+      "step": 395
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9629148244857788,
+      "learning_rate": 9.998496208824618e-05,
+      "loss": 4.3144,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.830991268157959,
+      "learning_rate": 9.99844566735124e-05,
+      "loss": 4.5015,
+      "step": 405
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5572060346603394,
+      "learning_rate": 9.998394290656172e-05,
+      "loss": 4.4563,
+      "step": 410
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2485480308532715,
+      "learning_rate": 9.998342078748002e-05,
+      "loss": 4.444,
+      "step": 415
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6104615926742554,
+      "learning_rate": 9.998289031635453e-05,
+      "loss": 4.4632,
+      "step": 420
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4561620950698853,
+      "learning_rate": 9.998235149327389e-05,
+      "loss": 4.497,
+      "step": 425
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5755102634429932,
+      "learning_rate": 9.998180431832816e-05,
+      "loss": 4.5424,
+      "step": 430
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5101748704910278,
+      "learning_rate": 9.998124879160872e-05,
+      "loss": 4.2844,
+      "step": 435
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4677684307098389,
+      "learning_rate": 9.998068491320844e-05,
+      "loss": 4.4009,
+      "step": 440
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2594773769378662,
+      "learning_rate": 9.998011268322152e-05,
+      "loss": 4.373,
+      "step": 445
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3999148607254028,
+      "learning_rate": 9.997953210174361e-05,
+      "loss": 4.3682,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3341504335403442,
+      "learning_rate": 9.997894316887167e-05,
+      "loss": 4.3541,
+      "step": 455
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8098266124725342,
+      "learning_rate": 9.997834588470414e-05,
+      "loss": 4.3316,
+      "step": 460
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.309643268585205,
+      "learning_rate": 9.997774024934082e-05,
+      "loss": 4.3708,
+      "step": 465
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6201887130737305,
+      "learning_rate": 9.997712626288292e-05,
+      "loss": 4.3227,
+      "step": 470
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3977891206741333,
+      "learning_rate": 9.997650392543303e-05,
+      "loss": 4.3275,
+      "step": 475
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.270277500152588,
+      "learning_rate": 9.997587323709512e-05,
+      "loss": 4.3605,
+      "step": 480
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3718878030776978,
+      "learning_rate": 9.997523419797459e-05,
+      "loss": 4.2566,
+      "step": 485
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.321706771850586,
+      "learning_rate": 9.997458680817824e-05,
+      "loss": 4.2997,
+      "step": 490
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9669241905212402,
+      "learning_rate": 9.997393106781421e-05,
+      "loss": 4.2287,
+      "step": 495
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3454854488372803,
+      "learning_rate": 9.99732669769921e-05,
+      "loss": 4.317,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7181518077850342,
+      "learning_rate": 9.997259453582286e-05,
+      "loss": 4.2565,
+      "step": 505
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4946560859680176,
+      "learning_rate": 9.997191374441887e-05,
+      "loss": 4.3073,
+      "step": 510
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3225387334823608,
+      "learning_rate": 9.997122460289387e-05,
+      "loss": 4.2733,
+      "step": 515
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9715029001235962,
+      "learning_rate": 9.997052711136301e-05,
+      "loss": 4.3079,
+      "step": 520
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4230276346206665,
+      "learning_rate": 9.996982126994287e-05,
+      "loss": 4.3568,
+      "step": 525
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0797390937805176,
+      "learning_rate": 9.996910707875136e-05,
+      "loss": 4.2322,
+      "step": 530
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.340152382850647,
+      "learning_rate": 9.996838453790784e-05,
+      "loss": 4.2986,
+      "step": 535
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.611376166343689,
+      "learning_rate": 9.996765364753302e-05,
+      "loss": 4.2446,
+      "step": 540
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6274476051330566,
+      "learning_rate": 9.996691440774906e-05,
+      "loss": 4.2255,
+      "step": 545
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2220962047576904,
+      "learning_rate": 9.996616681867945e-05,
+      "loss": 4.1849,
+      "step": 550
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.563598871231079,
+      "learning_rate": 9.996541088044913e-05,
+      "loss": 4.2059,
+      "step": 555
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.565755844116211,
+      "learning_rate": 9.996464659318442e-05,
+      "loss": 4.2583,
+      "step": 560
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4644912481307983,
+      "learning_rate": 9.996387395701302e-05,
+      "loss": 4.272,
+      "step": 565
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7817014455795288,
+      "learning_rate": 9.996309297206403e-05,
+      "loss": 4.1708,
+      "step": 570
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5674885511398315,
+      "learning_rate": 9.996230363846797e-05,
+      "loss": 4.2202,
+      "step": 575
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6613502502441406,
+      "learning_rate": 9.996150595635671e-05,
+      "loss": 4.26,
+      "step": 580
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4964529275894165,
+      "learning_rate": 9.996069992586355e-05,
+      "loss": 4.2586,
+      "step": 585
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2414133548736572,
+      "learning_rate": 9.995988554712318e-05,
+      "loss": 4.159,
+      "step": 590
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3225334882736206,
+      "learning_rate": 9.995906282027166e-05,
+      "loss": 4.2161,
+      "step": 595
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.283974528312683,
+      "learning_rate": 9.995823174544649e-05,
+      "loss": 4.2242,
+      "step": 600
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3114523887634277,
+      "learning_rate": 9.995739232278654e-05,
+      "loss": 4.1486,
+      "step": 605
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.217221975326538,
+      "learning_rate": 9.995654455243204e-05,
+      "loss": 4.1439,
+      "step": 610
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4537752866744995,
+      "learning_rate": 9.995568843452468e-05,
+      "loss": 4.2154,
+      "step": 615
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5249435901641846,
+      "learning_rate": 9.99548239692075e-05,
+      "loss": 4.1182,
+      "step": 620
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2458933591842651,
+      "learning_rate": 9.995395115662498e-05,
+      "loss": 4.129,
+      "step": 625
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1656845808029175,
+      "learning_rate": 9.995306999692293e-05,
+      "loss": 4.1478,
+      "step": 630
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2603402137756348,
+      "learning_rate": 9.995218049024858e-05,
+      "loss": 4.1329,
+      "step": 635
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3216552734375,
+      "learning_rate": 9.995128263675058e-05,
+      "loss": 4.0856,
+      "step": 640
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5108698606491089,
+      "learning_rate": 9.995037643657899e-05,
+      "loss": 4.1054,
+      "step": 645
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3153183460235596,
+      "learning_rate": 9.994946188988516e-05,
+      "loss": 4.1063,
+      "step": 650
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.448789358139038,
+      "learning_rate": 9.994853899682197e-05,
+      "loss": 4.0492,
+      "step": 655
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.255163311958313,
+      "learning_rate": 9.994760775754362e-05,
+      "loss": 4.1103,
+      "step": 660
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.418347716331482,
+      "learning_rate": 9.994666817220571e-05,
+      "loss": 4.0378,
+      "step": 665
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2833226919174194,
+      "learning_rate": 9.994572024096521e-05,
+      "loss": 4.0871,
+      "step": 670
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3747986555099487,
+      "learning_rate": 9.994476396398055e-05,
+      "loss": 4.0571,
+      "step": 675
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.312347650527954,
+      "learning_rate": 9.994379934141153e-05,
+      "loss": 4.0687,
+      "step": 680
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2285041809082031,
+      "learning_rate": 9.994282637341932e-05,
+      "loss": 4.1101,
+      "step": 685
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.099517822265625,
+      "learning_rate": 9.994184506016651e-05,
+      "loss": 4.0875,
+      "step": 690
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0488024950027466,
+      "learning_rate": 9.994085540181703e-05,
+      "loss": 4.1703,
+      "step": 695
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6460891962051392,
+      "learning_rate": 9.99398573985363e-05,
+      "loss": 4.0724,
+      "step": 700
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5542891025543213,
+      "learning_rate": 9.993885105049107e-05,
+      "loss": 4.0747,
+      "step": 705
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1527950763702393,
+      "learning_rate": 9.993783635784947e-05,
+      "loss": 4.1118,
+      "step": 710
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0955204963684082,
+      "learning_rate": 9.993681332078108e-05,
+      "loss": 4.0979,
+      "step": 715
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0570292472839355,
+      "learning_rate": 9.993578193945685e-05,
+      "loss": 4.0658,
+      "step": 720
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.403805136680603,
+      "learning_rate": 9.99347422140491e-05,
+      "loss": 4.0477,
+      "step": 725
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.275642991065979,
+      "learning_rate": 9.993369414473157e-05,
+      "loss": 4.0976,
+      "step": 730
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1618658304214478,
+      "learning_rate": 9.993263773167939e-05,
+      "loss": 4.1611,
+      "step": 735
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1634148359298706,
+      "learning_rate": 9.993157297506907e-05,
+      "loss": 3.9691,
+      "step": 740
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.260061264038086,
+      "learning_rate": 9.993049987507856e-05,
+      "loss": 4.0713,
+      "step": 745
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.157210350036621,
+      "learning_rate": 9.992941843188715e-05,
+      "loss": 3.9896,
+      "step": 750
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2721577882766724,
+      "learning_rate": 9.992832864567554e-05,
+      "loss": 4.0315,
+      "step": 755
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5931888818740845,
+      "learning_rate": 9.992723051662584e-05,
+      "loss": 4.0401,
+      "step": 760
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3395116329193115,
+      "learning_rate": 9.992612404492153e-05,
+      "loss": 4.0484,
+      "step": 765
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4800901412963867,
+      "learning_rate": 9.992500923074752e-05,
+      "loss": 4.0478,
+      "step": 770
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9852534532546997,
+      "learning_rate": 9.992388607429006e-05,
+      "loss": 4.0412,
+      "step": 775
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1356985569000244,
+      "learning_rate": 9.992275457573685e-05,
+      "loss": 3.9624,
+      "step": 780
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.229654312133789,
+      "learning_rate": 9.992161473527695e-05,
+      "loss": 3.9168,
+      "step": 785
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.193708896636963,
+      "learning_rate": 9.992046655310083e-05,
+      "loss": 3.9894,
+      "step": 790
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3792873620986938,
+      "learning_rate": 9.991931002940034e-05,
+      "loss": 3.9671,
+      "step": 795
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.169438362121582,
+      "learning_rate": 9.991814516436873e-05,
+      "loss": 4.0253,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1359736919403076,
+      "learning_rate": 9.991697195820065e-05,
+      "loss": 3.898,
+      "step": 805
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3887481689453125,
+      "learning_rate": 9.991579041109212e-05,
+      "loss": 4.026,
+      "step": 810
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3214277029037476,
+      "learning_rate": 9.991460052324061e-05,
+      "loss": 3.9373,
+      "step": 815
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0724738836288452,
+      "learning_rate": 9.991340229484491e-05,
+      "loss": 3.8542,
+      "step": 820
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.12058687210083,
+      "learning_rate": 9.991219572610526e-05,
+      "loss": 3.9061,
+      "step": 825
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0900886058807373,
+      "learning_rate": 9.991098081722325e-05,
+      "loss": 3.9291,
+      "step": 830
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1248831748962402,
+      "learning_rate": 9.990975756840193e-05,
+      "loss": 4.054,
+      "step": 835
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2869148254394531,
+      "learning_rate": 9.990852597984566e-05,
+      "loss": 3.9369,
+      "step": 840
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1658549308776855,
+      "learning_rate": 9.990728605176025e-05,
+      "loss": 3.9197,
+      "step": 845
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1743642091751099,
+      "learning_rate": 9.990603778435288e-05,
+      "loss": 3.93,
+      "step": 850
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0986660718917847,
+      "learning_rate": 9.990478117783214e-05,
+      "loss": 3.828,
+      "step": 855
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.00698184967041,
+      "learning_rate": 9.990351623240799e-05,
+      "loss": 3.9354,
+      "step": 860
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2176913022994995,
+      "learning_rate": 9.990224294829181e-05,
+      "loss": 4.0076,
+      "step": 865
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.18546462059021,
+      "learning_rate": 9.990096132569636e-05,
+      "loss": 3.992,
+      "step": 870
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0776242017745972,
+      "learning_rate": 9.98996713648358e-05,
+      "loss": 3.9186,
+      "step": 875
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1194149255752563,
+      "learning_rate": 9.989837306592567e-05,
+      "loss": 3.8018,
+      "step": 880
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1288461685180664,
+      "learning_rate": 9.989706642918291e-05,
+      "loss": 3.9639,
+      "step": 885
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0761008262634277,
+      "learning_rate": 9.989575145482583e-05,
+      "loss": 3.8745,
+      "step": 890
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4154248237609863,
+      "learning_rate": 9.989442814307422e-05,
+      "loss": 3.9412,
+      "step": 895
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.266314148902893,
+      "learning_rate": 9.989309649414914e-05,
+      "loss": 3.9154,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1292552947998047,
+      "learning_rate": 9.989175650827314e-05,
+      "loss": 3.8948,
+      "step": 905
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0503509044647217,
+      "learning_rate": 9.98904081856701e-05,
+      "loss": 3.8656,
+      "step": 910
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1102133989334106,
+      "learning_rate": 9.988905152656534e-05,
+      "loss": 3.9219,
+      "step": 915
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.239732027053833,
+      "learning_rate": 9.988768653118554e-05,
+      "loss": 3.9217,
+      "step": 920
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.046139121055603,
+      "learning_rate": 9.988631319975881e-05,
+      "loss": 3.8576,
+      "step": 925
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1388572454452515,
+      "learning_rate": 9.988493153251459e-05,
+      "loss": 3.8417,
+      "step": 930
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9642298817634583,
+      "learning_rate": 9.988354152968379e-05,
+      "loss": 3.7982,
+      "step": 935
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1193811893463135,
+      "learning_rate": 9.988214319149865e-05,
+      "loss": 3.8849,
+      "step": 940
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9981116056442261,
+      "learning_rate": 9.988073651819281e-05,
+      "loss": 3.9754,
+      "step": 945
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1638458967208862,
+      "learning_rate": 9.987932151000137e-05,
+      "loss": 3.844,
+      "step": 950
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1412110328674316,
+      "learning_rate": 9.987789816716073e-05,
+      "loss": 3.8951,
+      "step": 955
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9065037369728088,
+      "learning_rate": 9.987646648990876e-05,
+      "loss": 3.8415,
+      "step": 960
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.156842589378357,
+      "learning_rate": 9.987502647848466e-05,
+      "loss": 3.9489,
+      "step": 965
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5554940700531006,
+      "learning_rate": 9.987357813312905e-05,
+      "loss": 3.8335,
+      "step": 970
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9845486879348755,
+      "learning_rate": 9.987212145408395e-05,
+      "loss": 3.918,
+      "step": 975
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1541403532028198,
+      "learning_rate": 9.987065644159278e-05,
+      "loss": 3.8699,
+      "step": 980
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1157742738723755,
+      "learning_rate": 9.986918309590031e-05,
+      "loss": 3.8432,
+      "step": 985
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1114094257354736,
+      "learning_rate": 9.986770141725277e-05,
+      "loss": 3.7904,
+      "step": 990
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1606131792068481,
+      "learning_rate": 9.986621140589772e-05,
+      "loss": 3.8892,
+      "step": 995
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0417673587799072,
+      "learning_rate": 9.986471306208412e-05,
+      "loss": 3.8229,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0612916946411133,
+      "learning_rate": 9.986320638606235e-05,
+      "loss": 3.8372,
+      "step": 1005
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0083855390548706,
+      "learning_rate": 9.986169137808419e-05,
+      "loss": 3.7716,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0122841596603394,
+      "learning_rate": 9.986016803840275e-05,
+      "loss": 3.8423,
+      "step": 1015
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9446907043457031,
+      "learning_rate": 9.985863636727262e-05,
+      "loss": 3.7863,
+      "step": 1020
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9782904386520386,
+      "learning_rate": 9.985709636494971e-05,
+      "loss": 3.8002,
+      "step": 1025
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9217712879180908,
+      "learning_rate": 9.985554803169134e-05,
+      "loss": 3.9062,
+      "step": 1030
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1464534997940063,
+      "learning_rate": 9.985399136775627e-05,
+      "loss": 3.7285,
+      "step": 1035
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.141538143157959,
+      "learning_rate": 9.985242637340458e-05,
+      "loss": 3.9393,
+      "step": 1040
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1080266237258911,
+      "learning_rate": 9.985085304889778e-05,
+      "loss": 3.8413,
+      "step": 1045
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0593568086624146,
+      "learning_rate": 9.984927139449877e-05,
+      "loss": 3.8738,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9855124950408936,
+      "learning_rate": 9.984768141047184e-05,
+      "loss": 3.7909,
+      "step": 1055
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0388472080230713,
+      "learning_rate": 9.984608309708266e-05,
+      "loss": 3.755,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.986768901348114,
+      "learning_rate": 9.984447645459831e-05,
+      "loss": 3.877,
+      "step": 1065
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.111816167831421,
+      "learning_rate": 9.984286148328728e-05,
+      "loss": 3.8048,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9084323644638062,
+      "learning_rate": 9.984123818341937e-05,
+      "loss": 3.7669,
+      "step": 1075
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.083175778388977,
+      "learning_rate": 9.983960655526587e-05,
+      "loss": 3.812,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0170353651046753,
+      "learning_rate": 9.983796659909941e-05,
+      "loss": 3.8537,
+      "step": 1085
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.024672031402588,
+      "learning_rate": 9.983631831519401e-05,
+      "loss": 3.8547,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1080533266067505,
+      "learning_rate": 9.983466170382511e-05,
+      "loss": 3.781,
+      "step": 1095
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9686844348907471,
+      "learning_rate": 9.983299676526949e-05,
+      "loss": 3.7417,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9940996170043945,
+      "learning_rate": 9.98313234998054e-05,
+      "loss": 3.7718,
+      "step": 1105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0285981893539429,
+      "learning_rate": 9.982964190771242e-05,
+      "loss": 3.7209,
+      "step": 1110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9261362552642822,
+      "learning_rate": 9.98279519892715e-05,
+      "loss": 3.6878,
+      "step": 1115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.082834005355835,
+      "learning_rate": 9.982625374476508e-05,
+      "loss": 3.8352,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0107392072677612,
+      "learning_rate": 9.98245471744769e-05,
+      "loss": 3.7909,
+      "step": 1125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0110729932785034,
+      "learning_rate": 9.982283227869211e-05,
+      "loss": 3.7913,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0634263753890991,
+      "learning_rate": 9.982110905769729e-05,
+      "loss": 3.7178,
+      "step": 1135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9466355443000793,
+      "learning_rate": 9.981937751178037e-05,
+      "loss": 3.7233,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9739379286766052,
+      "learning_rate": 9.981763764123067e-05,
+      "loss": 3.7828,
+      "step": 1145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9577144980430603,
+      "learning_rate": 9.981588944633893e-05,
+      "loss": 3.8085,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9381738901138306,
+      "learning_rate": 9.981413292739727e-05,
+      "loss": 3.6912,
+      "step": 1155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.011684536933899,
+      "learning_rate": 9.981236808469922e-05,
+      "loss": 3.8247,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8699536323547363,
+      "learning_rate": 9.981059491853964e-05,
+      "loss": 3.7683,
+      "step": 1165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0118540525436401,
+      "learning_rate": 9.980881342921482e-05,
+      "loss": 3.8031,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.128758192062378,
+      "learning_rate": 9.980702361702246e-05,
+      "loss": 3.7555,
+      "step": 1175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9168882369995117,
+      "learning_rate": 9.980522548226162e-05,
+      "loss": 3.8392,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9790633916854858,
+      "learning_rate": 9.980341902523279e-05,
+      "loss": 3.7272,
+      "step": 1185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0279648303985596,
+      "learning_rate": 9.980160424623776e-05,
+      "loss": 3.831,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8011073470115662,
+      "learning_rate": 9.979978114557984e-05,
+      "loss": 3.7728,
+      "step": 1195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9939100742340088,
+      "learning_rate": 9.979794972356362e-05,
+      "loss": 3.5779,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.950244665145874,
+      "learning_rate": 9.979610998049516e-05,
+      "loss": 3.6712,
+      "step": 1205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8702495694160461,
+      "learning_rate": 9.979426191668182e-05,
+      "loss": 3.7341,
+      "step": 1210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9258893728256226,
+      "learning_rate": 9.979240553243246e-05,
+      "loss": 3.6993,
+      "step": 1215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9927293658256531,
+      "learning_rate": 9.979054082805726e-05,
+      "loss": 3.8192,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0064748525619507,
+      "learning_rate": 9.978866780386777e-05,
+      "loss": 3.7479,
+      "step": 1225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.901767373085022,
+      "learning_rate": 9.978678646017701e-05,
+      "loss": 3.7843,
+      "step": 1230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9215208888053894,
+      "learning_rate": 9.978489679729934e-05,
+      "loss": 3.6875,
+      "step": 1235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0209126472473145,
+      "learning_rate": 9.978299881555048e-05,
+      "loss": 3.68,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8332811594009399,
+      "learning_rate": 9.97810925152476e-05,
+      "loss": 3.7676,
+      "step": 1245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1346206665039062,
+      "learning_rate": 9.977917789670925e-05,
+      "loss": 3.69,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9856200814247131,
+      "learning_rate": 9.977725496025535e-05,
+      "loss": 3.6985,
+      "step": 1255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9143579006195068,
+      "learning_rate": 9.977532370620718e-05,
+      "loss": 3.7084,
+      "step": 1260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8938957452774048,
+      "learning_rate": 9.977338413488748e-05,
+      "loss": 3.5689,
+      "step": 1265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8601924180984497,
+      "learning_rate": 9.977143624662035e-05,
+      "loss": 3.6747,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1564587354660034,
+      "learning_rate": 9.976948004173125e-05,
+      "loss": 3.6629,
+      "step": 1275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9538741111755371,
+      "learning_rate": 9.976751552054706e-05,
+      "loss": 3.6907,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8944664597511292,
+      "learning_rate": 9.976554268339607e-05,
+      "loss": 3.6964,
+      "step": 1285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9314648509025574,
+      "learning_rate": 9.97635615306079e-05,
+      "loss": 3.6619,
+      "step": 1290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9331445097923279,
+      "learning_rate": 9.976157206251361e-05,
+      "loss": 3.7289,
+      "step": 1295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0160084962844849,
+      "learning_rate": 9.975957427944563e-05,
+      "loss": 3.7469,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8673502802848816,
+      "learning_rate": 9.975756818173778e-05,
+      "loss": 3.7092,
+      "step": 1305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8198760747909546,
+      "learning_rate": 9.975555376972529e-05,
+      "loss": 3.6425,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8011747002601624,
+      "learning_rate": 9.975353104374474e-05,
+      "loss": 3.6479,
+      "step": 1315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8020340800285339,
+      "learning_rate": 9.975150000413411e-05,
+      "loss": 3.6512,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0580570697784424,
+      "learning_rate": 9.97494606512328e-05,
+      "loss": 3.5737,
+      "step": 1325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.044822096824646,
+      "learning_rate": 9.974741298538158e-05,
+      "loss": 3.6988,
+      "step": 1330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.909848153591156,
+      "learning_rate": 9.974535700692257e-05,
+      "loss": 3.6237,
+      "step": 1335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7629563808441162,
+      "learning_rate": 9.974329271619938e-05,
+      "loss": 3.5832,
+      "step": 1340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9232847094535828,
+      "learning_rate": 9.97412201135569e-05,
+      "loss": 3.6308,
+      "step": 1345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.905464231967926,
+      "learning_rate": 9.973913919934147e-05,
+      "loss": 3.6569,
+      "step": 1350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.86399245262146,
+      "learning_rate": 9.973704997390079e-05,
+      "loss": 3.6178,
+      "step": 1355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9400821328163147,
+      "learning_rate": 9.973495243758398e-05,
+      "loss": 3.7282,
+      "step": 1360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.015331506729126,
