diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16833 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6633499170812603,
+  "eval_steps": 500,
+  "global_step": 12000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002763957987838585,
+      "grad_norm": 15.539673805236816,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 11.227,
+      "step": 5
+    },
+    {
+      "epoch": 0.000552791597567717,
+      "grad_norm": 16.613014221191406,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 11.1624,
+      "step": 10
+    },
+    {
+      "epoch": 0.0008291873963515755,
+      "grad_norm": 16.244783401489258,
+      "learning_rate": 1.5e-06,
+      "loss": 11.0487,
+      "step": 15
+    },
+    {
+      "epoch": 0.001105583195135434,
+      "grad_norm": 12.021380424499512,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 10.8436,
+      "step": 20
+    },
+    {
+      "epoch": 0.0013819789939192924,
+      "grad_norm": 11.764349937438965,
+      "learning_rate": 2.5e-06,
+      "loss": 10.637,
+      "step": 25
+    },
+    {
+      "epoch": 0.001658374792703151,
+      "grad_norm": 8.930374145507812,
+      "learning_rate": 3e-06,
+      "loss": 10.4451,
+      "step": 30
+    },
+    {
+      "epoch": 0.0019347705914870095,
+      "grad_norm": 6.179296016693115,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 10.2835,
+      "step": 35
+    },
+    {
+      "epoch": 0.002211166390270868,
+      "grad_norm": 5.235034942626953,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 10.1911,
+      "step": 40
+    },
+    {
+      "epoch": 0.0024875621890547263,
+      "grad_norm": 4.519562244415283,
+      "learning_rate": 4.5e-06,
+      "loss": 10.1875,
+      "step": 45
+    },
+    {
+      "epoch": 0.002763957987838585,
+      "grad_norm": 4.634215354919434,
+      "learning_rate": 5e-06,
+      "loss": 9.9655,
+      "step": 50
+    },
+    {
+      "epoch": 0.0030403537866224434,
+      "grad_norm": 4.817412853240967,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 9.9965,
+      "step": 55
+    },
+    {
+      "epoch": 0.003316749585406302,
+      "grad_norm": 4.490515232086182,
+      "learning_rate": 6e-06,
+      "loss": 9.9253,
+      "step": 60
+    },
+    {
+      "epoch": 0.0035931453841901604,
+      "grad_norm": 3.931199789047241,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 9.9455,
+      "step": 65
+    },
+    {
+      "epoch": 0.003869541182974019,
+      "grad_norm": 4.179060935974121,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 9.7904,
+      "step": 70
+    },
+    {
+      "epoch": 0.0041459369817578775,
+      "grad_norm": 4.031132698059082,
+      "learning_rate": 7.5e-06,
+      "loss": 9.7358,
+      "step": 75
+    },
+    {
+      "epoch": 0.004422332780541736,
+      "grad_norm": 3.7969465255737305,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 9.7051,
+      "step": 80
+    },
+    {
+      "epoch": 0.0046987285793255945,
+      "grad_norm": 4.057478904724121,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 9.6854,
+      "step": 85
+    },
+    {
+      "epoch": 0.004975124378109453,
+      "grad_norm": 4.059418201446533,
+      "learning_rate": 9e-06,
+      "loss": 9.6271,
+      "step": 90
+    },
+    {
+      "epoch": 0.005251520176893312,
+      "grad_norm": 4.01811408996582,
+      "learning_rate": 9.5e-06,
+      "loss": 9.614,
+      "step": 95
+    },
+    {
+      "epoch": 0.00552791597567717,
+      "grad_norm": 4.239936828613281,
+      "learning_rate": 1e-05,
+      "loss": 9.423,
+      "step": 100
+    },
+    {
+      "epoch": 0.005804311774461028,
+      "grad_norm": 3.8491415977478027,
+      "learning_rate": 1.05e-05,
+      "loss": 9.5076,
+      "step": 105
+    },
+    {
+      "epoch": 0.006080707573244887,
+      "grad_norm": 3.6000020503997803,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 9.5322,
+      "step": 110
+    },
+    {
+      "epoch": 0.006357103372028745,
+      "grad_norm": 5.558045387268066,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 9.2759,
+      "step": 115
+    },
+    {
+      "epoch": 0.006633499170812604,
+      "grad_norm": 4.2432637214660645,
+      "learning_rate": 1.2e-05,
+      "loss": 9.3174,
+      "step": 120
+    },
+    {
+      "epoch": 0.006909894969596462,
+      "grad_norm": 3.7061004638671875,
+      "learning_rate": 1.25e-05,
+      "loss": 9.3099,
+      "step": 125
+    },
+    {
+      "epoch": 0.007186290768380321,
+      "grad_norm": 3.863173007965088,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 9.0892,
+      "step": 130
+    },
+    {
+      "epoch": 0.007462686567164179,
+      "grad_norm": 4.161376953125,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 9.0976,
+      "step": 135
+    },
+    {
+      "epoch": 0.007739082365948038,
+      "grad_norm": 3.59145188331604,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 9.2131,
+      "step": 140
+    },
+    {
+      "epoch": 0.008015478164731896,
+      "grad_norm": 3.4447615146636963,
+      "learning_rate": 1.45e-05,
+      "loss": 8.9092,
+      "step": 145
+    },
+    {
+      "epoch": 0.008291873963515755,
+      "grad_norm": 3.227489471435547,
+      "learning_rate": 1.5e-05,
+      "loss": 8.927,
+      "step": 150
+    },
+    {
+      "epoch": 0.008568269762299612,
+      "grad_norm": 3.0332329273223877,
+      "learning_rate": 1.55e-05,
+      "loss": 8.8161,
+      "step": 155
+    },
+    {
+      "epoch": 0.008844665561083471,
+      "grad_norm": 4.128123760223389,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 8.8824,
+      "step": 160
+    },
+    {
+      "epoch": 0.00912106135986733,
+      "grad_norm": 3.798156499862671,
+      "learning_rate": 1.65e-05,
+      "loss": 8.7461,
+      "step": 165
+    },
+    {
+      "epoch": 0.009397457158651189,
+      "grad_norm": 4.008655548095703,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 8.5756,
+      "step": 170
+    },
+    {
+      "epoch": 0.009673852957435046,
+      "grad_norm": 3.501603603363037,
+      "learning_rate": 1.75e-05,
+      "loss": 8.7104,
+      "step": 175
+    },
+    {
+      "epoch": 0.009950248756218905,
+      "grad_norm": 3.0541610717773438,
+      "learning_rate": 1.8e-05,
+      "loss": 8.6184,
+      "step": 180
+    },
+    {
+      "epoch": 0.010226644555002764,
+      "grad_norm": 3.2599024772644043,
+      "learning_rate": 1.85e-05,
+      "loss": 8.571,
+      "step": 185
+    },
+    {
+      "epoch": 0.010503040353786623,
+      "grad_norm": 2.8244316577911377,
+      "learning_rate": 1.9e-05,
+      "loss": 8.5187,
+      "step": 190
+    },
+    {
+      "epoch": 0.01077943615257048,
+      "grad_norm": 3.0465221405029297,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 8.5168,
+      "step": 195
+    },
+    {
+      "epoch": 0.01105583195135434,
+      "grad_norm": 3.2354133129119873,
+      "learning_rate": 2e-05,
+      "loss": 8.2329,
+      "step": 200
+    },
+    {
+      "epoch": 0.011332227750138198,
+      "grad_norm": 2.9300291538238525,
+      "learning_rate": 2.05e-05,
+      "loss": 8.3095,
+      "step": 205
+    },
+    {
+      "epoch": 0.011608623548922056,
+      "grad_norm": 3.2358179092407227,
+      "learning_rate": 2.1e-05,
+      "loss": 8.3495,
+      "step": 210
+    },
+    {
+      "epoch": 0.011885019347705915,
+      "grad_norm": 3.006814479827881,
+      "learning_rate": 2.15e-05,
+      "loss": 8.068,
+      "step": 215
+    },
+    {
+      "epoch": 0.012161415146489774,
+      "grad_norm": 3.092186689376831,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 8.0844,
+      "step": 220
+    },
+    {
+      "epoch": 0.012437810945273632,
+      "grad_norm": 2.7867186069488525,
+      "learning_rate": 2.25e-05,
+      "loss": 7.7844,
+      "step": 225
+    },
+    {
+      "epoch": 0.01271420674405749,
+      "grad_norm": 3.151169538497925,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 8.0737,
+      "step": 230
+    },
+    {
+      "epoch": 0.012990602542841349,
+      "grad_norm": 2.7488925457000732,
+      "learning_rate": 2.35e-05,
+      "loss": 8.0835,
+      "step": 235
+    },
+    {
+      "epoch": 0.013266998341625208,
+      "grad_norm": 2.673896312713623,
+      "learning_rate": 2.4e-05,
+      "loss": 7.9807,
+      "step": 240
+    },
+    {
+      "epoch": 0.013543394140409067,
+      "grad_norm": 3.0482277870178223,
+      "learning_rate": 2.45e-05,
+      "loss": 7.7995,
+      "step": 245
+    },
+    {
+      "epoch": 0.013819789939192924,
+      "grad_norm": 2.6868321895599365,
+      "learning_rate": 2.5e-05,
+      "loss": 7.9454,
+      "step": 250
+    },
+    {
+      "epoch": 0.014096185737976783,
+      "grad_norm": 2.6580042839050293,
+      "learning_rate": 2.5500000000000003e-05,
+      "loss": 7.7582,
+      "step": 255
+    },
+    {
+      "epoch": 0.014372581536760642,
+      "grad_norm": 2.2857472896575928,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 7.784,
+      "step": 260
+    },
+    {
+      "epoch": 0.014648977335544499,
+      "grad_norm": 2.3364319801330566,
+      "learning_rate": 2.6500000000000004e-05,
+      "loss": 7.8675,
+      "step": 265
+    },
+    {
+      "epoch": 0.014925373134328358,
+      "grad_norm": 2.896152973175049,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 7.6433,
+      "step": 270
+    },
+    {
+      "epoch": 0.015201768933112217,
+      "grad_norm": 3.6049489974975586,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 7.4608,
+      "step": 275
+    },
+    {
+      "epoch": 0.015478164731896076,
+      "grad_norm": 2.9428746700286865,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 7.4888,
+      "step": 280
+    },
+    {
+      "epoch": 0.015754560530679935,
+      "grad_norm": 3.2422735691070557,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 7.4838,
+      "step": 285
+    },
+    {
+      "epoch": 0.016030956329463792,
+      "grad_norm": 2.6515588760375977,
+      "learning_rate": 2.9e-05,
+      "loss": 7.4845,
+      "step": 290
+    },
+    {
+      "epoch": 0.01630735212824765,
+      "grad_norm": 2.2527542114257812,
+      "learning_rate": 2.95e-05,
+      "loss": 7.679,
+      "step": 295
+    },
+    {
+      "epoch": 0.01658374792703151,
+      "grad_norm": 2.841431140899658,
+      "learning_rate": 3e-05,
+      "loss": 7.5644,
+      "step": 300
+    },
+    {
+      "epoch": 0.016860143725815367,
+      "grad_norm": 3.071385383605957,
+      "learning_rate": 3.05e-05,
+      "loss": 7.3476,
+      "step": 305
+    },
+    {
+      "epoch": 0.017136539524599224,
+      "grad_norm": 2.392310380935669,
+      "learning_rate": 3.1e-05,
+      "loss": 7.4821,
+      "step": 310
+    },
+    {
+      "epoch": 0.017412935323383085,
+      "grad_norm": 2.3450920581817627,
+      "learning_rate": 3.15e-05,
+      "loss": 7.6444,
+      "step": 315
+    },
+    {
+      "epoch": 0.017689331122166942,
+      "grad_norm": 2.5122437477111816,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 7.3695,
+      "step": 320
+    },
+    {
+      "epoch": 0.017965726920950803,
+      "grad_norm": 3.0981411933898926,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 7.2349,
+      "step": 325
+    },
+    {
+      "epoch": 0.01824212271973466,
+      "grad_norm": 2.4662463665008545,
+      "learning_rate": 3.3e-05,
+      "loss": 7.4194,
+      "step": 330
+    },
+    {
+      "epoch": 0.018518518518518517,
+      "grad_norm": 2.5492987632751465,
+      "learning_rate": 3.35e-05,
+      "loss": 7.2674,
+      "step": 335
+    },
+    {
+      "epoch": 0.018794914317302378,
+      "grad_norm": 2.4681410789489746,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 7.1909,
+      "step": 340
+    },
+    {
+      "epoch": 0.019071310116086235,
+      "grad_norm": 2.5183534622192383,
+      "learning_rate": 3.45e-05,
+      "loss": 7.3093,
+      "step": 345
+    },
+    {
+      "epoch": 0.019347705914870093,
+      "grad_norm": 2.3964405059814453,
+      "learning_rate": 3.5e-05,
+      "loss": 7.3742,
+      "step": 350
+    },
+    {
+      "epoch": 0.019624101713653953,
+      "grad_norm": 2.961442470550537,
+      "learning_rate": 3.55e-05,
+      "loss": 7.4874,
+      "step": 355
+    },
+    {
+      "epoch": 0.01990049751243781,
+      "grad_norm": 2.420823812484741,
+      "learning_rate": 3.6e-05,
+      "loss": 7.3551,
+      "step": 360
+    },
+    {
+      "epoch": 0.020176893311221668,
+      "grad_norm": 2.382369041442871,
+      "learning_rate": 3.65e-05,
+      "loss": 7.2977,
+      "step": 365
+    },
+    {
+      "epoch": 0.02045328911000553,
+      "grad_norm": 2.913240671157837,
+      "learning_rate": 3.7e-05,
+      "loss": 7.3576,
+      "step": 370
+    },
+    {
+      "epoch": 0.020729684908789386,
+      "grad_norm": 2.3969368934631348,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 7.096,
+      "step": 375
+    },
+    {
+      "epoch": 0.021006080707573246,
+      "grad_norm": 3.0736069679260254,
+      "learning_rate": 3.8e-05,
+      "loss": 7.1848,
+      "step": 380
+    },
+    {
+      "epoch": 0.021282476506357104,
+      "grad_norm": 2.2005414962768555,
+      "learning_rate": 3.85e-05,
+      "loss": 7.082,
+      "step": 385
+    },
+    {
+      "epoch": 0.02155887230514096,
+      "grad_norm": 3.3333756923675537,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 7.1535,
+      "step": 390
+    },
+    {
+      "epoch": 0.02183526810392482,
+      "grad_norm": 3.055220127105713,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 6.8787,
+      "step": 395
+    },
+    {
+      "epoch": 0.02211166390270868,
+      "grad_norm": 2.4127659797668457,
+      "learning_rate": 4e-05,
+      "loss": 6.9886,
+      "step": 400
+    },
+    {
+      "epoch": 0.022388059701492536,
+      "grad_norm": 2.8047995567321777,
+      "learning_rate": 4.05e-05,
+      "loss": 7.0981,
+      "step": 405
+    },
+    {
+      "epoch": 0.022664455500276397,
+      "grad_norm": 2.9609084129333496,
+      "learning_rate": 4.1e-05,
+      "loss": 6.9062,
+      "step": 410
+    },
+    {
+      "epoch": 0.022940851299060254,
+      "grad_norm": 3.0030362606048584,
+      "learning_rate": 4.15e-05,
+      "loss": 7.1212,
+      "step": 415
+    },
+    {
+      "epoch": 0.02321724709784411,
+      "grad_norm": 3.864861011505127,
+      "learning_rate": 4.2e-05,
+      "loss": 7.2378,
+      "step": 420
+    },
+    {
+      "epoch": 0.023493642896627972,
+      "grad_norm": 2.880798816680908,
+      "learning_rate": 4.25e-05,
+      "loss": 7.073,
+      "step": 425
+    },
+    {
+      "epoch": 0.02377003869541183,
+      "grad_norm": 2.903695821762085,
+      "learning_rate": 4.3e-05,
+      "loss": 7.0507,
+      "step": 430
+    },
+    {
+      "epoch": 0.02404643449419569,
+      "grad_norm": 3.056842088699341,
+      "learning_rate": 4.35e-05,
+      "loss": 6.843,
+      "step": 435
+    },
+    {
+      "epoch": 0.024322830292979547,
+      "grad_norm": 2.9644432067871094,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 7.0322,
+      "step": 440
+    },
+    {
+      "epoch": 0.024599226091763404,
+      "grad_norm": 2.783032178878784,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 7.0572,
+      "step": 445
+    },
+    {
+      "epoch": 0.024875621890547265,
+      "grad_norm": 2.3176393508911133,
+      "learning_rate": 4.5e-05,
+      "loss": 6.8992,
+      "step": 450
+    },
+    {
+      "epoch": 0.025152017689331122,
+      "grad_norm": 2.587258815765381,
+      "learning_rate": 4.55e-05,
+      "loss": 6.9982,
+      "step": 455
+    },
+    {
+      "epoch": 0.02542841348811498,
+      "grad_norm": 3.1604034900665283,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 6.8609,
+      "step": 460
+    },
+    {
+      "epoch": 0.02570480928689884,
+      "grad_norm": 3.7053439617156982,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 7.4181,
+      "step": 465
+    },
+    {
+      "epoch": 0.025981205085682697,
+      "grad_norm": 2.896162986755371,
+      "learning_rate": 4.7e-05,
+      "loss": 6.7783,
+      "step": 470
+    },
+    {
+      "epoch": 0.026257600884466555,
+      "grad_norm": 2.4549975395202637,
+      "learning_rate": 4.75e-05,
+      "loss": 6.8819,
+      "step": 475
+    },
+    {
+      "epoch": 0.026533996683250415,
+      "grad_norm": 2.6171348094940186,
+      "learning_rate": 4.8e-05,
+      "loss": 7.0355,
+      "step": 480
+    },
+    {
+      "epoch": 0.026810392482034272,
+      "grad_norm": 2.6561520099639893,
+      "learning_rate": 4.85e-05,
+      "loss": 6.8347,
+      "step": 485
+    },
+    {
+      "epoch": 0.027086788280818133,
+      "grad_norm": 2.9669582843780518,
+      "learning_rate": 4.9e-05,
+      "loss": 6.7402,
+      "step": 490
+    },
+    {
+      "epoch": 0.02736318407960199,
+      "grad_norm": 2.837339162826538,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 6.9979,
+      "step": 495
+    },
+    {
+      "epoch": 0.027639579878385848,
+      "grad_norm": 2.7479424476623535,
+      "learning_rate": 5e-05,
+      "loss": 6.8928,
+      "step": 500
+    },
+    {
+      "epoch": 0.02791597567716971,
+      "grad_norm": 3.010474920272827,
+      "learning_rate": 4.9985787379192726e-05,
+      "loss": 7.0164,
+      "step": 505
+    },
+    {
+      "epoch": 0.028192371475953566,
+      "grad_norm": 2.812869071960449,
+      "learning_rate": 4.997157475838545e-05,
+      "loss": 6.9878,
+      "step": 510
+    },
+    {
+      "epoch": 0.028468767274737423,
+      "grad_norm": 2.1367461681365967,
+      "learning_rate": 4.9957362137578174e-05,
+      "loss": 7.0706,
+      "step": 515
+    },
+    {
+      "epoch": 0.028745163073521283,
+      "grad_norm": 2.864809989929199,
+      "learning_rate": 4.994314951677089e-05,
+      "loss": 6.9731,
+      "step": 520
+    },
+    {
+      "epoch": 0.02902155887230514,
+      "grad_norm": 2.8308370113372803,
+      "learning_rate": 4.9928936895963616e-05,
+      "loss": 6.8747,
+      "step": 525
+    },
+    {
+      "epoch": 0.029297954671088998,
+      "grad_norm": 3.112752676010132,
+      "learning_rate": 4.9914724275156346e-05,
+      "loss": 6.858,
+      "step": 530
+    },
+    {
+      "epoch": 0.02957435046987286,
+      "grad_norm": 2.711458683013916,
+      "learning_rate": 4.9900511654349063e-05,
+      "loss": 7.0695,
+      "step": 535
+    },
+    {
+      "epoch": 0.029850746268656716,
+      "grad_norm": 3.2045886516571045,
+      "learning_rate": 4.988629903354179e-05,
+      "loss": 7.0511,
+      "step": 540
+    },
+    {
+      "epoch": 0.030127142067440577,
+      "grad_norm": 2.655226707458496,
+      "learning_rate": 4.987208641273451e-05,
+      "loss": 7.1596,
+      "step": 545
+    },
+    {
+      "epoch": 0.030403537866224434,
+      "grad_norm": 3.0165657997131348,
+      "learning_rate": 4.9857873791927235e-05,
+      "loss": 7.2406,
+      "step": 550
+    },
+    {
+      "epoch": 0.03067993366500829,
+      "grad_norm": 3.2889368534088135,
+      "learning_rate": 4.984366117111995e-05,
+      "loss": 7.1024,
+      "step": 555
+    },
+    {
+      "epoch": 0.03095632946379215,
+      "grad_norm": 3.375981569290161,
+      "learning_rate": 4.9829448550312677e-05,
+      "loss": 6.8454,
+      "step": 560
+    },
+    {
+      "epoch": 0.03123272526257601,
+      "grad_norm": 3.135676383972168,
+      "learning_rate": 4.981523592950541e-05,
+      "loss": 7.3552,
+      "step": 565
+    },
+    {
+      "epoch": 0.03150912106135987,
+      "grad_norm": 3.073650598526001,
+      "learning_rate": 4.9801023308698125e-05,
+      "loss": 7.1633,
+      "step": 570
+    },
+    {
+      "epoch": 0.03178551686014373,
+      "grad_norm": 2.4921205043792725,
+      "learning_rate": 4.978681068789085e-05,
+      "loss": 7.0104,
+      "step": 575
+    },
+    {
+      "epoch": 0.032061912658927584,
+      "grad_norm": 2.602283477783203,
+      "learning_rate": 4.977259806708357e-05,
+      "loss": 6.7988,
+      "step": 580
+    },
+    {
+      "epoch": 0.03233830845771144,
+      "grad_norm": 2.6181817054748535,
+      "learning_rate": 4.9758385446276296e-05,
+      "loss": 6.9489,
+      "step": 585
+    },
+    {
+      "epoch": 0.0326147042564953,
+      "grad_norm": 2.7632694244384766,
+      "learning_rate": 4.974417282546902e-05,
+      "loss": 6.8527,
+      "step": 590
+    },
+    {
+      "epoch": 0.03289110005527916,
+      "grad_norm": 2.530285120010376,
+      "learning_rate": 4.972996020466174e-05,
+      "loss": 6.902,
+      "step": 595
+    },
+    {
+      "epoch": 0.03316749585406302,
+      "grad_norm": 2.339693784713745,
+      "learning_rate": 4.971574758385447e-05,
+      "loss": 6.7824,
+      "step": 600
+    },
+    {
+      "epoch": 0.03344389165284688,
+      "grad_norm": 2.539701223373413,
+      "learning_rate": 4.9701534963047186e-05,
+      "loss": 6.7379,
+      "step": 605
+    },
+    {
+      "epoch": 0.033720287451630734,
+      "grad_norm": 2.586127758026123,
+      "learning_rate": 4.968732234223991e-05,
+      "loss": 7.1041,
+      "step": 610
+    },
+    {
+      "epoch": 0.03399668325041459,
+      "grad_norm": 2.9324231147766113,
+      "learning_rate": 4.9673109721432634e-05,
+      "loss": 6.778,
+      "step": 615
+    },
+    {
+      "epoch": 0.03427307904919845,
+      "grad_norm": 2.5098533630371094,
+      "learning_rate": 4.965889710062536e-05,
+      "loss": 6.8565,
+      "step": 620
+    },
+    {
+      "epoch": 0.03454947484798231,
+      "grad_norm": 3.0486905574798584,
+      "learning_rate": 4.964468447981808e-05,
+      "loss": 6.6843,
+      "step": 625
+    },
+    {
+      "epoch": 0.03482587064676617,
+      "grad_norm": 2.632387161254883,
+      "learning_rate": 4.96304718590108e-05,
+      "loss": 6.7764,
+      "step": 630
+    },
+    {
+      "epoch": 0.03510226644555003,
+      "grad_norm": 2.5123403072357178,
+      "learning_rate": 4.961625923820353e-05,
+      "loss": 6.7861,
+      "step": 635
+    },
+    {
+      "epoch": 0.035378662244333885,
+      "grad_norm": 2.678485631942749,
+      "learning_rate": 4.9602046617396253e-05,
+      "loss": 6.7484,
+      "step": 640
+    },
+    {
+      "epoch": 0.03565505804311774,
+      "grad_norm": 2.7005465030670166,
+      "learning_rate": 4.958783399658897e-05,
+      "loss": 6.7331,
+      "step": 645
+    },
+    {
+      "epoch": 0.035931453841901606,
+      "grad_norm": 2.8947432041168213,
+      "learning_rate": 4.95736213757817e-05,
+      "loss": 6.7002,
+      "step": 650
+    },
+    {
+      "epoch": 0.03620784964068546,
+      "grad_norm": 2.788076162338257,
+      "learning_rate": 4.955940875497442e-05,
+      "loss": 6.9371,
+      "step": 655
+    },
+    {
+      "epoch": 0.03648424543946932,
+      "grad_norm": 2.3308234214782715,
+      "learning_rate": 4.954519613416714e-05,
+      "loss": 6.7643,
+      "step": 660
+    },
+    {
+      "epoch": 0.03676064123825318,
+      "grad_norm": 2.0851082801818848,
+      "learning_rate": 4.953098351335986e-05,
+      "loss": 6.5921,
+      "step": 665
+    },
+    {
+      "epoch": 0.037037037037037035,
+      "grad_norm": 2.5905802249908447,
+      "learning_rate": 4.951677089255259e-05,
+      "loss": 6.7768,
+      "step": 670
+    },
+    {
+      "epoch": 0.03731343283582089,
+      "grad_norm": 2.8827717304229736,
+      "learning_rate": 4.9502558271745315e-05,
+      "loss": 6.7569,
+      "step": 675
+    },
+    {
+      "epoch": 0.037589828634604756,
+      "grad_norm": 4.079367160797119,
+      "learning_rate": 4.948834565093803e-05,
+      "loss": 6.8087,
+      "step": 680
+    },
+    {
+      "epoch": 0.037866224433388614,
+      "grad_norm": 2.45166277885437,
+      "learning_rate": 4.947413303013076e-05,
+      "loss": 6.7949,
+      "step": 685
+    },
+    {
+      "epoch": 0.03814262023217247,
+      "grad_norm": 3.1089086532592773,
+      "learning_rate": 4.945992040932348e-05,
+      "loss": 6.5695,
+      "step": 690
+    },
+    {
+      "epoch": 0.03841901603095633,