+      "learning_rate": 9.973284659074151e-05,
+      "loss": 3.7793,
+      "step": 1365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.344465732574463,
+      "learning_rate": 9.973073243372528e-05,
+      "loss": 3.5891,
+      "step": 1370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9030474424362183,
+      "learning_rate": 9.972860996688853e-05,
+      "loss": 3.605,
+      "step": 1375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9091427326202393,
+      "learning_rate": 9.972647919058595e-05,
+      "loss": 3.6551,
+      "step": 1380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9561840295791626,
+      "learning_rate": 9.972434010517358e-05,
+      "loss": 3.6338,
+      "step": 1385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9476805329322815,
+      "learning_rate": 9.972219271100882e-05,
+      "loss": 3.6646,
+      "step": 1390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8724963665008545,
+      "learning_rate": 9.972003700845052e-05,
+      "loss": 3.6233,
+      "step": 1395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8247894048690796,
+      "learning_rate": 9.971787299785888e-05,
+      "loss": 3.621,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7998039126396179,
+      "learning_rate": 9.97157006795955e-05,
+      "loss": 3.5927,
+      "step": 1405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9684839248657227,
+      "learning_rate": 9.971352005402338e-05,
+      "loss": 3.6307,
+      "step": 1410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9071174263954163,
+      "learning_rate": 9.971133112150688e-05,
+      "loss": 3.6069,
+      "step": 1415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.817969024181366,
+      "learning_rate": 9.970913388241177e-05,
+      "loss": 3.6413,
+      "step": 1420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.857496976852417,
+      "learning_rate": 9.970692833710517e-05,
+      "loss": 3.6274,
+      "step": 1425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8139945864677429,
+      "learning_rate": 9.970471448595566e-05,
+      "loss": 3.6797,
+      "step": 1430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8333423137664795,
+      "learning_rate": 9.970249232933315e-05,
+      "loss": 3.6256,
+      "step": 1435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.031546711921692,
+      "learning_rate": 9.970026186760896e-05,
+      "loss": 3.6935,
+      "step": 1440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9086576104164124,
+      "learning_rate": 9.969802310115577e-05,
+      "loss": 3.6312,
+      "step": 1445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9369786977767944,
+      "learning_rate": 9.96957760303477e-05,
+      "loss": 3.6322,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8668227195739746,
+      "learning_rate": 9.969352065556022e-05,
+      "loss": 3.7396,
+      "step": 1455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8558319807052612,
+      "learning_rate": 9.969125697717017e-05,
+      "loss": 3.6194,
+      "step": 1460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8007488250732422,
+      "learning_rate": 9.968898499555582e-05,
+      "loss": 3.5418,
+      "step": 1465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8372405767440796,
+      "learning_rate": 9.968670471109682e-05,
+      "loss": 3.5289,
+      "step": 1470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8030397891998291,
+      "learning_rate": 9.968441612417418e-05,
+      "loss": 3.6176,
+      "step": 1475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.827089250087738,
+      "learning_rate": 9.968211923517032e-05,
+      "loss": 3.5334,
+      "step": 1480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8236131072044373,
+      "learning_rate": 9.967981404446905e-05,
+      "loss": 3.6113,
+      "step": 1485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8784987330436707,
+      "learning_rate": 9.967750055245555e-05,
+      "loss": 3.5695,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.83465176820755,
+      "learning_rate": 9.967517875951638e-05,
+      "loss": 3.5725,
+      "step": 1495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8424170017242432,
+      "learning_rate": 9.967284866603955e-05,
+      "loss": 3.5106,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3840759992599487,
+      "learning_rate": 9.967051027241436e-05,
+      "loss": 3.6532,
+      "step": 1505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8444687724113464,
+      "learning_rate": 9.966816357903155e-05,
+      "loss": 3.5277,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7896638512611389,
+      "learning_rate": 9.96658085862833e-05,
+      "loss": 3.6335,
+      "step": 1515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8296946287155151,
+      "learning_rate": 9.966344529456305e-05,
+      "loss": 3.6824,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7958943247795105,
+      "learning_rate": 9.966107370426573e-05,
+      "loss": 3.6277,
+      "step": 1525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8976621627807617,
+      "learning_rate": 9.965869381578765e-05,
+      "loss": 3.6265,
+      "step": 1530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8068059086799622,
+      "learning_rate": 9.965630562952643e-05,
+      "loss": 3.5518,
+      "step": 1535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8068969249725342,
+      "learning_rate": 9.965390914588115e-05,
+      "loss": 3.5494,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8511276245117188,
+      "learning_rate": 9.965150436525225e-05,
+      "loss": 3.5863,
+      "step": 1545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7962851524353027,
+      "learning_rate": 9.964909128804159e-05,
+      "loss": 3.6683,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.806166410446167,
+      "learning_rate": 9.964666991465234e-05,
+      "loss": 3.622,
+      "step": 1555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8023310899734497,
+      "learning_rate": 9.964424024548914e-05,
+      "loss": 3.5681,
+      "step": 1560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9122126698493958,
+      "learning_rate": 9.964180228095793e-05,
+      "loss": 3.6593,
+      "step": 1565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7729650139808655,
+      "learning_rate": 9.963935602146612e-05,
+      "loss": 3.5379,
+      "step": 1570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7860742211341858,
+      "learning_rate": 9.963690146742248e-05,
+      "loss": 3.61,
+      "step": 1575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8527259230613708,
+      "learning_rate": 9.963443861923715e-05,
+      "loss": 3.6411,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7829805016517639,
+      "learning_rate": 9.963196747732164e-05,
+      "loss": 3.5349,
+      "step": 1585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7953242063522339,
+      "learning_rate": 9.96294880420889e-05,
+      "loss": 3.5565,
+      "step": 1590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9077381491661072,
+      "learning_rate": 9.96270003139532e-05,
+      "loss": 3.5149,
+      "step": 1595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8481189012527466,
+      "learning_rate": 9.962450429333026e-05,
+      "loss": 3.6406,
+      "step": 1600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.861329197883606,
+      "learning_rate": 9.962199998063715e-05,
+      "loss": 3.6645,
+      "step": 1605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8553574681282043,
+      "learning_rate": 9.961948737629231e-05,
+      "loss": 3.4426,
+      "step": 1610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9025221467018127,
+      "learning_rate": 9.961696648071561e-05,
+      "loss": 3.5863,
+      "step": 1615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7824848890304565,
+      "learning_rate": 9.961443729432828e-05,
+      "loss": 3.5548,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8887602686882019,
+      "learning_rate": 9.961189981755294e-05,
+      "loss": 3.6061,
+      "step": 1625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8720082640647888,
+      "learning_rate": 9.960935405081357e-05,
+      "loss": 3.5083,
+      "step": 1630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9271783828735352,
+      "learning_rate": 9.960679999453559e-05,
+      "loss": 3.6062,
+      "step": 1635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7855983376502991,
+      "learning_rate": 9.960423764914575e-05,
+      "loss": 3.4889,
+      "step": 1640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9302400946617126,
+      "learning_rate": 9.960166701507221e-05,
+      "loss": 3.5156,
+      "step": 1645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8552121520042419,
+      "learning_rate": 9.959908809274453e-05,
+      "loss": 3.5714,
+      "step": 1650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9043262600898743,
+      "learning_rate": 9.959650088259362e-05,
+      "loss": 3.5606,
+      "step": 1655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8065487742424011,
+      "learning_rate": 9.959390538505181e-05,
+      "loss": 3.5708,
+      "step": 1660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.942502498626709,
+      "learning_rate": 9.959130160055279e-05,
+      "loss": 3.5264,
+      "step": 1665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8328238129615784,
+      "learning_rate": 9.958868952953163e-05,
+      "loss": 3.5692,
+      "step": 1670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7958769202232361,
+      "learning_rate": 9.95860691724248e-05,
+      "loss": 3.5205,
+      "step": 1675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7748988270759583,
+      "learning_rate": 9.958344052967018e-05,
+      "loss": 3.4882,
+      "step": 1680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8796920776367188,
+      "learning_rate": 9.958080360170698e-05,
+      "loss": 3.4776,
+      "step": 1685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8776568174362183,
+      "learning_rate": 9.957815838897585e-05,
+      "loss": 3.4306,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.768972635269165,
+      "learning_rate": 9.957550489191874e-05,
+      "loss": 3.5405,
+      "step": 1695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8360869884490967,
+      "learning_rate": 9.957284311097908e-05,
+      "loss": 3.6362,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8310314416885376,
+      "learning_rate": 9.957017304660164e-05,
+      "loss": 3.535,
+      "step": 1705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9471503496170044,
+      "learning_rate": 9.956749469923258e-05,
+      "loss": 3.5473,
+      "step": 1710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8828099966049194,
+      "learning_rate": 9.956480806931942e-05,
+      "loss": 3.578,
+      "step": 1715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9010315537452698,
+      "learning_rate": 9.956211315731112e-05,
+      "loss": 3.5638,
+      "step": 1720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7857769727706909,
+      "learning_rate": 9.955940996365796e-05,
+      "loss": 3.5843,
+      "step": 1725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7853619456291199,
+      "learning_rate": 9.955669848881166e-05,
+      "loss": 3.4816,
+      "step": 1730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7379563450813293,
+      "learning_rate": 9.955397873322527e-05,
+      "loss": 3.5698,
+      "step": 1735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7403237223625183,
+      "learning_rate": 9.955125069735327e-05,
+      "loss": 3.516,
+      "step": 1740
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7524884343147278,
+      "learning_rate": 9.95485143816515e-05,
+      "loss": 3.5767,
+      "step": 1745
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8811848163604736,
+      "learning_rate": 9.95457697865772e-05,
+      "loss": 3.5968,
+      "step": 1750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8003664612770081,
+      "learning_rate": 9.954301691258894e-05,
+      "loss": 3.4212,
+      "step": 1755
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7701109647750854,
+      "learning_rate": 9.954025576014674e-05,
+      "loss": 3.519,
+      "step": 1760
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7772321105003357,
+      "learning_rate": 9.953748632971201e-05,
+      "loss": 3.648,
+      "step": 1765
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7890676856040955,
+      "learning_rate": 9.953470862174748e-05,
+      "loss": 3.5021,
+      "step": 1770
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.758638322353363,
+      "learning_rate": 9.953192263671728e-05,
+      "loss": 3.5261,
+      "step": 1775
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8380297422409058,
+      "learning_rate": 9.952912837508697e-05,
+      "loss": 3.4982,
+      "step": 1780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9461044669151306,
+      "learning_rate": 9.952632583732343e-05,
+      "loss": 3.4896,
+      "step": 1785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9555126428604126,
+      "learning_rate": 9.952351502389498e-05,
+      "loss": 3.5219,
+      "step": 1790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8303565979003906,
+      "learning_rate": 9.95206959352713e-05,
+      "loss": 3.5433,
+      "step": 1795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8429514765739441,
+      "learning_rate": 9.951786857192341e-05,
+      "loss": 3.4675,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8268835544586182,
+      "learning_rate": 9.951503293432378e-05,
+      "loss": 3.6183,
+      "step": 1805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.786369800567627,
+      "learning_rate": 9.951218902294625e-05,
+      "loss": 3.5544,
+      "step": 1810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8204576373100281,
+      "learning_rate": 9.9509336838266e-05,
+      "loss": 3.5327,
+      "step": 1815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8163040280342102,
+      "learning_rate": 9.950647638075963e-05,
+      "loss": 3.4644,
+      "step": 1820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7794016003608704,
+      "learning_rate": 9.950360765090511e-05,
+      "loss": 3.5335,
+      "step": 1825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7765195965766907,
+      "learning_rate": 9.950073064918179e-05,
+      "loss": 3.5244,
+      "step": 1830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7425858378410339,
+      "learning_rate": 9.949784537607042e-05,
+      "loss": 3.4522,
+      "step": 1835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8290202617645264,
+      "learning_rate": 9.949495183205311e-05,
+      "loss": 3.4792,
+      "step": 1840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7658949494361877,
+      "learning_rate": 9.949205001761334e-05,
+      "loss": 3.4914,
+      "step": 1845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7830814123153687,
+      "learning_rate": 9.948913993323603e-05,
+      "loss": 3.5549,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8343876600265503,
+      "learning_rate": 9.948622157940744e-05,
+      "loss": 3.3829,
+      "step": 1855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7723503708839417,
+      "learning_rate": 9.948329495661517e-05,
+      "loss": 3.4856,
+      "step": 1860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7451229095458984,
+      "learning_rate": 9.94803600653483e-05,
+      "loss": 3.4013,
+      "step": 1865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8351654410362244,
+      "learning_rate": 9.947741690609722e-05,
+      "loss": 3.5279,
+      "step": 1870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7712655663490295,
+      "learning_rate": 9.94744654793537e-05,
+      "loss": 3.4596,
+      "step": 1875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.818580687046051,
+      "learning_rate": 9.947150578561095e-05,
+      "loss": 3.4696,
+      "step": 1880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8177143335342407,
+      "learning_rate": 9.946853782536351e-05,
+      "loss": 3.5708,
+      "step": 1885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7996229529380798,
+      "learning_rate": 9.946556159910732e-05,
+      "loss": 3.5465,
+      "step": 1890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7518954277038574,
+      "learning_rate": 9.946257710733966e-05,
+      "loss": 3.4726,
+      "step": 1895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7876341938972473,
+      "learning_rate": 9.945958435055927e-05,
+      "loss": 3.4946,
+      "step": 1900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7520238161087036,
+      "learning_rate": 9.945658332926622e-05,
+      "loss": 3.4679,
+      "step": 1905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7998647689819336,
+      "learning_rate": 9.945357404396197e-05,
+      "loss": 3.4312,
+      "step": 1910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.836589515209198,
+      "learning_rate": 9.945055649514934e-05,
+      "loss": 3.4479,
+      "step": 1915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8144925236701965,
+      "learning_rate": 9.944753068333256e-05,
+      "loss": 3.502,
+      "step": 1920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7192187905311584,
+      "learning_rate": 9.944449660901725e-05,
+      "loss": 3.4519,
+      "step": 1925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8299037218093872,
+      "learning_rate": 9.944145427271036e-05,
+      "loss": 3.4393,
+      "step": 1930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7786000967025757,
+      "learning_rate": 9.94384036749203e-05,
+      "loss": 3.4382,
+      "step": 1935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8719263672828674,
+      "learning_rate": 9.943534481615677e-05,
+      "loss": 3.4513,
+      "step": 1940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7792828679084778,
+      "learning_rate": 9.943227769693091e-05,
+      "loss": 3.3969,
+      "step": 1945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7004716396331787,
+      "learning_rate": 9.942920231775524e-05,
+      "loss": 3.4047,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7165127992630005,
+      "learning_rate": 9.942611867914363e-05,
+      "loss": 3.4856,
+      "step": 1955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.792910099029541,
+      "learning_rate": 9.942302678161133e-05,
+      "loss": 3.4135,
+      "step": 1960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6870299577713013,
+      "learning_rate": 9.9419926625675e-05,
+      "loss": 3.4161,
+      "step": 1965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8193487524986267,
+      "learning_rate": 9.941681821185265e-05,
+      "loss": 3.4038,
+      "step": 1970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8124663233757019,
+      "learning_rate": 9.941370154066371e-05,
+      "loss": 3.3875,
+      "step": 1975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.728029191493988,
+      "learning_rate": 9.941057661262895e-05,
+      "loss": 3.3653,
+      "step": 1980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7501214146614075,
+      "learning_rate": 9.940744342827055e-05,
+      "loss": 3.3684,
+      "step": 1985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8016537427902222,
+      "learning_rate": 9.940430198811202e-05,
+      "loss": 3.4373,
+      "step": 1990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.828636109828949,
+      "learning_rate": 9.940115229267831e-05,
+      "loss": 3.4149,
+      "step": 1995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.791454553604126,
+      "learning_rate": 9.939799434249571e-05,
+      "loss": 3.4485,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8796780109405518,
+      "learning_rate": 9.93948281380919e-05,
+      "loss": 3.4762,
+      "step": 2005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7664965987205505,
+      "learning_rate": 9.939165367999597e-05,
+      "loss": 3.4844,
+      "step": 2010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7541711330413818,
+      "learning_rate": 9.938847096873831e-05,
+      "loss": 3.4752,
+      "step": 2015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8239563703536987,
+      "learning_rate": 9.938528000485078e-05,
+      "loss": 3.3884,
+      "step": 2020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7384853959083557,
+      "learning_rate": 9.938208078886655e-05,
+      "loss": 3.3374,
+      "step": 2025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7558611631393433,
+      "learning_rate": 9.937887332132023e-05,
+      "loss": 3.4445,
+      "step": 2030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7288622856140137,
+      "learning_rate": 9.937565760274776e-05,
+      "loss": 3.5747,
+      "step": 2035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7006734609603882,
+      "learning_rate": 9.937243363368645e-05,
+      "loss": 3.4287,
+      "step": 2040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7733967304229736,
+      "learning_rate": 9.936920141467505e-05,
+      "loss": 3.4599,
+      "step": 2045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7035228610038757,
+      "learning_rate": 9.936596094625364e-05,
+      "loss": 3.4613,
+      "step": 2050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7483477592468262,
+      "learning_rate": 9.936271222896367e-05,
+      "loss": 3.4375,
+      "step": 2055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6571712493896484,
+      "learning_rate": 9.935945526334803e-05,
+      "loss": 3.4323,
+      "step": 2060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7757775187492371,
+      "learning_rate": 9.935619004995089e-05,
+      "loss": 3.4619,
+      "step": 2065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8020771741867065,
+      "learning_rate": 9.935291658931791e-05,
+      "loss": 3.3676,
+      "step": 2070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7800244688987732,
+      "learning_rate": 9.934963488199604e-05,
+      "loss": 3.468,
+      "step": 2075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.822228193283081,
+      "learning_rate": 9.934634492853366e-05,
+      "loss": 3.4625,
+      "step": 2080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.759692907333374,
+      "learning_rate": 9.93430467294805e-05,
+      "loss": 3.4704,
+      "step": 2085
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7965198755264282,
+      "learning_rate": 9.933974028538768e-05,
+      "loss": 3.4194,
+      "step": 2090
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7468062043190002,
+      "learning_rate": 9.933642559680769e-05,
+      "loss": 3.4416,
+      "step": 2095
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7842198610305786,
+      "learning_rate": 9.93331026642944e-05,
+      "loss": 3.3782,
+      "step": 2100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7557224631309509,
+      "learning_rate": 9.932977148840307e-05,
+      "loss": 3.3766,
+      "step": 2105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7935373187065125,
+      "learning_rate": 9.93264320696903e-05,
+      "loss": 3.4447,
+      "step": 2110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7485852241516113,
+      "learning_rate": 9.932308440871413e-05,
+      "loss": 3.4347,
+      "step": 2115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7590314149856567,
+      "learning_rate": 9.931972850603394e-05,
+      "loss": 3.4009,
+      "step": 2120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.788416862487793,
+      "learning_rate": 9.931636436221048e-05,
+      "loss": 3.4979,
+      "step": 2125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8660835027694702,
+      "learning_rate": 9.931299197780586e-05,
+      "loss": 3.5175,
+      "step": 2130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7509318590164185,
+      "learning_rate": 9.930961135338361e-05,
+      "loss": 3.394,
+      "step": 2135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7240895628929138,
+      "learning_rate": 9.930622248950864e-05,
+      "loss": 3.2934,
+      "step": 2140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.735086977481842,
+      "learning_rate": 9.93028253867472e-05,
+      "loss": 3.4318,
+      "step": 2145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.735582709312439,
+      "learning_rate": 9.929942004566695e-05,
+      "loss": 3.3378,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7257517576217651,
+      "learning_rate": 9.929600646683686e-05,
+      "loss": 3.4089,
+      "step": 2155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7720511555671692,
+      "learning_rate": 9.929258465082738e-05,
+      "loss": 3.4862,
+      "step": 2160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8791617751121521,
+      "learning_rate": 9.928915459821027e-05,
+      "loss": 3.3811,
+      "step": 2165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7754148244857788,
+      "learning_rate": 9.928571630955865e-05,
+      "loss": 3.408,
+      "step": 2170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6992782354354858,
+      "learning_rate": 9.928226978544708e-05,
+      "loss": 3.3877,
+      "step": 2175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7253183126449585,
+      "learning_rate": 9.927881502645146e-05,
+      "loss": 3.337,
+      "step": 2180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7484565377235413,
+      "learning_rate": 9.927535203314904e-05,
+      "loss": 3.4433,
+      "step": 2185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7028684020042419,
+      "learning_rate": 9.927188080611849e-05,
+      "loss": 3.3066,
+      "step": 2190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7941984534263611,
+      "learning_rate": 9.926840134593984e-05,
+      "loss": 3.4578,
+      "step": 2195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7430591583251953,
+      "learning_rate": 9.92649136531945e-05,
+      "loss": 3.4135,
+      "step": 2200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7441457509994507,
+      "learning_rate": 9.926141772846525e-05,
+      "loss": 3.4251,
+      "step": 2205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7629740238189697,
+      "learning_rate": 9.925791357233623e-05,
+      "loss": 3.4154,
+      "step": 2210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7991272211074829,
+      "learning_rate": 9.925440118539298e-05,
+      "loss": 3.3413,
+      "step": 2215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7443709969520569,
+      "learning_rate": 9.925088056822241e-05,
+      "loss": 3.2458,
+      "step": 2220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6946473717689514,
+      "learning_rate": 9.924735172141281e-05,
+      "loss": 3.385,
+      "step": 2225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8029505014419556,
+      "learning_rate": 9.924381464555381e-05,
+      "loss": 3.339,
+      "step": 2230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8136982917785645,
+      "learning_rate": 9.924026934123647e-05,
+      "loss": 3.4344,
+      "step": 2235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7261207103729248,
+      "learning_rate": 9.92367158090532e-05,
+      "loss": 3.4305,
+      "step": 2240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7440099120140076,
+      "learning_rate": 9.923315404959775e-05,
+      "loss": 3.3642,
+      "step": 2245
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7527836561203003,
+      "learning_rate": 9.92295840634653e-05,
+      "loss": 3.4088,
+      "step": 2250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7050234079360962,
+      "learning_rate": 9.922600585125237e-05,
+      "loss": 3.4057,
+      "step": 2255
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7012151479721069,
+      "learning_rate": 9.922241941355688e-05,
+      "loss": 3.3754,
+      "step": 2260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7552962899208069,
+      "learning_rate": 9.921882475097811e-05,
+      "loss": 3.377,
+      "step": 2265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9649048447608948,
+      "learning_rate": 9.921522186411671e-05,
+      "loss": 3.3372,
+      "step": 2270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7405349612236023,
+      "learning_rate": 9.92116107535747e-05,
+      "loss": 3.2775,
+      "step": 2275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8034624457359314,
+      "learning_rate": 9.920799141995549e-05,
+      "loss": 3.4862,
+      "step": 2280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8035938143730164,
+      "learning_rate": 9.920436386386387e-05,
+      "loss": 3.3226,
+      "step": 2285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7336328029632568,
+      "learning_rate": 9.920072808590596e-05,
+      "loss": 3.4355,
+      "step": 2290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7250340580940247,
+      "learning_rate": 9.91970840866893e-05,
+      "loss": 3.3808,
+      "step": 2295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7266680598258972,
+      "learning_rate": 9.919343186682282e-05,
+      "loss": 3.3432,
+      "step": 2300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9230164885520935,
+      "learning_rate": 9.918977142691675e-05,
+      "loss": 3.3561,
+      "step": 2305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7719959020614624,
+      "learning_rate": 9.918610276758273e-05,
+      "loss": 3.3504,
+      "step": 2310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6781574487686157,
+      "learning_rate": 9.918242588943383e-05,
+      "loss": 3.3233,
+      "step": 2315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.704680323600769,
+      "learning_rate": 9.91787407930844e-05,
+      "loss": 3.3612,