+      "grad_norm": 2.3953475952148438,
+      "learning_rate": 4.9445707788516204e-05,
+      "loss": 6.7555,
+      "step": 695
+    },
+    {
+      "epoch": 0.038695411829740185,
+      "grad_norm": 3.1515231132507324,
+      "learning_rate": 4.943149516770893e-05,
+      "loss": 6.7233,
+      "step": 700
+    },
+    {
+      "epoch": 0.03897180762852405,
+      "grad_norm": 2.5655906200408936,
+      "learning_rate": 4.941728254690165e-05,
+      "loss": 6.8136,
+      "step": 705
+    },
+    {
+      "epoch": 0.03924820342730791,
+      "grad_norm": 2.360464096069336,
+      "learning_rate": 4.9403069926094376e-05,
+      "loss": 6.4072,
+      "step": 710
+    },
+    {
+      "epoch": 0.039524599226091764,
+      "grad_norm": 2.64111590385437,
+      "learning_rate": 4.938885730528709e-05,
+      "loss": 6.9907,
+      "step": 715
+    },
+    {
+      "epoch": 0.03980099502487562,
+      "grad_norm": 2.419311761856079,
+      "learning_rate": 4.9374644684479824e-05,
+      "loss": 6.6928,
+      "step": 720
+    },
+    {
+      "epoch": 0.04007739082365948,
+      "grad_norm": 3.4732820987701416,
+      "learning_rate": 4.936043206367255e-05,
+      "loss": 6.7389,
+      "step": 725
+    },
+    {
+      "epoch": 0.040353786622443336,
+      "grad_norm": 2.9171254634857178,
+      "learning_rate": 4.9346219442865265e-05,
+      "loss": 6.7079,
+      "step": 730
+    },
+    {
+      "epoch": 0.0406301824212272,
+      "grad_norm": 3.3903751373291016,
+      "learning_rate": 4.933200682205799e-05,
+      "loss": 6.4443,
+      "step": 735
+    },
+    {
+      "epoch": 0.04090657822001106,
+      "grad_norm": 3.01202130317688,
+      "learning_rate": 4.931779420125071e-05,
+      "loss": 7.024,
+      "step": 740
+    },
+    {
+      "epoch": 0.041182974018794914,
+      "grad_norm": 2.3956549167633057,
+      "learning_rate": 4.930358158044344e-05,
+      "loss": 6.4869,
+      "step": 745
+    },
+    {
+      "epoch": 0.04145936981757877,
+      "grad_norm": 2.6889407634735107,
+      "learning_rate": 4.928936895963616e-05,
+      "loss": 6.5392,
+      "step": 750
+    },
+    {
+      "epoch": 0.04173576561636263,
+      "grad_norm": 2.521833896636963,
+      "learning_rate": 4.9275156338828885e-05,
+      "loss": 6.5671,
+      "step": 755
+    },
+    {
+      "epoch": 0.04201216141514649,
+      "grad_norm": 2.741121530532837,
+      "learning_rate": 4.926094371802161e-05,
+      "loss": 6.6626,
+      "step": 760
+    },
+    {
+      "epoch": 0.04228855721393035,
+      "grad_norm": 2.8125641345977783,
+      "learning_rate": 4.9246731097214326e-05,
+      "loss": 6.961,
+      "step": 765
+    },
+    {
+      "epoch": 0.04256495301271421,
+      "grad_norm": 2.6478195190429688,
+      "learning_rate": 4.923251847640705e-05,
+      "loss": 6.6519,
+      "step": 770
+    },
+    {
+      "epoch": 0.042841348811498065,
+      "grad_norm": 2.3772130012512207,
+      "learning_rate": 4.9218305855599774e-05,
+      "loss": 6.5116,
+      "step": 775
+    },
+    {
+      "epoch": 0.04311774461028192,
+      "grad_norm": 2.921616554260254,
+      "learning_rate": 4.92040932347925e-05,
+      "loss": 6.7481,
+      "step": 780
+    },
+    {
+      "epoch": 0.04339414040906578,
+      "grad_norm": 2.6683406829833984,
+      "learning_rate": 4.918988061398522e-05,
+      "loss": 6.5871,
+      "step": 785
+    },
+    {
+      "epoch": 0.04367053620784964,
+      "grad_norm": 2.636958360671997,
+      "learning_rate": 4.9175667993177946e-05,
+      "loss": 6.353,
+      "step": 790
+    },
+    {
+      "epoch": 0.0439469320066335,
+      "grad_norm": 3.1544365882873535,
+      "learning_rate": 4.916145537237067e-05,
+      "loss": 6.7493,
+      "step": 795
+    },
+    {
+      "epoch": 0.04422332780541736,
+      "grad_norm": 3.5060462951660156,
+      "learning_rate": 4.914724275156339e-05,
+      "loss": 6.4544,
+      "step": 800
+    },
+    {
+      "epoch": 0.044499723604201215,
+      "grad_norm": 2.6805920600891113,
+      "learning_rate": 4.913303013075611e-05,
+      "loss": 6.736,
+      "step": 805
+    },
+    {
+      "epoch": 0.04477611940298507,
+      "grad_norm": 2.880324125289917,
+      "learning_rate": 4.911881750994884e-05,
+      "loss": 6.8972,
+      "step": 810
+    },
+    {
+      "epoch": 0.045052515201768936,
+      "grad_norm": 2.333005428314209,
+      "learning_rate": 4.910460488914156e-05,
+      "loss": 6.7566,
+      "step": 815
+    },
+    {
+      "epoch": 0.04532891100055279,
+      "grad_norm": 2.3347644805908203,
+      "learning_rate": 4.909039226833428e-05,
+      "loss": 6.4314,
+      "step": 820
+    },
+    {
+      "epoch": 0.04560530679933665,
+      "grad_norm": 3.022331476211548,
+      "learning_rate": 4.907617964752701e-05,
+      "loss": 6.773,
+      "step": 825
+    },
+    {
+      "epoch": 0.04588170259812051,
+      "grad_norm": 2.466000556945801,
+      "learning_rate": 4.906196702671973e-05,
+      "loss": 6.6999,
+      "step": 830
+    },
+    {
+      "epoch": 0.046158098396904365,
+      "grad_norm": 2.828768730163574,
+      "learning_rate": 4.9047754405912455e-05,
+      "loss": 6.92,
+      "step": 835
+    },
+    {
+      "epoch": 0.04643449419568822,
+      "grad_norm": 2.3771440982818604,
+      "learning_rate": 4.903354178510517e-05,
+      "loss": 6.8686,
+      "step": 840
+    },
+    {
+      "epoch": 0.046710889994472087,
+      "grad_norm": 2.7735869884490967,
+      "learning_rate": 4.90193291642979e-05,
+      "loss": 6.3304,
+      "step": 845
+    },
+    {
+      "epoch": 0.046987285793255944,
+      "grad_norm": 2.560903310775757,
+      "learning_rate": 4.900511654349062e-05,
+      "loss": 6.5665,
+      "step": 850
+    },
+    {
+      "epoch": 0.0472636815920398,
+      "grad_norm": 2.7346339225769043,
+      "learning_rate": 4.8990903922683344e-05,
+      "loss": 6.4082,
+      "step": 855
+    },
+    {
+      "epoch": 0.04754007739082366,
+      "grad_norm": 2.8105099201202393,
+      "learning_rate": 4.897669130187607e-05,
+      "loss": 6.55,
+      "step": 860
+    },
+    {
+      "epoch": 0.047816473189607515,
+      "grad_norm": 3.484123468399048,
+      "learning_rate": 4.896247868106879e-05,
+      "loss": 6.313,
+      "step": 865
+    },
+    {
+      "epoch": 0.04809286898839138,
+      "grad_norm": 2.645644426345825,
+      "learning_rate": 4.8948266060261516e-05,
+      "loss": 6.3278,
+      "step": 870
+    },
+    {
+      "epoch": 0.04836926478717524,
+      "grad_norm": 2.5998237133026123,
+      "learning_rate": 4.893405343945423e-05,
+      "loss": 6.6541,
+      "step": 875
+    },
+    {
+      "epoch": 0.048645660585959094,
+      "grad_norm": 2.4943838119506836,
+      "learning_rate": 4.8919840818646964e-05,
+      "loss": 6.6822,
+      "step": 880
+    },
+    {
+      "epoch": 0.04892205638474295,
+      "grad_norm": 3.156522035598755,
+      "learning_rate": 4.890562819783968e-05,
+      "loss": 6.5528,
+      "step": 885
+    },
+    {
+      "epoch": 0.04919845218352681,
+      "grad_norm": 3.012643337249756,
+      "learning_rate": 4.8891415577032405e-05,
+      "loss": 6.4346,
+      "step": 890
+    },
+    {
+      "epoch": 0.049474847982310666,
+      "grad_norm": 2.3925185203552246,
+      "learning_rate": 4.887720295622513e-05,
+      "loss": 6.7773,
+      "step": 895
+    },
+    {
+      "epoch": 0.04975124378109453,
+      "grad_norm": 2.846290349960327,
+      "learning_rate": 4.886299033541785e-05,
+      "loss": 6.3221,
+      "step": 900
+    },
+    {
+      "epoch": 0.05002763957987839,
+      "grad_norm": 2.715515375137329,
+      "learning_rate": 4.884877771461058e-05,
+      "loss": 6.4721,
+      "step": 905
+    },
+    {
+      "epoch": 0.050304035378662244,
+      "grad_norm": 2.458165407180786,
+      "learning_rate": 4.8834565093803294e-05,
+      "loss": 6.7841,
+      "step": 910
+    },
+    {
+      "epoch": 0.0505804311774461,
+      "grad_norm": 3.0126917362213135,
+      "learning_rate": 4.8820352472996025e-05,
+      "loss": 6.5939,
+      "step": 915
+    },
+    {
+      "epoch": 0.05085682697622996,
+      "grad_norm": 2.5035128593444824,
+      "learning_rate": 4.880613985218875e-05,
+      "loss": 6.4,
+      "step": 920
+    },
+    {
+      "epoch": 0.05113322277501382,
+      "grad_norm": 3.497965097427368,
+      "learning_rate": 4.8791927231381466e-05,
+      "loss": 6.7437,
+      "step": 925
+    },
+    {
+      "epoch": 0.05140961857379768,
+      "grad_norm": 2.88411021232605,
+      "learning_rate": 4.877771461057419e-05,
+      "loss": 6.4189,
+      "step": 930
+    },
+    {
+      "epoch": 0.05168601437258154,
+      "grad_norm": 2.9255616664886475,
+      "learning_rate": 4.8763501989766914e-05,
+      "loss": 6.7637,
+      "step": 935
+    },
+    {
+      "epoch": 0.051962410171365395,
+      "grad_norm": 3.0357139110565186,
+      "learning_rate": 4.874928936895964e-05,
+      "loss": 6.4721,
+      "step": 940
+    },
+    {
+      "epoch": 0.05223880597014925,
+      "grad_norm": 2.340343713760376,
+      "learning_rate": 4.873507674815236e-05,
+      "loss": 6.3291,
+      "step": 945
+    },
+    {
+      "epoch": 0.05251520176893311,
+      "grad_norm": 2.570775270462036,
+      "learning_rate": 4.8720864127345086e-05,
+      "loss": 6.2555,
+      "step": 950
+    },
+    {
+      "epoch": 0.05279159756771697,
+      "grad_norm": 2.6438934803009033,
+      "learning_rate": 4.870665150653781e-05,
+      "loss": 6.6452,
+      "step": 955
+    },
+    {
+      "epoch": 0.05306799336650083,
+      "grad_norm": 3.2586843967437744,
+      "learning_rate": 4.869243888573053e-05,
+      "loss": 6.6964,
+      "step": 960
+    },
+    {
+      "epoch": 0.05334438916528469,
+      "grad_norm": 2.744736433029175,
+      "learning_rate": 4.867822626492326e-05,
+      "loss": 6.3793,
+      "step": 965
+    },
+    {
+      "epoch": 0.053620784964068545,
+      "grad_norm": 2.517343044281006,
+      "learning_rate": 4.866401364411598e-05,
+      "loss": 6.6729,
+      "step": 970
+    },
+    {
+      "epoch": 0.0538971807628524,
+      "grad_norm": 2.725949287414551,
+      "learning_rate": 4.86498010233087e-05,
+      "loss": 6.341,
+      "step": 975
+    },
+    {
+      "epoch": 0.054173576561636266,
+      "grad_norm": 2.8635339736938477,
+      "learning_rate": 4.863558840250142e-05,
+      "loss": 6.4309,
+      "step": 980
+    },
+    {
+      "epoch": 0.054449972360420124,
+      "grad_norm": 2.70915150642395,
+      "learning_rate": 4.862137578169415e-05,
+      "loss": 6.3596,
+      "step": 985
+    },
+    {
+      "epoch": 0.05472636815920398,
+      "grad_norm": 3.162316083908081,
+      "learning_rate": 4.860716316088687e-05,
+      "loss": 6.6671,
+      "step": 990
+    },
+    {
+      "epoch": 0.05500276395798784,
+      "grad_norm": 2.084820508956909,
+      "learning_rate": 4.859295054007959e-05,
+      "loss": 6.5709,
+      "step": 995
+    },
+    {
+      "epoch": 0.055279159756771695,
+      "grad_norm": 2.8458499908447266,
+      "learning_rate": 4.857873791927232e-05,
+      "loss": 6.4753,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 3.0729644298553467,
+      "learning_rate": 4.856452529846504e-05,
+      "loss": 6.5452,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05583195135433942,
+      "grad_norm": 2.171372890472412,
+      "learning_rate": 4.855031267765776e-05,
+      "loss": 6.2128,
+      "step": 1010
+    },
+    {
+      "epoch": 0.056108347153123274,
+      "grad_norm": 2.624799966812134,
+      "learning_rate": 4.8536100056850484e-05,
+      "loss": 6.4788,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05638474295190713,
+      "grad_norm": 2.8430840969085693,
+      "learning_rate": 4.852188743604321e-05,
+      "loss": 6.3949,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05666113875069099,
+      "grad_norm": 3.119079351425171,
+      "learning_rate": 4.850767481523593e-05,
+      "loss": 6.6569,
+      "step": 1025
+    },
+    {
+      "epoch": 0.056937534549474846,
+      "grad_norm": 2.7463550567626953,
+      "learning_rate": 4.8493462194428656e-05,
+      "loss": 6.5224,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05721393034825871,
+      "grad_norm": 2.7679648399353027,
+      "learning_rate": 4.847924957362138e-05,
+      "loss": 6.5559,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05749032614704257,
+      "grad_norm": 3.098196268081665,
+      "learning_rate": 4.8465036952814104e-05,
+      "loss": 6.4804,
+      "step": 1040
+    },
+    {
+      "epoch": 0.057766721945826424,
+      "grad_norm": 2.649479866027832,
+      "learning_rate": 4.845082433200682e-05,
+      "loss": 6.553,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05804311774461028,
+      "grad_norm": 3.0194578170776367,
+      "learning_rate": 4.8436611711199545e-05,
+      "loss": 6.2782,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05831951354339414,
+      "grad_norm": 2.331939220428467,
+      "learning_rate": 4.8422399090392276e-05,
+      "loss": 6.4653,
+      "step": 1055
+    },
+    {
+      "epoch": 0.058595909342177996,
+      "grad_norm": 2.84271502494812,
+      "learning_rate": 4.840818646958499e-05,
+      "loss": 6.5789,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05887230514096186,
+      "grad_norm": 2.8978166580200195,
+      "learning_rate": 4.839397384877772e-05,
+      "loss": 6.495,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05914870093974572,
+      "grad_norm": 2.4874818325042725,
+      "learning_rate": 4.837976122797044e-05,
+      "loss": 6.5368,
+      "step": 1070
+    },
+    {
+      "epoch": 0.059425096738529574,
+      "grad_norm": 3.212247848510742,
+      "learning_rate": 4.8365548607163165e-05,
+      "loss": 6.3398,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05970149253731343,
+      "grad_norm": 2.3944737911224365,
+      "learning_rate": 4.835133598635589e-05,
+      "loss": 6.6459,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05997788833609729,
+      "grad_norm": 3.152069091796875,
+      "learning_rate": 4.8337123365548606e-05,
+      "loss": 6.2656,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06025428413488115,
+      "grad_norm": 3.0014097690582275,
+      "learning_rate": 4.832291074474134e-05,
+      "loss": 6.5379,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06053067993366501,
+      "grad_norm": 2.4241628646850586,
+      "learning_rate": 4.8308698123934054e-05,
+      "loss": 6.518,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06080707573244887,
+      "grad_norm": 2.194061040878296,
+      "learning_rate": 4.829448550312678e-05,
+      "loss": 6.2291,
+      "step": 1100
+    },
+    {
+      "epoch": 0.061083471531232725,
+      "grad_norm": 2.6672093868255615,
+      "learning_rate": 4.82802728823195e-05,
+      "loss": 6.2079,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06135986733001658,
+      "grad_norm": 2.9733340740203857,
+      "learning_rate": 4.8266060261512226e-05,
+      "loss": 6.7986,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06163626312880044,
+      "grad_norm": 2.6432945728302,
+      "learning_rate": 4.825184764070495e-05,
+      "loss": 6.3901,
+      "step": 1115
+    },
+    {
+      "epoch": 0.0619126589275843,
+      "grad_norm": 2.304767370223999,
+      "learning_rate": 4.823763501989767e-05,
+      "loss": 6.3117,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06218905472636816,
+      "grad_norm": 2.7202680110931396,
+      "learning_rate": 4.82234223990904e-05,
+      "loss": 6.5007,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06246545052515202,
+      "grad_norm": 2.201432943344116,
+      "learning_rate": 4.8209209778283115e-05,
+      "loss": 6.3724,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06274184632393588,
+      "grad_norm": 2.9685938358306885,
+      "learning_rate": 4.819499715747584e-05,
+      "loss": 6.4615,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06301824212271974,
+      "grad_norm": 2.2716915607452393,
+      "learning_rate": 4.818078453666856e-05,
+      "loss": 6.4481,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0632946379215036,
+      "grad_norm": 2.6046628952026367,
+      "learning_rate": 4.816657191586129e-05,
+      "loss": 6.4018,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06357103372028745,
+      "grad_norm": 3.6056673526763916,
+      "learning_rate": 4.815235929505401e-05,
+      "loss": 6.4416,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06384742951907131,
+      "grad_norm": 3.628232002258301,
+      "learning_rate": 4.813814667424673e-05,
+      "loss": 6.5809,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06412382531785517,
+      "grad_norm": 2.789034128189087,
+      "learning_rate": 4.812393405343946e-05,
+      "loss": 6.3966,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06440022111663903,
+      "grad_norm": 2.2794246673583984,
+      "learning_rate": 4.810972143263218e-05,
+      "loss": 6.3319,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06467661691542288,
+      "grad_norm": 2.4310152530670166,
+      "learning_rate": 4.80955088118249e-05,
+      "loss": 6.5324,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06495301271420674,
+      "grad_norm": 2.6283979415893555,
+      "learning_rate": 4.8081296191017624e-05,
+      "loss": 6.5129,
+      "step": 1175
+    },
+    {
+      "epoch": 0.0652294085129906,
+      "grad_norm": 2.8399784564971924,
+      "learning_rate": 4.806708357021035e-05,
+      "loss": 6.5845,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06550580431177445,
+      "grad_norm": 2.680716037750244,
+      "learning_rate": 4.805287094940307e-05,
+      "loss": 6.6554,
+      "step": 1185
+    },
+    {
+      "epoch": 0.06578220011055833,
+      "grad_norm": 3.5662529468536377,
+      "learning_rate": 4.8038658328595796e-05,
+      "loss": 6.3653,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06605859590934218,
+      "grad_norm": 2.429107666015625,
+      "learning_rate": 4.802444570778852e-05,
+      "loss": 6.2625,
+      "step": 1195
+    },
+    {
+      "epoch": 0.06633499170812604,
+      "grad_norm": 2.9469411373138428,
+      "learning_rate": 4.8010233086981244e-05,
+      "loss": 6.3618,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0666113875069099,
+      "grad_norm": 3.321293592453003,
+      "learning_rate": 4.799602046617396e-05,
+      "loss": 6.262,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06688778330569375,
+      "grad_norm": 2.542504072189331,
+      "learning_rate": 4.7981807845366685e-05,
+      "loss": 6.2755,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06716417910447761,
+      "grad_norm": 2.622391700744629,
+      "learning_rate": 4.796759522455941e-05,
+      "loss": 6.2732,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06744057490326147,
+      "grad_norm": 2.2811062335968018,
+      "learning_rate": 4.7953382603752133e-05,
+      "loss": 6.4282,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06771697070204533,
+      "grad_norm": 2.973114013671875,
+      "learning_rate": 4.793916998294486e-05,
+      "loss": 6.2921,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06799336650082918,
+      "grad_norm": 2.5322935581207275,
+      "learning_rate": 4.792495736213758e-05,
+      "loss": 6.2762,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06826976229961304,
+      "grad_norm": 3.3714489936828613,
+      "learning_rate": 4.7910744741330305e-05,
+      "loss": 6.4914,
+      "step": 1235
+    },
+    {
+      "epoch": 0.0685461580983969,
+      "grad_norm": 2.4342074394226074,
+      "learning_rate": 4.789653212052302e-05,
+      "loss": 6.1297,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06882255389718077,
+      "grad_norm": 3.002720594406128,
+      "learning_rate": 4.7882319499715747e-05,
+      "loss": 6.4413,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06909894969596463,
+      "grad_norm": 2.581346273422241,
+      "learning_rate": 4.786810687890848e-05,
+      "loss": 6.5842,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06937534549474848,
+      "grad_norm": 2.6828501224517822,
+      "learning_rate": 4.7853894258101195e-05,
+      "loss": 6.3015,
+      "step": 1255
+    },
+    {
+      "epoch": 0.06965174129353234,
+      "grad_norm": 3.318371295928955,
+      "learning_rate": 4.783968163729392e-05,
+      "loss": 6.561,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0699281370923162,
+      "grad_norm": 3.4630489349365234,
+      "learning_rate": 4.782546901648664e-05,
+      "loss": 6.7319,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07020453289110005,
+      "grad_norm": 3.342996597290039,
+      "learning_rate": 4.7811256395679366e-05,
+      "loss": 6.2168,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07048092868988391,
+      "grad_norm": 2.5235445499420166,
+      "learning_rate": 4.779704377487209e-05,
+      "loss": 6.5335,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07075732448866777,
+      "grad_norm": 3.2131083011627197,
+      "learning_rate": 4.7782831154064814e-05,
+      "loss": 6.4834,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07103372028745163,
+      "grad_norm": 2.924309015274048,
+      "learning_rate": 4.776861853325754e-05,
+      "loss": 6.3069,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07131011608623548,
+      "grad_norm": 2.966257333755493,
+      "learning_rate": 4.7754405912450256e-05,
+      "loss": 6.1896,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07158651188501934,
+      "grad_norm": 2.937412977218628,
+      "learning_rate": 4.774019329164298e-05,
+      "loss": 6.4074,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07186290768380321,
+      "grad_norm": 2.4649055004119873,
+      "learning_rate": 4.772598067083571e-05,
+      "loss": 6.4064,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07213930348258707,
+      "grad_norm": 2.4708802700042725,
+      "learning_rate": 4.771176805002843e-05,
+      "loss": 6.2426,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07241569928137093,
+      "grad_norm": 2.374051094055176,
+      "learning_rate": 4.769755542922115e-05,
+      "loss": 6.1917,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07269209508015478,
+      "grad_norm": 5.211256504058838,
+      "learning_rate": 4.7683342808413875e-05,
+      "loss": 6.0755,
+      "step": 1315
+    },
+    {
+      "epoch": 0.07296849087893864,
+      "grad_norm": 2.7231972217559814,
+      "learning_rate": 4.76691301876066e-05,
+      "loss": 6.2229,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0732448866777225,
+      "grad_norm": 2.7478833198547363,
+      "learning_rate": 4.765491756679932e-05,
+      "loss": 6.2957,
+      "step": 1325
+    },
+    {
+      "epoch": 0.07352128247650636,
+      "grad_norm": 2.573377847671509,
+      "learning_rate": 4.764070494599204e-05,
+      "loss": 6.1357,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07379767827529021,
+      "grad_norm": 2.7243764400482178,
+      "learning_rate": 4.762649232518477e-05,
+      "loss": 6.4513,
+      "step": 1335
+    },
+    {
+      "epoch": 0.07407407407407407,
+      "grad_norm": 2.9036333560943604,
+      "learning_rate": 4.761227970437749e-05,
+      "loss": 6.296,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07435046987285793,
+      "grad_norm": 3.135991096496582,
+      "learning_rate": 4.759806708357021e-05,
+      "loss": 6.4276,
+      "step": 1345
+    },
+    {
+      "epoch": 0.07462686567164178,
+      "grad_norm": 2.428114175796509,
+      "learning_rate": 4.7583854462762937e-05,
+      "loss": 6.2983,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07490326147042566,
+      "grad_norm": 2.220736026763916,
+      "learning_rate": 4.756964184195566e-05,
+      "loss": 6.2962,
+      "step": 1355
+    },
+    {
+      "epoch": 0.07517965726920951,
+      "grad_norm": 2.5993216037750244,
+      "learning_rate": 4.7555429221148385e-05,