+      "step": 2320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7366205453872681,
+      "learning_rate": 9.917504747915023e-05,
+      "loss": 3.4409,
+      "step": 2325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7304655909538269,
+      "learning_rate": 9.917134594824843e-05,
+      "loss": 3.3295,
+      "step": 2330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7136908769607544,
+      "learning_rate": 9.916763620099754e-05,
+      "loss": 3.2935,
+      "step": 2335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7670350074768066,
+      "learning_rate": 9.916391823801743e-05,
+      "loss": 3.3104,
+      "step": 2340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7044166922569275,
+      "learning_rate": 9.916019205992937e-05,
+      "loss": 3.4395,
+      "step": 2345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8116384148597717,
+      "learning_rate": 9.915645766735597e-05,
+      "loss": 3.4169,
+      "step": 2350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7117188572883606,
+      "learning_rate": 9.915271506092126e-05,
+      "loss": 3.3799,
+      "step": 2355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.732830286026001,
+      "learning_rate": 9.91489642412506e-05,
+      "loss": 3.4654,
+      "step": 2360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6808964014053345,
+      "learning_rate": 9.914520520897072e-05,
+      "loss": 3.2921,
+      "step": 2365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7430916428565979,
+      "learning_rate": 9.914143796470978e-05,
+      "loss": 3.3554,
+      "step": 2370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7446802854537964,
+      "learning_rate": 9.913766250909726e-05,
+      "loss": 3.3222,
+      "step": 2375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6790372133255005,
+      "learning_rate": 9.9133878842764e-05,
+      "loss": 3.3924,
+      "step": 2380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7498354911804199,
+      "learning_rate": 9.913008696634226e-05,
+      "loss": 3.2962,
+      "step": 2385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7527193427085876,
+      "learning_rate": 9.912628688046564e-05,
+      "loss": 3.273,
+      "step": 2390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7324744462966919,
+      "learning_rate": 9.912247858576914e-05,
+      "loss": 3.3401,
+      "step": 2395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6878606677055359,
+      "learning_rate": 9.911866208288906e-05,
+      "loss": 3.3445,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7372464537620544,
+      "learning_rate": 9.911483737246319e-05,
+      "loss": 3.3496,
+      "step": 2405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7022393941879272,
+      "learning_rate": 9.911100445513058e-05,
+      "loss": 3.4139,
+      "step": 2410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7131972908973694,
+      "learning_rate": 9.910716333153169e-05,
+      "loss": 3.3231,
+      "step": 2415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6918731331825256,
+      "learning_rate": 9.910331400230841e-05,
+      "loss": 3.3205,
+      "step": 2420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7217254638671875,
+      "learning_rate": 9.90994564681039e-05,
+      "loss": 3.3568,
+      "step": 2425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7195242047309875,
+      "learning_rate": 9.909559072956275e-05,
+      "loss": 3.4058,
+      "step": 2430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7110021710395813,
+      "learning_rate": 9.909171678733091e-05,
+      "loss": 3.3654,
+      "step": 2435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7080928087234497,
+      "learning_rate": 9.908783464205573e-05,
+      "loss": 3.2793,
+      "step": 2440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6980957984924316,
+      "learning_rate": 9.908394429438585e-05,
+      "loss": 3.301,
+      "step": 2445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.688827633857727,
+      "learning_rate": 9.908004574497139e-05,
+      "loss": 3.3625,
+      "step": 2450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6648275256156921,
+      "learning_rate": 9.907613899446372e-05,
+      "loss": 3.3314,
+      "step": 2455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6784229278564453,
+      "learning_rate": 9.90722240435157e-05,
+      "loss": 3.3703,
+      "step": 2460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6870729923248291,
+      "learning_rate": 9.906830089278148e-05,
+      "loss": 3.3604,
+      "step": 2465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7973740696907043,
+      "learning_rate": 9.906436954291661e-05,
+      "loss": 3.2797,
+      "step": 2470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7446298003196716,
+      "learning_rate": 9.906042999457798e-05,
+      "loss": 3.4091,
+      "step": 2475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7418840527534485,
+      "learning_rate": 9.905648224842391e-05,
+      "loss": 3.4256,
+      "step": 2480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7612913846969604,
+      "learning_rate": 9.905252630511405e-05,
+      "loss": 3.3844,
+      "step": 2485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7273263335227966,
+      "learning_rate": 9.904856216530938e-05,
+      "loss": 3.4116,
+      "step": 2490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7242324352264404,
+      "learning_rate": 9.904458982967235e-05,
+      "loss": 3.2981,
+      "step": 2495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7139917612075806,
+      "learning_rate": 9.904060929886668e-05,
+      "loss": 3.2878,
+      "step": 2500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6733770370483398,
+      "learning_rate": 9.903662057355751e-05,
+      "loss": 3.3887,
+      "step": 2505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7103335857391357,
+      "learning_rate": 9.903262365441137e-05,
+      "loss": 3.3138,
+      "step": 2510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7608098983764648,
+      "learning_rate": 9.90286185420961e-05,
+      "loss": 3.2506,
+      "step": 2515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.692154049873352,
+      "learning_rate": 9.902460523728094e-05,
+      "loss": 3.3113,
+      "step": 2520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7484728097915649,
+      "learning_rate": 9.902058374063651e-05,
+      "loss": 3.2914,
+      "step": 2525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.746238648891449,
+      "learning_rate": 9.90165540528348e-05,
+      "loss": 3.3557,
+      "step": 2530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7072189450263977,
+      "learning_rate": 9.901251617454913e-05,
+      "loss": 3.3233,
+      "step": 2535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7189834713935852,
+      "learning_rate": 9.900847010645422e-05,
+      "loss": 3.31,
+      "step": 2540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7310294508934021,
+      "learning_rate": 9.900441584922616e-05,
+      "loss": 3.3403,
+      "step": 2545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7556655406951904,
+      "learning_rate": 9.90003534035424e-05,
+      "loss": 3.4014,
+      "step": 2550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8121384978294373,
+      "learning_rate": 9.899628277008175e-05,
+      "loss": 3.365,
+      "step": 2555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6505420804023743,
+      "learning_rate": 9.899220394952441e-05,
+      "loss": 3.271,
+      "step": 2560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8079296350479126,
+      "learning_rate": 9.898811694255195e-05,
+      "loss": 3.4133,
+      "step": 2565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7313600182533264,
+      "learning_rate": 9.898402174984726e-05,
+      "loss": 3.3658,
+      "step": 2570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7520914077758789,
+      "learning_rate": 9.897991837209464e-05,
+      "loss": 3.2663,
+      "step": 2575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8224916458129883,
+      "learning_rate": 9.897580680997977e-05,
+      "loss": 3.2611,
+      "step": 2580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7937566041946411,
+      "learning_rate": 9.897168706418965e-05,
+      "loss": 3.3091,
+      "step": 2585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6934143304824829,
+      "learning_rate": 9.89675591354127e-05,
+      "loss": 3.2547,
+      "step": 2590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7341403961181641,
+      "learning_rate": 9.896342302433866e-05,
+      "loss": 3.2799,
+      "step": 2595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7666835188865662,
+      "learning_rate": 9.895927873165867e-05,
+      "loss": 3.3617,
+      "step": 2600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7247663736343384,
+      "learning_rate": 9.895512625806522e-05,
+      "loss": 3.2904,
+      "step": 2605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6859594583511353,
+      "learning_rate": 9.895096560425218e-05,
+      "loss": 3.3437,
+      "step": 2610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7189276814460754,
+      "learning_rate": 9.894679677091477e-05,
+      "loss": 3.3128,
+      "step": 2615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7319257855415344,
+      "learning_rate": 9.89426197587496e-05,
+      "loss": 3.3999,
+      "step": 2620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7438374161720276,
+      "learning_rate": 9.893843456845462e-05,
+      "loss": 3.3163,
+      "step": 2625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6758651733398438,
+      "learning_rate": 9.893424120072916e-05,
+      "loss": 3.3253,
+      "step": 2630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7067318558692932,
+      "learning_rate": 9.893003965627394e-05,
+      "loss": 3.3457,
+      "step": 2635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6741808652877808,
+      "learning_rate": 9.8925829935791e-05,
+      "loss": 3.2812,
+      "step": 2640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6706220507621765,
+      "learning_rate": 9.892161203998375e-05,
+      "loss": 3.343,
+      "step": 2645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7092440724372864,
+      "learning_rate": 9.891738596955703e-05,
+      "loss": 3.3328,
+      "step": 2650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7401556968688965,
+      "learning_rate": 9.891315172521696e-05,
+      "loss": 3.2183,
+      "step": 2655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7095944881439209,
+      "learning_rate": 9.890890930767111e-05,
+      "loss": 3.2984,
+      "step": 2660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7248217463493347,
+      "learning_rate": 9.890465871762832e-05,
+      "loss": 3.32,
+      "step": 2665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7243994474411011,
+      "learning_rate": 9.89003999557989e-05,
+      "loss": 3.312,
+      "step": 2670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6919440031051636,
+      "learning_rate": 9.889613302289443e-05,
+      "loss": 3.3569,
+      "step": 2675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7217360734939575,
+      "learning_rate": 9.889185791962792e-05,
+      "loss": 3.309,
+      "step": 2680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7249308228492737,
+      "learning_rate": 9.888757464671372e-05,
+      "loss": 3.3039,
+      "step": 2685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6634873151779175,
+      "learning_rate": 9.888328320486755e-05,
+      "loss": 3.3326,
+      "step": 2690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6858475208282471,
+      "learning_rate": 9.88789835948065e-05,
+      "loss": 3.2547,
+      "step": 2695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6670945882797241,
+      "learning_rate": 9.887467581724901e-05,
+      "loss": 3.2425,
+      "step": 2700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7184969782829285,
+      "learning_rate": 9.887035987291488e-05,
+      "loss": 3.384,
+      "step": 2705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7187245488166809,
+      "learning_rate": 9.88660357625253e-05,
+      "loss": 3.3667,
+      "step": 2710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7151724100112915,
+      "learning_rate": 9.886170348680284e-05,
+      "loss": 3.2858,
+      "step": 2715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6385703086853027,
+      "learning_rate": 9.885736304647139e-05,
+      "loss": 3.3021,
+      "step": 2720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6364243626594543,
+      "learning_rate": 9.885301444225619e-05,
+      "loss": 3.2837,
+      "step": 2725
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6725233793258667,
+      "learning_rate": 9.884865767488391e-05,
+      "loss": 3.261,
+      "step": 2730
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6616626381874084,
+      "learning_rate": 9.884429274508256e-05,
+      "loss": 3.4221,
+      "step": 2735
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6696721911430359,
+      "learning_rate": 9.883991965358146e-05,
+      "loss": 3.1439,
+      "step": 2740
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7198094725608826,
+      "learning_rate": 9.883553840111136e-05,
+      "loss": 3.3052,
+      "step": 2745
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7035394906997681,
+      "learning_rate": 9.883114898840437e-05,
+      "loss": 3.3631,
+      "step": 2750
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7049310207366943,
+      "learning_rate": 9.882675141619391e-05,
+      "loss": 3.3451,
+      "step": 2755
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6447438597679138,
+      "learning_rate": 9.882234568521482e-05,
+      "loss": 3.2856,
+      "step": 2760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6623960137367249,
+      "learning_rate": 9.881793179620328e-05,
+      "loss": 3.3314,
+      "step": 2765
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7173328399658203,
+      "learning_rate": 9.881350974989682e-05,
+      "loss": 3.2193,
+      "step": 2770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7123059034347534,
+      "learning_rate": 9.880907954703437e-05,
+      "loss": 3.2981,
+      "step": 2775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6808473467826843,
+      "learning_rate": 9.880464118835619e-05,
+      "loss": 3.3399,
+      "step": 2780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6751297116279602,
+      "learning_rate": 9.880019467460391e-05,
+      "loss": 3.3747,
+      "step": 2785
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6894731521606445,
+      "learning_rate": 9.879574000652052e-05,
+      "loss": 3.2083,
+      "step": 2790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.645969808101654,
+      "learning_rate": 9.87912771848504e-05,
+      "loss": 3.1975,
+      "step": 2795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7279164791107178,
+      "learning_rate": 9.878680621033925e-05,
+      "loss": 3.1256,
+      "step": 2800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6940102577209473,
+      "learning_rate": 9.878232708373415e-05,
+      "loss": 3.2655,
+      "step": 2805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7400137186050415,
+      "learning_rate": 9.877783980578357e-05,
+      "loss": 3.3473,
+      "step": 2810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7238442897796631,
+      "learning_rate": 9.877334437723728e-05,
+      "loss": 3.3062,
+      "step": 2815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6121941208839417,
+      "learning_rate": 9.87688407988465e-05,
+      "loss": 3.2928,
+      "step": 2820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6807535886764526,
+      "learning_rate": 9.876432907136372e-05,
+      "loss": 3.2904,
+      "step": 2825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6916024684906006,
+      "learning_rate": 9.875980919554287e-05,
+      "loss": 3.2693,
+      "step": 2830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.695193886756897,
+      "learning_rate": 9.875528117213915e-05,
+      "loss": 3.2031,
+      "step": 2835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7161367535591125,
+      "learning_rate": 9.875074500190922e-05,
+      "loss": 3.1934,
+      "step": 2840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6996201872825623,
+      "learning_rate": 9.874620068561104e-05,
+      "loss": 3.2371,
+      "step": 2845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6640839576721191,
+      "learning_rate": 9.874164822400396e-05,
+      "loss": 3.3108,
+      "step": 2850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6878724098205566,
+      "learning_rate": 9.873708761784866e-05,
+      "loss": 3.3985,
+      "step": 2855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7021346688270569,
+      "learning_rate": 9.873251886790722e-05,
+      "loss": 3.2611,
+      "step": 2860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6942867636680603,
+      "learning_rate": 9.872794197494307e-05,
+      "loss": 3.2419,
+      "step": 2865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.726254940032959,
+      "learning_rate": 9.872335693972094e-05,
+      "loss": 3.3666,
+      "step": 2870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7025614380836487,
+      "learning_rate": 9.871876376300702e-05,
+      "loss": 3.3,
+      "step": 2875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6845459938049316,
+      "learning_rate": 9.871416244556881e-05,
+      "loss": 3.1827,
+      "step": 2880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7301279306411743,
+      "learning_rate": 9.870955298817517e-05,
+      "loss": 3.3203,
+      "step": 2885
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7673966884613037,
+      "learning_rate": 9.87049353915963e-05,
+      "loss": 3.2354,
+      "step": 2890
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7274683713912964,
+      "learning_rate": 9.87003096566038e-05,
+      "loss": 3.4026,
+      "step": 2895
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6680096983909607,
+      "learning_rate": 9.869567578397062e-05,
+      "loss": 3.3123,
+      "step": 2900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6997368931770325,
+      "learning_rate": 9.869103377447105e-05,
+      "loss": 3.2531,
+      "step": 2905
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.674614667892456,
+      "learning_rate": 9.868638362888076e-05,
+      "loss": 3.3596,
+      "step": 2910
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6281630992889404,
+      "learning_rate": 9.868172534797678e-05,
+      "loss": 3.2692,
+      "step": 2915
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7349361777305603,
+      "learning_rate": 9.867705893253747e-05,
+      "loss": 3.2481,
+      "step": 2920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.692203938961029,
+      "learning_rate": 9.867238438334259e-05,
+      "loss": 3.2761,
+      "step": 2925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6389661431312561,
+      "learning_rate": 9.866770170117324e-05,
+      "loss": 3.2495,
+      "step": 2930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6601508259773254,
+      "learning_rate": 9.866301088681185e-05,
+      "loss": 3.3122,
+      "step": 2935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6330223679542542,
+      "learning_rate": 9.865831194104227e-05,
+      "loss": 3.2128,
+      "step": 2940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6822937726974487,
+      "learning_rate": 9.865360486464967e-05,
+      "loss": 3.3374,
+      "step": 2945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6770055890083313,
+      "learning_rate": 9.864888965842056e-05,
+      "loss": 3.246,
+      "step": 2950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6391374468803406,
+      "learning_rate": 9.864416632314287e-05,
+      "loss": 3.254,
+      "step": 2955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6947242021560669,
+      "learning_rate": 9.863943485960583e-05,
+      "loss": 3.301,
+      "step": 2960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7678213119506836,
+      "learning_rate": 9.863469526860005e-05,
+      "loss": 3.2587,
+      "step": 2965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6925395131111145,
+      "learning_rate": 9.862994755091752e-05,
+      "loss": 3.1832,
+      "step": 2970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6937146782875061,
+      "learning_rate": 9.862519170735154e-05,
+      "loss": 3.2617,
+      "step": 2975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6303795576095581,
+      "learning_rate": 9.86204277386968e-05,
+      "loss": 3.2288,
+      "step": 2980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6420886516571045,
+      "learning_rate": 9.861565564574935e-05,
+      "loss": 3.2738,
+      "step": 2985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7220498323440552,
+      "learning_rate": 9.861087542930658e-05,
+      "loss": 3.1792,
+      "step": 2990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8265786170959473,
+      "learning_rate": 9.860608709016726e-05,
+      "loss": 3.2049,
+      "step": 2995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6900822520256042,
+      "learning_rate": 9.860129062913147e-05,
+      "loss": 3.2084,
+      "step": 3000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6509965062141418,
+      "learning_rate": 9.859648604700072e-05,
+      "loss": 3.2476,
+      "step": 3005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8656744360923767,
+      "learning_rate": 9.859167334457781e-05,
+      "loss": 3.2641,
+      "step": 3010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6936625242233276,
+      "learning_rate": 9.858685252266696e-05,
+      "loss": 3.322,
+      "step": 3015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6918776631355286,
+      "learning_rate": 9.858202358207367e-05,
+      "loss": 3.139,
+      "step": 3020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.707094669342041,
+      "learning_rate": 9.857718652360486e-05,
+      "loss": 3.2844,
+      "step": 3025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7178537249565125,
+      "learning_rate": 9.857234134806878e-05,
+      "loss": 3.169,
+      "step": 3030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6739908456802368,
+      "learning_rate": 9.856748805627503e-05,
+      "loss": 3.25,
+      "step": 3035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6922372579574585,
+      "learning_rate": 9.856262664903462e-05,
+      "loss": 3.3107,
+      "step": 3040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6458384990692139,
+      "learning_rate": 9.855775712715981e-05,
+      "loss": 3.2646,
+      "step": 3045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6937810182571411,
+      "learning_rate": 9.855287949146432e-05,
+      "loss": 3.1089,
+      "step": 3050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.708445131778717,
+      "learning_rate": 9.854799374276318e-05,
+      "loss": 3.2554,
+      "step": 3055
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6433005332946777,
+      "learning_rate": 9.854309988187277e-05,
+      "loss": 3.2304,
+      "step": 3060
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6358730792999268,
+      "learning_rate": 9.853819790961086e-05,
+      "loss": 3.1303,
+      "step": 3065
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5910662412643433,
+      "learning_rate": 9.853328782679652e-05,
+      "loss": 3.1732,
+      "step": 3070
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6611315608024597,
+      "learning_rate": 9.852836963425021e-05,
+      "loss": 3.2462,
+      "step": 3075
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6739450097084045,
+      "learning_rate": 9.852344333279375e-05,
+      "loss": 3.1404,
+      "step": 3080
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6492950916290283,
+      "learning_rate": 9.851850892325032e-05,
+      "loss": 3.3213,
+      "step": 3085
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.668168842792511,
+      "learning_rate": 9.851356640644443e-05,
+      "loss": 3.1683,
+      "step": 3090
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6990168690681458,
+      "learning_rate": 9.850861578320196e-05,
+      "loss": 3.2769,
+      "step": 3095
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7037989497184753,
+      "learning_rate": 9.850365705435013e-05,
+      "loss": 3.2537,
+      "step": 3100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.665246844291687,
+      "learning_rate": 9.849869022071753e-05,
+      "loss": 3.1526,
+      "step": 3105
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6388899087905884,
+      "learning_rate": 9.849371528313411e-05,
+      "loss": 3.2363,
+      "step": 3110
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7189300656318665,
+      "learning_rate": 9.848873224243114e-05,
+      "loss": 3.0999,
+      "step": 3115
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7235907912254333,
+      "learning_rate": 9.848374109944128e-05,
+      "loss": 3.2508,
+      "step": 3120
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6510244011878967,
+      "learning_rate": 9.847874185499855e-05,
+      "loss": 3.2461,
+      "step": 3125
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6800297498703003,
+      "learning_rate": 9.847373450993828e-05,
+      "loss": 3.2225,
+      "step": 3130
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7299479246139526,
+      "learning_rate": 9.846871906509719e-05,
+      "loss": 3.2305,
+      "step": 3135
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6950435042381287,
+      "learning_rate": 9.846369552131334e-05,
+      "loss": 3.232,
+      "step": 3140
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6829844117164612,
+      "learning_rate": 9.845866387942614e-05,
+      "loss": 3.2625,
+      "step": 3145
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6509039998054504,
+      "learning_rate": 9.845362414027637e-05,
+      "loss": 3.1606,
+      "step": 3150
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6994816660881042,
+      "learning_rate": 9.844857630470614e-05,
+      "loss": 3.2491,
+      "step": 3155
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7125581502914429,
+      "learning_rate": 9.844352037355893e-05,
+      "loss": 3.2025,
+      "step": 3160
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6906768679618835,
+      "learning_rate": 9.843845634767957e-05,
+      "loss": 3.2156,
+      "step": 3165
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6400070190429688,
+      "learning_rate": 9.843338422791424e-05,
+      "loss": 3.3341,
+      "step": 3170
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6522157192230225,
+      "learning_rate": 9.842830401511048e-05,
+      "loss": 3.2248,
+      "step": 3175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7691868543624878,
+      "learning_rate": 9.842321571011714e-05,
+      "loss": 3.2055,
+      "step": 3180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6880500316619873,
+      "learning_rate": 9.841811931378448e-05,
+      "loss": 3.2013,
+      "step": 3185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6526750326156616,
+      "learning_rate": 9.841301482696413e-05,
+      "loss": 3.2802,
+      "step": 3190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6103237271308899,
+      "learning_rate": 9.840790225050896e-05,
+      "loss": 3.1417,
+      "step": 3195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6638354659080505,
+      "learning_rate": 9.840278158527328e-05,
+      "loss": 3.1694,
+      "step": 3200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6687728762626648,
+      "learning_rate": 9.839765283211278e-05,
+      "loss": 3.1514,
+      "step": 3205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6674135327339172,
+      "learning_rate": 9.83925159918844e-05,
+      "loss": 3.1811,
+      "step": 3210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6446532607078552,
+      "learning_rate": 9.838737106544651e-05,
+      "loss": 3.1256,
+      "step": 3215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6971847414970398,
+      "learning_rate": 9.838221805365881e-05,
+      "loss": 3.2874,
+      "step": 3220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1076183319091797,
+      "learning_rate": 9.837705695738236e-05,
+      "loss": 3.244,
+      "step": 3225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6273555755615234,
+      "learning_rate": 9.837188777747954e-05,
+      "loss": 3.283,
+      "step": 3230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6873939633369446,
+      "learning_rate": 9.836671051481411e-05,
+      "loss": 3.2541,
+      "step": 3235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6932997107505798,
+      "learning_rate": 9.836152517025116e-05,
+      "loss": 3.3517,
+      "step": 3240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6784059405326843,
+      "learning_rate": 9.835633174465715e-05,
+      "loss": 3.2127,
+      "step": 3245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6696476340293884,
+      "learning_rate": 9.83511302388999e-05,
+      "loss": 3.1993,
+      "step": 3250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6204133033752441,
+      "learning_rate": 9.834592065384853e-05,
+      "loss": 3.2364,
+      "step": 3255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6443333625793457,
+      "learning_rate": 9.834070299037357e-05,
+      "loss": 3.2183,
+      "step": 3260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7351436018943787,
+      "learning_rate": 9.833547724934687e-05,
+      "loss": 3.1399,
+      "step": 3265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6424255967140198,