+      "loss": 6.3815,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07545605306799337,
+      "grad_norm": 2.7360010147094727,
+      "learning_rate": 4.75412166003411e-05,
+      "loss": 6.3496,
+      "step": 1365
+    },
+    {
+      "epoch": 0.07573244886677723,
+      "grad_norm": 2.4501261711120605,
+      "learning_rate": 4.752700397953383e-05,
+      "loss": 6.3474,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07600884466556108,
+      "grad_norm": 2.4407808780670166,
+      "learning_rate": 4.751279135872655e-05,
+      "loss": 6.079,
+      "step": 1375
+    },
+    {
+      "epoch": 0.07628524046434494,
+      "grad_norm": 2.870891571044922,
+      "learning_rate": 4.7498578737919274e-05,
+      "loss": 6.009,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0765616362631288,
+      "grad_norm": 2.539259910583496,
+      "learning_rate": 4.7484366117112e-05,
+      "loss": 6.4406,
+      "step": 1385
+    },
+    {
+      "epoch": 0.07683803206191266,
+      "grad_norm": 3.2458794116973877,
+      "learning_rate": 4.747015349630472e-05,
+      "loss": 6.1957,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07711442786069651,
+      "grad_norm": 2.6678242683410645,
+      "learning_rate": 4.7455940875497446e-05,
+      "loss": 6.3314,
+      "step": 1395
+    },
+    {
+      "epoch": 0.07739082365948037,
+      "grad_norm": 2.6106460094451904,
+      "learning_rate": 4.744172825469016e-05,
+      "loss": 6.1362,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07766721945826423,
+      "grad_norm": 2.8481876850128174,
+      "learning_rate": 4.7427515633882894e-05,
+      "loss": 6.3231,
+      "step": 1405
+    },
+    {
+      "epoch": 0.0779436152570481,
+      "grad_norm": 2.367493152618408,
+      "learning_rate": 4.741330301307562e-05,
+      "loss": 6.0713,
+      "step": 1410
+    },
+    {
+      "epoch": 0.07822001105583196,
+      "grad_norm": 3.546706199645996,
+      "learning_rate": 4.7399090392268335e-05,
+      "loss": 6.1802,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07849640685461581,
+      "grad_norm": 2.659611701965332,
+      "learning_rate": 4.738487777146106e-05,
+      "loss": 6.2043,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07877280265339967,
+      "grad_norm": 2.9869518280029297,
+      "learning_rate": 4.737066515065378e-05,
+      "loss": 6.354,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07904919845218353,
+      "grad_norm": 2.5255351066589355,
+      "learning_rate": 4.735645252984651e-05,
+      "loss": 6.0689,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07932559425096739,
+      "grad_norm": 3.4141147136688232,
+      "learning_rate": 4.7342239909039224e-05,
+      "loss": 6.2125,
+      "step": 1435
+    },
+    {
+      "epoch": 0.07960199004975124,
+      "grad_norm": 2.526256799697876,
+      "learning_rate": 4.7328027288231955e-05,
+      "loss": 6.055,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0798783858485351,
+      "grad_norm": 2.8614776134490967,
+      "learning_rate": 4.731381466742468e-05,
+      "loss": 6.3068,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08015478164731896,
+      "grad_norm": 2.7898523807525635,
+      "learning_rate": 4.7299602046617396e-05,
+      "loss": 6.2226,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08043117744610281,
+      "grad_norm": 2.2415056228637695,
+      "learning_rate": 4.728538942581012e-05,
+      "loss": 6.1964,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08070757324488667,
+      "grad_norm": 3.079174280166626,
+      "learning_rate": 4.7271176805002844e-05,
+      "loss": 6.2307,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08098396904367054,
+      "grad_norm": 2.619187355041504,
+      "learning_rate": 4.725696418419557e-05,
+      "loss": 6.1481,
+      "step": 1465
+    },
+    {
+      "epoch": 0.0812603648424544,
+      "grad_norm": 3.067775249481201,
+      "learning_rate": 4.724275156338829e-05,
+      "loss": 6.3701,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08153676064123826,
+      "grad_norm": 2.4046452045440674,
+      "learning_rate": 4.7228538942581016e-05,
+      "loss": 6.1519,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08181315644002211,
+      "grad_norm": 2.766226291656494,
+      "learning_rate": 4.721432632177374e-05,
+      "loss": 6.1802,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08208955223880597,
+      "grad_norm": 2.1929092407226562,
+      "learning_rate": 4.720011370096646e-05,
+      "loss": 6.4416,
+      "step": 1485
+    },
+    {
+      "epoch": 0.08236594803758983,
+      "grad_norm": 2.9942455291748047,
+      "learning_rate": 4.718590108015918e-05,
+      "loss": 6.1566,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08264234383637369,
+      "grad_norm": 3.3760600090026855,
+      "learning_rate": 4.717168845935191e-05,
+      "loss": 6.2593,
+      "step": 1495
+    },
+    {
+      "epoch": 0.08291873963515754,
+      "grad_norm": 3.459946632385254,
+      "learning_rate": 4.715747583854463e-05,
+      "loss": 6.0409,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0831951354339414,
+      "grad_norm": 2.9691433906555176,
+      "learning_rate": 4.714326321773735e-05,
+      "loss": 6.2912,
+      "step": 1505
+    },
+    {
+      "epoch": 0.08347153123272526,
+      "grad_norm": 2.3945460319519043,
+      "learning_rate": 4.712905059693008e-05,
+      "loss": 6.3243,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08374792703150911,
+      "grad_norm": 3.2403388023376465,
+      "learning_rate": 4.71148379761228e-05,
+      "loss": 5.8076,
+      "step": 1515
+    },
+    {
+      "epoch": 0.08402432283029299,
+      "grad_norm": 2.4649736881256104,
+      "learning_rate": 4.710062535531552e-05,
+      "loss": 6.1595,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08430071862907684,
+      "grad_norm": 2.9593818187713623,
+      "learning_rate": 4.708641273450824e-05,
+      "loss": 6.1091,
+      "step": 1525
+    },
+    {
+      "epoch": 0.0845771144278607,
+      "grad_norm": 2.794445753097534,
+      "learning_rate": 4.707220011370097e-05,
+      "loss": 6.1391,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08485351022664456,
+      "grad_norm": 2.7484805583953857,
+      "learning_rate": 4.705798749289369e-05,
+      "loss": 6.0529,
+      "step": 1535
+    },
+    {
+      "epoch": 0.08512990602542841,
+      "grad_norm": 2.891587972640991,
+      "learning_rate": 4.7043774872086414e-05,
+      "loss": 6.1436,
+      "step": 1540
+    },
+    {
+      "epoch": 0.08540630182421227,
+      "grad_norm": 3.721700429916382,
+      "learning_rate": 4.702956225127914e-05,
+      "loss": 6.4103,
+      "step": 1545
+    },
+    {
+      "epoch": 0.08568269762299613,
+      "grad_norm": 2.4380359649658203,
+      "learning_rate": 4.701534963047186e-05,
+      "loss": 6.2798,
+      "step": 1550
+    },
+    {
+      "epoch": 0.08595909342177999,
+      "grad_norm": 3.160372257232666,
+      "learning_rate": 4.7001137009664586e-05,
+      "loss": 6.4028,
+      "step": 1555
+    },
+    {
+      "epoch": 0.08623548922056384,
+      "grad_norm": 2.3256590366363525,
+      "learning_rate": 4.69869243888573e-05,
+      "loss": 6.2657,
+      "step": 1560
+    },
+    {
+      "epoch": 0.0865118850193477,
+      "grad_norm": 2.191479444503784,
+      "learning_rate": 4.6972711768050034e-05,
+      "loss": 6.1302,
+      "step": 1565
+    },
+    {
+      "epoch": 0.08678828081813156,
+      "grad_norm": 2.7161953449249268,
+      "learning_rate": 4.695849914724275e-05,
+      "loss": 6.2056,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08706467661691543,
+      "grad_norm": 2.7655863761901855,
+      "learning_rate": 4.6944286526435475e-05,
+      "loss": 6.1198,
+      "step": 1575
+    },
+    {
+      "epoch": 0.08734107241569929,
+      "grad_norm": 2.158665180206299,
+      "learning_rate": 4.6930073905628206e-05,
+      "loss": 6.2007,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08761746821448314,
+      "grad_norm": 2.5755743980407715,
+      "learning_rate": 4.691586128482092e-05,
+      "loss": 6.2911,
+      "step": 1585
+    },
+    {
+      "epoch": 0.087893864013267,
+      "grad_norm": 2.5432612895965576,
+      "learning_rate": 4.690164866401365e-05,
+      "loss": 6.1612,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08817025981205086,
+      "grad_norm": 2.5009267330169678,
+      "learning_rate": 4.688743604320637e-05,
+      "loss": 6.3085,
+      "step": 1595
+    },
+    {
+      "epoch": 0.08844665561083472,
+      "grad_norm": 2.381110429763794,
+      "learning_rate": 4.6873223422399095e-05,
+      "loss": 6.211,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08872305140961857,
+      "grad_norm": 2.5844264030456543,
+      "learning_rate": 4.685901080159182e-05,
+      "loss": 6.0844,
+      "step": 1605
+    },
+    {
+      "epoch": 0.08899944720840243,
+      "grad_norm": 2.737116575241089,
+      "learning_rate": 4.6844798180784536e-05,
+      "loss": 6.2711,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08927584300718629,
+      "grad_norm": 2.8457653522491455,
+      "learning_rate": 4.683058555997727e-05,
+      "loss": 6.3523,
+      "step": 1615
+    },
+    {
+      "epoch": 0.08955223880597014,
+      "grad_norm": 3.002805471420288,
+      "learning_rate": 4.6816372939169984e-05,
+      "loss": 6.3546,
+      "step": 1620
+    },
+    {
+      "epoch": 0.089828634604754,
+      "grad_norm": 3.3119866847991943,
+      "learning_rate": 4.680216031836271e-05,
+      "loss": 6.3509,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09010503040353787,
+      "grad_norm": 2.4615001678466797,
+      "learning_rate": 4.678794769755543e-05,
+      "loss": 6.1708,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09038142620232173,
+      "grad_norm": 2.660804271697998,
+      "learning_rate": 4.6773735076748156e-05,
+      "loss": 6.1863,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09065782200110559,
+      "grad_norm": 2.9395289421081543,
+      "learning_rate": 4.675952245594088e-05,
+      "loss": 6.5699,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09093421779988944,
+      "grad_norm": 3.1004257202148438,
+      "learning_rate": 4.67453098351336e-05,
+      "loss": 6.1951,
+      "step": 1645
+    },
+    {
+      "epoch": 0.0912106135986733,
+      "grad_norm": 3.4309775829315186,
+      "learning_rate": 4.673109721432633e-05,
+      "loss": 5.9744,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09148700939745716,
+      "grad_norm": 2.9011757373809814,
+      "learning_rate": 4.6716884593519045e-05,
+      "loss": 6.1815,
+      "step": 1655
+    },
+    {
+      "epoch": 0.09176340519624102,
+      "grad_norm": 3.3524973392486572,
+      "learning_rate": 4.670267197271177e-05,
+      "loss": 6.1202,
+      "step": 1660
+    },
+    {
+      "epoch": 0.09203980099502487,
+      "grad_norm": 2.6807689666748047,
+      "learning_rate": 4.668845935190449e-05,
+      "loss": 5.9419,
+      "step": 1665
+    },
+    {
+      "epoch": 0.09231619679380873,
+      "grad_norm": 2.4009783267974854,
+      "learning_rate": 4.667424673109722e-05,
+      "loss": 6.2655,
+      "step": 1670
+    },
+    {
+      "epoch": 0.09259259259259259,
+      "grad_norm": 3.517772674560547,
+      "learning_rate": 4.666003411028994e-05,
+      "loss": 6.1189,
+      "step": 1675
+    },
+    {
+      "epoch": 0.09286898839137644,
+      "grad_norm": 2.677375555038452,
+      "learning_rate": 4.664582148948266e-05,
+      "loss": 6.2443,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09314538419016032,
+      "grad_norm": 3.299330949783325,
+      "learning_rate": 4.663160886867539e-05,
+      "loss": 5.8239,
+      "step": 1685
+    },
+    {
+      "epoch": 0.09342177998894417,
+      "grad_norm": 2.366581916809082,
+      "learning_rate": 4.661739624786811e-05,
+      "loss": 6.244,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09369817578772803,
+      "grad_norm": 2.2254860401153564,
+      "learning_rate": 4.660318362706083e-05,
+      "loss": 6.2237,
+      "step": 1695
+    },
+    {
+      "epoch": 0.09397457158651189,
+      "grad_norm": 2.3326756954193115,
+      "learning_rate": 4.6588971006253554e-05,
+      "loss": 6.008,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09425096738529574,
+      "grad_norm": 2.39668345451355,
+      "learning_rate": 4.657475838544628e-05,
+      "loss": 6.4698,
+      "step": 1705
+    },
+    {
+      "epoch": 0.0945273631840796,
+      "grad_norm": 2.79451584815979,
+      "learning_rate": 4.6560545764639e-05,
+      "loss": 6.0543,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09480375898286346,
+      "grad_norm": 3.0504002571105957,
+      "learning_rate": 4.6546333143831726e-05,
+      "loss": 5.9177,
+      "step": 1715
+    },
+    {
+      "epoch": 0.09508015478164732,
+      "grad_norm": 2.6137356758117676,
+      "learning_rate": 4.653212052302445e-05,
+      "loss": 6.0108,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09535655058043117,
+      "grad_norm": 2.320517063140869,
+      "learning_rate": 4.6517907902217174e-05,
+      "loss": 5.9592,
+      "step": 1725
+    },
+    {
+      "epoch": 0.09563294637921503,
+      "grad_norm": 2.4836413860321045,
+      "learning_rate": 4.650369528140989e-05,
+      "loss": 6.1646,
+      "step": 1730
+    },
+    {
+      "epoch": 0.09590934217799889,
+      "grad_norm": 2.3726565837860107,
+      "learning_rate": 4.6489482660602615e-05,
+      "loss": 6.0548,
+      "step": 1735
+    },
+    {
+      "epoch": 0.09618573797678276,
+      "grad_norm": 3.755805492401123,
+      "learning_rate": 4.647527003979534e-05,
+      "loss": 6.1781,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09646213377556662,
+      "grad_norm": 2.5394093990325928,
+      "learning_rate": 4.646105741898806e-05,
+      "loss": 5.9294,
+      "step": 1745
+    },
+    {
+      "epoch": 0.09673852957435047,
+      "grad_norm": 3.1130166053771973,
+      "learning_rate": 4.644684479818079e-05,
+      "loss": 6.0655,
+      "step": 1750
+    },
+    {
+      "epoch": 0.09701492537313433,
+      "grad_norm": 3.218590497970581,
+      "learning_rate": 4.643263217737351e-05,
+      "loss": 6.2316,
+      "step": 1755
+    },
+    {
+      "epoch": 0.09729132117191819,
+      "grad_norm": 3.100757598876953,
+      "learning_rate": 4.6418419556566235e-05,
+      "loss": 6.105,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09756771697070205,
+      "grad_norm": 2.9586379528045654,
+      "learning_rate": 4.640420693575895e-05,
+      "loss": 6.3479,
+      "step": 1765
+    },
+    {
+      "epoch": 0.0978441127694859,
+      "grad_norm": 2.569976806640625,
+      "learning_rate": 4.6389994314951676e-05,
+      "loss": 6.2001,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09812050856826976,
+      "grad_norm": 2.5895392894744873,
+      "learning_rate": 4.637578169414441e-05,
+      "loss": 6.2953,
+      "step": 1775
+    },
+    {
+      "epoch": 0.09839690436705362,
+      "grad_norm": 3.0334362983703613,
+      "learning_rate": 4.6361569073337124e-05,
+      "loss": 6.1325,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09867330016583747,
+      "grad_norm": 3.2410378456115723,
+      "learning_rate": 4.634735645252985e-05,
+      "loss": 6.1757,
+      "step": 1785
+    },
+    {
+      "epoch": 0.09894969596462133,
+      "grad_norm": 2.2880196571350098,
+      "learning_rate": 4.633314383172257e-05,
+      "loss": 6.1355,
+      "step": 1790
+    },
+    {
+      "epoch": 0.0992260917634052,
+      "grad_norm": 3.8285810947418213,
+      "learning_rate": 4.6318931210915296e-05,
+      "loss": 6.078,
+      "step": 1795
+    },
+    {
+      "epoch": 0.09950248756218906,
+      "grad_norm": 2.5488243103027344,
+      "learning_rate": 4.630471859010802e-05,
+      "loss": 5.956,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09977888336097292,
+      "grad_norm": 2.4518303871154785,
+      "learning_rate": 4.629050596930074e-05,
+      "loss": 6.0628,
+      "step": 1805
+    },
+    {
+      "epoch": 0.10005527915975677,
+      "grad_norm": 3.043971538543701,
+      "learning_rate": 4.627629334849347e-05,
+      "loss": 6.2971,
+      "step": 1810
+    },
+    {
+      "epoch": 0.10033167495854063,
+      "grad_norm": 3.071781873703003,
+      "learning_rate": 4.6262080727686185e-05,
+      "loss": 6.139,
+      "step": 1815
+    },
+    {
+      "epoch": 0.10060807075732449,
+      "grad_norm": 3.2873592376708984,
+      "learning_rate": 4.624786810687891e-05,
+      "loss": 6.1132,
+      "step": 1820
+    },
+    {
+      "epoch": 0.10088446655610835,
+      "grad_norm": 3.439819574356079,
+      "learning_rate": 4.623365548607163e-05,
+      "loss": 6.1686,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1011608623548922,
+      "grad_norm": 3.1802821159362793,
+      "learning_rate": 4.621944286526436e-05,
+      "loss": 5.9691,
+      "step": 1830
+    },
+    {
+      "epoch": 0.10143725815367606,
+      "grad_norm": 2.665397882461548,
+      "learning_rate": 4.620523024445708e-05,
+      "loss": 6.2404,
+      "step": 1835
+    },
+    {
+      "epoch": 0.10171365395245992,
+      "grad_norm": 2.8647611141204834,
+      "learning_rate": 4.61910176236498e-05,
+      "loss": 6.1963,
+      "step": 1840
+    },
+    {
+      "epoch": 0.10199004975124377,
+      "grad_norm": 3.0262868404388428,
+      "learning_rate": 4.617680500284253e-05,
+      "loss": 6.1664,
+      "step": 1845
+    },
+    {
+      "epoch": 0.10226644555002765,
+      "grad_norm": 3.05643367767334,
+      "learning_rate": 4.6162592382035246e-05,
+      "loss": 6.2904,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1025428413488115,
+      "grad_norm": 2.7000110149383545,
+      "learning_rate": 4.614837976122797e-05,
+      "loss": 6.0892,
+      "step": 1855
+    },
+    {
+      "epoch": 0.10281923714759536,
+      "grad_norm": 2.893402338027954,
+      "learning_rate": 4.6134167140420694e-05,
+      "loss": 5.8826,
+      "step": 1860
+    },
+    {
+      "epoch": 0.10309563294637922,
+      "grad_norm": 2.5244555473327637,
+      "learning_rate": 4.611995451961342e-05,
+      "loss": 5.9429,
+      "step": 1865
+    },
+    {
+      "epoch": 0.10337202874516307,
+      "grad_norm": 2.5196962356567383,
+      "learning_rate": 4.610574189880614e-05,
+      "loss": 5.8624,
+      "step": 1870
+    },
+    {
+      "epoch": 0.10364842454394693,
+      "grad_norm": 2.5912246704101562,
+      "learning_rate": 4.609152927799886e-05,
+      "loss": 6.1336,
+      "step": 1875
+    },
+    {
+      "epoch": 0.10392482034273079,
+      "grad_norm": 2.757858991622925,
+      "learning_rate": 4.607731665719159e-05,
+      "loss": 6.0385,
+      "step": 1880
+    },
+    {
+      "epoch": 0.10420121614151465,
+      "grad_norm": 2.6587064266204834,
+      "learning_rate": 4.6063104036384314e-05,
+      "loss": 6.0566,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1044776119402985,
+      "grad_norm": 3.415855646133423,
+      "learning_rate": 4.604889141557703e-05,
+      "loss": 6.4943,
+      "step": 1890
+    },
+    {
+      "epoch": 0.10475400773908236,
+      "grad_norm": 2.9792914390563965,
+      "learning_rate": 4.603467879476976e-05,
+      "loss": 5.8498,
+      "step": 1895
+    },
+    {
+      "epoch": 0.10503040353786622,
+      "grad_norm": 3.112447500228882,
+      "learning_rate": 4.602046617396248e-05,
+      "loss": 6.2611,
+      "step": 1900
+    },
+    {
+      "epoch": 0.10530679933665009,
+      "grad_norm": 2.8653833866119385,
+      "learning_rate": 4.60062535531552e-05,
+      "loss": 6.0917,
+      "step": 1905
+    },
+    {
+      "epoch": 0.10558319513543395,
+      "grad_norm": 2.7462995052337646,
+      "learning_rate": 4.599204093234793e-05,
+      "loss": 6.2313,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1058595909342178,
+      "grad_norm": 2.4381167888641357,
+      "learning_rate": 4.597782831154065e-05,
+      "loss": 6.1274,
+      "step": 1915
+    },
+    {
+      "epoch": 0.10613598673300166,
+      "grad_norm": 3.3550705909729004,
+      "learning_rate": 4.5963615690733375e-05,
+      "loss": 6.246,
+      "step": 1920
+    },
+    {
+      "epoch": 0.10641238253178552,
+      "grad_norm": 2.736597776412964,
+      "learning_rate": 4.594940306992609e-05,
+      "loss": 6.3576,
+      "step": 1925
+    },
+    {
+      "epoch": 0.10668877833056938,
+      "grad_norm": 2.4143283367156982,
+      "learning_rate": 4.593519044911882e-05,
+      "loss": 5.9825,
+      "step": 1930
+    },
+    {
+      "epoch": 0.10696517412935323,
+      "grad_norm": 2.622124195098877,
+      "learning_rate": 4.592097782831155e-05,
+      "loss": 6.1076,
+      "step": 1935
+    },
+    {
+      "epoch": 0.10724156992813709,
+      "grad_norm": 3.135380744934082,
+      "learning_rate": 4.5906765207504264e-05,
+      "loss": 6.0845,
+      "step": 1940
+    },
+    {
+      "epoch": 0.10751796572692095,
+      "grad_norm": 2.970335006713867,
+      "learning_rate": 4.589255258669699e-05,
+      "loss": 6.1974,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1077943615257048,
+      "grad_norm": 4.032550811767578,
+      "learning_rate": 4.587833996588971e-05,
+      "loss": 6.0646,
+      "step": 1950
+    },
+    {
+      "epoch": 0.10807075732448866,
+      "grad_norm": 2.632209300994873,
+      "learning_rate": 4.5864127345082436e-05,
+      "loss": 5.8096,
+      "step": 1955
+    },
+    {
+      "epoch": 0.10834715312327253,
+      "grad_norm": 2.648122549057007,
+      "learning_rate": 4.5849914724275154e-05,
+      "loss": 6.139,
+      "step": 1960
+    },
+    {
+      "epoch": 0.10862354892205639,
+      "grad_norm": 2.7293598651885986,
+      "learning_rate": 4.5835702103467884e-05,
+      "loss": 5.9382,
+      "step": 1965
+    },
+    {
+      "epoch": 0.10889994472084025,
+      "grad_norm": 2.997490167617798,
+      "learning_rate": 4.582148948266061e-05,
+      "loss": 6.1471,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1091763405196241,
+      "grad_norm": 3.3765926361083984,
+      "learning_rate": 4.5807276861853326e-05,
+      "loss": 6.2572,
+      "step": 1975
+    },
+    {
+      "epoch": 0.10945273631840796,
+      "grad_norm": 2.962548017501831,
+      "learning_rate": 4.579306424104605e-05,
+      "loss": 6.0159,
+      "step": 1980
+    },
+    {
+      "epoch": 0.10972913211719182,
+      "grad_norm": 2.896846294403076,
+      "learning_rate": 4.5778851620238773e-05,
+      "loss": 5.6825,
+      "step": 1985
+    },
+    {
+      "epoch": 0.11000552791597568,
+      "grad_norm": 3.3139047622680664,
+      "learning_rate": 4.57646389994315e-05,
+      "loss": 6.0216,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11028192371475953,
+      "grad_norm": 2.7148447036743164,
+      "learning_rate": 4.575042637862422e-05,
+      "loss": 6.1073,
+      "step": 1995
+    },
+    {
+      "epoch": 0.11055831951354339,
+      "grad_norm": 2.597428798675537,
+      "learning_rate": 4.5736213757816945e-05,
+      "loss": 6.2397,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11083471531232725,
+      "grad_norm": 3.059511184692383,
+      "learning_rate": 4.572200113700967e-05,
+      "loss": 6.2985,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 2.423731803894043,
+      "learning_rate": 4.570778851620239e-05,
+      "loss": 6.2662,
+      "step": 2010
+    },
+    {
+      "epoch": 0.11138750690989498,
+      "grad_norm": 2.665515899658203,
+      "learning_rate": 4.569357589539511e-05,
+      "loss": 6.1349,
+      "step": 2015
+    },
+    {
+      "epoch": 0.11166390270867883,
+      "grad_norm": 2.9362165927886963,
+      "learning_rate": 4.567936327458784e-05,
+      "loss": 6.1175,
+      "step": 2020
+    },
+    {
+      "epoch": 0.11194029850746269,
+      "grad_norm": 2.7762677669525146,
+      "learning_rate": 4.566515065378056e-05,