+      "learning_rate": 9.833024343164159e-05,
+      "loss": 3.2293,
+      "step": 3270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8383104205131531,
+      "learning_rate": 9.832500153813234e-05,
+      "loss": 3.2053,
+      "step": 3275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.683628261089325,
+      "learning_rate": 9.831975156969498e-05,
+      "loss": 3.1897,
+      "step": 3280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7020851969718933,
+      "learning_rate": 9.831449352720677e-05,
+      "loss": 3.2201,
+      "step": 3285
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7129231095314026,
+      "learning_rate": 9.830922741154631e-05,
+      "loss": 3.2146,
+      "step": 3290
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7093530297279358,
+      "learning_rate": 9.830395322359357e-05,
+      "loss": 3.164,
+      "step": 3295
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6937132477760315,
+      "learning_rate": 9.829867096422983e-05,
+      "loss": 3.2009,
+      "step": 3300
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6645590662956238,
+      "learning_rate": 9.829338063433773e-05,
+      "loss": 3.1923,
+      "step": 3305
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8908779621124268,
+      "learning_rate": 9.828808223480126e-05,
+      "loss": 3.2025,
+      "step": 3310
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7161830067634583,
+      "learning_rate": 9.828277576650578e-05,
+      "loss": 3.2582,
+      "step": 3315
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6130056977272034,
+      "learning_rate": 9.827746123033796e-05,
+      "loss": 3.3145,
+      "step": 3320
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6641003489494324,
+      "learning_rate": 9.827213862718587e-05,
+      "loss": 3.1725,
+      "step": 3325
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6423278450965881,
+      "learning_rate": 9.826680795793887e-05,
+      "loss": 3.2468,
+      "step": 3330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7068758010864258,
+      "learning_rate": 9.826146922348773e-05,
+      "loss": 3.2208,
+      "step": 3335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6306912302970886,
+      "learning_rate": 9.825612242472448e-05,
+      "loss": 3.1973,
+      "step": 3340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6844825148582458,
+      "learning_rate": 9.82507675625426e-05,
+      "loss": 3.2115,
+      "step": 3345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7078993916511536,
+      "learning_rate": 9.824540463783682e-05,
+      "loss": 3.1773,
+      "step": 3350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6761540174484253,
+      "learning_rate": 9.824003365150329e-05,
+      "loss": 3.2313,
+      "step": 3355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6909316182136536,
+      "learning_rate": 9.82346546044395e-05,
+      "loss": 3.2937,
+      "step": 3360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6783762574195862,
+      "learning_rate": 9.822926749754424e-05,
+      "loss": 3.2444,
+      "step": 3365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6544792056083679,
+      "learning_rate": 9.822387233171771e-05,
+      "loss": 3.1365,
+      "step": 3370
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.638312816619873,
+      "learning_rate": 9.821846910786136e-05,
+      "loss": 3.1913,
+      "step": 3375
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6626136898994446,
+      "learning_rate": 9.821305782687812e-05,
+      "loss": 3.188,
+      "step": 3380
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6832327246665955,
+      "learning_rate": 9.820763848967215e-05,
+      "loss": 3.259,
+      "step": 3385
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7239116430282593,
+      "learning_rate": 9.820221109714902e-05,
+      "loss": 3.1921,
+      "step": 3390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6268677711486816,
+      "learning_rate": 9.81967756502156e-05,
+      "loss": 3.231,
+      "step": 3395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6520034670829773,
+      "learning_rate": 9.819133214978017e-05,
+      "loss": 3.1732,
+      "step": 3400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7048172950744629,
+      "learning_rate": 9.818588059675229e-05,
+      "loss": 3.1825,
+      "step": 3405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6834260821342468,
+      "learning_rate": 9.818042099204292e-05,
+      "loss": 3.2314,
+      "step": 3410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6415216326713562,
+      "learning_rate": 9.817495333656433e-05,
+      "loss": 3.1701,
+      "step": 3415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6577292084693909,
+      "learning_rate": 9.816947763123013e-05,
+      "loss": 3.202,
+      "step": 3420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6985458731651306,
+      "learning_rate": 9.816399387695529e-05,
+      "loss": 3.2045,
+      "step": 3425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6625770926475525,
+      "learning_rate": 9.815850207465616e-05,
+      "loss": 3.2172,
+      "step": 3430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6343966126441956,
+      "learning_rate": 9.815300222525035e-05,
+      "loss": 3.1258,
+      "step": 3435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6446753144264221,
+      "learning_rate": 9.814749432965691e-05,
+      "loss": 3.1518,
+      "step": 3440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6606610417366028,
+      "learning_rate": 9.814197838879615e-05,
+      "loss": 3.1891,
+      "step": 3445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6140626072883606,
+      "learning_rate": 9.813645440358981e-05,
+      "loss": 3.1905,
+      "step": 3450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6814715266227722,
+      "learning_rate": 9.813092237496087e-05,
+      "loss": 3.1183,
+      "step": 3455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7064401507377625,
+      "learning_rate": 9.812538230383376e-05,
+      "loss": 3.1629,
+      "step": 3460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7968814373016357,
+      "learning_rate": 9.811983419113418e-05,
+      "loss": 3.2096,
+      "step": 3465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7050393223762512,
+      "learning_rate": 9.81142780377892e-05,
+      "loss": 3.2544,
+      "step": 3470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6132931709289551,
+      "learning_rate": 9.810871384472724e-05,
+      "loss": 3.238,
+      "step": 3475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6071353554725647,
+      "learning_rate": 9.810314161287806e-05,
+      "loss": 3.1788,
+      "step": 3480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6552751660346985,
+      "learning_rate": 9.809756134317276e-05,
+      "loss": 3.2166,
+      "step": 3485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6475270390510559,
+      "learning_rate": 9.809197303654376e-05,
+      "loss": 3.2188,
+      "step": 3490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6444715857505798,
+      "learning_rate": 9.808637669392487e-05,
+      "loss": 3.2661,
+      "step": 3495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6448482275009155,
+      "learning_rate": 9.808077231625119e-05,
+      "loss": 3.1941,
+      "step": 3500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.625289261341095,
+      "learning_rate": 9.807515990445922e-05,
+      "loss": 3.2009,
+      "step": 3505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6574786305427551,
+      "learning_rate": 9.806953945948676e-05,
+      "loss": 3.2079,
+      "step": 3510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7389100790023804,
+      "learning_rate": 9.806391098227297e-05,
+      "loss": 3.124,
+      "step": 3515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6082479357719421,
+      "learning_rate": 9.805827447375835e-05,
+      "loss": 3.092,
+      "step": 3520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6143671274185181,
+      "learning_rate": 9.805262993488473e-05,
+      "loss": 3.1652,
+      "step": 3525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6696958541870117,
+      "learning_rate": 9.804697736659532e-05,
+      "loss": 3.2125,
+      "step": 3530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6293437480926514,
+      "learning_rate": 9.804131676983458e-05,
+      "loss": 3.2424,
+      "step": 3535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6571451425552368,
+      "learning_rate": 9.803564814554844e-05,
+      "loss": 3.0912,
+      "step": 3540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6313217282295227,
+      "learning_rate": 9.802997149468408e-05,
+      "loss": 3.1664,
+      "step": 3545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6582979559898376,
+      "learning_rate": 9.802428681819003e-05,
+      "loss": 3.117,
+      "step": 3550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6158257722854614,
+      "learning_rate": 9.801859411701621e-05,
+      "loss": 3.2262,
+      "step": 3555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6781128644943237,
+      "learning_rate": 9.801289339211384e-05,
+      "loss": 3.2516,
+      "step": 3560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6261134147644043,
+      "learning_rate": 9.800718464443547e-05,
+      "loss": 3.2406,
+      "step": 3565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6096157431602478,
+      "learning_rate": 9.800146787493503e-05,
+      "loss": 3.1476,
+      "step": 3570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.668876051902771,
+      "learning_rate": 9.799574308456775e-05,
+      "loss": 3.1169,
+      "step": 3575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7436100840568542,
+      "learning_rate": 9.799001027429026e-05,
+      "loss": 3.1967,
+      "step": 3580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6563398241996765,
+      "learning_rate": 9.798426944506048e-05,
+      "loss": 3.134,
+      "step": 3585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.63784259557724,
+      "learning_rate": 9.797852059783766e-05,
+      "loss": 3.1217,
+      "step": 3590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6937146186828613,
+      "learning_rate": 9.79727637335824e-05,
+      "loss": 3.1341,
+      "step": 3595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6717680096626282,
+      "learning_rate": 9.79669988532567e-05,
+      "loss": 3.1444,
+      "step": 3600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6268311738967896,
+      "learning_rate": 9.79612259578238e-05,
+      "loss": 3.1094,
+      "step": 3605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6609651446342468,
+      "learning_rate": 9.795544504824836e-05,
+      "loss": 3.1905,
+      "step": 3610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6471235156059265,
+      "learning_rate": 9.794965612549635e-05,
+      "loss": 3.144,
+      "step": 3615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6305959820747375,
+      "learning_rate": 9.794385919053506e-05,
+      "loss": 3.1493,
+      "step": 3620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7570910453796387,
+      "learning_rate": 9.793805424433313e-05,
+      "loss": 3.2258,
+      "step": 3625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7368032336235046,
+      "learning_rate": 9.793224128786058e-05,
+      "loss": 3.2101,
+      "step": 3630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6152755618095398,
+      "learning_rate": 9.79264203220887e-05,
+      "loss": 3.1887,
+      "step": 3635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6257781386375427,
+      "learning_rate": 9.792059134799016e-05,
+      "loss": 3.0608,
+      "step": 3640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6451605558395386,
+      "learning_rate": 9.791475436653896e-05,
+      "loss": 3.0797,
+      "step": 3645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6791229844093323,
+      "learning_rate": 9.790890937871046e-05,
+      "loss": 3.1877,
+      "step": 3650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6172186732292175,
+      "learning_rate": 9.79030563854813e-05,
+      "loss": 3.1778,
+      "step": 3655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6723390817642212,
+      "learning_rate": 9.789719538782952e-05,
+      "loss": 3.1698,
+      "step": 3660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7430850267410278,
+      "learning_rate": 9.789132638673445e-05,
+      "loss": 3.1447,
+      "step": 3665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6657228469848633,
+      "learning_rate": 9.78854493831768e-05,
+      "loss": 3.1972,
+      "step": 3670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.620653510093689,
+      "learning_rate": 9.787956437813859e-05,
+      "loss": 3.1099,
+      "step": 3675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.621303379535675,
+      "learning_rate": 9.787367137260317e-05,
+      "loss": 3.103,
+      "step": 3680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6761735677719116,
+      "learning_rate": 9.786777036755525e-05,
+      "loss": 3.1539,
+      "step": 3685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.613307535648346,
+      "learning_rate": 9.786186136398085e-05,
+      "loss": 3.1688,
+      "step": 3690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6201168298721313,
+      "learning_rate": 9.785594436286738e-05,
+      "loss": 3.1426,
+      "step": 3695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6414672136306763,
+      "learning_rate": 9.78500193652035e-05,
+      "loss": 3.1375,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6555194854736328,
+      "learning_rate": 9.784408637197929e-05,
+      "loss": 3.1421,
+      "step": 3705
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6409485340118408,
+      "learning_rate": 9.783814538418612e-05,
+      "loss": 2.9721,
+      "step": 3710
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5833931565284729,
+      "learning_rate": 9.783219640281671e-05,
+      "loss": 3.0989,
+      "step": 3715
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6917932033538818,
+      "learning_rate": 9.78262394288651e-05,
+      "loss": 3.1739,
+      "step": 3720
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6427809000015259,
+      "learning_rate": 9.782027446332671e-05,
+      "loss": 3.2874,
+      "step": 3725
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0109460353851318,
+      "learning_rate": 9.781430150719822e-05,
+      "loss": 3.1655,
+      "step": 3730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6736050844192505,
+      "learning_rate": 9.780832056147772e-05,
+      "loss": 3.1941,
+      "step": 3735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6647701859474182,
+      "learning_rate": 9.78023316271646e-05,
+      "loss": 3.0777,
+      "step": 3740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6347657442092896,
+      "learning_rate": 9.779633470525957e-05,
+      "loss": 3.0666,
+      "step": 3745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6120061278343201,
+      "learning_rate": 9.77903297967647e-05,
+      "loss": 3.2301,
+      "step": 3750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7199821472167969,
+      "learning_rate": 9.778431690268341e-05,
+      "loss": 3.067,
+      "step": 3755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6637604236602783,
+      "learning_rate": 9.777829602402043e-05,
+      "loss": 3.0341,
+      "step": 3760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6265471577644348,
+      "learning_rate": 9.777226716178178e-05,
+      "loss": 3.1735,
+      "step": 3765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6056855916976929,
+      "learning_rate": 9.776623031697492e-05,
+      "loss": 3.1536,
+      "step": 3770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6232392191886902,
+      "learning_rate": 9.776018549060855e-05,
+      "loss": 3.1377,
+      "step": 3775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6995528936386108,
+      "learning_rate": 9.775413268369274e-05,
+      "loss": 3.1394,
+      "step": 3780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5970630645751953,
+      "learning_rate": 9.77480718972389e-05,
+      "loss": 3.077,
+      "step": 3785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6526083946228027,
+      "learning_rate": 9.774200313225976e-05,
+      "loss": 3.0191,
+      "step": 3790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5921361446380615,
+      "learning_rate": 9.773592638976939e-05,
+      "loss": 3.1994,
+      "step": 3795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6327505111694336,
+      "learning_rate": 9.77298416707832e-05,
+      "loss": 3.1781,
+      "step": 3800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6508796215057373,
+      "learning_rate": 9.772374897631791e-05,
+      "loss": 3.1881,
+      "step": 3805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6282609701156616,
+      "learning_rate": 9.771764830739159e-05,
+      "loss": 3.1379,
+      "step": 3810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6186907887458801,
+      "learning_rate": 9.771153966502365e-05,
+      "loss": 3.1894,
+      "step": 3815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6335012912750244,
+      "learning_rate": 9.77054230502348e-05,
+      "loss": 3.2136,
+      "step": 3820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6488093137741089,
+      "learning_rate": 9.769929846404711e-05,
+      "loss": 3.0878,
+      "step": 3825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7046669721603394,
+      "learning_rate": 9.769316590748401e-05,
+      "loss": 3.1325,
+      "step": 3830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3823482990264893,
+      "learning_rate": 9.768702538157018e-05,
+      "loss": 3.1018,
+      "step": 3835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7078198194503784,
+      "learning_rate": 9.76808768873317e-05,
+      "loss": 3.1249,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6842246055603027,
+      "learning_rate": 9.767472042579595e-05,
+      "loss": 3.1744,
+      "step": 3845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6205492615699768,
+      "learning_rate": 9.766855599799168e-05,
+      "loss": 3.1245,
+      "step": 3850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6463498473167419,
+      "learning_rate": 9.76623836049489e-05,
+      "loss": 3.2509,
+      "step": 3855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6388393044471741,
+      "learning_rate": 9.765620324769903e-05,
+      "loss": 3.1395,
+      "step": 3860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6021741032600403,
+      "learning_rate": 9.765001492727476e-05,
+      "loss": 3.1799,
+      "step": 3865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6463895440101624,
+      "learning_rate": 9.764381864471017e-05,
+      "loss": 3.1437,
+      "step": 3870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6303393840789795,
+      "learning_rate": 9.76376144010406e-05,
+      "loss": 3.0925,
+      "step": 3875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6446599364280701,
+      "learning_rate": 9.763140219730278e-05,
+      "loss": 3.1631,
+      "step": 3880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6532890200614929,
+      "learning_rate": 9.762518203453472e-05,
+      "loss": 3.1175,
+      "step": 3885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6147683262825012,
+      "learning_rate": 9.761895391377581e-05,
+      "loss": 3.117,
+      "step": 3890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6119741797447205,
+      "learning_rate": 9.761271783606674e-05,
+      "loss": 3.1498,
+      "step": 3895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6074857115745544,
+      "learning_rate": 9.760647380244952e-05,
+      "loss": 3.1026,
+      "step": 3900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6343613266944885,
+      "learning_rate": 9.760022181396753e-05,
+      "loss": 3.0389,
+      "step": 3905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6092501282691956,
+      "learning_rate": 9.759396187166543e-05,
+      "loss": 3.0966,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5676233768463135,
+      "learning_rate": 9.758769397658926e-05,
+      "loss": 3.1668,
+      "step": 3915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6449651122093201,
+      "learning_rate": 9.758141812978634e-05,
+      "loss": 3.0097,
+      "step": 3920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5896978378295898,
+      "learning_rate": 9.757513433230534e-05,
+      "loss": 3.1196,
+      "step": 3925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6278130412101746,
+      "learning_rate": 9.756884258519627e-05,
+      "loss": 3.095,
+      "step": 3930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6156342029571533,
+      "learning_rate": 9.756254288951046e-05,
+      "loss": 3.1703,
+      "step": 3935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6014965772628784,
+      "learning_rate": 9.755623524630055e-05,
+      "loss": 3.0652,
+      "step": 3940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.131996989250183,
+      "learning_rate": 9.754991965662055e-05,
+      "loss": 3.1273,
+      "step": 3945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6331797242164612,
+      "learning_rate": 9.754359612152574e-05,
+      "loss": 3.1662,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6510134935379028,
+      "learning_rate": 9.753726464207278e-05,
+      "loss": 3.1125,
+      "step": 3955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5750917792320251,
+      "learning_rate": 9.753092521931963e-05,
+      "loss": 3.1195,
+      "step": 3960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.654827356338501,
+      "learning_rate": 9.752457785432558e-05,
+      "loss": 3.1262,
+      "step": 3965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.654078483581543,
+      "learning_rate": 9.751822254815126e-05,
+      "loss": 3.1982,
+      "step": 3970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6338599920272827,
+      "learning_rate": 9.751185930185861e-05,
+      "loss": 3.1285,
+      "step": 3975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6408002972602844,
+      "learning_rate": 9.750548811651091e-05,
+      "loss": 3.0604,
+      "step": 3980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6109910011291504,
+      "learning_rate": 9.749910899317276e-05,
+      "loss": 3.0275,
+      "step": 3985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6385597586631775,
+      "learning_rate": 9.74927219329101e-05,
+      "loss": 3.11,
+      "step": 3990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6503455638885498,
+      "learning_rate": 9.748632693679016e-05,
+      "loss": 3.108,
+      "step": 3995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5951246619224548,
+      "learning_rate": 9.747992400588153e-05,
+      "loss": 3.0598,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.585462749004364,
+      "learning_rate": 9.747351314125413e-05,
+      "loss": 3.1363,
+      "step": 4005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6159416437149048,
+      "learning_rate": 9.746709434397916e-05,
+      "loss": 3.1042,
+      "step": 4010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6531094908714294,
+      "learning_rate": 9.746066761512921e-05,
+      "loss": 3.12,
+      "step": 4015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6289142966270447,
+      "learning_rate": 9.745423295577815e-05,
+      "loss": 3.1344,
+      "step": 4020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6866773366928101,
+      "learning_rate": 9.744779036700118e-05,
+      "loss": 3.2407,
+      "step": 4025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6542808413505554,
+      "learning_rate": 9.744133984987485e-05,
+      "loss": 3.1883,
+      "step": 4030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6818244457244873,
+      "learning_rate": 9.743488140547701e-05,
+      "loss": 3.1447,
+      "step": 4035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6424097418785095,
+      "learning_rate": 9.742841503488682e-05,
+      "loss": 3.1421,
+      "step": 4040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6345065236091614,
+      "learning_rate": 9.742194073918482e-05,
+      "loss": 3.144,
+      "step": 4045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.645403265953064,
+      "learning_rate": 9.741545851945282e-05,
+      "loss": 3.1778,
+      "step": 4050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.602630078792572,
+      "learning_rate": 9.740896837677398e-05,
+      "loss": 3.0316,
+      "step": 4055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6143417358398438,
+      "learning_rate": 9.740247031223279e-05,
+      "loss": 3.138,
+      "step": 4060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.643565833568573,
+      "learning_rate": 9.739596432691502e-05,
+      "loss": 3.0494,
+      "step": 4065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5763258934020996,
+      "learning_rate": 9.738945042190785e-05,
+      "loss": 3.1952,
+      "step": 4070
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6679778695106506,
+      "learning_rate": 9.738292859829968e-05,
+      "loss": 3.0807,
+      "step": 4075
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6908403635025024,
+      "learning_rate": 9.73763988571803e-05,
+      "loss": 3.0403,
+      "step": 4080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6412582993507385,
+      "learning_rate": 9.73698611996408e-05,
+      "loss": 3.1161,
+      "step": 4085
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5962137579917908,
+      "learning_rate": 9.736331562677361e-05,
+      "loss": 3.0424,
+      "step": 4090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6492986083030701,
+      "learning_rate": 9.735676213967247e-05,
+      "loss": 3.0848,
+      "step": 4095
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6263057589530945,
+      "learning_rate": 9.735020073943244e-05,
+      "loss": 3.0304,
+      "step": 4100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6632346510887146,
+      "learning_rate": 9.734363142714988e-05,
+      "loss": 3.1239,
+      "step": 4105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6600107550621033,
+      "learning_rate": 9.733705420392254e-05,
+      "loss": 3.0977,
+      "step": 4110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6258062124252319,
+      "learning_rate": 9.733046907084944e-05,
+      "loss": 3.0854,
+      "step": 4115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6360071897506714,
+      "learning_rate": 9.732387602903091e-05,
+      "loss": 3.1508,
+      "step": 4120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5659964084625244,
+      "learning_rate": 9.731727507956862e-05,
+      "loss": 3.0866,
+      "step": 4125
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.73382169008255,
+      "learning_rate": 9.73106662235656e-05,
+      "loss": 3.0781,
+      "step": 4130
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6236119866371155,
+      "learning_rate": 9.730404946212614e-05,
+      "loss": 3.099,
+      "step": 4135
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6347524523735046,
+      "learning_rate": 9.729742479635589e-05,
+      "loss": 3.0505,
+      "step": 4140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5888487696647644,
+      "learning_rate": 9.729079222736177e-05,
+      "loss": 3.1476,
+      "step": 4145
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6832713484764099,
+      "learning_rate": 9.728415175625211e-05,
+      "loss": 3.1666,
+      "step": 4150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5641723275184631,
+      "learning_rate": 9.727750338413647e-05,
+      "loss": 3.0604,
+      "step": 4155
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6226898431777954,
+      "learning_rate": 9.727084711212579e-05,
+      "loss": 3.0847,
+      "step": 4160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6184467673301697,
+      "learning_rate": 9.726418294133229e-05,
+      "loss": 3.0814,
+      "step": 4165
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6341397762298584,
+      "learning_rate": 9.725751087286954e-05,
+      "loss": 3.0311,
+      "step": 4170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6252658367156982,
+      "learning_rate": 9.725083090785242e-05,
+      "loss": 3.0592,
+      "step": 4175
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6152575016021729,
+      "learning_rate": 9.724414304739712e-05,
+      "loss": 3.1462,
+      "step": 4180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6440061330795288,
+      "learning_rate": 9.723744729262116e-05,
+      "loss": 3.1123,
+      "step": 4185
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.645527720451355,
+      "learning_rate": 9.723074364464337e-05,
+      "loss": 3.146,
+      "step": 4190
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6510219573974609,
+      "learning_rate": 9.722403210458393e-05,
+      "loss": 3.1032,
+      "step": 4195
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6001125574111938,
+      "learning_rate": 9.721731267356426e-05,
+      "loss": 3.0971,
+      "step": 4200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6256840825080872,
+      "learning_rate": 9.721058535270722e-05,
+      "loss": 3.0982,
+      "step": 4205
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6616936326026917,
+      "learning_rate": 9.720385014313686e-05,
+      "loss": 3.0733,
+      "step": 4210
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6059109568595886,
+      "learning_rate": 9.719710704597865e-05,
+      "loss": 2.9899,
+      "step": 4215
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0265027284622192,
+      "learning_rate": 9.719035606235931e-05,
+      "loss": 3.0437,
+      "step": 4220
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.632525622844696,
+      "learning_rate": 9.718359719340693e-05,
+      "loss": 3.1822,
+      "step": 4225
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6171260476112366,
+      "learning_rate": 9.717683044025086e-05,
+      "loss": 3.1204,
+      "step": 4230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5578847527503967,
+      "learning_rate": 9.717005580402184e-05,
+      "loss": 3.035,
+      "step": 4235
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6309463977813721,
+      "learning_rate": 9.716327328585183e-05,
+      "loss": 3.1622,
+      "step": 4240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6125911474227905,
+      "learning_rate": 9.715648288687421e-05,
+      "loss": 3.0241,
+      "step": 4245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8214325308799744,
+      "learning_rate": 9.714968460822363e-05,
+      "loss": 3.1525,
+      "step": 4250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5997372269630432,
+      "learning_rate": 9.714287845103602e-05,
+      "loss": 3.0268,
+      "step": 4255
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6563445329666138,
+      "learning_rate": 9.71360644164487e-05,
+      "loss": 3.1122,
+      "step": 4260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5990898609161377,
+      "learning_rate": 9.712924250560025e-05,
+      "loss": 3.0632,
+      "step": 4265
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6489446759223938,
+      "learning_rate": 9.712241271963059e-05,
+      "loss": 3.1445,
+      "step": 4270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6706987619400024,
+      "learning_rate": 9.711557505968097e-05,
+      "loss": 3.0258,
+      "step": 4275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6365798115730286,
+      "learning_rate": 9.710872952689391e-05,
+      "loss": 3.0447,
+      "step": 4280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5711104869842529,
+      "learning_rate": 9.71018761224133e-05,
+      "loss": 3.0913,
+      "step": 4285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6523838043212891,
+      "learning_rate": 9.709501484738429e-05,
+      "loss": 3.0405,
+      "step": 4290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5875277519226074,
+      "learning_rate": 9.70881457029534e-05,
+      "loss": 3.1342,
+      "step": 4295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.606128454208374,
+      "learning_rate": 9.708126869026841e-05,
+      "loss": 3.0624,
+      "step": 4300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5675714015960693,
+      "learning_rate": 9.707438381047846e-05,
+      "loss": 3.0668,
+      "step": 4305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5932487845420837,
+      "learning_rate": 9.7067491064734e-05,
+      "loss": 3.035,
+      "step": 4310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5474107265472412,
+      "learning_rate": 9.706059045418676e-05,
+      "loss": 3.0675,
+      "step": 4315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.609283447265625,
+      "learning_rate": 9.705368197998982e-05,
+      "loss": 3.0989,
+      "step": 4320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6039413213729858,
+      "learning_rate": 9.704676564329756e-05,
+      "loss": 3.0349,
+      "step": 4325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5953822135925293,
+      "learning_rate": 9.703984144526565e-05,
+      "loss": 3.0702,
+      "step": 4330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.597492516040802,
+      "learning_rate": 9.703290938705114e-05,
+      "loss": 3.0267,
+      "step": 4335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5713368654251099,
+      "learning_rate": 9.702596946981232e-05,
+      "loss": 3.1057,
+      "step": 4340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6499948501586914,
+      "learning_rate": 9.701902169470883e-05,
+      "loss": 3.0589,
+      "step": 4345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5947453379631042,
+      "learning_rate": 9.701206606290164e-05,
+      "loss": 3.0009,
+      "step": 4350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6248669028282166,
+      "learning_rate": 9.700510257555295e-05,
+      "loss": 3.1252,
+      "step": 4355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6154376864433289,
+      "learning_rate": 9.69981312338264e-05,
+      "loss": 3.0141,
+      "step": 4360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6073412299156189,
+      "learning_rate": 9.699115203888686e-05,
+      "loss": 3.083,
+      "step": 4365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5970540046691895,
+      "learning_rate": 9.69841649919005e-05,
+      "loss": 3.0533,
+      "step": 4370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.643619954586029,
+      "learning_rate": 9.697717009403485e-05,
+      "loss": 3.1047,
+      "step": 4375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6570290923118591,
+      "learning_rate": 9.697016734645873e-05,
+      "loss": 3.1369,
+      "step": 4380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.663725733757019,
+      "learning_rate": 9.696315675034227e-05,
+      "loss": 3.0773,
+      "step": 4385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6463174223899841,
+      "learning_rate": 9.69561383068569e-05,
+      "loss": 3.1116,
+      "step": 4390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6134265661239624,
+      "learning_rate": 9.694911201717541e-05,
+      "loss": 3.0863,
+      "step": 4395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6317304372787476,
+      "learning_rate": 9.694207788247186e-05,
+      "loss": 3.1819,
+      "step": 4400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5791853666305542,
+      "learning_rate": 9.693503590392159e-05,
+      "loss": 2.9822,
+      "step": 4405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6285654306411743,
+      "learning_rate": 9.692798608270132e-05,
+      "loss": 3.0867,
+      "step": 4410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.91982102394104,
+      "learning_rate": 9.692092841998904e-05,
+      "loss": 3.1132,
+      "step": 4415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7224445343017578,
+      "learning_rate": 9.691386291696407e-05,
+      "loss": 3.0606,
+      "step": 4420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6053935885429382,
+      "learning_rate": 9.690678957480702e-05,
+      "loss": 3.1155,
+      "step": 4425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.615533709526062,
+      "learning_rate": 9.689970839469981e-05,
+      "loss": 3.1091,
+      "step": 4430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5918856263160706,
+      "learning_rate": 9.689261937782571e-05,
+      "loss": 3.1038,
+      "step": 4435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6088792681694031,
+      "learning_rate": 9.688552252536924e-05,
+      "loss": 3.1005,
+      "step": 4440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6203143000602722,
+      "learning_rate": 9.687841783851626e-05,
+      "loss": 2.989,
+      "step": 4445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5926872491836548,
+      "learning_rate": 9.687130531845394e-05,
+      "loss": 3.162,
+      "step": 4450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5911262035369873,
+      "learning_rate": 9.686418496637076e-05,
+      "loss": 3.1295,
+      "step": 4455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.608135998249054,
+      "learning_rate": 9.685705678345652e-05,
+      "loss": 3.0324,
+      "step": 4460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.560269832611084,
+      "learning_rate": 9.684992077090227e-05,
+      "loss": 3.0199,
+      "step": 4465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5964846014976501,
+      "learning_rate": 9.684277692990045e-05,
+      "loss": 3.0405,
+      "step": 4470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5916455388069153,
+      "learning_rate": 9.683562526164475e-05,
+      "loss": 3.009,
+      "step": 4475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.650391697883606,
+      "learning_rate": 9.68284657673302e-05,
+      "loss": 3.13,
+      "step": 4480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5926150679588318,
+      "learning_rate": 9.682129844815311e-05,
+      "loss": 3.0809,
+      "step": 4485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6151012778282166,
+      "learning_rate": 9.681412330531112e-05,
+      "loss": 3.216,
+      "step": 4490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5863075852394104,
+      "learning_rate": 9.680694034000318e-05,
+      "loss": 3.2025,
+      "step": 4495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5777353048324585,
+      "learning_rate": 9.67997495534295e-05,
+      "loss": 3.0476,
+      "step": 4500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6134024858474731,
+      "learning_rate": 9.679255094679169e-05,
+      "loss": 3.0609,
+      "step": 4505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6142165660858154,
+      "learning_rate": 9.678534452129256e-05,
+      "loss": 3.0286,
+      "step": 4510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6046741008758545,
+      "learning_rate": 9.677813027813632e-05,
+      "loss": 3.0116,
+      "step": 4515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6047441363334656,
+      "learning_rate": 9.677090821852842e-05,
+      "loss": 3.0946,
+      "step": 4520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5957929491996765,
+      "learning_rate": 9.676367834367562e-05,
+      "loss": 2.9793,
+      "step": 4525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5568940043449402,
+      "learning_rate": 9.675644065478602e-05,
+      "loss": 3.0609,
+      "step": 4530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6303269267082214,
+      "learning_rate": 9.674919515306904e-05,
+      "loss": 3.1187,
+      "step": 4535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6048411130905151,
+      "learning_rate": 9.674194183973533e-05,
+      "loss": 2.9619,
+      "step": 4540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5991650223731995,
+      "learning_rate": 9.673468071599694e-05,
+      "loss": 2.9759,
+      "step": 4545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5938346982002258,
+      "learning_rate": 9.672741178306713e-05,
+      "loss": 3.1284,
+      "step": 4550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6247447729110718,
+      "learning_rate": 9.672013504216054e-05,
+      "loss": 3.0028,
+      "step": 4555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6094239354133606,
+      "learning_rate": 9.671285049449307e-05,
+      "loss": 3.0516,
+      "step": 4560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5746940970420837,
+      "learning_rate": 9.670555814128197e-05,
+      "loss": 2.9679,
+      "step": 4565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.60334712266922,
+      "learning_rate": 9.669825798374574e-05,
+      "loss": 3.0549,
+      "step": 4570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5870401263237,
+      "learning_rate": 9.66909500231042e-05,
+      "loss": 3.1055,
+      "step": 4575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5819172859191895,
+      "learning_rate": 9.668363426057852e-05,
+      "loss": 2.9642,
+      "step": 4580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6299050450325012,
+      "learning_rate": 9.66763106973911e-05,
+      "loss": 2.915,
+      "step": 4585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6059879064559937,
+      "learning_rate": 9.666897933476568e-05,
+      "loss": 3.0071,
+      "step": 4590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6717711687088013,
+      "learning_rate": 9.666164017392734e-05,
+      "loss": 3.0319,
+      "step": 4595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6143206357955933,
+      "learning_rate": 9.665429321610239e-05,
+      "loss": 3.1246,
+      "step": 4600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5944123268127441,
+      "learning_rate": 9.66469384625185e-05,
+      "loss": 3.0714,
+      "step": 4605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6126868724822998,
+      "learning_rate": 9.66395759144046e-05,
+      "loss": 3.1331,
+      "step": 4610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.569444477558136,
+      "learning_rate": 9.6632205572991e-05,
+      "loss": 3.0565,
+      "step": 4615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6019296646118164,
+      "learning_rate": 9.662482743950918e-05,
+      "loss": 3.045,
+      "step": 4620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.591454267501831,
+      "learning_rate": 9.661744151519206e-05,
+      "loss": 2.991,
+      "step": 4625
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6686195731163025,
+      "learning_rate": 9.661004780127373e-05,
+      "loss": 3.0972,
+      "step": 4630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.614673376083374,
+      "learning_rate": 9.660264629898974e-05,
+      "loss": 2.993,
+      "step": 4635
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5894550085067749,
+      "learning_rate": 9.65952370095768e-05,
+      "loss": 2.981,
+      "step": 4640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5858415365219116,
+      "learning_rate": 9.6587819934273e-05,
+      "loss": 3.0827,
+      "step": 4645
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6210569739341736,
+      "learning_rate": 9.658039507431768e-05,
+      "loss": 2.9882,
+      "step": 4650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5756790041923523,
+      "learning_rate": 9.657296243095151e-05,
+      "loss": 3.0755,
+      "step": 4655
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5973567366600037,
+      "learning_rate": 9.656552200541648e-05,
+      "loss": 3.0352,
+      "step": 4660
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6366568207740784,
+      "learning_rate": 9.655807379895583e-05,
+      "loss": 3.0411,
+      "step": 4665
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6964821815490723,
+      "learning_rate": 9.655061781281413e-05,
+      "loss": 2.9624,
+      "step": 4670
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6142486929893494,
+      "learning_rate": 9.654315404823729e-05,
+      "loss": 3.097,
+      "step": 4675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5889878869056702,
+      "learning_rate": 9.653568250647242e-05,
+      "loss": 3.1204,
+      "step": 4680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5988968014717102,
+      "learning_rate": 9.652820318876802e-05,
+      "loss": 2.9721,
+      "step": 4685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6125725507736206,
+      "learning_rate": 9.652071609637385e-05,
+      "loss": 3.0752,
+      "step": 4690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5981602668762207,
+      "learning_rate": 9.651322123054097e-05,
+      "loss": 2.9526,
+      "step": 4695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6357486844062805,
+      "learning_rate": 9.650571859252175e-05,
+      "loss": 3.0438,
+      "step": 4700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5893916487693787,
+      "learning_rate": 9.649820818356985e-05,
+      "loss": 3.0221,
+      "step": 4705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5716902017593384,
+      "learning_rate": 9.649069000494025e-05,
+      "loss": 3.0524,
+      "step": 4710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6035239696502686,
+      "learning_rate": 9.648316405788916e-05,
+      "loss": 3.0724,
+      "step": 4715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5987858772277832,
+      "learning_rate": 9.647563034367419e-05,
+      "loss": 2.971,
+      "step": 4720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5953534245491028,
+      "learning_rate": 9.646808886355417e-05,
+      "loss": 3.0368,
+      "step": 4725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6075890064239502,
+      "learning_rate": 9.646053961878925e-05,
+      "loss": 3.0426,
+      "step": 4730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6410869359970093,
+      "learning_rate": 9.64529826106409e-05,
+      "loss": 3.1187,
+      "step": 4735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5855988264083862,
+      "learning_rate": 9.644541784037186e-05,
+      "loss": 3.0326,
+      "step": 4740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6307476162910461,
+      "learning_rate": 9.643784530924616e-05,
+      "loss": 3.0489,
+      "step": 4745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.648952066898346,
+      "learning_rate": 9.643026501852916e-05,
+      "loss": 3.0136,
+      "step": 4750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5945757627487183,
+      "learning_rate": 9.64226769694875e-05,
+      "loss": 3.0947,
+      "step": 4755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6080991625785828,
+      "learning_rate": 9.641508116338908e-05,
+      "loss": 2.9928,
+      "step": 4760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6081297993659973,
+      "learning_rate": 9.640747760150319e-05,
+      "loss": 3.0517,
+      "step": 4765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6354160904884338,
+      "learning_rate": 9.639986628510032e-05,
+      "loss": 3.0584,
+      "step": 4770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5998715162277222,
+      "learning_rate": 9.639224721545229e-05,
+      "loss": 2.996,
+      "step": 4775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6090976595878601,
+      "learning_rate": 9.638462039383223e-05,
+      "loss": 3.0373,
+      "step": 4780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6166030764579773,
+      "learning_rate": 9.637698582151454e-05,
+      "loss": 3.032,
+      "step": 4785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.589093804359436,
+      "learning_rate": 9.636934349977493e-05,
+      "loss": 3.016,
+      "step": 4790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6204208135604858,
+      "learning_rate": 9.636169342989043e-05,
+      "loss": 2.9571,
+      "step": 4795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6287553310394287,
+      "learning_rate": 9.635403561313932e-05,
+      "loss": 3.1261,
+      "step": 4800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.717161238193512,
+      "learning_rate": 9.634637005080119e-05,
+      "loss": 2.9692,
+      "step": 4805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6154970526695251,
+      "learning_rate": 9.633869674415693e-05,
+      "loss": 3.004,
+      "step": 4810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7264851927757263,
+      "learning_rate": 9.633101569448873e-05,
+      "loss": 2.9752,
+      "step": 4815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5853413939476013,
+      "learning_rate": 9.632332690308004e-05,
+      "loss": 3.0189,
+      "step": 4820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6311028599739075,
+      "learning_rate": 9.631563037121565e-05,
+      "loss": 2.9473,
+      "step": 4825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5712893009185791,
+      "learning_rate": 9.63079261001816e-05,
+      "loss": 2.911,
+      "step": 4830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6190448999404907,
+      "learning_rate": 9.630021409126529e-05,
+      "loss": 2.9949,
+      "step": 4835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6113419532775879,
+      "learning_rate": 9.629249434575531e-05,
+      "loss": 3.0369,
+      "step": 4840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6142393350601196,
+      "learning_rate": 9.628476686494164e-05,
+      "loss": 2.9459,
+      "step": 4845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6480656266212463,
+      "learning_rate": 9.627703165011551e-05,
+      "loss": 3.049,
+      "step": 4850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6011250615119934,
+      "learning_rate": 9.626928870256944e-05,
+      "loss": 3.0967,
+      "step": 4855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5870364308357239,
+      "learning_rate": 9.626153802359724e-05,
+      "loss": 3.0707,
+      "step": 4860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6466168165206909,
+      "learning_rate": 9.625377961449402e-05,
+      "loss": 3.0524,
+      "step": 4865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5873356461524963,
+      "learning_rate": 9.62460134765562e-05,
+      "loss": 3.0864,
+      "step": 4870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6134336590766907,
+      "learning_rate": 9.623823961108145e-05,
+      "loss": 3.0023,
+      "step": 4875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5789670348167419,
+      "learning_rate": 9.623045801936876e-05,
+      "loss": 3.1161,
+      "step": 4880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5540945529937744,
+      "learning_rate": 9.622266870271842e-05,
+      "loss": 2.9856,
+      "step": 4885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6214756965637207,
+      "learning_rate": 9.6214871662432e-05,
+      "loss": 3.0294,
+      "step": 4890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5459863543510437,
+      "learning_rate": 9.620706689981232e-05,
+      "loss": 3.0793,
+      "step": 4895
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5777967572212219,
+      "learning_rate": 9.619925441616354e-05,
+      "loss": 3.0197,
+      "step": 4900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3430885076522827,
+      "learning_rate": 9.619143421279112e-05,
+      "loss": 2.9659,
+      "step": 4905
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.591229259967804,
+      "learning_rate": 9.618360629100176e-05,
+      "loss": 3.0135,
+      "step": 4910
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5897184014320374,
+      "learning_rate": 9.61757706521035e-05,
+      "loss": 3.1034,
+      "step": 4915
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5979292988777161,
+      "learning_rate": 9.616792729740564e-05,
+      "loss": 3.0249,
+      "step": 4920
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5851757526397705,
+      "learning_rate": 9.616007622821877e-05,
+      "loss": 3.0762,
+      "step": 4925
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5663010478019714,
+      "learning_rate": 9.615221744585476e-05,
+      "loss": 2.9226,
+      "step": 4930
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.553331196308136,
+      "learning_rate": 9.614435095162679e-05,
+      "loss": 2.9607,
+      "step": 4935
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5576261878013611,
+      "learning_rate": 9.613647674684936e-05,
+      "loss": 2.9852,
+      "step": 4940
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5791786313056946,
+      "learning_rate": 9.612859483283818e-05,
+      "loss": 3.0739,
+      "step": 4945
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5969935059547424,
+      "learning_rate": 9.612070521091029e-05,
+      "loss": 3.0378,
+      "step": 4950
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.578628659248352,
+      "learning_rate": 9.611280788238403e-05,
+      "loss": 2.9717,
+      "step": 4955
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6108145713806152,
+      "learning_rate": 9.610490284857901e-05,
+      "loss": 3.0665,
+      "step": 4960
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5779094099998474,
+      "learning_rate": 9.609699011081613e-05,
+      "loss": 3.01,
+      "step": 4965
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5818458795547485,
+      "learning_rate": 9.608906967041757e-05,
+      "loss": 2.9787,
+      "step": 4970
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5818671584129333,
+      "learning_rate": 9.608114152870684e-05,
+      "loss": 2.9014,
+      "step": 4975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5762121677398682,
+      "learning_rate": 9.607320568700866e-05,
+      "loss": 3.0292,
+      "step": 4980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5528386235237122,
+      "learning_rate": 9.606526214664911e-05,
+      "loss": 2.9871,
+      "step": 4985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5721897482872009,
+      "learning_rate": 9.605731090895551e-05,
+      "loss": 2.9007,
+      "step": 4990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5522501468658447,
+      "learning_rate": 9.604935197525649e-05,
+      "loss": 3.0394,
+      "step": 4995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5523527264595032,
+      "learning_rate": 9.604138534688196e-05,
+      "loss": 2.9343,
+      "step": 5000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5490130186080933,
+      "learning_rate": 9.60334110251631e-05,
+      "loss": 3.0215,
+      "step": 5005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6023851037025452,
+      "learning_rate": 9.602542901143241e-05,
+      "loss": 3.0138,
+      "step": 5010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5751814842224121,
+      "learning_rate": 9.601743930702363e-05,
+      "loss": 3.0093,
+      "step": 5015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6199713945388794,
+      "learning_rate": 9.600944191327184e-05,
+      "loss": 3.1001,
+      "step": 5020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5751425623893738,
+      "learning_rate": 9.600143683151335e-05,
+      "loss": 3.1229,
+      "step": 5025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6376388669013977,
+      "learning_rate": 9.59934240630858e-05,
+      "loss": 3.0371,
+      "step": 5030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6093132495880127,
+      "learning_rate": 9.598540360932806e-05,
+      "loss": 2.952,
+      "step": 5035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6021653413772583,
+      "learning_rate": 9.597737547158035e-05,
+      "loss": 3.0717,
+      "step": 5040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5976895689964294,
+      "learning_rate": 9.596933965118413e-05,
+      "loss": 2.9929,
+      "step": 5045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5661852955818176,
+      "learning_rate": 9.596129614948217e-05,
+      "loss": 2.9715,
+      "step": 5050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6110250949859619,
+      "learning_rate": 9.595324496781849e-05,
+      "loss": 2.9821,
+      "step": 5055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5849717259407043,
+      "learning_rate": 9.594518610753841e-05,
+      "loss": 2.9477,
+      "step": 5060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6186911463737488,
+      "learning_rate": 9.593711956998855e-05,
+      "loss": 3.0132,
+      "step": 5065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5757038593292236,
+      "learning_rate": 9.59290453565168e-05,
+      "loss": 3.0242,
+      "step": 5070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6154767274856567,
+      "learning_rate": 9.592096346847232e-05,
+      "loss": 2.9959,
+      "step": 5075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5663397312164307,
+      "learning_rate": 9.591287390720555e-05,
+      "loss": 2.9557,
+      "step": 5080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6402688026428223,
+      "learning_rate": 9.590477667406828e-05,
+      "loss": 2.8717,
+      "step": 5085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5841224193572998,
+      "learning_rate": 9.589667177041347e-05,
+      "loss": 3.0737,
+      "step": 5090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6521735191345215,
+      "learning_rate": 9.588855919759544e-05,
+      "loss": 2.8883,
+      "step": 5095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5969855189323425,
+      "learning_rate": 9.588043895696978e-05,
+      "loss": 2.9863,
+      "step": 5100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.588037371635437,
+      "learning_rate": 9.587231104989334e-05,
+      "loss": 2.9906,
+      "step": 5105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5970036387443542,
+      "learning_rate": 9.586417547772426e-05,
+      "loss": 3.1028,
+      "step": 5110
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5990484952926636,
+      "learning_rate": 9.585603224182196e-05,
+      "loss": 2.971,
+      "step": 5115
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5897454023361206,
+      "learning_rate": 9.584788134354718e-05,
+      "loss": 2.9152,
+      "step": 5120
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5886784195899963,
+      "learning_rate": 9.583972278426186e-05,
+      "loss": 3.073,
+      "step": 5125
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6416183710098267,
+      "learning_rate": 9.583155656532929e-05,
+      "loss": 2.9778,
+      "step": 5130
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6140429377555847,
+      "learning_rate": 9.582338268811402e-05,
+      "loss": 3.021,
+      "step": 5135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5906372666358948,
+      "learning_rate": 9.581520115398185e-05,
+      "loss": 2.9714,
+      "step": 5140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5692211389541626,
+      "learning_rate": 9.580701196429991e-05,
+      "loss": 2.7904,
+      "step": 5145
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.570862352848053,
+      "learning_rate": 9.579881512043655e-05,
+      "loss": 3.0642,
+      "step": 5150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5729061365127563,
+      "learning_rate": 9.579061062376146e-05,
+      "loss": 3.037,
+      "step": 5155
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6067327857017517,
+      "learning_rate": 9.578239847564559e-05,
+      "loss": 2.9341,
+      "step": 5160
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5570735335350037,
+      "learning_rate": 9.577417867746113e-05,
+      "loss": 2.9904,