+      "loss": 5.9836,
+      "step": 2025
+    },
+    {
+      "epoch": 0.11221669430624655,
+      "grad_norm": 3.224057674407959,
+      "learning_rate": 4.565093803297328e-05,
+      "loss": 6.0051,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1124930901050304,
+      "grad_norm": 3.2076215744018555,
+      "learning_rate": 4.5636725412166007e-05,
+      "loss": 6.1282,
+      "step": 2035
+    },
+    {
+      "epoch": 0.11276948590381426,
+      "grad_norm": 2.4749698638916016,
+      "learning_rate": 4.562251279135873e-05,
+      "loss": 6.0949,
+      "step": 2040
+    },
+    {
+      "epoch": 0.11304588170259812,
+      "grad_norm": 2.6076836585998535,
+      "learning_rate": 4.5608300170551454e-05,
+      "loss": 6.1536,
+      "step": 2045
+    },
+    {
+      "epoch": 0.11332227750138198,
+      "grad_norm": 2.7696752548217773,
+      "learning_rate": 4.559408754974417e-05,
+      "loss": 6.0666,
+      "step": 2050
+    },
+    {
+      "epoch": 0.11359867330016583,
+      "grad_norm": 2.964919328689575,
+      "learning_rate": 4.55798749289369e-05,
+      "loss": 6.0302,
+      "step": 2055
+    },
+    {
+      "epoch": 0.11387506909894969,
+      "grad_norm": 2.8098812103271484,
+      "learning_rate": 4.556566230812962e-05,
+      "loss": 6.0762,
+      "step": 2060
+    },
+    {
+      "epoch": 0.11415146489773355,
+      "grad_norm": 3.016932725906372,
+      "learning_rate": 4.5551449687322344e-05,
+      "loss": 5.993,
+      "step": 2065
+    },
+    {
+      "epoch": 0.11442786069651742,
+      "grad_norm": 2.170443058013916,
+      "learning_rate": 4.553723706651507e-05,
+      "loss": 6.0171,
+      "step": 2070
+    },
+    {
+      "epoch": 0.11470425649530128,
+      "grad_norm": 2.617326259613037,
+      "learning_rate": 4.552302444570779e-05,
+      "loss": 5.916,
+      "step": 2075
+    },
+    {
+      "epoch": 0.11498065229408513,
+      "grad_norm": 2.2992942333221436,
+      "learning_rate": 4.5508811824900516e-05,
+      "loss": 6.2716,
+      "step": 2080
+    },
+    {
+      "epoch": 0.11525704809286899,
+      "grad_norm": 2.6933460235595703,
+      "learning_rate": 4.549459920409323e-05,
+      "loss": 6.0187,
+      "step": 2085
+    },
+    {
+      "epoch": 0.11553344389165285,
+      "grad_norm": 2.333048105239868,
+      "learning_rate": 4.5480386583285964e-05,
+      "loss": 6.0797,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1158098396904367,
+      "grad_norm": 2.7053120136260986,
+      "learning_rate": 4.546617396247868e-05,
+      "loss": 6.3752,
+      "step": 2095
+    },
+    {
+      "epoch": 0.11608623548922056,
+      "grad_norm": 2.7705891132354736,
+      "learning_rate": 4.5451961341671405e-05,
+      "loss": 5.9062,
+      "step": 2100
+    },
+    {
+      "epoch": 0.11636263128800442,
+      "grad_norm": 2.7516586780548096,
+      "learning_rate": 4.543774872086413e-05,
+      "loss": 5.8142,
+      "step": 2105
+    },
+    {
+      "epoch": 0.11663902708678828,
+      "grad_norm": 3.1628334522247314,
+      "learning_rate": 4.542353610005685e-05,
+      "loss": 6.151,
+      "step": 2110
+    },
+    {
+      "epoch": 0.11691542288557213,
+      "grad_norm": 2.7654128074645996,
+      "learning_rate": 4.540932347924958e-05,
+      "loss": 6.0071,
+      "step": 2115
+    },
+    {
+      "epoch": 0.11719181868435599,
+      "grad_norm": 2.2299458980560303,
+      "learning_rate": 4.5395110858442294e-05,
+      "loss": 5.9453,
+      "step": 2120
+    },
+    {
+      "epoch": 0.11746821448313986,
+      "grad_norm": 2.7742066383361816,
+      "learning_rate": 4.5380898237635025e-05,
+      "loss": 6.0043,
+      "step": 2125
+    },
+    {
+      "epoch": 0.11774461028192372,
+      "grad_norm": 2.242952585220337,
+      "learning_rate": 4.536668561682775e-05,
+      "loss": 5.8624,
+      "step": 2130
+    },
+    {
+      "epoch": 0.11802100608070758,
+      "grad_norm": 3.177030086517334,
+      "learning_rate": 4.5352472996020466e-05,
+      "loss": 5.9175,
+      "step": 2135
+    },
+    {
+      "epoch": 0.11829740187949143,
+      "grad_norm": 3.0436911582946777,
+      "learning_rate": 4.533826037521319e-05,
+      "loss": 6.0026,
+      "step": 2140
+    },
+    {
+      "epoch": 0.11857379767827529,
+      "grad_norm": 3.5062520503997803,
+      "learning_rate": 4.5324047754405914e-05,
+      "loss": 6.2474,
+      "step": 2145
+    },
+    {
+      "epoch": 0.11885019347705915,
+      "grad_norm": 2.291909694671631,
+      "learning_rate": 4.530983513359864e-05,
+      "loss": 6.2062,
+      "step": 2150
+    },
+    {
+      "epoch": 0.119126589275843,
+      "grad_norm": 2.3270986080169678,
+      "learning_rate": 4.529562251279136e-05,
+      "loss": 6.2895,
+      "step": 2155
+    },
+    {
+      "epoch": 0.11940298507462686,
+      "grad_norm": 3.0153403282165527,
+      "learning_rate": 4.5281409891984086e-05,
+      "loss": 6.1369,
+      "step": 2160
+    },
+    {
+      "epoch": 0.11967938087341072,
+      "grad_norm": 2.496553897857666,
+      "learning_rate": 4.526719727117681e-05,
+      "loss": 6.1861,
+      "step": 2165
+    },
+    {
+      "epoch": 0.11995577667219458,
+      "grad_norm": 2.5240025520324707,
+      "learning_rate": 4.525298465036953e-05,
+      "loss": 6.3099,
+      "step": 2170
+    },
+    {
+      "epoch": 0.12023217247097844,
+      "grad_norm": 2.7190871238708496,
+      "learning_rate": 4.523877202956225e-05,
+      "loss": 5.737,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1205085682697623,
+      "grad_norm": 2.7631752490997314,
+      "learning_rate": 4.5224559408754975e-05,
+      "loss": 6.1848,
+      "step": 2180
+    },
+    {
+      "epoch": 0.12078496406854616,
+      "grad_norm": 2.1841232776641846,
+      "learning_rate": 4.52103467879477e-05,
+      "loss": 6.0631,
+      "step": 2185
+    },
+    {
+      "epoch": 0.12106135986733002,
+      "grad_norm": 2.531573534011841,
+      "learning_rate": 4.519613416714042e-05,
+      "loss": 6.1404,
+      "step": 2190
+    },
+    {
+      "epoch": 0.12133775566611388,
+      "grad_norm": 3.5655975341796875,
+      "learning_rate": 4.518192154633315e-05,
+      "loss": 6.1925,
+      "step": 2195
+    },
+    {
+      "epoch": 0.12161415146489774,
+      "grad_norm": 2.4978582859039307,
+      "learning_rate": 4.516770892552587e-05,
+      "loss": 5.9768,
+      "step": 2200
+    },
+    {
+      "epoch": 0.12189054726368159,
+      "grad_norm": 2.800384759902954,
+      "learning_rate": 4.515349630471859e-05,
+      "loss": 6.1647,
+      "step": 2205
+    },
+    {
+      "epoch": 0.12216694306246545,
+      "grad_norm": 3.2375316619873047,
+      "learning_rate": 4.513928368391132e-05,
+      "loss": 5.9686,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1224433388612493,
+      "grad_norm": 3.0477945804595947,
+      "learning_rate": 4.512507106310404e-05,
+      "loss": 5.8481,
+      "step": 2215
+    },
+    {
+      "epoch": 0.12271973466003316,
+      "grad_norm": 2.7675116062164307,
+      "learning_rate": 4.511085844229676e-05,
+      "loss": 6.1965,
+      "step": 2220
+    },
+    {
+      "epoch": 0.12299613045881702,
+      "grad_norm": 3.220231056213379,
+      "learning_rate": 4.5096645821489484e-05,
+      "loss": 6.0191,
+      "step": 2225
+    },
+    {
+      "epoch": 0.12327252625760088,
+      "grad_norm": 2.473461389541626,
+      "learning_rate": 4.508243320068221e-05,
+      "loss": 5.7706,
+      "step": 2230
+    },
+    {
+      "epoch": 0.12354892205638475,
+      "grad_norm": 2.950104236602783,
+      "learning_rate": 4.506822057987493e-05,
+      "loss": 6.0164,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1238253178551686,
+      "grad_norm": 3.0107226371765137,
+      "learning_rate": 4.5054007959067656e-05,
+      "loss": 5.9676,
+      "step": 2240
+    },
+    {
+      "epoch": 0.12410171365395246,
+      "grad_norm": 3.3894078731536865,
+      "learning_rate": 4.503979533826038e-05,
+      "loss": 5.9496,
+      "step": 2245
+    },
+    {
+      "epoch": 0.12437810945273632,
+      "grad_norm": 2.7371933460235596,
+      "learning_rate": 4.5025582717453104e-05,
+      "loss": 5.9917,
+      "step": 2250
+    },
+    {
+      "epoch": 0.12465450525152018,
+      "grad_norm": 2.829718589782715,
+      "learning_rate": 4.501137009664582e-05,
+      "loss": 5.881,
+      "step": 2255
+    },
+    {
+      "epoch": 0.12493090105030404,
+      "grad_norm": 2.5616791248321533,
+      "learning_rate": 4.4997157475838545e-05,
+      "loss": 5.8512,
+      "step": 2260
+    },
+    {
+      "epoch": 0.1252072968490879,
+      "grad_norm": 2.7623887062072754,
+      "learning_rate": 4.4982944855031276e-05,
+      "loss": 6.1254,
+      "step": 2265
+    },
+    {
+      "epoch": 0.12548369264787176,
+      "grad_norm": 2.115161895751953,
+      "learning_rate": 4.496873223422399e-05,
+      "loss": 5.9236,
+      "step": 2270
+    },
+    {
+      "epoch": 0.12576008844665562,
+      "grad_norm": 2.32071590423584,
+      "learning_rate": 4.495451961341672e-05,
+      "loss": 5.8425,
+      "step": 2275
+    },
+    {
+      "epoch": 0.12603648424543948,
+      "grad_norm": 3.4130492210388184,
+      "learning_rate": 4.494030699260944e-05,
+      "loss": 5.8947,
+      "step": 2280
+    },
+    {
+      "epoch": 0.12631288004422334,
+      "grad_norm": 2.7116572856903076,
+      "learning_rate": 4.4926094371802165e-05,
+      "loss": 6.2554,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1265892758430072,
+      "grad_norm": 2.960740804672241,
+      "learning_rate": 4.491188175099488e-05,
+      "loss": 5.835,
+      "step": 2290
+    },
+    {
+      "epoch": 0.12686567164179105,
+      "grad_norm": 2.641920804977417,
+      "learning_rate": 4.4897669130187606e-05,
+      "loss": 6.0452,
+      "step": 2295
+    },
+    {
+      "epoch": 0.1271420674405749,
+      "grad_norm": 2.40816068649292,
+      "learning_rate": 4.488345650938034e-05,
+      "loss": 6.1682,
+      "step": 2300
+    },
+    {
+      "epoch": 0.12741846323935876,
+      "grad_norm": 2.953784942626953,
+      "learning_rate": 4.4869243888573054e-05,
+      "loss": 5.8982,
+      "step": 2305
+    },
+    {
+      "epoch": 0.12769485903814262,
+      "grad_norm": 2.7458655834198,
+      "learning_rate": 4.485503126776578e-05,
+      "loss": 5.7274,
+      "step": 2310
+    },
+    {
+      "epoch": 0.12797125483692648,
+      "grad_norm": 2.4384260177612305,
+      "learning_rate": 4.48408186469585e-05,
+      "loss": 5.8491,
+      "step": 2315
+    },
+    {
+      "epoch": 0.12824765063571034,
+      "grad_norm": 3.5330193042755127,
+      "learning_rate": 4.4826606026151226e-05,
+      "loss": 6.1105,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1285240464344942,
+      "grad_norm": 3.598749876022339,
+      "learning_rate": 4.481239340534395e-05,
+      "loss": 5.7398,
+      "step": 2325
+    },
+    {
+      "epoch": 0.12880044223327805,
+      "grad_norm": 2.857062339782715,
+      "learning_rate": 4.479818078453667e-05,
+      "loss": 6.0652,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1290768380320619,
+      "grad_norm": 2.578871250152588,
+      "learning_rate": 4.47839681637294e-05,
+      "loss": 5.9814,
+      "step": 2335
+    },
+    {
+      "epoch": 0.12935323383084577,
+      "grad_norm": 3.088256597518921,
+      "learning_rate": 4.4769755542922115e-05,
+      "loss": 5.8343,
+      "step": 2340
+    },
+    {
+      "epoch": 0.12962962962962962,
+      "grad_norm": 2.48786997795105,
+      "learning_rate": 4.475554292211484e-05,
+      "loss": 5.9264,
+      "step": 2345
+    },
+    {
+      "epoch": 0.12990602542841348,
+      "grad_norm": 2.87296199798584,
+      "learning_rate": 4.474133030130756e-05,
+      "loss": 5.8988,
+      "step": 2350
+    },
+    {
+      "epoch": 0.13018242122719734,
+      "grad_norm": 3.202390193939209,
+      "learning_rate": 4.472711768050029e-05,
+      "loss": 5.7166,
+      "step": 2355
+    },
+    {
+      "epoch": 0.1304588170259812,
+      "grad_norm": 3.6477270126342773,
+      "learning_rate": 4.471290505969301e-05,
+      "loss": 5.7914,
+      "step": 2360
+    },
+    {
+      "epoch": 0.13073521282476505,
+      "grad_norm": 2.6237668991088867,
+      "learning_rate": 4.469869243888573e-05,
+      "loss": 5.9484,
+      "step": 2365
+    },
+    {
+      "epoch": 0.1310116086235489,
+      "grad_norm": 2.782205104827881,
+      "learning_rate": 4.468447981807846e-05,
+      "loss": 6.0338,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1312880044223328,
+      "grad_norm": 2.7849576473236084,
+      "learning_rate": 4.467026719727118e-05,
+      "loss": 6.1244,
+      "step": 2375
+    },
+    {
+      "epoch": 0.13156440022111665,
+      "grad_norm": 2.8187060356140137,
+      "learning_rate": 4.46560545764639e-05,
+      "loss": 5.6802,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1318407960199005,
+      "grad_norm": 2.4502339363098145,
+      "learning_rate": 4.4641841955656624e-05,
+      "loss": 5.9567,
+      "step": 2385
+    },
+    {
+      "epoch": 0.13211719181868437,
+      "grad_norm": 2.743391275405884,
+      "learning_rate": 4.462762933484935e-05,
+      "loss": 5.737,
+      "step": 2390
+    },
+    {
+      "epoch": 0.13239358761746822,
+      "grad_norm": 2.580054998397827,
+      "learning_rate": 4.461341671404207e-05,
+      "loss": 5.7443,
+      "step": 2395
+    },
+    {
+      "epoch": 0.13266998341625208,
+      "grad_norm": 3.157968282699585,
+      "learning_rate": 4.459920409323479e-05,
+      "loss": 6.083,
+      "step": 2400
+    },
+    {
+      "epoch": 0.13294637921503594,
+      "grad_norm": 2.9599556922912598,
+      "learning_rate": 4.458499147242752e-05,
+      "loss": 6.0339,
+      "step": 2405
+    },
+    {
+      "epoch": 0.1332227750138198,
+      "grad_norm": 3.3958144187927246,
+      "learning_rate": 4.4570778851620244e-05,
+      "loss": 5.9857,
+      "step": 2410
+    },
+    {
+      "epoch": 0.13349917081260365,
+      "grad_norm": 3.3407669067382812,
+      "learning_rate": 4.455656623081296e-05,
+      "loss": 5.7594,
+      "step": 2415
+    },
+    {
+      "epoch": 0.1337755666113875,
+      "grad_norm": 2.816286563873291,
+      "learning_rate": 4.4542353610005685e-05,
+      "loss": 6.0003,
+      "step": 2420
+    },
+    {
+      "epoch": 0.13405196241017137,
+      "grad_norm": 2.553955554962158,
+      "learning_rate": 4.452814098919841e-05,
+      "loss": 6.0085,
+      "step": 2425
+    },
+    {
+      "epoch": 0.13432835820895522,
+      "grad_norm": 3.3297877311706543,
+      "learning_rate": 4.451392836839113e-05,
+      "loss": 5.8029,
+      "step": 2430
+    },
+    {
+      "epoch": 0.13460475400773908,
+      "grad_norm": 2.530499219894409,
+      "learning_rate": 4.449971574758386e-05,
+      "loss": 5.9603,
+      "step": 2435
+    },
+    {
+      "epoch": 0.13488114980652294,
+      "grad_norm": 2.859200954437256,
+      "learning_rate": 4.448550312677658e-05,
+      "loss": 5.9611,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1351575456053068,
+      "grad_norm": 3.111984968185425,
+      "learning_rate": 4.4471290505969305e-05,
+      "loss": 5.9601,
+      "step": 2445
+    },
+    {
+      "epoch": 0.13543394140409065,
+      "grad_norm": 2.6724886894226074,
+      "learning_rate": 4.445707788516202e-05,
+      "loss": 5.9028,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1357103372028745,
+      "grad_norm": 2.909550666809082,
+      "learning_rate": 4.4442865264354746e-05,
+      "loss": 6.1483,
+      "step": 2455
+    },
+    {
+      "epoch": 0.13598673300165837,
+      "grad_norm": 2.8380892276763916,
+      "learning_rate": 4.442865264354748e-05,
+      "loss": 5.9103,
+      "step": 2460
+    },
+    {
+      "epoch": 0.13626312880044222,
+      "grad_norm": 2.6073994636535645,
+      "learning_rate": 4.4414440022740194e-05,
+      "loss": 5.9974,
+      "step": 2465
+    },
+    {
+      "epoch": 0.13653952459922608,
+      "grad_norm": 3.0573458671569824,
+      "learning_rate": 4.440022740193292e-05,
+      "loss": 5.9833,
+      "step": 2470
+    },
+    {
+      "epoch": 0.13681592039800994,
+      "grad_norm": 2.835435152053833,
+      "learning_rate": 4.438601478112564e-05,
+      "loss": 5.9305,
+      "step": 2475
+    },
+    {
+      "epoch": 0.1370923161967938,
+      "grad_norm": 2.9409842491149902,
+      "learning_rate": 4.4371802160318366e-05,
+      "loss": 6.0945,
+      "step": 2480
+    },
+    {
+      "epoch": 0.13736871199557768,
+      "grad_norm": 2.7987194061279297,
+      "learning_rate": 4.435758953951108e-05,
+      "loss": 6.0813,
+      "step": 2485
+    },
+    {
+      "epoch": 0.13764510779436154,
+      "grad_norm": 2.926170825958252,
+      "learning_rate": 4.434337691870381e-05,
+      "loss": 6.1433,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1379215035931454,
+      "grad_norm": 2.812005043029785,
+      "learning_rate": 4.432916429789654e-05,
+      "loss": 6.1284,
+      "step": 2495
+    },
+    {
+      "epoch": 0.13819789939192925,
+      "grad_norm": 4.218148231506348,
+      "learning_rate": 4.4314951677089255e-05,
+      "loss": 6.2463,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1384742951907131,
+      "grad_norm": 3.2735049724578857,
+      "learning_rate": 4.430073905628198e-05,
+      "loss": 6.3637,
+      "step": 2505
+    },
+    {
+      "epoch": 0.13875069098949697,
+      "grad_norm": 2.975898265838623,
+      "learning_rate": 4.42865264354747e-05,
+      "loss": 5.8744,
+      "step": 2510
+    },
+    {
+      "epoch": 0.13902708678828082,
+      "grad_norm": 2.5579276084899902,
+      "learning_rate": 4.427231381466743e-05,
+      "loss": 5.8671,
+      "step": 2515
+    },
+    {
+      "epoch": 0.13930348258706468,
+      "grad_norm": 2.9788095951080322,
+      "learning_rate": 4.425810119386015e-05,
+      "loss": 5.7345,
+      "step": 2520
+    },
+    {
+      "epoch": 0.13957987838584854,
+      "grad_norm": 2.781212568283081,
+      "learning_rate": 4.4243888573052875e-05,
+      "loss": 6.0089,
+      "step": 2525
+    },
+    {
+      "epoch": 0.1398562741846324,
+      "grad_norm": 3.1490135192871094,
+      "learning_rate": 4.42296759522456e-05,
+      "loss": 5.9384,
+      "step": 2530
+    },
+    {
+      "epoch": 0.14013266998341625,
+      "grad_norm": 3.2188096046447754,
+      "learning_rate": 4.4215463331438316e-05,
+      "loss": 6.0045,
+      "step": 2535
+    },
+    {
+      "epoch": 0.1404090657822001,
+      "grad_norm": 3.2777388095855713,
+      "learning_rate": 4.420125071063104e-05,
+      "loss": 5.956,
+      "step": 2540
+    },
+    {
+      "epoch": 0.14068546158098397,
+      "grad_norm": 3.3653669357299805,
+      "learning_rate": 4.418703808982377e-05,
+      "loss": 5.9132,
+      "step": 2545
+    },
+    {
+      "epoch": 0.14096185737976782,
+      "grad_norm": 2.6247451305389404,
+      "learning_rate": 4.417282546901649e-05,
+      "loss": 5.8262,
+      "step": 2550
+    },
+    {
+      "epoch": 0.14123825317855168,
+      "grad_norm": 2.3984429836273193,
+      "learning_rate": 4.415861284820921e-05,
+      "loss": 5.8723,
+      "step": 2555
+    },
+    {
+      "epoch": 0.14151464897733554,
+      "grad_norm": 2.4766428470611572,
+      "learning_rate": 4.4144400227401936e-05,
+      "loss": 6.0487,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1417910447761194,
+      "grad_norm": 2.6634018421173096,
+      "learning_rate": 4.413018760659466e-05,
+      "loss": 6.0009,
+      "step": 2565
+    },
+    {
+      "epoch": 0.14206744057490325,
+      "grad_norm": 3.0467846393585205,
+      "learning_rate": 4.4115974985787384e-05,
+      "loss": 6.0557,
+      "step": 2570
+    },
+    {
+      "epoch": 0.1423438363736871,
+      "grad_norm": 2.9271202087402344,
+      "learning_rate": 4.41017623649801e-05,
+      "loss": 5.7928,
+      "step": 2575
+    },
+    {
+      "epoch": 0.14262023217247097,
+      "grad_norm": 2.9135117530822754,
+      "learning_rate": 4.408754974417283e-05,
+      "loss": 5.6288,
+      "step": 2580
+    },
+    {
+      "epoch": 0.14289662797125482,
+      "grad_norm": 2.667640209197998,
+      "learning_rate": 4.407333712336555e-05,
+      "loss": 5.9423,
+      "step": 2585
+    },
+    {
+      "epoch": 0.14317302377003868,
+      "grad_norm": 2.59460711479187,
+      "learning_rate": 4.405912450255827e-05,
+      "loss": 6.1813,
+      "step": 2590
+    },
+    {
+      "epoch": 0.14344941956882257,
+      "grad_norm": 3.593357801437378,
+      "learning_rate": 4.4044911881751e-05,
+      "loss": 5.8255,
+      "step": 2595
+    },
+    {
+      "epoch": 0.14372581536760642,
+      "grad_norm": 2.5840964317321777,
+      "learning_rate": 4.403069926094372e-05,
+      "loss": 5.9867,
+      "step": 2600
+    },
+    {
+      "epoch": 0.14400221116639028,
+      "grad_norm": 3.780487060546875,
+      "learning_rate": 4.4016486640136445e-05,
+      "loss": 5.734,
+      "step": 2605
+    },
+    {
+      "epoch": 0.14427860696517414,
+      "grad_norm": 2.3756935596466064,
+      "learning_rate": 4.400227401932916e-05,
+      "loss": 6.0119,
+      "step": 2610
+    },
+    {
+      "epoch": 0.144555002763958,
+      "grad_norm": 2.420318126678467,
+      "learning_rate": 4.398806139852189e-05,
+      "loss": 5.8429,
+      "step": 2615
+    },
+    {
+      "epoch": 0.14483139856274185,
+      "grad_norm": 2.4723665714263916,
+      "learning_rate": 4.397384877771461e-05,
+      "loss": 5.9102,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1451077943615257,
+      "grad_norm": 2.8111989498138428,
+      "learning_rate": 4.3959636156907334e-05,
+      "loss": 5.796,
+      "step": 2625
+    },
+    {
+      "epoch": 0.14538419016030957,
+      "grad_norm": 2.4505462646484375,
+      "learning_rate": 4.394542353610006e-05,
+      "loss": 5.7925,
+      "step": 2630
+    },
+    {
+      "epoch": 0.14566058595909342,
+      "grad_norm": 2.9195985794067383,
+      "learning_rate": 4.393121091529278e-05,
+      "loss": 5.9991,
+      "step": 2635
+    },
+    {
+      "epoch": 0.14593698175787728,
+      "grad_norm": 3.3594954013824463,
+      "learning_rate": 4.3916998294485506e-05,
+      "loss": 6.0823,
+      "step": 2640
+    },
+    {
+      "epoch": 0.14621337755666114,
+      "grad_norm": 2.6641616821289062,
+      "learning_rate": 4.3902785673678224e-05,
+      "loss": 6.0646,
+      "step": 2645
+    },
+    {
+      "epoch": 0.146489773355445,
+      "grad_norm": 3.2339580059051514,
+      "learning_rate": 4.3888573052870954e-05,
+      "loss": 6.2217,
+      "step": 2650
+    },
+    {
+      "epoch": 0.14676616915422885,
+      "grad_norm": 2.6006035804748535,
+      "learning_rate": 4.387436043206368e-05,
+      "loss": 5.8923,
+      "step": 2655
+    },
+    {
+      "epoch": 0.1470425649530127,
+      "grad_norm": 2.918951988220215,
+      "learning_rate": 4.3860147811256395e-05,
+      "loss": 5.8821,
+      "step": 2660
+    },
+    {
+      "epoch": 0.14731896075179657,
+      "grad_norm": 3.5756337642669678,
+      "learning_rate": 4.384593519044912e-05,
+      "loss": 5.9689,
+      "step": 2665
+    },
+    {
+      "epoch": 0.14759535655058043,
+      "grad_norm": 2.472301483154297,
+      "learning_rate": 4.3831722569641843e-05,
+      "loss": 6.1298,
+      "step": 2670
+    },
+    {
+      "epoch": 0.14787175234936428,
+      "grad_norm": 3.1976962089538574,
+      "learning_rate": 4.381750994883457e-05,
+      "loss": 5.9328,
+      "step": 2675
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 2.672852039337158,
+      "learning_rate": 4.380329732802729e-05,
+      "loss": 5.7058,
+      "step": 2680
+    },
+    {
+      "epoch": 0.148424543946932,
+      "grad_norm": 2.8173885345458984,
+      "learning_rate": 4.3789084707220015e-05,
+      "loss": 5.9733,
+      "step": 2685
+    },
+    {
+      "epoch": 0.14870093974571585,
+      "grad_norm": 2.9812614917755127,