+      "step": 5165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5639290809631348,
+      "learning_rate": 9.576595123058158e-05,
+      "loss": 3.0199,
+      "step": 5170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5590497255325317,
+      "learning_rate": 9.575771613638175e-05,
+      "loss": 3.0315,
+      "step": 5175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5961495041847229,
+      "learning_rate": 9.574947339623765e-05,
+      "loss": 2.961,
+      "step": 5180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.569841206073761,
+      "learning_rate": 9.574122301152665e-05,
+      "loss": 2.9692,
+      "step": 5185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5561034083366394,
+      "learning_rate": 9.573296498362733e-05,
+      "loss": 3.0087,
+      "step": 5190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5449041128158569,
+      "learning_rate": 9.572469931391959e-05,
+      "loss": 2.9811,
+      "step": 5195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6024003624916077,
+      "learning_rate": 9.571642600378459e-05,
+      "loss": 3.0579,
+      "step": 5200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5973739624023438,
+      "learning_rate": 9.570814505460475e-05,
+      "loss": 2.9412,
+      "step": 5205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5798643231391907,
+      "learning_rate": 9.569985646776382e-05,
+      "loss": 2.9128,
+      "step": 5210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5798316597938538,
+      "learning_rate": 9.569156024464677e-05,
+      "loss": 2.9783,
+      "step": 5215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.58490389585495,
+      "learning_rate": 9.568325638663989e-05,
+      "loss": 2.9545,
+      "step": 5220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5917816162109375,
+      "learning_rate": 9.567494489513068e-05,
+      "loss": 3.0486,
+      "step": 5225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5882901549339294,
+      "learning_rate": 9.5666625771508e-05,
+      "loss": 3.0563,
+      "step": 5230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5690311193466187,
+      "learning_rate": 9.565829901716194e-05,
+      "loss": 2.9503,
+      "step": 5235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5425330996513367,
+      "learning_rate": 9.564996463348384e-05,
+      "loss": 3.047,
+      "step": 5240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5781831741333008,
+      "learning_rate": 9.564162262186638e-05,
+      "loss": 2.9943,
+      "step": 5245
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6207824945449829,
+      "learning_rate": 9.563327298370347e-05,
+      "loss": 2.9075,
+      "step": 5250
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5939603447914124,
+      "learning_rate": 9.56249157203903e-05,
+      "loss": 2.9768,
+      "step": 5255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5579645037651062,
+      "learning_rate": 9.561655083332333e-05,
+      "loss": 2.9762,
+      "step": 5260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5610690712928772,
+      "learning_rate": 9.56081783239003e-05,
+      "loss": 2.9885,
+      "step": 5265
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5744231343269348,
+      "learning_rate": 9.559979819352024e-05,
+      "loss": 2.9734,
+      "step": 5270
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6599621772766113,
+      "learning_rate": 9.559141044358342e-05,
+      "loss": 2.9398,
+      "step": 5275
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.610150158405304,
+      "learning_rate": 9.558301507549143e-05,
+      "loss": 3.0204,
+      "step": 5280
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6108394265174866,
+      "learning_rate": 9.557461209064708e-05,
+      "loss": 2.8804,
+      "step": 5285
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6205480098724365,
+      "learning_rate": 9.556620149045447e-05,
+      "loss": 3.0109,
+      "step": 5290
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5954495668411255,
+      "learning_rate": 9.555778327631903e-05,
+      "loss": 2.9744,
+      "step": 5295
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.59709632396698,
+      "learning_rate": 9.554935744964734e-05,
+      "loss": 3.0464,
+      "step": 5300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5658520460128784,
+      "learning_rate": 9.55409240118474e-05,
+      "loss": 2.9096,
+      "step": 5305
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5825792551040649,
+      "learning_rate": 9.553248296432835e-05,
+      "loss": 2.962,
+      "step": 5310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5943558812141418,
+      "learning_rate": 9.552403430850068e-05,
+      "loss": 2.9326,
+      "step": 5315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5840505361557007,
+      "learning_rate": 9.551557804577612e-05,
+      "loss": 2.9484,
+      "step": 5320
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5731467008590698,
+      "learning_rate": 9.550711417756767e-05,
+      "loss": 3.0095,
+      "step": 5325
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6134483218193054,
+      "learning_rate": 9.549864270528965e-05,
+      "loss": 3.022,
+      "step": 5330
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5741324424743652,
+      "learning_rate": 9.549016363035758e-05,
+      "loss": 2.9431,
+      "step": 5335
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5801990032196045,
+      "learning_rate": 9.548167695418828e-05,
+      "loss": 3.0646,
+      "step": 5340
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5719301104545593,
+      "learning_rate": 9.547318267819986e-05,
+      "loss": 3.0219,
+      "step": 5345
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5628625750541687,
+      "learning_rate": 9.546468080381167e-05,
+      "loss": 2.9833,
+      "step": 5350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5710899233818054,
+      "learning_rate": 9.545617133244436e-05,
+      "loss": 2.9587,
+      "step": 5355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5758590698242188,
+      "learning_rate": 9.544765426551981e-05,
+      "loss": 3.0817,
+      "step": 5360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5717889070510864,
+      "learning_rate": 9.543912960446117e-05,
+      "loss": 2.9171,
+      "step": 5365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5406732559204102,
+      "learning_rate": 9.543059735069293e-05,
+      "loss": 2.9743,
+      "step": 5370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7118645906448364,
+      "learning_rate": 9.542205750564077e-05,
+      "loss": 2.9945,
+      "step": 5375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5385851860046387,
+      "learning_rate": 9.541351007073167e-05,
+      "loss": 3.0186,
+      "step": 5380
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5722284317016602,
+      "learning_rate": 9.540495504739387e-05,
+      "loss": 2.9661,
+      "step": 5385
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5985246896743774,
+      "learning_rate": 9.53963924370569e-05,
+      "loss": 2.9337,
+      "step": 5390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5782822370529175,
+      "learning_rate": 9.538782224115151e-05,
+      "loss": 2.8609,
+      "step": 5395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6297509074211121,
+      "learning_rate": 9.537924446110976e-05,
+      "loss": 3.0064,
+      "step": 5400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5725342631340027,
+      "learning_rate": 9.537065909836497e-05,
+      "loss": 2.9231,
+      "step": 5405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5984869003295898,
+      "learning_rate": 9.536206615435174e-05,
+      "loss": 3.0435,
+      "step": 5410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5797191858291626,
+      "learning_rate": 9.535346563050591e-05,
+      "loss": 2.9765,
+      "step": 5415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6115691661834717,
+      "learning_rate": 9.534485752826456e-05,
+      "loss": 2.9914,
+      "step": 5420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6625143885612488,
+      "learning_rate": 9.53362418490661e-05,
+      "loss": 2.9495,
+      "step": 5425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6165873408317566,
+      "learning_rate": 9.53276185943502e-05,
+      "loss": 2.8682,
+      "step": 5430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6059501767158508,
+      "learning_rate": 9.531898776555773e-05,
+      "loss": 3.0199,
+      "step": 5435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6219189167022705,
+      "learning_rate": 9.531034936413091e-05,
+      "loss": 2.9767,
+      "step": 5440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5619981288909912,
+      "learning_rate": 9.530170339151314e-05,
+      "loss": 2.9727,
+      "step": 5445
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5867578387260437,
+      "learning_rate": 9.529304984914917e-05,
+      "loss": 2.9941,
+      "step": 5450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6011172533035278,
+      "learning_rate": 9.528438873848496e-05,
+      "loss": 2.9431,
+      "step": 5455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6187102198600769,
+      "learning_rate": 9.527572006096773e-05,
+      "loss": 2.9786,
+      "step": 5460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5892276167869568,
+      "learning_rate": 9.526704381804601e-05,
+      "loss": 3.033,
+      "step": 5465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7618337869644165,
+      "learning_rate": 9.525836001116957e-05,
+      "loss": 2.978,
+      "step": 5470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5733631253242493,
+      "learning_rate": 9.524966864178943e-05,
+      "loss": 2.966,
+      "step": 5475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5610578060150146,
+      "learning_rate": 9.524096971135789e-05,
+      "loss": 3.0258,
+      "step": 5480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.557549238204956,
+      "learning_rate": 9.52322632213285e-05,
+      "loss": 3.0223,
+      "step": 5485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6017749309539795,
+      "learning_rate": 9.52235491731561e-05,
+      "loss": 3.0331,
+      "step": 5490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5682252049446106,
+      "learning_rate": 9.521482756829675e-05,
+      "loss": 2.9914,
+      "step": 5495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6039074063301086,
+      "learning_rate": 9.520609840820782e-05,
+      "loss": 2.98,
+      "step": 5500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5930071473121643,
+      "learning_rate": 9.51973616943479e-05,
+      "loss": 2.8607,
+      "step": 5505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0643894672393799,
+      "learning_rate": 9.518861742817689e-05,
+      "loss": 2.9617,
+      "step": 5510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5629966259002686,
+      "learning_rate": 9.51798656111559e-05,
+      "loss": 2.9368,
+      "step": 5515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5865014791488647,
+      "learning_rate": 9.517110624474733e-05,
+      "loss": 2.9648,
+      "step": 5520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5893402695655823,
+      "learning_rate": 9.516233933041485e-05,
+      "loss": 2.885,
+      "step": 5525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6322982907295227,
+      "learning_rate": 9.515356486962337e-05,
+      "loss": 2.9061,
+      "step": 5530
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5798627138137817,
+      "learning_rate": 9.514478286383907e-05,
+      "loss": 2.9145,
+      "step": 5535
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5427137017250061,
+      "learning_rate": 9.513599331452939e-05,
+      "loss": 2.9662,
+      "step": 5540
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5554309487342834,
+      "learning_rate": 9.512719622316304e-05,
+      "loss": 2.9788,
+      "step": 5545
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6033381223678589,
+      "learning_rate": 9.511839159120997e-05,
+      "loss": 2.9885,
+      "step": 5550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5634362697601318,
+      "learning_rate": 9.51095794201414e-05,
+      "loss": 2.9552,
+      "step": 5555
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5524389147758484,
+      "learning_rate": 9.510075971142983e-05,
+      "loss": 2.8438,
+      "step": 5560
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5727753043174744,
+      "learning_rate": 9.509193246654896e-05,
+      "loss": 2.9143,
+      "step": 5565
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6197867393493652,
+      "learning_rate": 9.508309768697384e-05,
+      "loss": 3.0105,
+      "step": 5570
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5803563594818115,
+      "learning_rate": 9.507425537418068e-05,
+      "loss": 2.8933,
+      "step": 5575
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5719127655029297,
+      "learning_rate": 9.506540552964704e-05,
+      "loss": 2.8481,
+      "step": 5580
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5865441560745239,
+      "learning_rate": 9.505654815485168e-05,
+      "loss": 2.9553,
+      "step": 5585
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5539926886558533,
+      "learning_rate": 9.504768325127464e-05,
+      "loss": 2.8761,
+      "step": 5590
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5409018993377686,
+      "learning_rate": 9.503881082039719e-05,
+      "loss": 2.9554,
+      "step": 5595
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5662526488304138,
+      "learning_rate": 9.50299308637019e-05,
+      "loss": 2.904,
+      "step": 5600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5841567516326904,
+      "learning_rate": 9.502104338267257e-05,
+      "loss": 2.9878,
+      "step": 5605
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6460723280906677,
+      "learning_rate": 9.501214837879428e-05,
+      "loss": 2.8437,
+      "step": 5610
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5596894025802612,
+      "learning_rate": 9.500324585355333e-05,
+      "loss": 2.9185,
+      "step": 5615
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5684837698936462,
+      "learning_rate": 9.499433580843732e-05,
+      "loss": 2.9706,
+      "step": 5620
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5364347100257874,
+      "learning_rate": 9.498541824493506e-05,
+      "loss": 2.9233,
+      "step": 5625
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5619500279426575,
+      "learning_rate": 9.497649316453667e-05,
+      "loss": 2.9103,
+      "step": 5630
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5582467317581177,
+      "learning_rate": 9.496756056873348e-05,
+      "loss": 2.9542,
+      "step": 5635
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5239951014518738,
+      "learning_rate": 9.49586204590181e-05,
+      "loss": 2.9525,
+      "step": 5640
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5564820766448975,
+      "learning_rate": 9.494967283688438e-05,
+      "loss": 2.9737,
+      "step": 5645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5779933929443359,
+      "learning_rate": 9.494071770382745e-05,
+      "loss": 2.9515,
+      "step": 5650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.590217649936676,
+      "learning_rate": 9.493175506134367e-05,
+      "loss": 2.941,
+      "step": 5655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5666140913963318,
+      "learning_rate": 9.492278491093067e-05,
+      "loss": 3.0097,
+      "step": 5660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5883950591087341,
+      "learning_rate": 9.491380725408731e-05,
+      "loss": 2.9412,
+      "step": 5665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5530170202255249,
+      "learning_rate": 9.490482209231374e-05,
+      "loss": 2.9032,
+      "step": 5670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5779628753662109,
+      "learning_rate": 9.489582942711135e-05,
+      "loss": 2.9609,
+      "step": 5675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5664241313934326,
+      "learning_rate": 9.488682925998277e-05,
+      "loss": 2.9989,
+      "step": 5680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.585910975933075,
+      "learning_rate": 9.48778215924319e-05,
+      "loss": 2.9451,
+      "step": 5685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6111571788787842,
+      "learning_rate": 9.48688064259639e-05,
+      "loss": 2.8197,
+      "step": 5690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6127909421920776,
+      "learning_rate": 9.485978376208513e-05,
+      "loss": 2.9609,
+      "step": 5695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5551013946533203,
+      "learning_rate": 9.485075360230328e-05,
+      "loss": 2.8675,
+      "step": 5700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5652710199356079,
+      "learning_rate": 9.484171594812728e-05,
+      "loss": 3.0067,
+      "step": 5705
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5753379464149475,
+      "learning_rate": 9.483267080106723e-05,
+      "loss": 2.923,
+      "step": 5710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5468341708183289,
+      "learning_rate": 9.482361816263456e-05,
+      "loss": 2.9638,
+      "step": 5715
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5311121344566345,
+      "learning_rate": 9.481455803434195e-05,
+      "loss": 2.9313,
+      "step": 5720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5821118354797363,
+      "learning_rate": 9.48054904177033e-05,
+      "loss": 2.9242,
+      "step": 5725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5618748068809509,
+      "learning_rate": 9.479641531423377e-05,
+      "loss": 3.0829,
+      "step": 5730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5709080696105957,
+      "learning_rate": 9.47873327254498e-05,
+      "loss": 2.958,
+      "step": 5735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6024799942970276,
+      "learning_rate": 9.477824265286904e-05,
+      "loss": 2.8649,
+      "step": 5740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6166125535964966,
+      "learning_rate": 9.476914509801039e-05,
+      "loss": 2.9787,
+      "step": 5745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5704414248466492,
+      "learning_rate": 9.476004006239406e-05,
+      "loss": 2.9025,
+      "step": 5750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5633816123008728,
+      "learning_rate": 9.475092754754142e-05,
+      "loss": 2.9855,
+      "step": 5755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5753917098045349,
+      "learning_rate": 9.474180755497517e-05,
+      "loss": 2.8495,
+      "step": 5760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5372222065925598,
+      "learning_rate": 9.473268008621922e-05,
+      "loss": 2.9963,
+      "step": 5765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.553519606590271,
+      "learning_rate": 9.472354514279874e-05,
+      "loss": 2.8932,
+      "step": 5770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.559740424156189,
+      "learning_rate": 9.471440272624012e-05,
+      "loss": 2.9063,
+      "step": 5775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.604865550994873,
+      "learning_rate": 9.470525283807105e-05,
+      "loss": 2.9772,
+      "step": 5780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5670210123062134,
+      "learning_rate": 9.469609547982044e-05,
+      "loss": 2.8899,
+      "step": 5785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5659230947494507,
+      "learning_rate": 9.468693065301843e-05,
+      "loss": 2.9068,
+      "step": 5790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5749325752258301,
+      "learning_rate": 9.467775835919646e-05,
+      "loss": 2.8546,
+      "step": 5795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5506710410118103,
+      "learning_rate": 9.466857859988715e-05,
+      "loss": 2.9286,
+      "step": 5800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5292921662330627,
+      "learning_rate": 9.465939137662444e-05,
+      "loss": 2.8695,
+      "step": 5805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5484473705291748,
+      "learning_rate": 9.465019669094346e-05,
+      "loss": 2.8881,
+      "step": 5810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5465819835662842,
+      "learning_rate": 9.464099454438062e-05,
+      "loss": 2.8951,
+      "step": 5815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5700329542160034,
+      "learning_rate": 9.463178493847355e-05,
+      "loss": 2.9297,
+      "step": 5820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6440792083740234,
+      "learning_rate": 9.462256787476113e-05,
+      "loss": 2.9494,
+      "step": 5825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.552334189414978,
+      "learning_rate": 9.461334335478354e-05,
+      "loss": 2.8999,
+      "step": 5830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5664337873458862,
+      "learning_rate": 9.460411138008214e-05,
+      "loss": 2.9573,
+      "step": 5835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.631130039691925,
+      "learning_rate": 9.459487195219954e-05,
+      "loss": 2.942,
+      "step": 5840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5636876225471497,
+      "learning_rate": 9.458562507267963e-05,
+      "loss": 2.878,
+      "step": 5845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5582862496376038,
+      "learning_rate": 9.457637074306754e-05,
+      "loss": 2.9972,
+      "step": 5850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5719597339630127,
+      "learning_rate": 9.456710896490961e-05,
+      "loss": 2.9024,
+      "step": 5855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5531226992607117,
+      "learning_rate": 9.455783973975348e-05,
+      "loss": 2.9179,
+      "step": 5860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5812441110610962,
+      "learning_rate": 9.454856306914797e-05,
+      "loss": 3.0274,
+      "step": 5865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5888627171516418,
+      "learning_rate": 9.453927895464321e-05,
+      "loss": 2.8952,
+      "step": 5870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5892916917800903,
+      "learning_rate": 9.452998739779051e-05,
+      "loss": 3.0804,
+      "step": 5875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6010014414787292,
+      "learning_rate": 9.452068840014246e-05,
+      "loss": 2.9492,
+      "step": 5880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5906623005867004,
+      "learning_rate": 9.451138196325291e-05,
+      "loss": 2.8768,
+      "step": 5885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5793782472610474,
+      "learning_rate": 9.45020680886769e-05,
+      "loss": 2.9501,
+      "step": 5890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6063160300254822,
+      "learning_rate": 9.449274677797078e-05,
+      "loss": 2.8959,
+      "step": 5895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5469617247581482,
+      "learning_rate": 9.448341803269208e-05,
+      "loss": 2.9931,
+      "step": 5900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5838633179664612,
+      "learning_rate": 9.44740818543996e-05,
+      "loss": 2.8648,
+      "step": 5905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5361950397491455,
+      "learning_rate": 9.44647382446534e-05,
+      "loss": 2.9166,
+      "step": 5910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5754554271697998,
+      "learning_rate": 9.445538720501474e-05,
+      "loss": 2.9628,
+      "step": 5915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5997181534767151,
+      "learning_rate": 9.444602873704616e-05,
+      "loss": 2.8761,
+      "step": 5920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5525051951408386,
+      "learning_rate": 9.443666284231141e-05,
+      "loss": 3.0209,
+      "step": 5925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5555387735366821,
+      "learning_rate": 9.44272895223755e-05,
+      "loss": 2.8965,
+      "step": 5930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7443075180053711,
+      "learning_rate": 9.441790877880469e-05,
+      "loss": 2.9419,
+      "step": 5935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.561226487159729,
+      "learning_rate": 9.440852061316645e-05,
+      "loss": 2.9545,
+      "step": 5940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5551393032073975,
+      "learning_rate": 9.439912502702952e-05,
+      "loss": 2.9028,
+      "step": 5945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5862253904342651,
+      "learning_rate": 9.438972202196387e-05,
+      "loss": 2.8389,
+      "step": 5950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.568202793598175,
+      "learning_rate": 9.43803115995407e-05,
+      "loss": 3.002,
+      "step": 5955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.568098247051239,
+      "learning_rate": 9.437089376133246e-05,
+      "loss": 2.9388,
+      "step": 5960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.557004451751709,
+      "learning_rate": 9.436146850891283e-05,
+      "loss": 3.0554,
+      "step": 5965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5677062273025513,
+      "learning_rate": 9.435203584385673e-05,
+      "loss": 2.886,
+      "step": 5970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5853590369224548,
+      "learning_rate": 9.434259576774034e-05,
+      "loss": 2.9415,
+      "step": 5975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5628269910812378,
+      "learning_rate": 9.433314828214103e-05,
+      "loss": 2.9017,
+      "step": 5980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5658286213874817,
+      "learning_rate": 9.43236933886375e-05,
+      "loss": 2.9206,
+      "step": 5985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5133522748947144,
+      "learning_rate": 9.431423108880957e-05,
+      "loss": 2.9659,
+      "step": 5990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5464906692504883,
+      "learning_rate": 9.430476138423838e-05,
+      "loss": 2.9085,
+      "step": 5995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6164173483848572,
+      "learning_rate": 9.429528427650627e-05,
+      "loss": 2.9178,
+      "step": 6000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5737559199333191,
+      "learning_rate": 9.428579976719684e-05,
+      "loss": 2.8456,
+      "step": 6005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.559264600276947,
+      "learning_rate": 9.427630785789491e-05,
+      "loss": 2.9421,
+      "step": 6010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5611038208007812,
+      "learning_rate": 9.426680855018654e-05,
+      "loss": 2.8726,
+      "step": 6015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5311176180839539,
+      "learning_rate": 9.425730184565907e-05,
+      "loss": 2.9288,
+      "step": 6020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5495334267616272,
+      "learning_rate": 9.424778774590096e-05,
+      "loss": 2.8964,
+      "step": 6025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.547717273235321,
+      "learning_rate": 9.423826625250204e-05,
+      "loss": 2.849,
+      "step": 6030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5711108446121216,
+      "learning_rate": 9.422873736705329e-05,
+      "loss": 2.9985,
+      "step": 6035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5532501339912415,
+      "learning_rate": 9.421920109114696e-05,
+      "loss": 2.8916,
+      "step": 6040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6609427332878113,
+      "learning_rate": 9.420965742637653e-05,
+      "loss": 2.9519,
+      "step": 6045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6078450679779053,
+      "learning_rate": 9.420010637433669e-05,
+      "loss": 2.8446,
+      "step": 6050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.553199827671051,
+      "learning_rate": 9.419054793662342e-05,
+      "loss": 2.9268,
+      "step": 6055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5539187788963318,
+      "learning_rate": 9.418098211483387e-05,
+      "loss": 2.877,
+      "step": 6060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5353770852088928,
+      "learning_rate": 9.417140891056646e-05,
+      "loss": 2.9979,
+      "step": 6065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5734512805938721,
+      "learning_rate": 9.416182832542085e-05,
+      "loss": 2.901,
+      "step": 6070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5867028832435608,
+      "learning_rate": 9.41522403609979e-05,
+      "loss": 2.8712,
+      "step": 6075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5963934659957886,
+      "learning_rate": 9.414264501889975e-05,
+      "loss": 2.9033,
+      "step": 6080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5971652865409851,
+      "learning_rate": 9.41330423007297e-05,
+      "loss": 2.8521,
+      "step": 6085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5670334100723267,
+      "learning_rate": 9.412343220809239e-05,
+      "loss": 2.9093,
+      "step": 6090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5659826993942261,
+      "learning_rate": 9.41138147425936e-05,
+      "loss": 2.815,
+      "step": 6095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5430881381034851,
+      "learning_rate": 9.410418990584034e-05,
+      "loss": 2.9488,
+      "step": 6100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5578214526176453,
+      "learning_rate": 9.409455769944093e-05,
+      "loss": 2.844,
+      "step": 6105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5729809403419495,
+      "learning_rate": 9.408491812500486e-05,
+      "loss": 2.9097,
+      "step": 6110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5481757521629333,
+      "learning_rate": 9.407527118414287e-05,
+      "loss": 2.9359,
+      "step": 6115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5715023875236511,
+      "learning_rate": 9.406561687846693e-05,
+      "loss": 2.9505,
+      "step": 6120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6070970296859741,
+      "learning_rate": 9.405595520959025e-05,
+      "loss": 2.8366,
+      "step": 6125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5430773496627808,
+      "learning_rate": 9.404628617912723e-05,
+      "loss": 2.849,
+      "step": 6130
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5314534902572632,
+      "learning_rate": 9.403660978869353e-05,
+      "loss": 2.8414,
+      "step": 6135
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5773774981498718,
+      "learning_rate": 9.402692603990608e-05,
+      "loss": 2.8643,
+      "step": 6140
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5885508060455322,
+      "learning_rate": 9.401723493438295e-05,
+      "loss": 2.9234,
+      "step": 6145
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5338242650032043,
+      "learning_rate": 9.400753647374352e-05,
+      "loss": 2.8933,
+      "step": 6150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5901104211807251,
+      "learning_rate": 9.399783065960835e-05,
+      "loss": 2.984,
+      "step": 6155
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5759269595146179,
+      "learning_rate": 9.398811749359924e-05,
+      "loss": 2.9413,
+      "step": 6160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5674792528152466,
+      "learning_rate": 9.397839697733924e-05,
+      "loss": 2.9684,
+      "step": 6165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5751272439956665,
+      "learning_rate": 9.396866911245261e-05,
+      "loss": 2.916,
+      "step": 6170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5756967663764954,
+      "learning_rate": 9.395893390056484e-05,
+      "loss": 2.9159,
+      "step": 6175
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5597189664840698,
+      "learning_rate": 9.394919134330264e-05,
+      "loss": 2.9128,
+      "step": 6180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5353022813796997,
+      "learning_rate": 9.393944144229396e-05,
+      "loss": 2.9865,
+      "step": 6185
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5555561184883118,
+      "learning_rate": 9.392968419916797e-05,
+      "loss": 2.8467,
+      "step": 6190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.509667694568634,
+      "learning_rate": 9.391991961555507e-05,
+      "loss": 2.8448,
+      "step": 6195
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.561540961265564,
+      "learning_rate": 9.39101476930869e-05,
+      "loss": 2.8963,
+      "step": 6200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5637139678001404,
+      "learning_rate": 9.390036843339629e-05,
+      "loss": 2.8807,
+      "step": 6205
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6163474321365356,
+      "learning_rate": 9.389058183811732e-05,
+      "loss": 2.9649,
+      "step": 6210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5708631873130798,
+      "learning_rate": 9.38807879088853e-05,
+      "loss": 2.9521,
+      "step": 6215
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5705059170722961,
+      "learning_rate": 9.387098664733676e-05,
+      "loss": 3.0278,
+      "step": 6220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6119548082351685,
+      "learning_rate": 9.386117805510945e-05,
+      "loss": 2.8872,
+      "step": 6225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5603761076927185,
+      "learning_rate": 9.385136213384237e-05,
+      "loss": 2.9311,
+      "step": 6230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5855942964553833,
+      "learning_rate": 9.38415388851757e-05,
+      "loss": 2.9476,
+      "step": 6235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55571448802948,
+      "learning_rate": 9.383170831075086e-05,
+      "loss": 2.8068,
+      "step": 6240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5550623536109924,
+      "learning_rate": 9.382187041221054e-05,
+      "loss": 2.8769,
+      "step": 6245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.570541501045227,
+      "learning_rate": 9.381202519119858e-05,
+      "loss": 2.8834,
+      "step": 6250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5700785517692566,
+      "learning_rate": 9.38021726493601e-05,
+      "loss": 2.9295,
+      "step": 6255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5558483600616455,
+      "learning_rate": 9.37923127883414e-05,
+      "loss": 2.975,
+      "step": 6260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51957106590271,
+      "learning_rate": 9.378244560979007e-05,
+      "loss": 2.8167,
+      "step": 6265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5505496859550476,
+      "learning_rate": 9.377257111535482e-05,
+      "loss": 3.0015,
+      "step": 6270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5620570182800293,
+      "learning_rate": 9.376268930668569e-05,
+      "loss": 2.8813,
+      "step": 6275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5273504257202148,
+      "learning_rate": 9.375280018543387e-05,
+      "loss": 2.9017,
+      "step": 6280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5470396876335144,
+      "learning_rate": 9.374290375325181e-05,
+      "loss": 2.9271,
+      "step": 6285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5683776140213013,
+      "learning_rate": 9.373300001179316e-05,
+      "loss": 2.8335,
+      "step": 6290
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5375957489013672,
+      "learning_rate": 9.37230889627128e-05,
+      "loss": 2.8706,
+      "step": 6295
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5238027572631836,
+      "learning_rate": 9.371317060766682e-05,
+      "loss": 2.8603,
+      "step": 6300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5620431900024414,
+      "learning_rate": 9.370324494831254e-05,
+      "loss": 2.8766,
+      "step": 6305
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5202515125274658,
+      "learning_rate": 9.369331198630852e-05,
+      "loss": 2.9133,
+      "step": 6310
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5206733345985413,
+      "learning_rate": 9.36833717233145e-05,
+      "loss": 2.9402,
+      "step": 6315
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5413232445716858,
+      "learning_rate": 9.367342416099147e-05,
+      "loss": 2.8496,
+      "step": 6320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5489947199821472,
+      "learning_rate": 9.366346930100163e-05,
+      "loss": 2.9663,
+      "step": 6325
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6418606042861938,
+      "learning_rate": 9.36535071450084e-05,
+      "loss": 2.956,
+      "step": 6330
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5444689989089966,
+      "learning_rate": 9.364353769467643e-05,
+      "loss": 2.8778,
+      "step": 6335
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5354869961738586,
+      "learning_rate": 9.363356095167156e-05,
+      "loss": 2.97,
+      "step": 6340
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5355974435806274,
+      "learning_rate": 9.362357691766088e-05,
+      "loss": 2.8608,
+      "step": 6345
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5451033711433411,
+      "learning_rate": 9.361358559431266e-05,
+      "loss": 2.8492,
+      "step": 6350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5956314206123352,
+      "learning_rate": 9.360358698329646e-05,
+      "loss": 2.8953,
+      "step": 6355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6001141667366028,
+      "learning_rate": 9.359358108628298e-05,
+      "loss": 2.8803,
+      "step": 6360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5375651121139526,
+      "learning_rate": 9.358356790494416e-05,
+      "loss": 2.8805,
+      "step": 6365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5422669053077698,
+      "learning_rate": 9.357354744095317e-05,
+      "loss": 2.9723,
+      "step": 6370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6271029114723206,
+      "learning_rate": 9.35635196959844e-05,
+      "loss": 2.841,
+      "step": 6375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5451935529708862,
+      "learning_rate": 9.355348467171347e-05,
+      "loss": 2.9503,
+      "step": 6380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5685514807701111,
+      "learning_rate": 9.354344236981716e-05,
+      "loss": 2.9715,
+      "step": 6385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5677580237388611,
+      "learning_rate": 9.353339279197351e-05,
+      "loss": 2.9603,
+      "step": 6390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5922720432281494,
+      "learning_rate": 9.352333593986176e-05,
+      "loss": 2.9009,
+      "step": 6395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.553828239440918,
+      "learning_rate": 9.35132718151624e-05,
+      "loss": 2.8676,
+      "step": 6400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5401423573493958,
+      "learning_rate": 9.35032004195571e-05,
+      "loss": 2.9743,
+      "step": 6405
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5475908517837524,
+      "learning_rate": 9.349312175472873e-05,
+      "loss": 2.8204,
+      "step": 6410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5914902091026306,
+      "learning_rate": 9.348303582236142e-05,
+      "loss": 2.9112,
+      "step": 6415
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5125696063041687,
+      "learning_rate": 9.347294262414047e-05,
+      "loss": 2.8516,
+      "step": 6420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5975127220153809,
+      "learning_rate": 9.346284216175245e-05,
+      "loss": 2.8703,
+      "step": 6425
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5590592622756958,
+      "learning_rate": 9.345273443688509e-05,
+      "loss": 2.8579,
+      "step": 6430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.541729748249054,
+      "learning_rate": 9.344261945122734e-05,
+      "loss": 2.8057,
+      "step": 6435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5591338872909546,
+      "learning_rate": 9.343249720646939e-05,
+      "loss": 2.9377,
+      "step": 6440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5825858116149902,
+      "learning_rate": 9.342236770430264e-05,
+      "loss": 2.9228,
+      "step": 6445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5661072731018066,
+      "learning_rate": 9.341223094641967e-05,
+      "loss": 2.9055,
+      "step": 6450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5213721394538879,
+      "learning_rate": 9.340208693451431e-05,
+      "loss": 2.8554,
+      "step": 6455
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5403419137001038,
+      "learning_rate": 9.339193567028157e-05,
+      "loss": 2.8815,
+      "step": 6460
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5399751663208008,
+      "learning_rate": 9.338177715541772e-05,
+      "loss": 2.8742,
+      "step": 6465
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6029462218284607,
+      "learning_rate": 9.337161139162019e-05,
+      "loss": 2.9429,
+      "step": 6470
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5680956244468689,
+      "learning_rate": 9.336143838058761e-05,
+      "loss": 2.9135,
+      "step": 6475
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5575125217437744,
+      "learning_rate": 9.335125812401993e-05,
+      "loss": 2.9406,
+      "step": 6480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5209574699401855,
+      "learning_rate": 9.334107062361816e-05,
+      "loss": 2.9302,
+      "step": 6485
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6443679332733154,
+      "learning_rate": 9.333087588108462e-05,
+      "loss": 2.9296,
+      "step": 6490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5513743162155151,
+      "learning_rate": 9.332067389812282e-05,
+      "loss": 2.8619,
+      "step": 6495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5454977750778198,
+      "learning_rate": 9.331046467643746e-05,
+      "loss": 2.8198,
+      "step": 6500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5524970293045044,
+      "learning_rate": 9.330024821773448e-05,
+      "loss": 2.9063,
+      "step": 6505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5268482565879822,
+      "learning_rate": 9.329002452372102e-05,
+      "loss": 2.8008,
+      "step": 6510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5384489893913269,
+      "learning_rate": 9.327979359610537e-05,
+      "loss": 2.8573,
+      "step": 6515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5311044454574585,
+      "learning_rate": 9.326955543659714e-05,
+      "loss": 2.931,
+      "step": 6520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5381263494491577,
+      "learning_rate": 9.325931004690706e-05,
+      "loss": 2.787,
+      "step": 6525
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5417134165763855,
+      "learning_rate": 9.324905742874709e-05,
+      "loss": 2.9022,
+      "step": 6530
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5814573168754578,
+      "learning_rate": 9.323879758383042e-05,
+      "loss": 2.9384,
+      "step": 6535
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.545272707939148,
+      "learning_rate": 9.322853051387141e-05,
+      "loss": 2.9372,
+      "step": 6540
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5291239023208618,
+      "learning_rate": 9.321825622058568e-05,
+      "loss": 2.9435,
+      "step": 6545
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5472492575645447,
+      "learning_rate": 9.320797470569001e-05,
+      "loss": 2.8316,
+      "step": 6550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5657878518104553,
+      "learning_rate": 9.319768597090242e-05,
+      "loss": 2.9356,
+      "step": 6555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6310515999794006,
+      "learning_rate": 9.318739001794208e-05,
+      "loss": 2.9812,
+      "step": 6560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6317135691642761,
+      "learning_rate": 9.317708684852942e-05,
+      "loss": 2.8366,
+      "step": 6565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5348029136657715,
+      "learning_rate": 9.31667764643861e-05,
+      "loss": 2.9113,
+      "step": 6570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5741630792617798,
+      "learning_rate": 9.31564588672349e-05,
+      "loss": 2.8358,
+      "step": 6575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5488686561584473,
+      "learning_rate": 9.314613405879986e-05,
+      "loss": 2.8389,
+      "step": 6580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5394400358200073,
+      "learning_rate": 9.313580204080623e-05,
+      "loss": 2.8868,
+      "step": 6585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5457869172096252,
+      "learning_rate": 9.312546281498047e-05,
+      "loss": 2.8924,
+      "step": 6590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5890506505966187,
+      "learning_rate": 9.311511638305018e-05,
+      "loss": 2.8109,
+      "step": 6595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5238346457481384,
+      "learning_rate": 9.310476274674425e-05,
+      "loss": 2.907,
+      "step": 6600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5219760537147522,
+      "learning_rate": 9.30944019077927e-05,
+      "loss": 2.8129,
+      "step": 6605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5474616289138794,
+      "learning_rate": 9.30840338679268e-05,
+      "loss": 2.8082,
+      "step": 6610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5662362575531006,
+      "learning_rate": 9.307365862887905e-05,
+      "loss": 2.913,
+      "step": 6615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8119179010391235,
+      "learning_rate": 9.306327619238306e-05,
+      "loss": 2.876,
+      "step": 6620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5428465604782104,
+      "learning_rate": 9.305288656017372e-05,
+      "loss": 2.7774,
+      "step": 6625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6504521369934082,
+      "learning_rate": 9.304248973398711e-05,
+      "loss": 2.9171,
+      "step": 6630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5630825757980347,
+      "learning_rate": 9.303208571556048e-05,
+      "loss": 2.8065,
+      "step": 6635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5979588627815247,
+      "learning_rate": 9.30216745066323e-05,
+      "loss": 2.8975,
+      "step": 6640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5782840847969055,
+      "learning_rate": 9.301125610894226e-05,
+      "loss": 2.8699,
+      "step": 6645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5811892151832581,
+      "learning_rate": 9.300083052423124e-05,
+      "loss": 3.0316,
+      "step": 6650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.56390380859375,
+      "learning_rate": 9.29903977542413e-05,
+      "loss": 2.8826,
+      "step": 6655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5908696055412292,
+      "learning_rate": 9.297995780071571e-05,
+      "loss": 2.9401,
+      "step": 6660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5418515205383301,
+      "learning_rate": 9.296951066539899e-05,
+      "loss": 2.8647,
+      "step": 6665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5257374048233032,
+      "learning_rate": 9.295905635003677e-05,
+      "loss": 2.8311,
+      "step": 6670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5564547777175903,
+      "learning_rate": 9.294859485637595e-05,
+      "loss": 2.8893,
+      "step": 6675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.516284704208374,
+      "learning_rate": 9.29381261861646e-05,
+      "loss": 2.8918,
+      "step": 6680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5883457064628601,
+      "learning_rate": 9.292765034115198e-05,
+      "loss": 2.8268,
+      "step": 6685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5426115393638611,
+      "learning_rate": 9.291716732308862e-05,
+      "loss": 2.8346,
+      "step": 6690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5587972402572632,
+      "learning_rate": 9.290667713372613e-05,
+      "loss": 2.878,
+      "step": 6695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5451183319091797,
+      "learning_rate": 9.289617977481741e-05,
+      "loss": 2.9365,
+      "step": 6700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5555893182754517,
+      "learning_rate": 9.288567524811654e-05,
+      "loss": 2.8256,
+      "step": 6705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5587037801742554,
+      "learning_rate": 9.287516355537876e-05,
+      "loss": 2.7969,
+      "step": 6710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5337013602256775,
+      "learning_rate": 9.286464469836056e-05,
+      "loss": 2.8981,
+      "step": 6715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5971136689186096,
+      "learning_rate": 9.28541186788196e-05,
+      "loss": 2.8801,
+      "step": 6720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5782601237297058,
+      "learning_rate": 9.284358549851471e-05,
+      "loss": 2.8215,
+      "step": 6725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5351136922836304,
+      "learning_rate": 9.283304515920598e-05,
+      "loss": 2.8395,
+      "step": 6730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5437111258506775,
+      "learning_rate": 9.282249766265464e-05,
+      "loss": 2.8448,
+      "step": 6735
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5917949080467224,
+      "learning_rate": 9.281194301062315e-05,
+      "loss": 2.9128,
+      "step": 6740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5078500509262085,
+      "learning_rate": 9.280138120487514e-05,
+      "loss": 2.8121,
+      "step": 6745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6040711998939514,
+      "learning_rate": 9.279081224717547e-05,
+      "loss": 2.8444,
+      "step": 6750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5629652142524719,
+      "learning_rate": 9.278023613929014e-05,
+      "loss": 2.8227,
+      "step": 6755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5609646439552307,
+      "learning_rate": 9.276965288298642e-05,
+      "loss": 2.8349,
+      "step": 6760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.530739426612854,
+      "learning_rate": 9.27590624800327e-05,
+      "loss": 2.8951,
+      "step": 6765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5921422839164734,
+      "learning_rate": 9.27484649321986e-05,
+      "loss": 2.8401,
+      "step": 6770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5754078030586243,
+      "learning_rate": 9.273786024125496e-05,
+      "loss": 2.9617,
+      "step": 6775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5602753162384033,
+      "learning_rate": 9.272724840897375e-05,
+      "loss": 2.8907,
+      "step": 6780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6170728802680969,
+      "learning_rate": 9.271662943712818e-05,
+      "loss": 2.8151,
+      "step": 6785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5295075178146362,
+      "learning_rate": 9.270600332749264e-05,
+      "loss": 2.8476,
+      "step": 6790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5158007144927979,
+      "learning_rate": 9.269537008184273e-05,
+      "loss": 2.9172,
+      "step": 6795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5209262371063232,
+      "learning_rate": 9.268472970195519e-05,
+      "loss": 2.8557,
+      "step": 6800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5245833396911621,
+      "learning_rate": 9.267408218960803e-05,
+      "loss": 2.8591,
+      "step": 6805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5542829036712646,
+      "learning_rate": 9.266342754658036e-05,
+      "loss": 2.8769,
+      "step": 6810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5721807479858398,
+      "learning_rate": 9.265276577465259e-05,
+      "loss": 2.8484,
+      "step": 6815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5594983100891113,
+      "learning_rate": 9.264209687560621e-05,
+      "loss": 2.8968,
+      "step": 6820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.562432050704956,
+      "learning_rate": 9.263142085122399e-05,
+      "loss": 2.8075,
+      "step": 6825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4970925450325012,
+      "learning_rate": 9.262073770328983e-05,
+      "loss": 2.773,
+      "step": 6830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5462027192115784,
+      "learning_rate": 9.261004743358886e-05,
+      "loss": 2.9124,
+      "step": 6835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5766883492469788,
+      "learning_rate": 9.259935004390737e-05,
+      "loss": 2.7665,
+      "step": 6840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5428696870803833,
+      "learning_rate": 9.258864553603286e-05,
+      "loss": 2.8391,
+      "step": 6845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5348033905029297,
+      "learning_rate": 9.257793391175402e-05,
+      "loss": 2.9228,
+      "step": 6850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5248808860778809,
+      "learning_rate": 9.256721517286073e-05,
+      "loss": 2.8814,
+      "step": 6855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6151602268218994,
+      "learning_rate": 9.255648932114403e-05,
+      "loss": 2.9425,
+      "step": 6860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.588208019733429,
+      "learning_rate": 9.254575635839618e-05,
+      "loss": 2.8275,
+      "step": 6865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5270939469337463,
+      "learning_rate": 9.253501628641061e-05,
+      "loss": 2.8899,
+      "step": 6870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5379821062088013,
+      "learning_rate": 9.252426910698198e-05,
+      "loss": 2.8064,
+      "step": 6875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5831663012504578,
+      "learning_rate": 9.251351482190604e-05,
+      "loss": 2.8433,
+      "step": 6880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5916703343391418,
+      "learning_rate": 9.250275343297985e-05,
+      "loss": 2.8036,
+      "step": 6885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6007803082466125,
+      "learning_rate": 9.249198494200158e-05,
+      "loss": 2.9261,
+      "step": 6890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.634512186050415,
+      "learning_rate": 9.24812093507706e-05,
+      "loss": 2.7927,
+      "step": 6895
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6148631572723389,
+      "learning_rate": 9.247042666108747e-05,
+      "loss": 2.8211,
+      "step": 6900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5134620666503906,
+      "learning_rate": 9.245963687475395e-05,
+      "loss": 2.8484,
+      "step": 6905
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5442782044410706,
+      "learning_rate": 9.244883999357293e-05,
+      "loss": 3.0275,
+      "step": 6910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5433021187782288,
+      "learning_rate": 9.24380360193486e-05,
+      "loss": 2.8397,
+      "step": 6915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5350810885429382,
+      "learning_rate": 9.242722495388621e-05,
+      "loss": 2.834,
+      "step": 6920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.540776789188385,
+      "learning_rate": 9.241640679899226e-05,
+      "loss": 2.85,
+      "step": 6925
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5657823085784912,
+      "learning_rate": 9.240558155647443e-05,
+      "loss": 2.7983,
+      "step": 6930
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5670983791351318,
+      "learning_rate": 9.239474922814157e-05,
+      "loss": 2.7851,
+      "step": 6935
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5339034199714661,
+      "learning_rate": 9.238390981580375e-05,
+      "loss": 2.8806,
+      "step": 6940
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5140305161476135,
+      "learning_rate": 9.237306332127215e-05,
+      "loss": 2.8801,
+      "step": 6945
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5349153876304626,
+      "learning_rate": 9.236220974635923e-05,
+      "loss": 2.7963,
+      "step": 6950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5407957434654236,
+      "learning_rate": 9.235134909287853e-05,
+      "loss": 2.776,
+      "step": 6955
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5694733262062073,
+      "learning_rate": 9.234048136264485e-05,
+      "loss": 2.8545,
+      "step": 6960
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5330975651741028,
+      "learning_rate": 9.232960655747416e-05,
+      "loss": 2.8609,
+      "step": 6965
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5424284338951111,
+      "learning_rate": 9.231872467918358e-05,
+      "loss": 2.8309,
+      "step": 6970
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5230695009231567,
+      "learning_rate": 9.230783572959143e-05,
+      "loss": 2.8484,
+      "step": 6975
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5917704105377197,
+      "learning_rate": 9.229693971051723e-05,
+      "loss": 2.9078,
+      "step": 6980
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5976876616477966,
+      "learning_rate": 9.228603662378166e-05,
+      "loss": 2.8314,
+      "step": 6985
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5881170630455017,
+      "learning_rate": 9.227512647120656e-05,
+      "loss": 2.9145,
+      "step": 6990
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5368829369544983,
+      "learning_rate": 9.2264209254615e-05,
+      "loss": 2.8994,
+      "step": 6995
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5506414771080017,
+      "learning_rate": 9.225328497583122e-05,
+      "loss": 2.8926,
+      "step": 7000
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5483494400978088,
+      "learning_rate": 9.224235363668058e-05,
+      "loss": 2.8229,
+      "step": 7005
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5191677808761597,
+      "learning_rate": 9.223141523898972e-05,
+      "loss": 2.7731,
+      "step": 7010
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5449312329292297,
+      "learning_rate": 9.222046978458639e-05,
+      "loss": 2.7811,
+      "step": 7015
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5310829877853394,
+      "learning_rate": 9.22095172752995e-05,
+      "loss": 2.8496,
+      "step": 7020
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5556698441505432,
+      "learning_rate": 9.21985577129592e-05,
+      "loss": 2.8468,
+      "step": 7025
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5422739386558533,
+      "learning_rate": 9.218759109939681e-05,
+      "loss": 2.6908,
+      "step": 7030
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5606404542922974,
+      "learning_rate": 9.217661743644477e-05,
+      "loss": 2.8992,
+      "step": 7035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6334173679351807,
+      "learning_rate": 9.216563672593677e-05,
+      "loss": 2.904,
+      "step": 7040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6318129301071167,
+      "learning_rate": 9.215464896970763e-05,
+      "loss": 2.8923,
+      "step": 7045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.565613329410553,
+      "learning_rate": 9.214365416959336e-05,
+      "loss": 2.8645,
+      "step": 7050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5209188461303711,
+      "learning_rate": 9.213265232743119e-05,
+      "loss": 2.8299,
+      "step": 7055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5941889882087708,
+      "learning_rate": 9.212164344505942e-05,