+      "learning_rate": 4.377487208641274e-05,
+      "loss": 5.9916,
+      "step": 2690
+    },
+    {
+      "epoch": 0.1489773355444997,
+      "grad_norm": 2.7552757263183594,
+      "learning_rate": 4.3760659465605457e-05,
+      "loss": 5.4035,
+      "step": 2695
+    },
+    {
+      "epoch": 0.14925373134328357,
+      "grad_norm": 2.684389114379883,
+      "learning_rate": 4.374644684479818e-05,
+      "loss": 5.9404,
+      "step": 2700
+    },
+    {
+      "epoch": 0.14953012714206745,
+      "grad_norm": 3.3270440101623535,
+      "learning_rate": 4.3732234223990905e-05,
+      "loss": 5.9081,
+      "step": 2705
+    },
+    {
+      "epoch": 0.1498065229408513,
+      "grad_norm": 3.212458848953247,
+      "learning_rate": 4.371802160318363e-05,
+      "loss": 6.2057,
+      "step": 2710
+    },
+    {
+      "epoch": 0.15008291873963517,
+      "grad_norm": 2.9287021160125732,
+      "learning_rate": 4.370380898237635e-05,
+      "loss": 6.2109,
+      "step": 2715
+    },
+    {
+      "epoch": 0.15035931453841903,
+      "grad_norm": 3.4137637615203857,
+      "learning_rate": 4.3689596361569076e-05,
+      "loss": 5.805,
+      "step": 2720
+    },
+    {
+      "epoch": 0.15063571033720288,
+      "grad_norm": 2.5240910053253174,
+      "learning_rate": 4.36753837407618e-05,
+      "loss": 5.8864,
+      "step": 2725
+    },
+    {
+      "epoch": 0.15091210613598674,
+      "grad_norm": 3.12414288520813,
+      "learning_rate": 4.366117111995452e-05,
+      "loss": 6.0336,
+      "step": 2730
+    },
+    {
+      "epoch": 0.1511885019347706,
+      "grad_norm": 3.1242873668670654,
+      "learning_rate": 4.364695849914724e-05,
+      "loss": 6.021,
+      "step": 2735
+    },
+    {
+      "epoch": 0.15146489773355445,
+      "grad_norm": 3.223360538482666,
+      "learning_rate": 4.363274587833997e-05,
+      "loss": 6.0656,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1517412935323383,
+      "grad_norm": 3.3459601402282715,
+      "learning_rate": 4.361853325753269e-05,
+      "loss": 6.0403,
+      "step": 2745
+    },
+    {
+      "epoch": 0.15201768933112217,
+      "grad_norm": 2.8621153831481934,
+      "learning_rate": 4.3604320636725414e-05,
+      "loss": 5.6734,
+      "step": 2750
+    },
+    {
+      "epoch": 0.15229408512990603,
+      "grad_norm": 2.841547966003418,
+      "learning_rate": 4.359010801591814e-05,
+      "loss": 5.8039,
+      "step": 2755
+    },
+    {
+      "epoch": 0.15257048092868988,
+      "grad_norm": 2.655369758605957,
+      "learning_rate": 4.357589539511086e-05,
+      "loss": 5.8627,
+      "step": 2760
+    },
+    {
+      "epoch": 0.15284687672747374,
+      "grad_norm": 3.5070695877075195,
+      "learning_rate": 4.3561682774303586e-05,
+      "loss": 5.7962,
+      "step": 2765
+    },
+    {
+      "epoch": 0.1531232725262576,
+      "grad_norm": 2.460430860519409,
+      "learning_rate": 4.35474701534963e-05,
+      "loss": 5.9058,
+      "step": 2770
+    },
+    {
+      "epoch": 0.15339966832504145,
+      "grad_norm": 2.9735209941864014,
+      "learning_rate": 4.3533257532689033e-05,
+      "loss": 5.9459,
+      "step": 2775
+    },
+    {
+      "epoch": 0.1536760641238253,
+      "grad_norm": 2.6049184799194336,
+      "learning_rate": 4.351904491188175e-05,
+      "loss": 5.8831,
+      "step": 2780
+    },
+    {
+      "epoch": 0.15395245992260917,
+      "grad_norm": 3.0773181915283203,
+      "learning_rate": 4.3504832291074475e-05,
+      "loss": 5.8567,
+      "step": 2785
+    },
+    {
+      "epoch": 0.15422885572139303,
+      "grad_norm": 2.9767990112304688,
+      "learning_rate": 4.34906196702672e-05,
+      "loss": 5.884,
+      "step": 2790
+    },
+    {
+      "epoch": 0.15450525152017688,
+      "grad_norm": 2.9856009483337402,
+      "learning_rate": 4.347640704945992e-05,
+      "loss": 5.7991,
+      "step": 2795
+    },
+    {
+      "epoch": 0.15478164731896074,
+      "grad_norm": 2.7694013118743896,
+      "learning_rate": 4.3462194428652647e-05,
+      "loss": 5.9117,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1550580431177446,
+      "grad_norm": 2.524625778198242,
+      "learning_rate": 4.3447981807845364e-05,
+      "loss": 5.8089,
+      "step": 2805
+    },
+    {
+      "epoch": 0.15533443891652846,
+      "grad_norm": 2.265674591064453,
+      "learning_rate": 4.3433769187038095e-05,
+      "loss": 5.7669,
+      "step": 2810
+    },
+    {
+      "epoch": 0.15561083471531234,
+      "grad_norm": 2.9891350269317627,
+      "learning_rate": 4.341955656623081e-05,
+      "loss": 5.7408,
+      "step": 2815
+    },
+    {
+      "epoch": 0.1558872305140962,
+      "grad_norm": 2.8526082038879395,
+      "learning_rate": 4.3405343945423536e-05,
+      "loss": 5.5601,
+      "step": 2820
+    },
+    {
+      "epoch": 0.15616362631288006,
+      "grad_norm": 2.755467653274536,
+      "learning_rate": 4.3391131324616266e-05,
+      "loss": 5.6325,
+      "step": 2825
+    },
+    {
+      "epoch": 0.1564400221116639,
+      "grad_norm": 3.075253963470459,
+      "learning_rate": 4.3376918703808984e-05,
+      "loss": 6.0243,
+      "step": 2830
+    },
+    {
+      "epoch": 0.15671641791044777,
+      "grad_norm": 2.6494526863098145,
+      "learning_rate": 4.336270608300171e-05,
+      "loss": 5.7588,
+      "step": 2835
+    },
+    {
+      "epoch": 0.15699281370923163,
+      "grad_norm": 3.440871238708496,
+      "learning_rate": 4.334849346219443e-05,
+      "loss": 5.9682,
+      "step": 2840
+    },
+    {
+      "epoch": 0.15726920950801548,
+      "grad_norm": 3.129068374633789,
+      "learning_rate": 4.3334280841387156e-05,
+      "loss": 5.8397,
+      "step": 2845
+    },
+    {
+      "epoch": 0.15754560530679934,
+      "grad_norm": 2.8854775428771973,
+      "learning_rate": 4.332006822057988e-05,
+      "loss": 5.6775,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1578220011055832,
+      "grad_norm": 2.4169180393218994,
+      "learning_rate": 4.33058555997726e-05,
+      "loss": 5.7348,
+      "step": 2855
+    },
+    {
+      "epoch": 0.15809839690436706,
+      "grad_norm": 2.6656081676483154,
+      "learning_rate": 4.329164297896533e-05,
+      "loss": 5.5804,
+      "step": 2860
+    },
+    {
+      "epoch": 0.1583747927031509,
+      "grad_norm": 2.1529061794281006,
+      "learning_rate": 4.3277430358158045e-05,
+      "loss": 6.0426,
+      "step": 2865
+    },
+    {
+      "epoch": 0.15865118850193477,
+      "grad_norm": 2.7012405395507812,
+      "learning_rate": 4.326321773735077e-05,
+      "loss": 5.8865,
+      "step": 2870
+    },
+    {
+      "epoch": 0.15892758430071863,
+      "grad_norm": 3.113957643508911,
+      "learning_rate": 4.324900511654349e-05,
+      "loss": 5.7905,
+      "step": 2875
+    },
+    {
+      "epoch": 0.15920398009950248,
+      "grad_norm": 2.826045274734497,
+      "learning_rate": 4.323479249573622e-05,
+      "loss": 5.8868,
+      "step": 2880
+    },
+    {
+      "epoch": 0.15948037589828634,
+      "grad_norm": 2.5294294357299805,
+      "learning_rate": 4.322057987492894e-05,
+      "loss": 5.6514,
+      "step": 2885
+    },
+    {
+      "epoch": 0.1597567716970702,
+      "grad_norm": 2.923967123031616,
+      "learning_rate": 4.320636725412166e-05,
+      "loss": 6.0134,
+      "step": 2890
+    },
+    {
+      "epoch": 0.16003316749585406,
+      "grad_norm": 3.1587777137756348,
+      "learning_rate": 4.319215463331439e-05,
+      "loss": 6.1809,
+      "step": 2895
+    },
+    {
+      "epoch": 0.1603095632946379,
+      "grad_norm": 2.6380045413970947,
+      "learning_rate": 4.317794201250711e-05,
+      "loss": 5.828,
+      "step": 2900
+    },
+    {
+      "epoch": 0.16058595909342177,
+      "grad_norm": 3.320361852645874,
+      "learning_rate": 4.316372939169983e-05,
+      "loss": 5.8606,
+      "step": 2905
+    },
+    {
+      "epoch": 0.16086235489220563,
+      "grad_norm": 2.4899818897247314,
+      "learning_rate": 4.3149516770892554e-05,
+      "loss": 5.7003,
+      "step": 2910
+    },
+    {
+      "epoch": 0.16113875069098949,
+      "grad_norm": 3.366610050201416,
+      "learning_rate": 4.313530415008528e-05,
+      "loss": 5.7348,
+      "step": 2915
+    },
+    {
+      "epoch": 0.16141514648977334,
+      "grad_norm": 3.16973876953125,
+      "learning_rate": 4.3121091529278e-05,
+      "loss": 5.6955,
+      "step": 2920
+    },
+    {
+      "epoch": 0.16169154228855723,
+      "grad_norm": 3.0917766094207764,
+      "learning_rate": 4.310687890847072e-05,
+      "loss": 5.8221,
+      "step": 2925
+    },
+    {
+      "epoch": 0.16196793808734108,
+      "grad_norm": 3.0360071659088135,
+      "learning_rate": 4.309266628766345e-05,
+      "loss": 5.7854,
+      "step": 2930
+    },
+    {
+      "epoch": 0.16224433388612494,
+      "grad_norm": 2.421694040298462,
+      "learning_rate": 4.3078453666856174e-05,
+      "loss": 5.5865,
+      "step": 2935
+    },
+    {
+      "epoch": 0.1625207296849088,
+      "grad_norm": 2.4142487049102783,
+      "learning_rate": 4.306424104604889e-05,
+      "loss": 5.6557,
+      "step": 2940
+    },
+    {
+      "epoch": 0.16279712548369266,
+      "grad_norm": 2.854003429412842,
+      "learning_rate": 4.3050028425241615e-05,
+      "loss": 5.8642,
+      "step": 2945
+    },
+    {
+      "epoch": 0.1630735212824765,
+      "grad_norm": 2.4394283294677734,
+      "learning_rate": 4.303581580443434e-05,
+      "loss": 5.5967,
+      "step": 2950
+    },
+    {
+      "epoch": 0.16334991708126037,
+      "grad_norm": 3.362593650817871,
+      "learning_rate": 4.302160318362706e-05,
+      "loss": 5.9702,
+      "step": 2955
+    },
+    {
+      "epoch": 0.16362631288004423,
+      "grad_norm": 2.2404420375823975,
+      "learning_rate": 4.300739056281979e-05,
+      "loss": 5.6981,
+      "step": 2960
+    },
+    {
+      "epoch": 0.16390270867882809,
+      "grad_norm": 2.7826569080352783,
+      "learning_rate": 4.299317794201251e-05,
+      "loss": 5.7644,
+      "step": 2965
+    },
+    {
+      "epoch": 0.16417910447761194,
+      "grad_norm": 2.6472930908203125,
+      "learning_rate": 4.2978965321205235e-05,
+      "loss": 6.105,
+      "step": 2970
+    },
+    {
+      "epoch": 0.1644555002763958,
+      "grad_norm": 3.248873472213745,
+      "learning_rate": 4.296475270039795e-05,
+      "loss": 5.7841,
+      "step": 2975
+    },
+    {
+      "epoch": 0.16473189607517966,
+      "grad_norm": 3.790102958679199,
+      "learning_rate": 4.2950540079590676e-05,
+      "loss": 5.7012,
+      "step": 2980
+    },
+    {
+      "epoch": 0.16500829187396351,
+      "grad_norm": 3.1310534477233887,
+      "learning_rate": 4.293632745878341e-05,
+      "loss": 5.7268,
+      "step": 2985
+    },
+    {
+      "epoch": 0.16528468767274737,
+      "grad_norm": 2.8928165435791016,
+      "learning_rate": 4.2922114837976124e-05,
+      "loss": 5.7898,
+      "step": 2990
+    },
+    {
+      "epoch": 0.16556108347153123,
+      "grad_norm": 2.956012010574341,
+      "learning_rate": 4.290790221716885e-05,
+      "loss": 6.1025,
+      "step": 2995
+    },
+    {
+      "epoch": 0.16583747927031509,
+      "grad_norm": 3.146536111831665,
+      "learning_rate": 4.289368959636157e-05,
+      "loss": 5.8821,
+      "step": 3000
+    },
+    {
+      "epoch": 0.16611387506909894,
+      "grad_norm": 3.373002767562866,
+      "learning_rate": 4.2879476975554296e-05,
+      "loss": 5.433,
+      "step": 3005
+    },
+    {
+      "epoch": 0.1663902708678828,
+      "grad_norm": 3.2501721382141113,
+      "learning_rate": 4.286526435474702e-05,
+      "loss": 6.0184,
+      "step": 3010
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 2.656000852584839,
+      "learning_rate": 4.285105173393974e-05,
+      "loss": 5.8359,
+      "step": 3015
+    },
+    {
+      "epoch": 0.16694306246545051,
+      "grad_norm": 3.513477325439453,
+      "learning_rate": 4.283683911313247e-05,
+      "loss": 6.0025,
+      "step": 3020
+    },
+    {
+      "epoch": 0.16721945826423437,
+      "grad_norm": 3.080552816390991,
+      "learning_rate": 4.2822626492325185e-05,
+      "loss": 5.7096,
+      "step": 3025
+    },
+    {
+      "epoch": 0.16749585406301823,
+      "grad_norm": 2.202702045440674,
+      "learning_rate": 4.280841387151791e-05,
+      "loss": 5.9594,
+      "step": 3030
+    },
+    {
+      "epoch": 0.16777224986180211,
+      "grad_norm": 2.7904155254364014,
+      "learning_rate": 4.279420125071063e-05,
+      "loss": 5.749,
+      "step": 3035
+    },
+    {
+      "epoch": 0.16804864566058597,
+      "grad_norm": 3.466552495956421,
+      "learning_rate": 4.277998862990336e-05,
+      "loss": 5.8486,
+      "step": 3040
+    },
+    {
+      "epoch": 0.16832504145936983,
+      "grad_norm": 2.64650297164917,
+      "learning_rate": 4.276577600909608e-05,
+      "loss": 5.8657,
+      "step": 3045
+    },
+    {
+      "epoch": 0.16860143725815369,
+      "grad_norm": 3.313481569290161,
+      "learning_rate": 4.27515633882888e-05,
+      "loss": 5.5925,
+      "step": 3050
+    },
+    {
+      "epoch": 0.16887783305693754,
+      "grad_norm": 3.433079957962036,
+      "learning_rate": 4.273735076748153e-05,
+      "loss": 5.6123,
+      "step": 3055
+    },
+    {
+      "epoch": 0.1691542288557214,
+      "grad_norm": 3.088146686553955,
+      "learning_rate": 4.2723138146674246e-05,
+      "loss": 5.6197,
+      "step": 3060
+    },
+    {
+      "epoch": 0.16943062465450526,
+      "grad_norm": 2.703796863555908,
+      "learning_rate": 4.270892552586697e-05,
+      "loss": 5.8104,
+      "step": 3065
+    },
+    {
+      "epoch": 0.16970702045328911,
+      "grad_norm": 2.511690139770508,
+      "learning_rate": 4.2694712905059694e-05,
+      "loss": 5.5404,
+      "step": 3070
+    },
+    {
+      "epoch": 0.16998341625207297,
+      "grad_norm": 2.678618907928467,
+      "learning_rate": 4.268050028425242e-05,
+      "loss": 5.8477,
+      "step": 3075
+    },
+    {
+      "epoch": 0.17025981205085683,
+      "grad_norm": 2.9002177715301514,
+      "learning_rate": 4.266628766344514e-05,
+      "loss": 5.907,
+      "step": 3080
+    },
+    {
+      "epoch": 0.1705362078496407,
+      "grad_norm": 3.1903774738311768,
+      "learning_rate": 4.265207504263786e-05,
+      "loss": 5.8601,
+      "step": 3085
+    },
+    {
+      "epoch": 0.17081260364842454,
+      "grad_norm": 3.3715593814849854,
+      "learning_rate": 4.263786242183059e-05,
+      "loss": 5.8326,
+      "step": 3090
+    },
+    {
+      "epoch": 0.1710889994472084,
+      "grad_norm": 2.6727895736694336,
+      "learning_rate": 4.2623649801023314e-05,
+      "loss": 5.6508,
+      "step": 3095
+    },
+    {
+      "epoch": 0.17136539524599226,
+      "grad_norm": 2.429774761199951,
+      "learning_rate": 4.260943718021603e-05,
+      "loss": 5.7209,
+      "step": 3100
+    },
+    {
+      "epoch": 0.17164179104477612,
+      "grad_norm": 2.418609619140625,
+      "learning_rate": 4.2595224559408755e-05,
+      "loss": 5.6107,
+      "step": 3105
+    },
+    {
+      "epoch": 0.17191818684355997,
+      "grad_norm": 3.0767714977264404,
+      "learning_rate": 4.258101193860148e-05,
+      "loss": 5.6858,
+      "step": 3110
+    },
+    {
+      "epoch": 0.17219458264234383,
+      "grad_norm": 3.297687530517578,
+      "learning_rate": 4.25667993177942e-05,
+      "loss": 5.9651,
+      "step": 3115
+    },
+    {
+      "epoch": 0.1724709784411277,
+      "grad_norm": 3.491100549697876,
+      "learning_rate": 4.255258669698693e-05,
+      "loss": 5.6576,
+      "step": 3120
+    },
+    {
+      "epoch": 0.17274737423991154,
+      "grad_norm": 2.744009256362915,
+      "learning_rate": 4.253837407617965e-05,
+      "loss": 5.6868,
+      "step": 3125
+    },
+    {
+      "epoch": 0.1730237700386954,
+      "grad_norm": 3.076695680618286,
+      "learning_rate": 4.2524161455372375e-05,
+      "loss": 5.7896,
+      "step": 3130
+    },
+    {
+      "epoch": 0.17330016583747926,
+      "grad_norm": 2.9809587001800537,
+      "learning_rate": 4.250994883456509e-05,
+      "loss": 5.7975,
+      "step": 3135
+    },
+    {
+      "epoch": 0.17357656163626312,
+      "grad_norm": 2.666238307952881,
+      "learning_rate": 4.249573621375782e-05,
+      "loss": 6.0512,
+      "step": 3140
+    },
+    {
+      "epoch": 0.173852957435047,
+      "grad_norm": 2.905074119567871,
+      "learning_rate": 4.248152359295054e-05,
+      "loss": 5.5453,
+      "step": 3145
+    },
+    {
+      "epoch": 0.17412935323383086,
+      "grad_norm": 3.2428863048553467,
+      "learning_rate": 4.2467310972143264e-05,
+      "loss": 5.5774,
+      "step": 3150
+    },
+    {
+      "epoch": 0.17440574903261472,
+      "grad_norm": 2.5885746479034424,
+      "learning_rate": 4.245309835133599e-05,
+      "loss": 5.8335,
+      "step": 3155
+    },
+    {
+      "epoch": 0.17468214483139857,
+      "grad_norm": 2.6241211891174316,
+      "learning_rate": 4.243888573052871e-05,
+      "loss": 5.8666,
+      "step": 3160
+    },
+    {
+      "epoch": 0.17495854063018243,
+      "grad_norm": 2.9486637115478516,
+      "learning_rate": 4.2424673109721436e-05,
+      "loss": 5.4234,
+      "step": 3165
+    },
+    {
+      "epoch": 0.1752349364289663,
+      "grad_norm": 3.1195502281188965,
+      "learning_rate": 4.241046048891415e-05,
+      "loss": 5.7667,
+      "step": 3170
+    },
+    {
+      "epoch": 0.17551133222775014,
+      "grad_norm": 2.9421496391296387,
+      "learning_rate": 4.2396247868106884e-05,
+      "loss": 5.6917,
+      "step": 3175
+    },
+    {
+      "epoch": 0.175787728026534,
+      "grad_norm": 3.021730422973633,
+      "learning_rate": 4.238203524729961e-05,
+      "loss": 6.0067,
+      "step": 3180
+    },
+    {
+      "epoch": 0.17606412382531786,
+      "grad_norm": 2.6247682571411133,
+      "learning_rate": 4.2367822626492325e-05,
+      "loss": 5.7843,
+      "step": 3185
+    },
+    {
+      "epoch": 0.17634051962410172,
+      "grad_norm": 2.8806023597717285,
+      "learning_rate": 4.235361000568505e-05,
+      "loss": 5.8293,
+      "step": 3190
+    },
+    {
+      "epoch": 0.17661691542288557,
+      "grad_norm": 3.3191425800323486,
+      "learning_rate": 4.233939738487777e-05,
+      "loss": 5.8644,
+      "step": 3195
+    },
+    {
+      "epoch": 0.17689331122166943,
+      "grad_norm": 2.656522274017334,
+      "learning_rate": 4.23251847640705e-05,
+      "loss": 5.8543,
+      "step": 3200
+    },
+    {
+      "epoch": 0.1771697070204533,
+      "grad_norm": 3.2881433963775635,
+      "learning_rate": 4.231097214326322e-05,
+      "loss": 5.5148,
+      "step": 3205
+    },
+    {
+      "epoch": 0.17744610281923714,
+      "grad_norm": 2.557159423828125,
+      "learning_rate": 4.2296759522455945e-05,
+      "loss": 5.8617,
+      "step": 3210
+    },
+    {
+      "epoch": 0.177722498618021,
+      "grad_norm": 3.0635502338409424,
+      "learning_rate": 4.228254690164867e-05,
+      "loss": 5.7063,
+      "step": 3215
+    },
+    {
+      "epoch": 0.17799889441680486,
+      "grad_norm": 2.6538784503936768,
+      "learning_rate": 4.2268334280841386e-05,
+      "loss": 5.6792,
+      "step": 3220
+    },
+    {
+      "epoch": 0.17827529021558872,
+      "grad_norm": 3.1617372035980225,
+      "learning_rate": 4.225412166003411e-05,
+      "loss": 5.9972,
+      "step": 3225
+    },
+    {
+      "epoch": 0.17855168601437257,
+      "grad_norm": 2.5315380096435547,
+      "learning_rate": 4.223990903922684e-05,
+      "loss": 5.7918,
+      "step": 3230
+    },
+    {
+      "epoch": 0.17882808181315643,
+      "grad_norm": 3.2303788661956787,
+      "learning_rate": 4.222569641841956e-05,
+      "loss": 5.7051,
+      "step": 3235
+    },
+    {
+      "epoch": 0.1791044776119403,
+      "grad_norm": 2.689042568206787,
+      "learning_rate": 4.221148379761228e-05,
+      "loss": 5.7993,
+      "step": 3240
+    },
+    {
+      "epoch": 0.17938087341072415,
+      "grad_norm": 2.5231776237487793,
+      "learning_rate": 4.2197271176805006e-05,
+      "loss": 5.9223,
+      "step": 3245
+    },
+    {
+      "epoch": 0.179657269209508,
+      "grad_norm": 3.4282543659210205,
+      "learning_rate": 4.218305855599773e-05,
+      "loss": 5.6935,
+      "step": 3250
+    },
+    {
+      "epoch": 0.1799336650082919,
+      "grad_norm": 2.655792236328125,
+      "learning_rate": 4.216884593519045e-05,
+      "loss": 5.6622,
+      "step": 3255
+    },
+    {
+      "epoch": 0.18021006080707574,
+      "grad_norm": 3.699481248855591,
+      "learning_rate": 4.215463331438317e-05,
+      "loss": 5.6104,
+      "step": 3260
+    },
+    {
+      "epoch": 0.1804864566058596,
+      "grad_norm": 2.7211077213287354,
+      "learning_rate": 4.21404206935759e-05,
+      "loss": 5.7032,
+      "step": 3265
+    },
+    {
+      "epoch": 0.18076285240464346,
+      "grad_norm": 3.4669954776763916,
+      "learning_rate": 4.212620807276862e-05,
+      "loss": 5.4362,
+      "step": 3270
+    },
+    {
+      "epoch": 0.18103924820342732,
+      "grad_norm": 2.743119716644287,
+      "learning_rate": 4.211199545196134e-05,
+      "loss": 5.5355,
+      "step": 3275
+    },
+    {
+      "epoch": 0.18131564400221117,
+      "grad_norm": 2.8650245666503906,
+      "learning_rate": 4.209778283115407e-05,
+      "loss": 5.6457,
+      "step": 3280
+    },
+    {
+      "epoch": 0.18159203980099503,
+      "grad_norm": 2.683289051055908,
+      "learning_rate": 4.208357021034679e-05,
+      "loss": 5.8724,
+      "step": 3285
+    },
+    {
+      "epoch": 0.1818684355997789,
+      "grad_norm": 2.849991798400879,
+      "learning_rate": 4.2069357589539515e-05,
+      "loss": 5.8989,
+      "step": 3290
+    },
+    {
+      "epoch": 0.18214483139856275,
+      "grad_norm": 3.0860819816589355,
+      "learning_rate": 4.205514496873223e-05,
+      "loss": 5.9586,
+      "step": 3295
+    },
+    {
+      "epoch": 0.1824212271973466,
+      "grad_norm": 3.3663172721862793,
+      "learning_rate": 4.204093234792496e-05,
+      "loss": 5.7115,
+      "step": 3300
+    },
+    {
+      "epoch": 0.18269762299613046,
+      "grad_norm": 3.262286901473999,
+      "learning_rate": 4.202671972711768e-05,
+      "loss": 5.9303,
+      "step": 3305
+    },
+    {
+      "epoch": 0.18297401879491432,
+      "grad_norm": 2.6756162643432617,
+      "learning_rate": 4.2012507106310404e-05,
+      "loss": 6.0514,
+      "step": 3310
+    },
+    {
+      "epoch": 0.18325041459369817,
+      "grad_norm": 2.981628179550171,
+      "learning_rate": 4.199829448550313e-05,
+      "loss": 5.5975,
+      "step": 3315
+    },
+    {
+      "epoch": 0.18352681039248203,
+      "grad_norm": 3.121140956878662,
+      "learning_rate": 4.198408186469585e-05,
+      "loss": 5.8209,
+      "step": 3320
+    },
+    {
+      "epoch": 0.1838032061912659,
+      "grad_norm": 2.5124523639678955,
+      "learning_rate": 4.1969869243888576e-05,
+      "loss": 5.7161,
+      "step": 3325
+    },
+    {
+      "epoch": 0.18407960199004975,
+      "grad_norm": 2.389613389968872,
+      "learning_rate": 4.1955656623081294e-05,
+      "loss": 5.5582,
+      "step": 3330
+    },
+    {
+      "epoch": 0.1843559977888336,
+      "grad_norm": 3.512432336807251,
+      "learning_rate": 4.1941444002274024e-05,
+      "loss": 5.5329,
+      "step": 3335
+    },
+    {
+      "epoch": 0.18463239358761746,
+      "grad_norm": 2.90018367767334,
+      "learning_rate": 4.192723138146675e-05,
+      "loss": 5.9161,
+      "step": 3340
+    },
+    {
+      "epoch": 0.18490878938640132,
+      "grad_norm": 3.013446092605591,
+      "learning_rate": 4.1913018760659465e-05,
+      "loss": 5.9448,
+      "step": 3345
+    },
+    {
+      "epoch": 0.18518518518518517,
+      "grad_norm": 2.8273327350616455,
+      "learning_rate": 4.189880613985219e-05,
+      "loss": 5.5355,
+      "step": 3350
+    },
+    {
+      "epoch": 0.18546158098396903,