+      "loss": 2.993,
+      "step": 7060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5412931442260742,
+      "learning_rate": 9.211062752431765e-05,
+      "loss": 2.8088,
+      "step": 7065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5895765423774719,
+      "learning_rate": 9.209960456704656e-05,
+      "loss": 2.8068,
+      "step": 7070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5596634745597839,
+      "learning_rate": 9.208857457508809e-05,
+      "loss": 2.8987,
+      "step": 7075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5735235810279846,
+      "learning_rate": 9.207753755028526e-05,
+      "loss": 2.8749,
+      "step": 7080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5558637380599976,
+      "learning_rate": 9.206649349448235e-05,
+      "loss": 2.9051,
+      "step": 7085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5749927759170532,
+      "learning_rate": 9.205544240952477e-05,
+      "loss": 2.8806,
+      "step": 7090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5449641346931458,
+      "learning_rate": 9.204438429725911e-05,
+      "loss": 2.8354,
+      "step": 7095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5233917832374573,
+      "learning_rate": 9.203331915953314e-05,
+      "loss": 2.8518,
+      "step": 7100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6108044981956482,
+      "learning_rate": 9.202224699819584e-05,
+      "loss": 2.8308,
+      "step": 7105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5810139179229736,
+      "learning_rate": 9.201116781509726e-05,
+      "loss": 2.8212,
+      "step": 7110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.576529860496521,
+      "learning_rate": 9.200008161208875e-05,
+      "loss": 2.9373,
+      "step": 7115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5298864245414734,
+      "learning_rate": 9.198898839102273e-05,
+      "loss": 2.821,
+      "step": 7120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5164542198181152,
+      "learning_rate": 9.197788815375287e-05,
+      "loss": 2.8589,
+      "step": 7125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.579471230506897,
+      "learning_rate": 9.196678090213395e-05,
+      "loss": 2.7931,
+      "step": 7130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234999656677246,
+      "learning_rate": 9.195566663802198e-05,
+      "loss": 2.8561,
+      "step": 7135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.541429340839386,
+      "learning_rate": 9.194454536327408e-05,
+      "loss": 2.9099,
+      "step": 7140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5331956148147583,
+      "learning_rate": 9.19334170797486e-05,
+      "loss": 2.7444,
+      "step": 7145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5507737398147583,
+      "learning_rate": 9.192228178930503e-05,
+      "loss": 2.8387,
+      "step": 7150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.292999744415283,
+      "learning_rate": 9.191113949380402e-05,
+      "loss": 2.8551,
+      "step": 7155
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5325794219970703,
+      "learning_rate": 9.189999019510742e-05,
+      "loss": 2.8291,
+      "step": 7160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5165179967880249,
+      "learning_rate": 9.188883389507823e-05,
+      "loss": 2.7658,
+      "step": 7165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5215129852294922,
+      "learning_rate": 9.187767059558064e-05,
+      "loss": 2.8811,
+      "step": 7170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5586404800415039,
+      "learning_rate": 9.186650029847998e-05,
+      "loss": 2.7863,
+      "step": 7175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5327104926109314,
+      "learning_rate": 9.185532300564276e-05,
+      "loss": 2.8333,
+      "step": 7180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5372282266616821,
+      "learning_rate": 9.184413871893668e-05,
+      "loss": 2.7879,
+      "step": 7185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7631470561027527,
+      "learning_rate": 9.183294744023058e-05,
+      "loss": 2.8815,
+      "step": 7190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5490337014198303,
+      "learning_rate": 9.18217491713945e-05,
+      "loss": 2.8833,
+      "step": 7195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5564482808113098,
+      "learning_rate": 9.181054391429962e-05,
+      "loss": 2.8098,
+      "step": 7200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5813400149345398,
+      "learning_rate": 9.179933167081828e-05,
+      "loss": 2.9471,
+      "step": 7205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5282649993896484,
+      "learning_rate": 9.178811244282404e-05,
+      "loss": 2.8019,
+      "step": 7210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5374988317489624,
+      "learning_rate": 9.177688623219154e-05,
+      "loss": 2.9053,
+      "step": 7215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.509655237197876,
+      "learning_rate": 9.176565304079669e-05,
+      "loss": 2.8354,
+      "step": 7220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5160524845123291,
+      "learning_rate": 9.175441287051648e-05,
+      "loss": 2.7691,
+      "step": 7225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5125441551208496,
+      "learning_rate": 9.174316572322912e-05,
+      "loss": 2.86,
+      "step": 7230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5570948719978333,
+      "learning_rate": 9.173191160081395e-05,
+      "loss": 2.9894,
+      "step": 7235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5379411578178406,
+      "learning_rate": 9.17206505051515e-05,
+      "loss": 2.817,
+      "step": 7240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.541541576385498,
+      "learning_rate": 9.170938243812349e-05,
+      "loss": 2.7718,
+      "step": 7245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.543175995349884,
+      "learning_rate": 9.16981074016127e-05,
+      "loss": 2.7586,
+      "step": 7250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5533475875854492,
+      "learning_rate": 9.16868253975032e-05,
+      "loss": 2.8442,
+      "step": 7255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6192885041236877,
+      "learning_rate": 9.167553642768017e-05,
+      "loss": 2.9001,
+      "step": 7260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5266651511192322,
+      "learning_rate": 9.166424049402993e-05,
+      "loss": 2.9491,
+      "step": 7265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6244593262672424,
+      "learning_rate": 9.165293759844e-05,
+      "loss": 2.7588,
+      "step": 7270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5242811441421509,
+      "learning_rate": 9.164162774279905e-05,
+      "loss": 2.8553,
+      "step": 7275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5429643988609314,
+      "learning_rate": 9.163031092899693e-05,
+      "loss": 2.8308,
+      "step": 7280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5625463128089905,
+      "learning_rate": 9.161898715892463e-05,
+      "loss": 2.9105,
+      "step": 7285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5612660050392151,
+      "learning_rate": 9.160765643447429e-05,
+      "loss": 2.8136,
+      "step": 7290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5403995513916016,
+      "learning_rate": 9.159631875753925e-05,
+      "loss": 2.898,
+      "step": 7295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6147511005401611,
+      "learning_rate": 9.1584974130014e-05,
+      "loss": 2.7274,
+      "step": 7300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6268966794013977,
+      "learning_rate": 9.157362255379419e-05,
+      "loss": 2.8683,
+      "step": 7305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5353759527206421,
+      "learning_rate": 9.156226403077657e-05,
+      "loss": 2.8109,
+      "step": 7310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5682114958763123,
+      "learning_rate": 9.155089856285917e-05,
+      "loss": 2.8289,
+      "step": 7315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5466665625572205,
+      "learning_rate": 9.153952615194111e-05,
+      "loss": 2.8257,
+      "step": 7320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5085710883140564,
+      "learning_rate": 9.152814679992265e-05,
+      "loss": 2.7832,
+      "step": 7325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6378833651542664,
+      "learning_rate": 9.151676050870526e-05,
+      "loss": 2.7981,
+      "step": 7330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5237205028533936,
+      "learning_rate": 9.150536728019154e-05,
+      "loss": 2.8683,
+      "step": 7335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.546835720539093,
+      "learning_rate": 9.149396711628523e-05,
+      "loss": 2.8506,
+      "step": 7340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5346294045448303,
+      "learning_rate": 9.14825600188913e-05,
+      "loss": 2.8396,
+      "step": 7345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5240088105201721,
+      "learning_rate": 9.147114598991583e-05,
+      "loss": 2.8856,
+      "step": 7350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5487208962440491,
+      "learning_rate": 9.145972503126603e-05,
+      "loss": 2.8328,
+      "step": 7355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5299774408340454,
+      "learning_rate": 9.14482971448503e-05,
+      "loss": 2.8321,
+      "step": 7360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5197721123695374,
+      "learning_rate": 9.143686233257824e-05,
+      "loss": 2.8433,
+      "step": 7365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5689733624458313,
+      "learning_rate": 9.142542059636053e-05,
+      "loss": 2.7861,
+      "step": 7370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.521331787109375,
+      "learning_rate": 9.141397193810903e-05,
+      "loss": 2.8419,
+      "step": 7375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6614023447036743,
+      "learning_rate": 9.14025163597368e-05,
+      "loss": 2.8256,
+      "step": 7380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7999863624572754,
+      "learning_rate": 9.1391053863158e-05,
+      "loss": 2.8721,
+      "step": 7385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5129424929618835,
+      "learning_rate": 9.137958445028799e-05,
+      "loss": 2.8369,
+      "step": 7390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5174964070320129,
+      "learning_rate": 9.136810812304326e-05,
+      "loss": 2.6901,
+      "step": 7395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5271317362785339,
+      "learning_rate": 9.135662488334147e-05,
+      "loss": 2.7799,
+      "step": 7400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5830038785934448,
+      "learning_rate": 9.13451347331014e-05,
+      "loss": 2.7278,
+      "step": 7405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.556191086769104,
+      "learning_rate": 9.133363767424303e-05,
+      "loss": 2.923,
+      "step": 7410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5488478541374207,
+      "learning_rate": 9.132213370868748e-05,
+      "loss": 2.8137,
+      "step": 7415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5441930890083313,
+      "learning_rate": 9.131062283835701e-05,
+      "loss": 2.804,
+      "step": 7420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.532233476638794,
+      "learning_rate": 9.129910506517503e-05,
+      "loss": 2.8233,
+      "step": 7425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5188441872596741,
+      "learning_rate": 9.128758039106616e-05,
+      "loss": 2.8715,
+      "step": 7430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5375588536262512,
+      "learning_rate": 9.127604881795609e-05,
+      "loss": 2.8213,
+      "step": 7435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5606820583343506,
+      "learning_rate": 9.126451034777173e-05,
+      "loss": 2.7711,
+      "step": 7440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5495339632034302,
+      "learning_rate": 9.12529649824411e-05,
+      "loss": 2.9294,
+      "step": 7445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5420064330101013,
+      "learning_rate": 9.124141272389337e-05,
+      "loss": 2.7936,
+      "step": 7450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5830769538879395,
+      "learning_rate": 9.122985357405892e-05,
+      "loss": 2.8309,
+      "step": 7455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5509148240089417,
+      "learning_rate": 9.121828753486921e-05,
+      "loss": 2.8605,
+      "step": 7460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5194603800773621,
+      "learning_rate": 9.120671460825691e-05,
+      "loss": 2.8417,
+      "step": 7465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5391156673431396,
+      "learning_rate": 9.119513479615577e-05,
+      "loss": 2.887,
+      "step": 7470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.522241473197937,
+      "learning_rate": 9.118354810050079e-05,
+      "loss": 2.8063,
+      "step": 7475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5224587321281433,
+      "learning_rate": 9.117195452322801e-05,
+      "loss": 2.8037,
+      "step": 7480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5074285864830017,
+      "learning_rate": 9.11603540662747e-05,
+      "loss": 2.8296,
+      "step": 7485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5290721654891968,
+      "learning_rate": 9.114874673157927e-05,
+      "loss": 2.7955,
+      "step": 7490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5563904643058777,
+      "learning_rate": 9.113713252108123e-05,
+      "loss": 2.7988,
+      "step": 7495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.495453804731369,
+      "learning_rate": 9.112551143672127e-05,
+      "loss": 2.8734,
+      "step": 7500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.587644100189209,
+      "learning_rate": 9.111388348044125e-05,
+      "loss": 2.8995,
+      "step": 7505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5910301804542542,
+      "learning_rate": 9.110224865418417e-05,
+      "loss": 2.895,
+      "step": 7510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5398467779159546,
+      "learning_rate": 9.109060695989413e-05,
+      "loss": 2.7004,
+      "step": 7515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8639580011367798,
+      "learning_rate": 9.107895839951643e-05,
+      "loss": 2.8701,
+      "step": 7520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6013845205307007,
+      "learning_rate": 9.106730297499752e-05,
+      "loss": 2.8753,
+      "step": 7525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6004367470741272,
+      "learning_rate": 9.105564068828497e-05,
+      "loss": 2.8129,
+      "step": 7530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5059248208999634,
+      "learning_rate": 9.104397154132747e-05,
+      "loss": 2.783,
+      "step": 7535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5806273818016052,
+      "learning_rate": 9.103229553607493e-05,
+      "loss": 2.8278,
+      "step": 7540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5287309885025024,
+      "learning_rate": 9.102061267447835e-05,
+      "loss": 2.7703,
+      "step": 7545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5301191806793213,
+      "learning_rate": 9.100892295848989e-05,
+      "loss": 2.8761,
+      "step": 7550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5276175141334534,
+      "learning_rate": 9.099722639006287e-05,
+      "loss": 2.9171,
+      "step": 7555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5188120603561401,
+      "learning_rate": 9.098552297115175e-05,
+      "loss": 2.9049,
+      "step": 7560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5469840168952942,
+      "learning_rate": 9.097381270371212e-05,
+      "loss": 2.8379,
+      "step": 7565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5636797547340393,
+      "learning_rate": 9.09620955897007e-05,
+      "loss": 2.7849,
+      "step": 7570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5288849472999573,
+      "learning_rate": 9.09503716310754e-05,
+      "loss": 2.7627,
+      "step": 7575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5372269153594971,
+      "learning_rate": 9.093864082979525e-05,
+      "loss": 2.8875,
+      "step": 7580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5337250232696533,
+      "learning_rate": 9.092690318782042e-05,
+      "loss": 2.7838,
+      "step": 7585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5452127456665039,
+      "learning_rate": 9.091515870711221e-05,
+      "loss": 2.8349,
+      "step": 7590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.527032732963562,
+      "learning_rate": 9.09034073896331e-05,
+      "loss": 2.8127,
+      "step": 7595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5359138250350952,
+      "learning_rate": 9.08916492373467e-05,
+      "loss": 2.6947,
+      "step": 7600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5325815081596375,
+      "learning_rate": 9.087988425221772e-05,
+      "loss": 2.8661,
+      "step": 7605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5182110071182251,
+      "learning_rate": 9.086811243621206e-05,
+      "loss": 2.8391,
+      "step": 7610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5781309604644775,
+      "learning_rate": 9.085633379129677e-05,
+      "loss": 2.8295,
+      "step": 7615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5888844728469849,
+      "learning_rate": 9.084454831943998e-05,
+      "loss": 2.8006,
+      "step": 7620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.531403124332428,
+      "learning_rate": 9.083275602261101e-05,
+      "loss": 2.8714,
+      "step": 7625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5550820231437683,
+      "learning_rate": 9.082095690278034e-05,
+      "loss": 2.8629,
+      "step": 7630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5457887649536133,
+      "learning_rate": 9.080915096191951e-05,
+      "loss": 2.8286,
+      "step": 7635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5800537467002869,
+      "learning_rate": 9.07973382020013e-05,
+      "loss": 2.8973,
+      "step": 7640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5341840982437134,
+      "learning_rate": 9.078551862499953e-05,
+      "loss": 2.7635,
+      "step": 7645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4948440194129944,
+      "learning_rate": 9.077369223288923e-05,
+      "loss": 2.8161,
+      "step": 7650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5851250886917114,
+      "learning_rate": 9.076185902764656e-05,
+      "loss": 2.8624,
+      "step": 7655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5164834856987,
+      "learning_rate": 9.075001901124878e-05,
+      "loss": 2.8077,
+      "step": 7660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5397130250930786,
+      "learning_rate": 9.073817218567431e-05,
+      "loss": 2.8605,
+      "step": 7665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5070657134056091,
+      "learning_rate": 9.072631855290275e-05,
+      "loss": 2.8076,
+      "step": 7670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5288485288619995,
+      "learning_rate": 9.071445811491475e-05,
+      "loss": 2.9258,
+      "step": 7675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5369570851325989,
+      "learning_rate": 9.070259087369218e-05,
+      "loss": 2.8185,
+      "step": 7680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5198988914489746,
+      "learning_rate": 9.069071683121799e-05,
+      "loss": 2.8083,
+      "step": 7685
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5447361469268799,
+      "learning_rate": 9.067883598947628e-05,
+      "loss": 2.8564,
+      "step": 7690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5372093319892883,
+      "learning_rate": 9.066694835045233e-05,
+      "loss": 2.8344,
+      "step": 7695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5328856706619263,
+      "learning_rate": 9.06550539161325e-05,
+      "loss": 2.8692,
+      "step": 7700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5340718030929565,
+      "learning_rate": 9.06431526885043e-05,
+      "loss": 2.7373,
+      "step": 7705
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6088284850120544,
+      "learning_rate": 9.063124466955639e-05,
+      "loss": 2.8487,
+      "step": 7710
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5233157277107239,
+      "learning_rate": 9.061932986127857e-05,
+      "loss": 2.7417,
+      "step": 7715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5209876894950867,
+      "learning_rate": 9.060740826566172e-05,
+      "loss": 2.8215,
+      "step": 7720
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5358843207359314,
+      "learning_rate": 9.059547988469792e-05,
+      "loss": 2.8297,
+      "step": 7725
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5036966800689697,
+      "learning_rate": 9.058354472038037e-05,
+      "loss": 2.682,
+      "step": 7730
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5447607636451721,
+      "learning_rate": 9.057160277470337e-05,
+      "loss": 2.7941,
+      "step": 7735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5820391774177551,
+      "learning_rate": 9.055965404966238e-05,
+      "loss": 2.7844,
+      "step": 7740
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.541779100894928,
+      "learning_rate": 9.054769854725399e-05,
+      "loss": 2.8905,
+      "step": 7745
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5248153209686279,
+      "learning_rate": 9.053573626947593e-05,
+      "loss": 2.851,
+      "step": 7750
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5728263854980469,
+      "learning_rate": 9.052376721832702e-05,
+      "loss": 2.7795,
+      "step": 7755
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5240707993507385,
+      "learning_rate": 9.051179139580727e-05,
+      "loss": 2.8491,
+      "step": 7760
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5872864127159119,
+      "learning_rate": 9.049980880391782e-05,
+      "loss": 2.9385,
+      "step": 7765
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.545243501663208,
+      "learning_rate": 9.048781944466087e-05,
+      "loss": 2.7936,
+      "step": 7770
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5171115398406982,
+      "learning_rate": 9.04758233200398e-05,
+      "loss": 2.7078,
+      "step": 7775
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5274750590324402,
+      "learning_rate": 9.046382043205914e-05,
+      "loss": 2.7646,
+      "step": 7780
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.514201283454895,
+      "learning_rate": 9.045181078272453e-05,
+      "loss": 2.8071,
+      "step": 7785
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5764077305793762,
+      "learning_rate": 9.043979437404273e-05,
+      "loss": 2.7963,
+      "step": 7790
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.551925539970398,
+      "learning_rate": 9.042777120802164e-05,
+      "loss": 2.8139,
+      "step": 7795
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5414553284645081,
+      "learning_rate": 9.041574128667027e-05,
+      "loss": 2.7531,
+      "step": 7800
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.541358470916748,
+      "learning_rate": 9.040370461199879e-05,
+      "loss": 2.8217,
+      "step": 7805
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5357571244239807,
+      "learning_rate": 9.039166118601849e-05,
+      "loss": 2.8039,
+      "step": 7810
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5271549820899963,
+      "learning_rate": 9.037961101074175e-05,
+      "loss": 2.7316,
+      "step": 7815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5658634305000305,
+      "learning_rate": 9.036755408818217e-05,
+      "loss": 2.8135,
+      "step": 7820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5812297463417053,
+      "learning_rate": 9.035549042035434e-05,
+      "loss": 2.8535,
+      "step": 7825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5334533452987671,
+      "learning_rate": 9.03434200092741e-05,
+      "loss": 2.8095,
+      "step": 7830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.550849199295044,
+      "learning_rate": 9.033134285695837e-05,
+      "loss": 2.9269,
+      "step": 7835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5268474221229553,
+      "learning_rate": 9.031925896542521e-05,
+      "loss": 2.8311,
+      "step": 7840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5316630601882935,
+      "learning_rate": 9.030716833669375e-05,
+      "loss": 2.8665,
+      "step": 7845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5346488356590271,
+      "learning_rate": 9.029507097278433e-05,
+      "loss": 2.7669,
+      "step": 7850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5898863673210144,
+      "learning_rate": 9.028296687571834e-05,
+      "loss": 2.8992,
+      "step": 7855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5821563601493835,
+      "learning_rate": 9.027085604751835e-05,
+      "loss": 2.6869,
+      "step": 7860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.512016773223877,
+      "learning_rate": 9.025873849020805e-05,
+      "loss": 2.8073,
+      "step": 7865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6339289546012878,
+      "learning_rate": 9.024661420581222e-05,
+      "loss": 2.756,
+      "step": 7870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5226809978485107,
+      "learning_rate": 9.023448319635676e-05,
+      "loss": 2.8248,
+      "step": 7875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.032348871231079,
+      "learning_rate": 9.022234546386877e-05,
+      "loss": 2.758,
+      "step": 7880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5232419967651367,
+      "learning_rate": 9.021020101037637e-05,
+      "loss": 2.8176,
+      "step": 7885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5480360984802246,
+      "learning_rate": 9.019804983790888e-05,
+      "loss": 2.7988,
+      "step": 7890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5680719017982483,
+      "learning_rate": 9.018589194849672e-05,
+      "loss": 2.8056,
+      "step": 7895
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5436723828315735,
+      "learning_rate": 9.01737273441714e-05,
+      "loss": 2.7914,
+      "step": 7900
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6553776860237122,
+      "learning_rate": 9.016155602696559e-05,
+      "loss": 2.8646,
+      "step": 7905
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5426666736602783,
+      "learning_rate": 9.01493779989131e-05,
+      "loss": 2.758,
+      "step": 7910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5348551273345947,
+      "learning_rate": 9.013719326204881e-05,
+      "loss": 2.773,
+      "step": 7915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5224864482879639,
+      "learning_rate": 9.012500181840874e-05,
+      "loss": 2.8222,
+      "step": 7920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5326815843582153,
+      "learning_rate": 9.011280367003003e-05,
+      "loss": 2.8439,
+      "step": 7925
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.527374804019928,
+      "learning_rate": 9.010059881895099e-05,
+      "loss": 2.7859,
+      "step": 7930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5139932036399841,
+      "learning_rate": 9.008838726721094e-05,
+      "loss": 2.8646,
+      "step": 7935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.49312150478363037,
+      "learning_rate": 9.007616901685042e-05,
+      "loss": 2.8657,
+      "step": 7940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5010385513305664,
+      "learning_rate": 9.006394406991108e-05,
+      "loss": 2.7699,
+      "step": 7945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5484030842781067,
+      "learning_rate": 9.00517124284356e-05,
+      "loss": 2.7366,
+      "step": 7950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5252638459205627,
+      "learning_rate": 9.003947409446789e-05,
+      "loss": 2.7718,
+      "step": 7955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5393187999725342,
+      "learning_rate": 9.002722907005292e-05,
+      "loss": 2.7715,
+      "step": 7960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5177910327911377,
+      "learning_rate": 9.001497735723679e-05,
+      "loss": 2.7457,
+      "step": 7965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.621920645236969,
+      "learning_rate": 9.000271895806669e-05,
+      "loss": 2.7298,
+      "step": 7970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.49828028678894043,
+      "learning_rate": 8.9990453874591e-05,
+      "loss": 2.7588,
+      "step": 7975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.497603178024292,
+      "learning_rate": 8.997818210885912e-05,
+      "loss": 2.7177,
+      "step": 7980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5158050656318665,
+      "learning_rate": 8.996590366292164e-05,
+      "loss": 2.8138,
+      "step": 7985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.525467574596405,
+      "learning_rate": 8.995361853883026e-05,
+      "loss": 2.8319,
+      "step": 7990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5405153632164001,
+      "learning_rate": 8.994132673863774e-05,
+      "loss": 2.8167,
+      "step": 7995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.543835461139679,
+      "learning_rate": 8.992902826439802e-05,
+      "loss": 2.8324,
+      "step": 8000
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 38527,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 2.126010100010189e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}