+      "grad_norm": 3.1768956184387207,
+      "learning_rate": 4.1884593519044913e-05,
+      "loss": 6.1034,
+      "step": 3355
+    },
+    {
+      "epoch": 0.1857379767827529,
+      "grad_norm": 3.083205461502075,
+      "learning_rate": 4.187038089823764e-05,
+      "loss": 5.7557,
+      "step": 3360
+    },
+    {
+      "epoch": 0.18601437258153677,
+      "grad_norm": 3.252182960510254,
+      "learning_rate": 4.1856168277430355e-05,
+      "loss": 5.6727,
+      "step": 3365
+    },
+    {
+      "epoch": 0.18629076838032063,
+      "grad_norm": 2.590454578399658,
+      "learning_rate": 4.1841955656623085e-05,
+      "loss": 6.1319,
+      "step": 3370
+    },
+    {
+      "epoch": 0.1865671641791045,
+      "grad_norm": 2.937279224395752,
+      "learning_rate": 4.182774303581581e-05,
+      "loss": 5.8309,
+      "step": 3375
+    },
+    {
+      "epoch": 0.18684355997788835,
+      "grad_norm": 2.746546745300293,
+      "learning_rate": 4.1813530415008527e-05,
+      "loss": 5.6252,
+      "step": 3380
+    },
+    {
+      "epoch": 0.1871199557766722,
+      "grad_norm": 3.1793739795684814,
+      "learning_rate": 4.179931779420125e-05,
+      "loss": 5.7403,
+      "step": 3385
+    },
+    {
+      "epoch": 0.18739635157545606,
+      "grad_norm": 3.3896684646606445,
+      "learning_rate": 4.1785105173393974e-05,
+      "loss": 6.023,
+      "step": 3390
+    },
+    {
+      "epoch": 0.18767274737423992,
+      "grad_norm": 2.4643993377685547,
+      "learning_rate": 4.17708925525867e-05,
+      "loss": 5.9285,
+      "step": 3395
+    },
+    {
+      "epoch": 0.18794914317302377,
+      "grad_norm": 2.6043946743011475,
+      "learning_rate": 4.175667993177942e-05,
+      "loss": 5.655,
+      "step": 3400
+    },
+    {
+      "epoch": 0.18822553897180763,
+      "grad_norm": 2.694716691970825,
+      "learning_rate": 4.1742467310972146e-05,
+      "loss": 5.7065,
+      "step": 3405
+    },
+    {
+      "epoch": 0.1885019347705915,
+      "grad_norm": 2.6867024898529053,
+      "learning_rate": 4.172825469016487e-05,
+      "loss": 5.8445,
+      "step": 3410
+    },
+    {
+      "epoch": 0.18877833056937535,
+      "grad_norm": 2.5469698905944824,
+      "learning_rate": 4.171404206935759e-05,
+      "loss": 5.7215,
+      "step": 3415
+    },
+    {
+      "epoch": 0.1890547263681592,
+      "grad_norm": 2.410991668701172,
+      "learning_rate": 4.169982944855031e-05,
+      "loss": 5.4893,
+      "step": 3420
+    },
+    {
+      "epoch": 0.18933112216694306,
+      "grad_norm": 3.104326009750366,
+      "learning_rate": 4.168561682774304e-05,
+      "loss": 5.4989,
+      "step": 3425
+    },
+    {
+      "epoch": 0.18960751796572692,
+      "grad_norm": 2.4258954524993896,
+      "learning_rate": 4.167140420693576e-05,
+      "loss": 5.6876,
+      "step": 3430
+    },
+    {
+      "epoch": 0.18988391376451078,
+      "grad_norm": 2.748441219329834,
+      "learning_rate": 4.1657191586128484e-05,
+      "loss": 5.5542,
+      "step": 3435
+    },
+    {
+      "epoch": 0.19016030956329463,
+      "grad_norm": 2.7649223804473877,
+      "learning_rate": 4.164297896532121e-05,
+      "loss": 6.136,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1904367053620785,
+      "grad_norm": 2.782585620880127,
+      "learning_rate": 4.162876634451393e-05,
+      "loss": 5.882,
+      "step": 3445
+    },
+    {
+      "epoch": 0.19071310116086235,
+      "grad_norm": 2.9212682247161865,
+      "learning_rate": 4.161455372370665e-05,
+      "loss": 5.9311,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1909894969596462,
+      "grad_norm": 3.2894344329833984,
+      "learning_rate": 4.160034110289938e-05,
+      "loss": 5.734,
+      "step": 3455
+    },
+    {
+      "epoch": 0.19126589275843006,
+      "grad_norm": 2.266057014465332,
+      "learning_rate": 4.1586128482092103e-05,
+      "loss": 5.4366,
+      "step": 3460
+    },
+    {
+      "epoch": 0.19154228855721392,
+      "grad_norm": 3.4635226726531982,
+      "learning_rate": 4.157191586128482e-05,
+      "loss": 5.5668,
+      "step": 3465
+    },
+    {
+      "epoch": 0.19181868435599778,
+      "grad_norm": 2.7193620204925537,
+      "learning_rate": 4.1557703240477545e-05,
+      "loss": 5.8679,
+      "step": 3470
+    },
+    {
+      "epoch": 0.19209508015478166,
+      "grad_norm": 2.5179836750030518,
+      "learning_rate": 4.154349061967027e-05,
+      "loss": 5.5888,
+      "step": 3475
+    },
+    {
+      "epoch": 0.19237147595356552,
+      "grad_norm": 2.735239267349243,
+      "learning_rate": 4.152927799886299e-05,
+      "loss": 5.6179,
+      "step": 3480
+    },
+    {
+      "epoch": 0.19264787175234938,
+      "grad_norm": 2.676177501678467,
+      "learning_rate": 4.1515065378055717e-05,
+      "loss": 5.7376,
+      "step": 3485
+    },
+    {
+      "epoch": 0.19292426755113323,
+      "grad_norm": 3.584538221359253,
+      "learning_rate": 4.150085275724844e-05,
+      "loss": 5.9845,
+      "step": 3490
+    },
+    {
+      "epoch": 0.1932006633499171,
+      "grad_norm": 3.5439796447753906,
+      "learning_rate": 4.1486640136441164e-05,
+      "loss": 5.8218,
+      "step": 3495
+    },
+    {
+      "epoch": 0.19347705914870095,
+      "grad_norm": 3.0446224212646484,
+      "learning_rate": 4.147242751563388e-05,
+      "loss": 5.5155,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1937534549474848,
+      "grad_norm": 2.8608455657958984,
+      "learning_rate": 4.1458214894826606e-05,
+      "loss": 5.9078,
+      "step": 3505
+    },
+    {
+      "epoch": 0.19402985074626866,
+      "grad_norm": 3.2207865715026855,
+      "learning_rate": 4.1444002274019336e-05,
+      "loss": 5.7971,
+      "step": 3510
+    },
+    {
+      "epoch": 0.19430624654505252,
+      "grad_norm": 3.01859188079834,
+      "learning_rate": 4.1429789653212054e-05,
+      "loss": 5.7289,
+      "step": 3515
+    },
+    {
+      "epoch": 0.19458264234383638,
+      "grad_norm": 3.373577117919922,
+      "learning_rate": 4.141557703240478e-05,
+      "loss": 5.3903,
+      "step": 3520
+    },
+    {
+      "epoch": 0.19485903814262023,
+      "grad_norm": 2.7322349548339844,
+      "learning_rate": 4.14013644115975e-05,
+      "loss": 5.3543,
+      "step": 3525
+    },
+    {
+      "epoch": 0.1951354339414041,
+      "grad_norm": 2.5114071369171143,
+      "learning_rate": 4.1387151790790226e-05,
+      "loss": 5.7759,
+      "step": 3530
+    },
+    {
+      "epoch": 0.19541182974018795,
+      "grad_norm": 2.954470634460449,
+      "learning_rate": 4.137293916998295e-05,
+      "loss": 5.8468,
+      "step": 3535
+    },
+    {
+      "epoch": 0.1956882255389718,
+      "grad_norm": 2.818599224090576,
+      "learning_rate": 4.135872654917567e-05,
+      "loss": 5.4117,
+      "step": 3540
+    },
+    {
+      "epoch": 0.19596462133775566,
+      "grad_norm": 3.0054855346679688,
+      "learning_rate": 4.13445139283684e-05,
+      "loss": 5.8437,
+      "step": 3545
+    },
+    {
+      "epoch": 0.19624101713653952,
+      "grad_norm": 2.919210433959961,
+      "learning_rate": 4.1330301307561115e-05,
+      "loss": 5.9989,
+      "step": 3550
+    },
+    {
+      "epoch": 0.19651741293532338,
+      "grad_norm": 2.6916286945343018,
+      "learning_rate": 4.131608868675384e-05,
+      "loss": 5.8415,
+      "step": 3555
+    },
+    {
+      "epoch": 0.19679380873410723,
+      "grad_norm": 2.8010470867156982,
+      "learning_rate": 4.130187606594656e-05,
+      "loss": 5.5132,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1970702045328911,
+      "grad_norm": 3.2105109691619873,
+      "learning_rate": 4.128766344513929e-05,
+      "loss": 5.7389,
+      "step": 3565
+    },
+    {
+      "epoch": 0.19734660033167495,
+      "grad_norm": 2.8323161602020264,
+      "learning_rate": 4.127345082433201e-05,
+      "loss": 5.5191,
+      "step": 3570
+    },
+    {
+      "epoch": 0.1976229961304588,
+      "grad_norm": 2.6223807334899902,
+      "learning_rate": 4.125923820352473e-05,
+      "loss": 5.6713,
+      "step": 3575
+    },
+    {
+      "epoch": 0.19789939192924266,
+      "grad_norm": 3.166440963745117,
+      "learning_rate": 4.124502558271746e-05,
+      "loss": 5.4685,
+      "step": 3580
+    },
+    {
+      "epoch": 0.19817578772802655,
+      "grad_norm": 2.5099334716796875,
+      "learning_rate": 4.1230812961910176e-05,
+      "loss": 5.7664,
+      "step": 3585
+    },
+    {
+      "epoch": 0.1984521835268104,
+      "grad_norm": 2.3826889991760254,
+      "learning_rate": 4.12166003411029e-05,
+      "loss": 5.4451,
+      "step": 3590
+    },
+    {
+      "epoch": 0.19872857932559426,
+      "grad_norm": 2.5310139656066895,
+      "learning_rate": 4.1202387720295624e-05,
+      "loss": 5.748,
+      "step": 3595
+    },
+    {
+      "epoch": 0.19900497512437812,
+      "grad_norm": 2.9146652221679688,
+      "learning_rate": 4.118817509948835e-05,
+      "loss": 5.4966,
+      "step": 3600
+    },
+    {
+      "epoch": 0.19928137092316198,
+      "grad_norm": 2.7887117862701416,
+      "learning_rate": 4.117396247868107e-05,
+      "loss": 5.6988,
+      "step": 3605
+    },
+    {
+      "epoch": 0.19955776672194583,
+      "grad_norm": 2.6834566593170166,
+      "learning_rate": 4.115974985787379e-05,
+      "loss": 5.4005,
+      "step": 3610
+    },
+    {
+      "epoch": 0.1998341625207297,
+      "grad_norm": 2.7212436199188232,
+      "learning_rate": 4.114553723706652e-05,
+      "loss": 5.4062,
+      "step": 3615
+    },
+    {
+      "epoch": 0.20011055831951355,
+      "grad_norm": 2.5989837646484375,
+      "learning_rate": 4.1131324616259244e-05,
+      "loss": 5.7674,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2003869541182974,
+      "grad_norm": 2.8077232837677,
+      "learning_rate": 4.111711199545196e-05,
+      "loss": 5.7709,
+      "step": 3625
+    },
+    {
+      "epoch": 0.20066334991708126,
+      "grad_norm": 2.3831634521484375,
+      "learning_rate": 4.1102899374644685e-05,
+      "loss": 5.6026,
+      "step": 3630
+    },
+    {
+      "epoch": 0.20093974571586512,
+      "grad_norm": 3.917057991027832,
+      "learning_rate": 4.108868675383741e-05,
+      "loss": 5.6208,
+      "step": 3635
+    },
+    {
+      "epoch": 0.20121614151464898,
+      "grad_norm": 2.841247081756592,
+      "learning_rate": 4.107447413303013e-05,
+      "loss": 5.6398,
+      "step": 3640
+    },
+    {
+      "epoch": 0.20149253731343283,
+      "grad_norm": 2.5820930004119873,
+      "learning_rate": 4.106026151222286e-05,
+      "loss": 5.7411,
+      "step": 3645
+    },
+    {
+      "epoch": 0.2017689331122167,
+      "grad_norm": 2.7610599994659424,
+      "learning_rate": 4.104604889141558e-05,
+      "loss": 5.5423,
+      "step": 3650
+    },
+    {
+      "epoch": 0.20204532891100055,
+      "grad_norm": 3.1269173622131348,
+      "learning_rate": 4.1031836270608305e-05,
+      "loss": 5.5471,
+      "step": 3655
+    },
+    {
+      "epoch": 0.2023217247097844,
+      "grad_norm": 2.523005247116089,
+      "learning_rate": 4.101762364980102e-05,
+      "loss": 5.8834,
+      "step": 3660
+    },
+    {
+      "epoch": 0.20259812050856826,
+      "grad_norm": 2.9817779064178467,
+      "learning_rate": 4.1003411028993746e-05,
+      "loss": 5.7866,
+      "step": 3665
+    },
+    {
+      "epoch": 0.20287451630735212,
+      "grad_norm": 2.738621950149536,
+      "learning_rate": 4.098919840818647e-05,
+      "loss": 5.7785,
+      "step": 3670
+    },
+    {
+      "epoch": 0.20315091210613598,
+      "grad_norm": 2.758222818374634,
+      "learning_rate": 4.0974985787379194e-05,
+      "loss": 5.4694,
+      "step": 3675
+    },
+    {
+      "epoch": 0.20342730790491984,
+      "grad_norm": 3.3197033405303955,
+      "learning_rate": 4.096077316657192e-05,
+      "loss": 5.6767,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2037037037037037,
+      "grad_norm": 3.1457207202911377,
+      "learning_rate": 4.094656054576464e-05,
+      "loss": 5.8161,
+      "step": 3685
+    },
+    {
+      "epoch": 0.20398009950248755,
+      "grad_norm": 2.59710955619812,
+      "learning_rate": 4.0932347924957366e-05,
+      "loss": 5.3963,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2042564953012714,
+      "grad_norm": 3.7148916721343994,
+      "learning_rate": 4.091813530415008e-05,
+      "loss": 5.7489,
+      "step": 3695
+    },
+    {
+      "epoch": 0.2045328911000553,
+      "grad_norm": 2.7009363174438477,
+      "learning_rate": 4.090392268334281e-05,
+      "loss": 5.619,
+      "step": 3700
+    },
+    {
+      "epoch": 0.20480928689883915,
+      "grad_norm": 2.5814390182495117,
+      "learning_rate": 4.088971006253554e-05,
+      "loss": 5.4951,
+      "step": 3705
+    },
+    {
+      "epoch": 0.205085682697623,
+      "grad_norm": 2.773533344268799,
+      "learning_rate": 4.0875497441728255e-05,
+      "loss": 5.763,
+      "step": 3710
+    },
+    {
+      "epoch": 0.20536207849640686,
+      "grad_norm": 3.0854074954986572,
+      "learning_rate": 4.086128482092098e-05,
+      "loss": 5.3959,
+      "step": 3715
+    },
+    {
+      "epoch": 0.20563847429519072,
+      "grad_norm": 2.8978331089019775,
+      "learning_rate": 4.08470722001137e-05,
+      "loss": 5.7336,
+      "step": 3720
+    },
+    {
+      "epoch": 0.20591487009397458,
+      "grad_norm": 2.4324569702148438,
+      "learning_rate": 4.083285957930643e-05,
+      "loss": 5.832,
+      "step": 3725
+    },
+    {
+      "epoch": 0.20619126589275844,
+      "grad_norm": 2.3226120471954346,
+      "learning_rate": 4.081864695849915e-05,
+      "loss": 5.6816,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2064676616915423,
+      "grad_norm": 2.733042001724243,
+      "learning_rate": 4.080443433769187e-05,
+      "loss": 5.6433,
+      "step": 3735
+    },
+    {
+      "epoch": 0.20674405749032615,
+      "grad_norm": 2.787104845046997,
+      "learning_rate": 4.07902217168846e-05,
+      "loss": 5.7111,
+      "step": 3740
+    },
+    {
+      "epoch": 0.20702045328911,
+      "grad_norm": 2.508944511413574,
+      "learning_rate": 4.0776009096077316e-05,
+      "loss": 5.7261,
+      "step": 3745
+    },
+    {
+      "epoch": 0.20729684908789386,
+      "grad_norm": 3.51177978515625,
+      "learning_rate": 4.076179647527004e-05,
+      "loss": 5.3499,
+      "step": 3750
+    },
+    {
+      "epoch": 0.20757324488667772,
+      "grad_norm": 2.7052156925201416,
+      "learning_rate": 4.074758385446277e-05,
+      "loss": 5.7292,
+      "step": 3755
+    },
+    {
+      "epoch": 0.20784964068546158,
+      "grad_norm": 2.662250280380249,
+      "learning_rate": 4.073337123365549e-05,
+      "loss": 5.4517,
+      "step": 3760
+    },
+    {
+      "epoch": 0.20812603648424544,
+      "grad_norm": 2.365370750427246,
+      "learning_rate": 4.071915861284821e-05,
+      "loss": 5.7968,
+      "step": 3765
+    },
+    {
+      "epoch": 0.2084024322830293,
+      "grad_norm": 3.118715763092041,
+      "learning_rate": 4.0704945992040936e-05,
+      "loss": 5.5743,
+      "step": 3770
+    },
+    {
+      "epoch": 0.20867882808181315,
+      "grad_norm": 3.200096607208252,
+      "learning_rate": 4.069073337123366e-05,
+      "loss": 6.0694,
+      "step": 3775
+    },
+    {
+      "epoch": 0.208955223880597,
+      "grad_norm": 3.215636730194092,
+      "learning_rate": 4.067652075042638e-05,
+      "loss": 5.6844,
+      "step": 3780
+    },
+    {
+      "epoch": 0.20923161967938086,
+      "grad_norm": 3.387874126434326,
+      "learning_rate": 4.06623081296191e-05,
+      "loss": 5.5597,
+      "step": 3785
+    },
+    {
+      "epoch": 0.20950801547816472,
+      "grad_norm": 3.392089605331421,
+      "learning_rate": 4.064809550881183e-05,
+      "loss": 5.8506,
+      "step": 3790
+    },
+    {
+      "epoch": 0.20978441127694858,
+      "grad_norm": 2.8729498386383057,
+      "learning_rate": 4.063388288800455e-05,
+      "loss": 5.4329,
+      "step": 3795
+    },
+    {
+      "epoch": 0.21006080707573244,
+      "grad_norm": 2.7015790939331055,
+      "learning_rate": 4.061967026719727e-05,
+      "loss": 5.6762,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2103372028745163,
+      "grad_norm": 3.2836618423461914,
+      "learning_rate": 4.060545764639e-05,
+      "loss": 5.6025,
+      "step": 3805
+    },
+    {
+      "epoch": 0.21061359867330018,
+      "grad_norm": 2.447291612625122,
+      "learning_rate": 4.059124502558272e-05,
+      "loss": 5.6676,
+      "step": 3810
+    },
+    {
+      "epoch": 0.21088999447208404,
+      "grad_norm": 2.9781346321105957,
+      "learning_rate": 4.0577032404775445e-05,
+      "loss": 5.6131,
+      "step": 3815
+    },
+    {
+      "epoch": 0.2111663902708679,
+      "grad_norm": 3.1656429767608643,
+      "learning_rate": 4.056281978396816e-05,
+      "loss": 5.7342,
+      "step": 3820
+    },
+    {
+      "epoch": 0.21144278606965175,
+      "grad_norm": 2.7425482273101807,
+      "learning_rate": 4.054860716316089e-05,
+      "loss": 5.6741,
+      "step": 3825
+    },
+    {
+      "epoch": 0.2117191818684356,
+      "grad_norm": 3.253185510635376,
+      "learning_rate": 4.053439454235361e-05,
+      "loss": 5.801,
+      "step": 3830
+    },
+    {
+      "epoch": 0.21199557766721946,
+      "grad_norm": 3.0883877277374268,
+      "learning_rate": 4.0520181921546334e-05,
+      "loss": 5.6809,
+      "step": 3835
+    },
+    {
+      "epoch": 0.21227197346600332,
+      "grad_norm": 3.222590446472168,
+      "learning_rate": 4.050596930073906e-05,
+      "loss": 5.4639,
+      "step": 3840
+    },
+    {
+      "epoch": 0.21254836926478718,
+      "grad_norm": 2.510165214538574,
+      "learning_rate": 4.049175667993178e-05,
+      "loss": 5.6433,
+      "step": 3845
+    },
+    {
+      "epoch": 0.21282476506357104,
+      "grad_norm": 3.1436879634857178,
+      "learning_rate": 4.0477544059124506e-05,
+      "loss": 5.3783,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2131011608623549,
+      "grad_norm": 2.9798269271850586,
+      "learning_rate": 4.046333143831722e-05,
+      "loss": 5.6385,
+      "step": 3855
+    },
+    {
+      "epoch": 0.21337755666113875,
+      "grad_norm": 2.9167640209198,
+      "learning_rate": 4.0449118817509954e-05,
+      "loss": 5.6914,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2136539524599226,
+      "grad_norm": 3.467259645462036,
+      "learning_rate": 4.043490619670268e-05,
+      "loss": 5.8466,
+      "step": 3865
+    },
+    {
+      "epoch": 0.21393034825870647,
+      "grad_norm": 2.689114809036255,
+      "learning_rate": 4.0420693575895395e-05,
+      "loss": 5.9034,
+      "step": 3870
+    },
+    {
+      "epoch": 0.21420674405749032,
+      "grad_norm": 2.7007036209106445,
+      "learning_rate": 4.040648095508812e-05,
+      "loss": 5.6707,
+      "step": 3875
+    },
+    {
+      "epoch": 0.21448313985627418,
+      "grad_norm": 2.5844995975494385,
+      "learning_rate": 4.039226833428084e-05,
+      "loss": 5.726,
+      "step": 3880
+    },
+    {
+      "epoch": 0.21475953565505804,
+      "grad_norm": 2.7026453018188477,
+      "learning_rate": 4.037805571347357e-05,
+      "loss": 5.6363,
+      "step": 3885
+    },
+    {
+      "epoch": 0.2150359314538419,
+      "grad_norm": 3.0205512046813965,
+      "learning_rate": 4.0363843092666284e-05,
+      "loss": 5.6149,
+      "step": 3890
+    },
+    {
+      "epoch": 0.21531232725262575,
+      "grad_norm": 2.615713596343994,
+      "learning_rate": 4.0349630471859015e-05,
+      "loss": 5.4971,
+      "step": 3895
+    },
+    {
+      "epoch": 0.2155887230514096,
+      "grad_norm": 3.4436938762664795,
+      "learning_rate": 4.033541785105174e-05,
+      "loss": 5.9492,
+      "step": 3900
+    },
+    {
+      "epoch": 0.21586511885019347,
+      "grad_norm": 2.6737794876098633,
+      "learning_rate": 4.0321205230244456e-05,
+      "loss": 5.5133,
+      "step": 3905
+    },
+    {
+      "epoch": 0.21614151464897732,
+      "grad_norm": 2.6840312480926514,
+      "learning_rate": 4.030699260943718e-05,
+      "loss": 5.9368,
+      "step": 3910
+    },
+    {
+      "epoch": 0.21641791044776118,
+      "grad_norm": 2.652055501937866,
+      "learning_rate": 4.0292779988629904e-05,
+      "loss": 5.7533,
+      "step": 3915
+    },
+    {
+      "epoch": 0.21669430624654507,
+      "grad_norm": 3.5088555812835693,
+      "learning_rate": 4.027856736782263e-05,
+      "loss": 5.6949,
+      "step": 3920
+    },
+    {
+      "epoch": 0.21697070204532892,
+      "grad_norm": 3.445812940597534,
+      "learning_rate": 4.026435474701535e-05,
+      "loss": 5.5864,
+      "step": 3925
+    },
+    {
+      "epoch": 0.21724709784411278,
+      "grad_norm": 3.1001298427581787,
+      "learning_rate": 4.0250142126208076e-05,
+      "loss": 5.8767,
+      "step": 3930
+    },
+    {
+      "epoch": 0.21752349364289664,
+      "grad_norm": 2.6372926235198975,
+      "learning_rate": 4.02359295054008e-05,
+      "loss": 5.6745,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2177998894416805,
+      "grad_norm": 2.3404037952423096,
+      "learning_rate": 4.022171688459352e-05,
+      "loss": 5.6526,
+      "step": 3940
+    },
+    {
+      "epoch": 0.21807628524046435,
+      "grad_norm": 3.181204080581665,
+      "learning_rate": 4.020750426378624e-05,
+      "loss": 5.5995,
+      "step": 3945
+    },
+    {
+      "epoch": 0.2183526810392482,
+      "grad_norm": 3.0466437339782715,
+      "learning_rate": 4.019329164297897e-05,
+      "loss": 5.7707,
+      "step": 3950
+    },
+    {
+      "epoch": 0.21862907683803207,
+      "grad_norm": 2.544200897216797,
+      "learning_rate": 4.017907902217169e-05,
+      "loss": 5.9233,
+      "step": 3955
+    },
+    {
+      "epoch": 0.21890547263681592,
+      "grad_norm": 3.109133720397949,
+      "learning_rate": 4.016486640136441e-05,
+      "loss": 5.7499,
+      "step": 3960
+    },
+    {
+      "epoch": 0.21918186843559978,
+      "grad_norm": 2.555316209793091,
+      "learning_rate": 4.015065378055714e-05,
+      "loss": 5.3721,
+      "step": 3965
+    },
+    {
+      "epoch": 0.21945826423438364,
+      "grad_norm": 2.7728629112243652,
+      "learning_rate": 4.013644115974986e-05,
+      "loss": 5.7565,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2197346600331675,
+      "grad_norm": 3.0431509017944336,
+      "learning_rate": 4.0122228538942585e-05,
+      "loss": 5.5652,
+      "step": 3975
+    },
+    {
+      "epoch": 0.22001105583195135,
+      "grad_norm": 3.0363245010375977,
+      "learning_rate": 4.01080159181353e-05,
+      "loss": 5.6372,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2202874516307352,
+      "grad_norm": 2.699867010116577,
+      "learning_rate": 4.009380329732803e-05,
+      "loss": 5.5497,
+      "step": 3985
+    },
+    {
+      "epoch": 0.22056384742951907,
+      "grad_norm": 2.5388708114624023,
+      "learning_rate": 4.007959067652075e-05,
+      "loss": 5.4853,
+      "step": 3990
+    },
+    {
+      "epoch": 0.22084024322830292,
+      "grad_norm": 2.6334452629089355,
+      "learning_rate": 4.0065378055713474e-05,
+      "loss": 5.5215,
+      "step": 3995
+    },
+    {
+      "epoch": 0.22111663902708678,
+      "grad_norm": 2.968801498413086,
+      "learning_rate": 4.00511654349062e-05,
+      "loss": 5.5953,
+      "step": 4000
+    },
+    {
+      "epoch": 0.22139303482587064,
+      "grad_norm": 3.2419137954711914,
+      "learning_rate": 4.003695281409892e-05,
+      "loss": 5.8652,
+      "step": 4005
+    },
+    {
+      "epoch": 0.2216694306246545,
+      "grad_norm": 3.5117664337158203,
+      "learning_rate": 4.0022740193291646e-05,
+      "loss": 5.1776,
+      "step": 4010
+    },
+    {
+      "epoch": 0.22194582642343835,
+      "grad_norm": 3.309300661087036,
+      "learning_rate": 4.0008527572484363e-05,
+      "loss": 5.9688,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 3.4260222911834717,
+      "learning_rate": 3.9994314951677094e-05,
+      "loss": 5.6959,
+      "step": 4020
+    },
+    {
+      "epoch": 0.22249861802100607,
+      "grad_norm": 2.736858606338501,
+      "learning_rate": 3.998010233086981e-05,
+      "loss": 5.69,
+      "step": 4025
+    },
+    {
+      "epoch": 0.22277501381978995,
+      "grad_norm": 3.037052869796753,
+      "learning_rate": 3.9965889710062535e-05,
+      "loss": 5.3564,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2230514096185738,
+      "grad_norm": 2.519865036010742,
+      "learning_rate": 3.995167708925526e-05,
+      "loss": 5.6226,
+      "step": 4035
+    },
+    {
+      "epoch": 0.22332780541735767,
+      "grad_norm": 2.619724988937378,
+      "learning_rate": 3.993746446844798e-05,
+      "loss": 5.8328,
+      "step": 4040
+    },
+    {
+      "epoch": 0.22360420121614152,
+      "grad_norm": 2.6823482513427734,
+      "learning_rate": 3.992325184764071e-05,
+      "loss": 5.2537,
+      "step": 4045
+    },
+    {
+      "epoch": 0.22388059701492538,
+      "grad_norm": 2.5863189697265625,
+      "learning_rate": 3.9909039226833425e-05,
+      "loss": 5.139,
+      "step": 4050
+    },
+    {
+      "epoch": 0.22415699281370924,
+      "grad_norm": 2.9364101886749268,
+      "learning_rate": 3.9894826606026155e-05,
+      "loss": 5.544,
+      "step": 4055
+    },
+    {
+      "epoch": 0.2244333886124931,
+      "grad_norm": 2.504804849624634,
+      "learning_rate": 3.988061398521888e-05,
+      "loss": 5.7221,
+      "step": 4060
+    },
+    {
+      "epoch": 0.22470978441127695,
+      "grad_norm": 2.4577889442443848,
+      "learning_rate": 3.9866401364411596e-05,
+      "loss": 5.6893,
+      "step": 4065
+    },
+    {
+      "epoch": 0.2249861802100608,
+      "grad_norm": 2.8523504734039307,
+      "learning_rate": 3.985218874360433e-05,
+      "loss": 5.4065,
+      "step": 4070
+    },
+    {
+      "epoch": 0.22526257600884467,
+      "grad_norm": 2.8694982528686523,
+      "learning_rate": 3.9837976122797044e-05,
+      "loss": 5.7412,
+      "step": 4075
+    },
+    {
+      "epoch": 0.22553897180762852,
+      "grad_norm": 2.2998769283294678,
+      "learning_rate": 3.982376350198977e-05,
+      "loss": 5.3882,
+      "step": 4080
+    },
+    {
+      "epoch": 0.22581536760641238,
+      "grad_norm": 2.82647442817688,
+      "learning_rate": 3.980955088118249e-05,
+      "loss": 5.5734,
+      "step": 4085
+    },
+    {
+      "epoch": 0.22609176340519624,
+      "grad_norm": 2.7505252361297607,
+      "learning_rate": 3.9795338260375216e-05,
+      "loss": 5.7028,
+      "step": 4090
+    },
+    {
+      "epoch": 0.2263681592039801,
+      "grad_norm": 3.391613483428955,
+      "learning_rate": 3.978112563956794e-05,
+      "loss": 5.731,
+      "step": 4095
+    },
+    {
+      "epoch": 0.22664455500276395,
+      "grad_norm": 3.5600905418395996,
+      "learning_rate": 3.976691301876066e-05,
+      "loss": 5.5761,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2269209508015478,
+      "grad_norm": 3.087388515472412,
+      "learning_rate": 3.975270039795339e-05,
+      "loss": 6.0301,
+      "step": 4105
+    },
+    {
+      "epoch": 0.22719734660033167,
+      "grad_norm": 2.921119213104248,
+      "learning_rate": 3.9738487777146106e-05,
+      "loss": 5.5959,
+      "step": 4110
+    },
+    {
+      "epoch": 0.22747374239911552,
+      "grad_norm": 3.0585193634033203,
+      "learning_rate": 3.972427515633883e-05,
+      "loss": 5.61,
+      "step": 4115
+    },
+    {
+      "epoch": 0.22775013819789938,
+      "grad_norm": 3.3914947509765625,
+      "learning_rate": 3.9710062535531553e-05,
+      "loss": 5.3884,
+      "step": 4120
+    },
+    {
+      "epoch": 0.22802653399668324,
+      "grad_norm": 3.621385097503662,
+      "learning_rate": 3.969584991472428e-05,
+      "loss": 5.7627,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2283029297954671,
+      "grad_norm": 2.755340099334717,
+      "learning_rate": 3.9681637293917e-05,
+      "loss": 5.6035,
+      "step": 4130
+    },
+    {
+      "epoch": 0.22857932559425095,
+      "grad_norm": 2.5260558128356934,
+      "learning_rate": 3.966742467310972e-05,
+      "loss": 5.795,
+      "step": 4135
+    },
+    {
+      "epoch": 0.22885572139303484,
+      "grad_norm": 2.8096625804901123,
+      "learning_rate": 3.965321205230245e-05,
+      "loss": 5.4867,
+      "step": 4140
+    },
+    {
+      "epoch": 0.2291321171918187,
+      "grad_norm": 2.795567274093628,
+      "learning_rate": 3.963899943149517e-05,
+      "loss": 5.6046,
+      "step": 4145
+    },
+    {
+      "epoch": 0.22940851299060255,
+      "grad_norm": 2.43609881401062,
+      "learning_rate": 3.962478681068789e-05,
+      "loss": 5.4769,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2296849087893864,
+      "grad_norm": 3.2818362712860107,
+      "learning_rate": 3.9610574189880615e-05,
+      "loss": 5.4739,
+      "step": 4155
+    },
+    {
+      "epoch": 0.22996130458817027,
+      "grad_norm": 3.0869808197021484,
+      "learning_rate": 3.959636156907334e-05,
+      "loss": 5.6502,
+      "step": 4160
+    },
+    {
+      "epoch": 0.23023770038695412,
+      "grad_norm": 2.792592763900757,
+      "learning_rate": 3.958214894826606e-05,
+      "loss": 5.707,
+      "step": 4165
+    },
+    {
+      "epoch": 0.23051409618573798,
+      "grad_norm": 3.079249858856201,
+      "learning_rate": 3.9567936327458786e-05,
+      "loss": 5.8627,
+      "step": 4170
+    },
+    {
+      "epoch": 0.23079049198452184,
+      "grad_norm": 2.5989654064178467,
+      "learning_rate": 3.955372370665151e-05,
+      "loss": 5.6885,
+      "step": 4175
+    },
+    {
+      "epoch": 0.2310668877833057,
+      "grad_norm": 2.6450164318084717,
+      "learning_rate": 3.9539511085844234e-05,
+      "loss": 5.4694,
+      "step": 4180
+    },
+    {
+      "epoch": 0.23134328358208955,
+      "grad_norm": 2.6688590049743652,
+      "learning_rate": 3.952529846503695e-05,
+      "loss": 5.5172,
+      "step": 4185
+    },
+    {
+      "epoch": 0.2316196793808734,
+      "grad_norm": 2.876706600189209,
+      "learning_rate": 3.9511085844229676e-05,
+      "loss": 5.6633,
+      "step": 4190
+    },
+    {
+      "epoch": 0.23189607517965727,
+      "grad_norm": 2.682499408721924,
+      "learning_rate": 3.9496873223422406e-05,
+      "loss": 5.6945,
+      "step": 4195
+    },
+    {
+      "epoch": 0.23217247097844113,
+      "grad_norm": 2.5121068954467773,
+      "learning_rate": 3.9482660602615124e-05,
+      "loss": 5.6058,
+      "step": 4200
+    },
+    {
+      "epoch": 0.23244886677722498,
+      "grad_norm": 3.3784987926483154,
+      "learning_rate": 3.946844798180785e-05,
+      "loss": 5.4177,
+      "step": 4205
+    },
+    {
+      "epoch": 0.23272526257600884,
+      "grad_norm": 2.7231504917144775,
+      "learning_rate": 3.945423536100057e-05,
+      "loss": 5.3509,
+      "step": 4210
+    },
+    {
+      "epoch": 0.2330016583747927,
+      "grad_norm": 3.0387542247772217,
+      "learning_rate": 3.9440022740193296e-05,
+      "loss": 5.7707,
+      "step": 4215
+    },
+    {
+      "epoch": 0.23327805417357655,
+      "grad_norm": 2.9296815395355225,
+      "learning_rate": 3.942581011938601e-05,
+      "loss": 5.7334,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2335544499723604,
+      "grad_norm": 2.396350622177124,
+      "learning_rate": 3.941159749857874e-05,
+      "loss": 5.254,
+      "step": 4225
+    },
+    {
+      "epoch": 0.23383084577114427,
+      "grad_norm": 3.36370587348938,
+      "learning_rate": 3.939738487777147e-05,
+      "loss": 5.8167,
+      "step": 4230
+    },
+    {
+      "epoch": 0.23410724156992813,
+      "grad_norm": 3.5464701652526855,
+      "learning_rate": 3.9383172256964185e-05,
+      "loss": 5.742,
+      "step": 4235
+    },
+    {
+      "epoch": 0.23438363736871198,
+      "grad_norm": 2.342013120651245,
+      "learning_rate": 3.936895963615691e-05,
+      "loss": 5.5515,
+      "step": 4240
+    },
+    {
+      "epoch": 0.23466003316749584,
+      "grad_norm": 2.8633322715759277,
+      "learning_rate": 3.935474701534963e-05,
+      "loss": 5.4822,
+      "step": 4245
+    },
+    {
+      "epoch": 0.23493642896627973,
+      "grad_norm": 2.7258920669555664,
+      "learning_rate": 3.9340534394542357e-05,
+      "loss": 5.2667,
+      "step": 4250
+    },
+    {
+      "epoch": 0.23521282476506358,
+      "grad_norm": 2.6543519496917725,
+      "learning_rate": 3.932632177373508e-05,
+      "loss": 5.5115,
+      "step": 4255
+    },
+    {
+      "epoch": 0.23548922056384744,
+      "grad_norm": 3.5072028636932373,
+      "learning_rate": 3.93121091529278e-05,
+      "loss": 5.4936,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2357656163626313,
+      "grad_norm": 2.486929178237915,
+      "learning_rate": 3.929789653212053e-05,
+      "loss": 5.719,
+      "step": 4265
+    },
+    {
+      "epoch": 0.23604201216141515,
+      "grad_norm": 3.5531532764434814,
+      "learning_rate": 3.9283683911313246e-05,
+      "loss": 5.6424,
+      "step": 4270
+    },
+    {
+      "epoch": 0.236318407960199,
+      "grad_norm": 3.207571506500244,
+      "learning_rate": 3.926947129050597e-05,
+      "loss": 5.6972,
+      "step": 4275
+    },
+    {
+      "epoch": 0.23659480375898287,
+      "grad_norm": 3.538468599319458,
+      "learning_rate": 3.9255258669698694e-05,
+      "loss": 5.8425,
+      "step": 4280
+    },
+    {
+      "epoch": 0.23687119955776673,
+      "grad_norm": 3.163320541381836,
+      "learning_rate": 3.924104604889142e-05,
+      "loss": 5.2799,
+      "step": 4285
+    },
+    {
+      "epoch": 0.23714759535655058,
+      "grad_norm": 4.0374274253845215,
+      "learning_rate": 3.922683342808414e-05,
+      "loss": 5.5183,
+      "step": 4290
+    },
+    {
+      "epoch": 0.23742399115533444,
+      "grad_norm": 3.0995869636535645,
+      "learning_rate": 3.921262080727686e-05,
+      "loss": 5.7765,
+      "step": 4295
+    },
+    {
+      "epoch": 0.2377003869541183,
+      "grad_norm": 2.8097715377807617,
+      "learning_rate": 3.919840818646959e-05,
+      "loss": 5.7519,
+      "step": 4300
+    },
+    {
+      "epoch": 0.23797678275290216,
+      "grad_norm": 2.7227861881256104,
+      "learning_rate": 3.9184195565662314e-05,
+      "loss": 5.3365,
+      "step": 4305
+    },
+    {
+      "epoch": 0.238253178551686,
+      "grad_norm": 2.6679046154022217,
+      "learning_rate": 3.916998294485503e-05,
+      "loss": 5.4229,
+      "step": 4310
+    },
+    {
+      "epoch": 0.23852957435046987,
+      "grad_norm": 2.4163801670074463,
+      "learning_rate": 3.9155770324047755e-05,
+      "loss": 5.1811,
+      "step": 4315
+    },
+    {
+      "epoch": 0.23880597014925373,
+      "grad_norm": 2.996711254119873,
+      "learning_rate": 3.914155770324048e-05,
+      "loss": 5.8349,
+      "step": 4320
+    },
+    {
+      "epoch": 0.23908236594803758,
+      "grad_norm": 2.91500186920166,
+      "learning_rate": 3.91273450824332e-05,
+      "loss": 5.3502,
+      "step": 4325
+    },
+    {
+      "epoch": 0.23935876174682144,
+      "grad_norm": 2.5672600269317627,
+      "learning_rate": 3.911313246162592e-05,
+      "loss": 5.4671,
+      "step": 4330
+    },
+    {
+      "epoch": 0.2396351575456053,
+      "grad_norm": 2.848611831665039,
+      "learning_rate": 3.909891984081865e-05,
+      "loss": 5.6904,
+      "step": 4335
+    },
+    {
+      "epoch": 0.23991155334438916,
+      "grad_norm": 3.451052665710449,
+      "learning_rate": 3.9084707220011375e-05,
+      "loss": 5.4361,
+      "step": 4340
+    },
+    {
+      "epoch": 0.240187949143173,
+      "grad_norm": 2.8216664791107178,
+      "learning_rate": 3.907049459920409e-05,
+      "loss": 5.6986,
+      "step": 4345
+    },
+    {
+      "epoch": 0.24046434494195687,
+      "grad_norm": 2.856027841567993,
+      "learning_rate": 3.9056281978396816e-05,
+      "loss": 5.9277,
+      "step": 4350
+    },
+    {
+      "epoch": 0.24074074074074073,
+      "grad_norm": 2.87872576713562,
+      "learning_rate": 3.904206935758954e-05,
+      "loss": 5.6129,
+      "step": 4355
+    },
+    {
+      "epoch": 0.2410171365395246,
+      "grad_norm": 2.844909906387329,
+      "learning_rate": 3.9027856736782264e-05,
+      "loss": 5.5358,
+      "step": 4360
+    },
+    {
+      "epoch": 0.24129353233830847,
+      "grad_norm": 3.371220111846924,
+      "learning_rate": 3.901364411597499e-05,
+      "loss": 5.2788,
+      "step": 4365
+    },
+    {
+      "epoch": 0.24156992813709233,
+      "grad_norm": 2.4404332637786865,
+      "learning_rate": 3.899943149516771e-05,
+      "loss": 5.4483,
+      "step": 4370
+    },
+    {
+      "epoch": 0.24184632393587618,
+      "grad_norm": 2.876338481903076,
+      "learning_rate": 3.8985218874360436e-05,
+      "loss": 5.4887,
+      "step": 4375
+    },
+    {
+      "epoch": 0.24212271973466004,
+      "grad_norm": 3.4181478023529053,
+      "learning_rate": 3.897100625355315e-05,
+      "loss": 5.6298,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2423991155334439,
+      "grad_norm": 2.8636157512664795,
+      "learning_rate": 3.8956793632745884e-05,
+      "loss": 5.7623,
+      "step": 4385
+    },
+    {
+      "epoch": 0.24267551133222776,
+      "grad_norm": 2.881701707839966,
+      "learning_rate": 3.894258101193861e-05,
+      "loss": 5.5819,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2429519071310116,
+      "grad_norm": 2.676478385925293,
+      "learning_rate": 3.8928368391131325e-05,
+      "loss": 5.5775,
+      "step": 4395
+    },
+    {
+      "epoch": 0.24322830292979547,
+      "grad_norm": 3.2318267822265625,
+      "learning_rate": 3.891415577032405e-05,
+      "loss": 5.416,
+      "step": 4400
+    },
+    {
+      "epoch": 0.24350469872857933,
+      "grad_norm": 2.6567542552948,
+      "learning_rate": 3.889994314951677e-05,
+      "loss": 5.6167,
+      "step": 4405
+    },
+    {
+      "epoch": 0.24378109452736318,
+      "grad_norm": 2.6973955631256104,
+      "learning_rate": 3.88857305287095e-05,
+      "loss": 5.5814,
+      "step": 4410
+    },
+    {
+      "epoch": 0.24405749032614704,
+      "grad_norm": 2.4810473918914795,
+      "learning_rate": 3.8871517907902214e-05,
+      "loss": 5.61,
+      "step": 4415
+    },
+    {
+      "epoch": 0.2443338861249309,
+      "grad_norm": 2.4649670124053955,
+      "learning_rate": 3.8857305287094945e-05,
+      "loss": 5.3495,
+      "step": 4420
+    },
+    {
+      "epoch": 0.24461028192371476,
+      "grad_norm": 2.682745933532715,
+      "learning_rate": 3.884309266628767e-05,
+      "loss": 5.498,
+      "step": 4425
+    },
+    {
+      "epoch": 0.2448866777224986,
+      "grad_norm": 3.384434700012207,
+      "learning_rate": 3.8828880045480386e-05,
+      "loss": 5.4192,
+      "step": 4430
+    },
+    {
+      "epoch": 0.24516307352128247,
+      "grad_norm": 3.0535390377044678,
+      "learning_rate": 3.881466742467311e-05,
+      "loss": 5.4866,
+      "step": 4435
+    },
+    {
+      "epoch": 0.24543946932006633,
+      "grad_norm": 2.8525378704071045,
+      "learning_rate": 3.8800454803865834e-05,
+      "loss": 5.5854,
+      "step": 4440
+    },
+    {
+      "epoch": 0.24571586511885019,
+      "grad_norm": 2.7716307640075684,
+      "learning_rate": 3.878624218305856e-05,
+      "loss": 5.6654,
+      "step": 4445
+    },
+    {
+      "epoch": 0.24599226091763404,
+      "grad_norm": 2.2745914459228516,
+      "learning_rate": 3.877202956225128e-05,
+      "loss": 5.5521,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2462686567164179,
+      "grad_norm": 3.1492974758148193,
+      "learning_rate": 3.8757816941444006e-05,
+      "loss": 5.5735,
+      "step": 4455
+    },
+    {
+      "epoch": 0.24654505251520176,
+      "grad_norm": 3.3391618728637695,
+      "learning_rate": 3.874360432063673e-05,
+      "loss": 5.3698,
+      "step": 4460
+    },
+    {
+      "epoch": 0.24682144831398561,
+      "grad_norm": 2.884638547897339,
+      "learning_rate": 3.872939169982945e-05,
+      "loss": 5.7967,
+      "step": 4465
+    },
+    {
+      "epoch": 0.2470978441127695,
+      "grad_norm": 2.4962894916534424,
+      "learning_rate": 3.871517907902217e-05,
+      "loss": 5.2797,
+      "step": 4470
+    },
+    {
+      "epoch": 0.24737423991155336,
+      "grad_norm": 3.132397413253784,
+      "learning_rate": 3.87009664582149e-05,
+      "loss": 5.6922,
+      "step": 4475
+    },
+    {
+      "epoch": 0.2476506357103372,
+      "grad_norm": 2.939412832260132,
+      "learning_rate": 3.868675383740762e-05,
+      "loss": 5.7016,
+      "step": 4480
+    },
+    {
+      "epoch": 0.24792703150912107,
+      "grad_norm": 2.6651360988616943,
+      "learning_rate": 3.867254121660034e-05,
+      "loss": 5.4948,
+      "step": 4485
+    },
+    {
+      "epoch": 0.24820342730790493,
+      "grad_norm": 3.104663372039795,
+      "learning_rate": 3.865832859579307e-05,
+      "loss": 5.6978,
+      "step": 4490
+    },
+    {
+      "epoch": 0.24847982310668879,
+      "grad_norm": 3.229151725769043,
+      "learning_rate": 3.864411597498579e-05,
+      "loss": 5.7317,
+      "step": 4495
+    },
+    {
+      "epoch": 0.24875621890547264,
+      "grad_norm": 2.4405629634857178,
+      "learning_rate": 3.8629903354178515e-05,
+      "loss": 5.7138,
+      "step": 4500
+    },
+    {
+      "epoch": 0.2490326147042565,
+      "grad_norm": 2.657789707183838,
+      "learning_rate": 3.861569073337123e-05,
+      "loss": 5.674,
+      "step": 4505
+    },
+    {
+      "epoch": 0.24930901050304036,
+      "grad_norm": 2.536699056625366,
+      "learning_rate": 3.860147811256396e-05,
+      "loss": 5.6521,
+      "step": 4510
+    },
+    {
+      "epoch": 0.24958540630182421,
+      "grad_norm": 2.947270631790161,
+      "learning_rate": 3.858726549175668e-05,
+      "loss": 5.6656,
+      "step": 4515
+    },
+    {
+      "epoch": 0.24986180210060807,
+      "grad_norm": 3.1712615489959717,
+      "learning_rate": 3.8573052870949404e-05,
+      "loss": 5.4181,
+      "step": 4520
+    },
+    {
+      "epoch": 0.25013819789939196,
+      "grad_norm": 2.742157459259033,
+      "learning_rate": 3.855884025014213e-05,
+      "loss": 5.5457,
+      "step": 4525
+    },
+    {
+      "epoch": 0.2504145936981758,
+      "grad_norm": 2.4692294597625732,
+      "learning_rate": 3.854462762933485e-05,
+      "loss": 5.5266,
+      "step": 4530
+    },
+    {
+      "epoch": 0.25069098949695967,
+      "grad_norm": 2.7221648693084717,
+      "learning_rate": 3.8530415008527576e-05,
+      "loss": 5.6376,
+      "step": 4535
+    },
+    {
+      "epoch": 0.25096738529574353,
+      "grad_norm": 2.809128999710083,
+      "learning_rate": 3.851620238772029e-05,
+      "loss": 5.4378,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2512437810945274,
+      "grad_norm": 2.9492316246032715,
+      "learning_rate": 3.8501989766913024e-05,
+      "loss": 5.7847,
+      "step": 4545
+    },
+    {
+      "epoch": 0.25152017689331124,
+      "grad_norm": 3.0182037353515625,
+      "learning_rate": 3.848777714610574e-05,
+      "loss": 5.6268,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2517965726920951,
+      "grad_norm": 3.395892381668091,
+      "learning_rate": 3.8473564525298465e-05,
+      "loss": 5.5913,
+      "step": 4555
+    },
+    {
+      "epoch": 0.25207296849087896,
+      "grad_norm": 2.6461169719696045,
+      "learning_rate": 3.845935190449119e-05,
+      "loss": 5.655,
+      "step": 4560
+    },
+    {
+      "epoch": 0.2523493642896628,
+      "grad_norm": 3.7275054454803467,
+      "learning_rate": 3.844513928368391e-05,
+      "loss": 5.5196,
+      "step": 4565
+    },
+    {
+      "epoch": 0.25262576008844667,
+      "grad_norm": 3.1424355506896973,
+      "learning_rate": 3.843092666287664e-05,
+      "loss": 5.3785,
+      "step": 4570
+    },
+    {
+      "epoch": 0.25290215588723053,
+      "grad_norm": 2.984804153442383,
+      "learning_rate": 3.8416714042069354e-05,
+      "loss": 5.4338,
+      "step": 4575
+    },
+    {
+      "epoch": 0.2531785516860144,
+      "grad_norm": 3.411898136138916,
+      "learning_rate": 3.8402501421262085e-05,
+      "loss": 5.7189,
+      "step": 4580
+    },
+    {
+      "epoch": 0.25345494748479824,
+      "grad_norm": 3.0697154998779297,
+      "learning_rate": 3.838828880045481e-05,
+      "loss": 5.554,
+      "step": 4585
+    },
+    {
+      "epoch": 0.2537313432835821,
+      "grad_norm": 2.617682695388794,
+      "learning_rate": 3.8374076179647526e-05,
+      "loss": 5.4068,
+      "step": 4590
+    },
+    {
+      "epoch": 0.25400773908236596,
+      "grad_norm": 2.9888644218444824,
+      "learning_rate": 3.835986355884025e-05,
+      "loss": 5.9172,
+      "step": 4595
+    },
+    {
+      "epoch": 0.2542841348811498,
+      "grad_norm": 2.7901175022125244,
+      "learning_rate": 3.8345650938032974e-05,
+      "loss": 5.3799,
+      "step": 4600
+    },
+    {
+      "epoch": 0.25456053067993367,
+      "grad_norm": 2.9775073528289795,
+      "learning_rate": 3.83314383172257e-05,
+      "loss": 5.9651,
+      "step": 4605
+    },
+    {
+      "epoch": 0.25483692647871753,
+      "grad_norm": 3.188081979751587,
+      "learning_rate": 3.831722569641842e-05,
+      "loss": 5.5412,
+      "step": 4610
+    },
+    {
+      "epoch": 0.2551133222775014,
+      "grad_norm": 3.1348631381988525,
+      "learning_rate": 3.8303013075611146e-05,
+      "loss": 5.4277,
+      "step": 4615
+    },
+    {
+      "epoch": 0.25538971807628524,
+      "grad_norm": 2.817936658859253,
+      "learning_rate": 3.828880045480387e-05,
+      "loss": 5.6873,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2556661138750691,
+      "grad_norm": 2.883748769760132,
+      "learning_rate": 3.827458783399659e-05,
+      "loss": 5.4839,
+      "step": 4625
+    },
+    {
+      "epoch": 0.25594250967385296,
+      "grad_norm": 2.654115915298462,
+      "learning_rate": 3.826037521318931e-05,
+      "loss": 5.7442,
+      "step": 4630
+    },
+    {
+      "epoch": 0.2562189054726368,
+      "grad_norm": 2.976658821105957,
+      "learning_rate": 3.8246162592382035e-05,
+      "loss": 5.374,
+      "step": 4635
+    },
+    {
+      "epoch": 0.2564953012714207,
+      "grad_norm": 2.345407009124756,
+      "learning_rate": 3.823194997157476e-05,
+      "loss": 5.551,
+      "step": 4640
+    },
+    {
+      "epoch": 0.25677169707020453,
+      "grad_norm": 3.0773231983184814,
+      "learning_rate": 3.821773735076748e-05,
+      "loss": 5.7161,
+      "step": 4645
+    },
+    {
+      "epoch": 0.2570480928689884,
+      "grad_norm": 2.6813673973083496,
+      "learning_rate": 3.820352472996021e-05,
+      "loss": 5.2693,
+      "step": 4650
+    },
+    {
+      "epoch": 0.25732448866777224,
+      "grad_norm": 2.9697906970977783,
+      "learning_rate": 3.818931210915293e-05,
+      "loss": 5.2863,
+      "step": 4655
+    },
+    {
+      "epoch": 0.2576008844665561,
+      "grad_norm": 2.9038257598876953,
+      "learning_rate": 3.817509948834565e-05,
+      "loss": 5.4955,
+      "step": 4660
+    },
+    {
+      "epoch": 0.25787728026533996,
+      "grad_norm": 3.2474894523620605,
+      "learning_rate": 3.816088686753837e-05,
+      "loss": 5.316,
+      "step": 4665
+    },
+    {
+      "epoch": 0.2581536760641238,
+      "grad_norm": 3.7227532863616943,
+      "learning_rate": 3.81466742467311e-05,
+      "loss": 5.6569,
+      "step": 4670
+    },
+    {
+      "epoch": 0.2584300718629077,
+      "grad_norm": 3.429311513900757,
+      "learning_rate": 3.813246162592382e-05,
+      "loss": 5.6594,
+      "step": 4675
+    },
+    {
+      "epoch": 0.25870646766169153,
+      "grad_norm": 2.9430253505706787,
+      "learning_rate": 3.8118249005116544e-05,
+      "loss": 5.3997,
+      "step": 4680
+    },
+    {
+      "epoch": 0.2589828634604754,
+      "grad_norm": 3.2652957439422607,
+      "learning_rate": 3.810403638430927e-05,
+      "loss": 5.7445,
+      "step": 4685
+    },
+    {
+      "epoch": 0.25925925925925924,
+      "grad_norm": 3.255427360534668,
+      "learning_rate": 3.808982376350199e-05,
+      "loss": 5.3571,
+      "step": 4690
+    },
+    {
+      "epoch": 0.2595356550580431,
+      "grad_norm": 2.7767574787139893,
+      "learning_rate": 3.8075611142694716e-05,
+      "loss": 5.5315,
+      "step": 4695
+    },
+    {
+      "epoch": 0.25981205085682696,
+      "grad_norm": 2.555504560470581,
+      "learning_rate": 3.806139852188744e-05,
+      "loss": 5.4223,
+      "step": 4700
+    },
+    {
+      "epoch": 0.2600884466556108,
+      "grad_norm": 2.6958086490631104,
+      "learning_rate": 3.8047185901080164e-05,
+      "loss": 5.5566,
+      "step": 4705
+    },
+    {
+      "epoch": 0.2603648424543947,
+      "grad_norm": 3.072174310684204,
+      "learning_rate": 3.803297328027288e-05,
+      "loss": 5.9995,
+      "step": 4710
+    },
+    {
+      "epoch": 0.26064123825317853,
+      "grad_norm": 3.545971632003784,
+      "learning_rate": 3.8018760659465605e-05,
+      "loss": 5.5504,
+      "step": 4715
+    },
+    {
+      "epoch": 0.2609176340519624,
+      "grad_norm": 2.8304150104522705,
+      "learning_rate": 3.8004548038658336e-05,
+      "loss": 5.6472,
+      "step": 4720
+    },
+    {
+      "epoch": 0.26119402985074625,
+      "grad_norm": 3.0823028087615967,
+      "learning_rate": 3.799033541785105e-05,
+      "loss": 5.3301,
+      "step": 4725
+    },
+    {
+      "epoch": 0.2614704256495301,
+      "grad_norm": 3.2477657794952393,
+      "learning_rate": 3.797612279704378e-05,
+      "loss": 5.4662,
+      "step": 4730
+    },
+    {
+      "epoch": 0.26174682144831396,
+      "grad_norm": 2.6965558528900146,
+      "learning_rate": 3.79619101762365e-05,
+      "loss": 5.2598,
+      "step": 4735
+    },
+    {
+      "epoch": 0.2620232172470978,
+      "grad_norm": 3.41056752204895,
+      "learning_rate": 3.7947697555429225e-05,
+      "loss": 5.4817,
+      "step": 4740
+    },
+    {
+      "epoch": 0.26229961304588173,
+      "grad_norm": 3.00512433052063,
+      "learning_rate": 3.793348493462194e-05,
+      "loss": 5.3896,
+      "step": 4745
+    },
+    {
+      "epoch": 0.2625760088446656,
+      "grad_norm": 3.316678047180176,
+      "learning_rate": 3.7919272313814666e-05,
+      "loss": 6.0564,
+      "step": 4750
+    },
+    {
+      "epoch": 0.26285240464344944,
+      "grad_norm": 2.8839213848114014,
+      "learning_rate": 3.79050596930074e-05,
+      "loss": 5.715,
+      "step": 4755
+    },
+    {
+      "epoch": 0.2631288004422333,
+      "grad_norm": 3.264241933822632,
+      "learning_rate": 3.7890847072200114e-05,
+      "loss": 5.5688,
+      "step": 4760
+    },
+    {
+      "epoch": 0.26340519624101716,
+      "grad_norm": 2.51202130317688,
+      "learning_rate": 3.787663445139284e-05,
+      "loss": 5.4865,
+      "step": 4765
+    },
+    {
+      "epoch": 0.263681592039801,
+      "grad_norm": 2.940673828125,
+      "learning_rate": 3.786242183058556e-05,
+      "loss": 5.6147,
+      "step": 4770
+    },
+    {
+      "epoch": 0.2639579878385849,
+      "grad_norm": 2.740130662918091,
+      "learning_rate": 3.7848209209778286e-05,
+      "loss": 5.7488,
+      "step": 4775
+    },
+    {
+      "epoch": 0.26423438363736873,
+      "grad_norm": 2.5281081199645996,
+      "learning_rate": 3.783399658897101e-05,
+      "loss": 5.7751,
+      "step": 4780
+    },
+    {
+      "epoch": 0.2645107794361526,
+      "grad_norm": 2.7616381645202637,
+      "learning_rate": 3.781978396816373e-05,
+      "loss": 5.5114,
+      "step": 4785
+    },
+    {
+      "epoch": 0.26478717523493644,
+      "grad_norm": 3.112661838531494,
+      "learning_rate": 3.780557134735646e-05,
+      "loss": 5.7249,
+      "step": 4790
+    },
+    {
+      "epoch": 0.2650635710337203,
+      "grad_norm": 2.2833783626556396,
+      "learning_rate": 3.7791358726549175e-05,
+      "loss": 5.0357,
+      "step": 4795
+    },
+    {
+      "epoch": 0.26533996683250416,
+      "grad_norm": 2.7864673137664795,
+      "learning_rate": 3.77771461057419e-05,
+      "loss": 5.68,
+      "step": 4800
+    },
+    {
+      "epoch": 0.265616362631288,
+      "grad_norm": 2.8897244930267334,
+      "learning_rate": 3.7762933484934623e-05,
+      "loss": 5.5812,
+      "step": 4805
+    },
+    {
+      "epoch": 0.2658927584300719,
+      "grad_norm": 3.157487154006958,
+      "learning_rate": 3.774872086412735e-05,
+      "loss": 5.5694,
+      "step": 4810
+    },
+    {
+      "epoch": 0.26616915422885573,
+      "grad_norm": 3.268732786178589,
+      "learning_rate": 3.773450824332007e-05,
+      "loss": 5.3811,
+      "step": 4815
+    },
+    {
+      "epoch": 0.2664455500276396,
+      "grad_norm": 2.7463014125823975,
+      "learning_rate": 3.772029562251279e-05,
+      "loss": 5.4335,
+      "step": 4820
+    },
+    {
+      "epoch": 0.26672194582642345,
+      "grad_norm": 3.34234619140625,
+      "learning_rate": 3.770608300170552e-05,
+      "loss": 5.401,
+      "step": 4825
+    },
+    {
+      "epoch": 0.2669983416252073,
+      "grad_norm": 4.338949680328369,
+      "learning_rate": 3.769187038089824e-05,
+      "loss": 5.5645,
+      "step": 4830
+    },
+    {
+      "epoch": 0.26727473742399116,
+      "grad_norm": 2.9800288677215576,
+      "learning_rate": 3.767765776009096e-05,
+      "loss": 5.1976,
+      "step": 4835
+    },
+    {
+      "epoch": 0.267551133222775,
+      "grad_norm": 3.4728004932403564,
+      "learning_rate": 3.7663445139283685e-05,
+      "loss": 5.6093,
+      "step": 4840
+    },
+    {
+      "epoch": 0.2678275290215589,
+      "grad_norm": 3.496291399002075,
+      "learning_rate": 3.764923251847641e-05,
+      "loss": 5.5747,
+      "step": 4845
+    },
+    {
+      "epoch": 0.26810392482034273,
+      "grad_norm": 2.850144386291504,
+      "learning_rate": 3.763501989766913e-05,
+      "loss": 5.1295,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2683803206191266,
+      "grad_norm": 2.8431968688964844,
+      "learning_rate": 3.762080727686185e-05,
+      "loss": 5.5363,
+      "step": 4855
+    },
+    {
+      "epoch": 0.26865671641791045,
+      "grad_norm": 2.7617273330688477,
+      "learning_rate": 3.760659465605458e-05,
+      "loss": 5.7338,
+      "step": 4860
+    },
+    {
+      "epoch": 0.2689331122166943,
+      "grad_norm": 2.990050792694092,
+      "learning_rate": 3.7592382035247304e-05,
+      "loss": 5.2071,
+      "step": 4865
+    },
+    {
+      "epoch": 0.26920950801547816,
+      "grad_norm": 3.1524643898010254,
+      "learning_rate": 3.757816941444002e-05,
+      "loss": 5.6822,
+      "step": 4870
+    },
+    {
+      "epoch": 0.269485903814262,
+      "grad_norm": 2.6048810482025146,
+      "learning_rate": 3.7563956793632746e-05,
+      "loss": 5.6043,
+      "step": 4875
+    },
+    {
+      "epoch": 0.2697622996130459,
+      "grad_norm": 2.8759994506835938,
+      "learning_rate": 3.754974417282547e-05,
+      "loss": 5.4891,
+      "step": 4880
+    },
+    {
+      "epoch": 0.27003869541182973,
+      "grad_norm": 2.8228535652160645,
+      "learning_rate": 3.7535531552018194e-05,
+      "loss": 5.5018,
+      "step": 4885
+    },
+    {
+      "epoch": 0.2703150912106136,
+      "grad_norm": 3.0626003742218018,
+      "learning_rate": 3.752131893121092e-05,
+      "loss": 5.3939,
+      "step": 4890
+    },
+    {
+      "epoch": 0.27059148700939745,
+      "grad_norm": 2.5288095474243164,
+      "learning_rate": 3.750710631040364e-05,
+      "loss": 5.6604,
+      "step": 4895
+    },
+    {
+      "epoch": 0.2708678828081813,
+      "grad_norm": 2.9323790073394775,
+      "learning_rate": 3.7492893689596365e-05,
+      "loss": 5.4378,
+      "step": 4900
+    },
+    {
+      "epoch": 0.27114427860696516,
+      "grad_norm": 2.815352439880371,
+      "learning_rate": 3.747868106878908e-05,
+      "loss": 5.6052,
+      "step": 4905
+    },
+    {
+      "epoch": 0.271420674405749,
+      "grad_norm": 2.981050968170166,
+      "learning_rate": 3.746446844798181e-05,
+      "loss": 5.5689,
+      "step": 4910
+    },
+    {
+      "epoch": 0.2716970702045329,
+      "grad_norm": 3.062026262283325,
+      "learning_rate": 3.745025582717454e-05,
+      "loss": 5.4102,
+      "step": 4915
+    },
+    {
+      "epoch": 0.27197346600331673,
+      "grad_norm": 3.0991098880767822,
+      "learning_rate": 3.7436043206367255e-05,
+      "loss": 5.4325,
+      "step": 4920
+    },
+    {
+      "epoch": 0.2722498618021006,
+      "grad_norm": 3.309856414794922,
+      "learning_rate": 3.742183058555998e-05,
+      "loss": 5.2731,
+      "step": 4925
+    },
+    {
+      "epoch": 0.27252625760088445,
+      "grad_norm": 3.4550788402557373,
+      "learning_rate": 3.74076179647527e-05,
+      "loss": 5.5482,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2728026533996683,
+      "grad_norm": 2.121727705001831,
+      "learning_rate": 3.7393405343945427e-05,
+      "loss": 5.5073,
+      "step": 4935
+    },
+    {
+      "epoch": 0.27307904919845216,
+      "grad_norm": 2.894728660583496,
+      "learning_rate": 3.737919272313815e-05,
+      "loss": 5.693,
+      "step": 4940
+    },
+    {
+      "epoch": 0.273355444997236,
+      "grad_norm": 3.626741647720337,
+      "learning_rate": 3.736498010233087e-05,
+      "loss": 5.7241,
+      "step": 4945
+    },
+    {
+      "epoch": 0.2736318407960199,
+      "grad_norm": 2.86633563041687,
+      "learning_rate": 3.73507674815236e-05,
+      "loss": 5.4346,
+      "step": 4950
+    },
+    {
+      "epoch": 0.27390823659480373,
+      "grad_norm": 3.4413692951202393,
+      "learning_rate": 3.7336554860716316e-05,
+      "loss": 5.4228,
+      "step": 4955
+    },
+    {
+      "epoch": 0.2741846323935876,
+      "grad_norm": 2.800236225128174,
+      "learning_rate": 3.732234223990904e-05,
+      "loss": 5.3015,
+      "step": 4960
+    },
+    {
+      "epoch": 0.2744610281923715,
+      "grad_norm": 3.2357637882232666,
+      "learning_rate": 3.7308129619101764e-05,
+      "loss": 5.7801,
+      "step": 4965
+    },
+    {
+      "epoch": 0.27473742399115536,
+      "grad_norm": 2.5633909702301025,
+      "learning_rate": 3.729391699829449e-05,
+      "loss": 5.6144,
+      "step": 4970
+    },
+    {
+      "epoch": 0.2750138197899392,
+      "grad_norm": 2.8940200805664062,
+      "learning_rate": 3.727970437748721e-05,
+      "loss": 5.442,
+      "step": 4975
+    },
+    {
+      "epoch": 0.2752902155887231,
+      "grad_norm": 3.922053337097168,
+      "learning_rate": 3.726549175667993e-05,
+      "loss": 5.4147,
+      "step": 4980
+    },
+    {
+      "epoch": 0.27556661138750693,
+      "grad_norm": 3.393669366836548,
+      "learning_rate": 3.725127913587266e-05,
+      "loss": 5.6151,
+      "step": 4985
+    },
+    {
+      "epoch": 0.2758430071862908,
+      "grad_norm": 3.1016077995300293,
+      "learning_rate": 3.723706651506538e-05,
+      "loss": 5.5034,
+      "step": 4990
+    },
+    {
+      "epoch": 0.27611940298507465,
+      "grad_norm": 4.4078145027160645,
+      "learning_rate": 3.72228538942581e-05,
+      "loss": 5.5826,
+      "step": 4995
+    },
+    {
+      "epoch": 0.2763957987838585,
+      "grad_norm": 3.1527042388916016,
+      "learning_rate": 3.720864127345083e-05,
+      "loss": 5.4405,
+      "step": 5000
+    },
+    {
+      "epoch": 0.27667219458264236,
+      "grad_norm": 3.0095982551574707,
+      "learning_rate": 3.719442865264355e-05,
+      "loss": 5.6309,
+      "step": 5005
+    },
+    {
+      "epoch": 0.2769485903814262,
+      "grad_norm": 3.327120542526245,
+      "learning_rate": 3.718021603183627e-05,
+      "loss": 5.5869,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2772249861802101,
+      "grad_norm": 3.148468017578125,
+      "learning_rate": 3.7166003411029e-05,
+      "loss": 5.3626,
+      "step": 5015
+    },
+    {
+      "epoch": 0.27750138197899393,
+      "grad_norm": 2.774047374725342,
+      "learning_rate": 3.715179079022172e-05,
+      "loss": 5.0038,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2777777777777778,
+      "grad_norm": 2.5972516536712646,
+      "learning_rate": 3.7137578169414445e-05,
+      "loss": 5.8164,
+      "step": 5025
+    },
+    {
+      "epoch": 0.27805417357656165,
+      "grad_norm": 3.2580530643463135,
+      "learning_rate": 3.712336554860716e-05,
+      "loss": 5.5117,
+      "step": 5030
+    },
+    {
+      "epoch": 0.2783305693753455,
+      "grad_norm": 2.4557392597198486,
+      "learning_rate": 3.710915292779989e-05,
+      "loss": 5.4633,
+      "step": 5035
+    },
+    {
+      "epoch": 0.27860696517412936,
+      "grad_norm": 2.765587091445923,
+      "learning_rate": 3.709494030699261e-05,
+      "loss": 5.3116,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2788833609729132,
+      "grad_norm": 3.075885057449341,
+      "learning_rate": 3.7080727686185334e-05,
+      "loss": 5.4564,
+      "step": 5045
+    },
+    {
+      "epoch": 0.2791597567716971,
+      "grad_norm": 2.8484435081481934,
+      "learning_rate": 3.706651506537806e-05,
+      "loss": 5.5371,
+      "step": 5050
+    },
+    {
+      "epoch": 0.27943615257048093,
+      "grad_norm": 2.929783344268799,
+      "learning_rate": 3.705230244457078e-05,
+      "loss": 5.4535,
+      "step": 5055
+    },
+    {
+      "epoch": 0.2797125483692648,
+      "grad_norm": 2.6549274921417236,
+      "learning_rate": 3.7038089823763506e-05,
+      "loss": 5.4736,
+      "step": 5060
+    },
+    {
+      "epoch": 0.27998894416804865,
+      "grad_norm": 3.388282537460327,
+      "learning_rate": 3.702387720295622e-05,
+      "loss": 5.3436,
+      "step": 5065
+    },
+    {
+      "epoch": 0.2802653399668325,
+      "grad_norm": 3.1960010528564453,
+      "learning_rate": 3.7009664582148954e-05,
+      "loss": 5.8104,
+      "step": 5070
+    },
+    {
+      "epoch": 0.28054173576561636,
+      "grad_norm": 2.6687440872192383,
+      "learning_rate": 3.699545196134167e-05,
+      "loss": 5.2787,
+      "step": 5075
+    },
+    {
+      "epoch": 0.2808181315644002,
+      "grad_norm": 3.2794923782348633,
+      "learning_rate": 3.6981239340534395e-05,
+      "loss": 5.5187,
+      "step": 5080
+    },
+    {
+      "epoch": 0.2810945273631841,
+      "grad_norm": 2.8185513019561768,
+      "learning_rate": 3.696702671972712e-05,
+      "loss": 5.3756,
+      "step": 5085
+    },
+    {
+      "epoch": 0.28137092316196793,
+      "grad_norm": 3.3090620040893555,
+      "learning_rate": 3.695281409891984e-05,
+      "loss": 5.6135,
+      "step": 5090
+    },
+    {
+      "epoch": 0.2816473189607518,
+      "grad_norm": 2.823322057723999,
+      "learning_rate": 3.693860147811257e-05,
+      "loss": 5.7067,
+      "step": 5095
+    },
+    {
+      "epoch": 0.28192371475953565,
+      "grad_norm": 4.680220603942871,
+      "learning_rate": 3.6924388857305284e-05,
+      "loss": 5.3231,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2822001105583195,
+      "grad_norm": 3.1721339225769043,
+      "learning_rate": 3.6910176236498015e-05,
+      "loss": 5.3679,
+      "step": 5105
+    },
+    {
+      "epoch": 0.28247650635710336,
+      "grad_norm": 3.343784809112549,
+      "learning_rate": 3.689596361569074e-05,
+      "loss": 5.5471,
+      "step": 5110
+    },
+    {
+      "epoch": 0.2827529021558872,
+      "grad_norm": 2.900912284851074,
+      "learning_rate": 3.6881750994883456e-05,
+      "loss": 5.4448,
+      "step": 5115
+    },
+    {
+      "epoch": 0.2830292979546711,
+      "grad_norm": 2.725116729736328,
+      "learning_rate": 3.686753837407618e-05,
+      "loss": 5.1778,
+      "step": 5120
+    },
+    {
+      "epoch": 0.28330569375345493,
+      "grad_norm": 2.9266819953918457,
+      "learning_rate": 3.6853325753268904e-05,
+      "loss": 5.706,
+      "step": 5125
+    },
+    {
+      "epoch": 0.2835820895522388,
+      "grad_norm": 3.6640381813049316,
+      "learning_rate": 3.683911313246163e-05,
+      "loss": 5.355,
+      "step": 5130
+    },
+    {
+      "epoch": 0.28385848535102265,
+      "grad_norm": 3.3055994510650635,
+      "learning_rate": 3.682490051165435e-05,
+      "loss": 5.2927,
+      "step": 5135
+    },
+    {
+      "epoch": 0.2841348811498065,
+      "grad_norm": 3.171983242034912,
+      "learning_rate": 3.6810687890847076e-05,
+      "loss": 5.625,
+      "step": 5140
+    },
+    {
+      "epoch": 0.28441127694859036,
+      "grad_norm": 3.0071423053741455,
+      "learning_rate": 3.67964752700398e-05,
+      "loss": 5.2521,
+      "step": 5145
+    },
+    {
+      "epoch": 0.2846876727473742,
+      "grad_norm": 2.843132972717285,
+      "learning_rate": 3.678226264923252e-05,
+      "loss": 5.6159,
+      "step": 5150
+    },
+    {
+      "epoch": 0.2849640685461581,
+      "grad_norm": 2.8676323890686035,
+      "learning_rate": 3.676805002842524e-05,
+      "loss": 5.4452,
+      "step": 5155
+    },
+    {
+      "epoch": 0.28524046434494194,
+      "grad_norm": 2.7447266578674316,
+      "learning_rate": 3.675383740761797e-05,
+      "loss": 5.4009,
+      "step": 5160
+    },
+    {
+      "epoch": 0.2855168601437258,
+      "grad_norm": 3.2596964836120605,
+      "learning_rate": 3.673962478681069e-05,
+      "loss": 5.7716,
+      "step": 5165
+    },
+    {
+      "epoch": 0.28579325594250965,
+      "grad_norm": 2.8645823001861572,
+      "learning_rate": 3.672541216600341e-05,
+      "loss": 5.5613,
+      "step": 5170
+    },
+    {
+      "epoch": 0.2860696517412935,
+      "grad_norm": 2.947704315185547,
+      "learning_rate": 3.671119954519614e-05,
+      "loss": 5.343,
+      "step": 5175
+    },
+    {
+      "epoch": 0.28634604754007736,
+      "grad_norm": 3.6739652156829834,
+      "learning_rate": 3.669698692438886e-05,
+      "loss": 5.3991,
+      "step": 5180
+    },
+    {
+      "epoch": 0.2866224433388613,
+      "grad_norm": 3.433134078979492,
+      "learning_rate": 3.668277430358158e-05,
+      "loss": 5.6598,
+      "step": 5185
+    },
+    {
+      "epoch": 0.28689883913764513,
+      "grad_norm": 2.977222204208374,
+      "learning_rate": 3.66685616827743e-05,
+      "loss": 5.4816,
+      "step": 5190
+    },
+    {
+      "epoch": 0.287175234936429,
+      "grad_norm": 2.5906851291656494,
+      "learning_rate": 3.665434906196703e-05,
+      "loss": 5.3199,
+      "step": 5195
+    },
+    {
+      "epoch": 0.28745163073521285,
+      "grad_norm": 2.8117191791534424,
+      "learning_rate": 3.664013644115975e-05,
+      "loss": 5.4231,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2877280265339967,
+      "grad_norm": 2.9381093978881836,
+      "learning_rate": 3.6625923820352474e-05,
+      "loss": 5.4396,
+      "step": 5205
+    },
+    {
+      "epoch": 0.28800442233278056,
+      "grad_norm": 3.0122227668762207,
+      "learning_rate": 3.66117111995452e-05,
+      "loss": 5.2438,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2882808181315644,
+      "grad_norm": 2.6538515090942383,
+      "learning_rate": 3.659749857873792e-05,
+      "loss": 5.3428,
+      "step": 5215
+    },
+    {
+      "epoch": 0.2885572139303483,
+      "grad_norm": 3.465743064880371,
+      "learning_rate": 3.6583285957930646e-05,
+      "loss": 5.4583,
+      "step": 5220
+    },
+    {
+      "epoch": 0.28883360972913213,
+      "grad_norm": 3.0137925148010254,
+      "learning_rate": 3.656907333712336e-05,
+      "loss": 5.6031,
+      "step": 5225
+    },
+    {
+      "epoch": 0.289110005527916,
+      "grad_norm": 3.7536206245422363,
+      "learning_rate": 3.6554860716316094e-05,
+      "loss": 5.3189,
+      "step": 5230
+    },
+    {
+      "epoch": 0.28938640132669985,
+      "grad_norm": 2.687974691390991,
+      "learning_rate": 3.654064809550881e-05,
+      "loss": 5.1424,
+      "step": 5235
+    },
+    {
+      "epoch": 0.2896627971254837,
+      "grad_norm": 2.5427982807159424,
+      "learning_rate": 3.6526435474701535e-05,
+      "loss": 5.5922,
+      "step": 5240
+    },
+    {
+      "epoch": 0.28993919292426756,
+      "grad_norm": 2.6604325771331787,
+      "learning_rate": 3.651222285389426e-05,
+      "loss": 5.4598,
+      "step": 5245
+    },
+    {
+      "epoch": 0.2902155887230514,
+      "grad_norm": 2.8820483684539795,
+      "learning_rate": 3.649801023308698e-05,
+      "loss": 5.3329,
+      "step": 5250
+    },
+    {
+      "epoch": 0.2904919845218353,
+      "grad_norm": 4.090969085693359,
+      "learning_rate": 3.648379761227971e-05,
+      "loss": 5.5375,
+      "step": 5255
+    },
+    {
+      "epoch": 0.29076838032061914,
+      "grad_norm": 3.3297054767608643,
+      "learning_rate": 3.6469584991472424e-05,
+      "loss": 5.217,
+      "step": 5260
+    },
+    {
+      "epoch": 0.291044776119403,
+      "grad_norm": 2.8751707077026367,
+      "learning_rate": 3.6455372370665155e-05,
+      "loss": 5.3424,
+      "step": 5265
+    },
+    {
+      "epoch": 0.29132117191818685,
+      "grad_norm": 2.618412971496582,
+      "learning_rate": 3.644115974985788e-05,
+      "loss": 5.4186,
+      "step": 5270
+    },
+    {
+      "epoch": 0.2915975677169707,
+      "grad_norm": 3.5012338161468506,
+      "learning_rate": 3.6426947129050596e-05,
+      "loss": 5.5651,
+      "step": 5275
+    },
+    {
+      "epoch": 0.29187396351575456,
+      "grad_norm": 3.1529078483581543,
+      "learning_rate": 3.641273450824333e-05,
+      "loss": 5.4826,
+      "step": 5280
+    },
+    {
+      "epoch": 0.2921503593145384,
+      "grad_norm": 3.4742085933685303,
+      "learning_rate": 3.6398521887436044e-05,
+      "loss": 5.3871,
+      "step": 5285
+    },
+    {
+      "epoch": 0.2924267551133223,
+      "grad_norm": 3.666706085205078,
+      "learning_rate": 3.638430926662877e-05,
+      "loss": 5.6392,
+      "step": 5290
+    },
+    {
+      "epoch": 0.29270315091210614,
+      "grad_norm": 3.182141065597534,
+      "learning_rate": 3.6370096645821485e-05,
+      "loss": 5.3455,
+      "step": 5295
+    },
+    {
+      "epoch": 0.29297954671089,
+      "grad_norm": 3.0457582473754883,
+      "learning_rate": 3.6355884025014216e-05,
+      "loss": 5.7264,
+      "step": 5300
+    },
+    {
+      "epoch": 0.29325594250967385,
+      "grad_norm": 2.9654111862182617,
+      "learning_rate": 3.634167140420694e-05,
+      "loss": 5.3143,
+      "step": 5305
+    },
+    {
+      "epoch": 0.2935323383084577,
+      "grad_norm": 3.9714043140411377,
+      "learning_rate": 3.632745878339966e-05,
+      "loss": 5.6,
+      "step": 5310
+    },
+    {
+      "epoch": 0.29380873410724156,
+      "grad_norm": 3.186901092529297,
+      "learning_rate": 3.631324616259239e-05,
+      "loss": 5.4677,
+      "step": 5315
+    },
+    {
+      "epoch": 0.2940851299060254,
+      "grad_norm": 3.43259334564209,
+      "learning_rate": 3.6299033541785105e-05,
+      "loss": 5.5161,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2943615257048093,
+      "grad_norm": 2.980299711227417,
+      "learning_rate": 3.628482092097783e-05,
+      "loss": 5.4279,
+      "step": 5325
+    },
+    {
+      "epoch": 0.29463792150359314,
+      "grad_norm": 2.5553932189941406,
+      "learning_rate": 3.627060830017055e-05,
+      "loss": 5.2727,
+      "step": 5330
+    },
+    {
+      "epoch": 0.294914317302377,
+      "grad_norm": 2.9207799434661865,
+      "learning_rate": 3.625639567936328e-05,
+      "loss": 5.5978,
+      "step": 5335
+    },
+    {
+      "epoch": 0.29519071310116085,
+      "grad_norm": 2.6809563636779785,
+      "learning_rate": 3.6242183058556e-05,
+      "loss": 5.6545,
+      "step": 5340
+    },
+    {
+      "epoch": 0.2954671088999447,
+      "grad_norm": 3.1138181686401367,
+      "learning_rate": 3.622797043774872e-05,
+      "loss": 5.3426,
+      "step": 5345
+    },
+    {
+      "epoch": 0.29574350469872857,
+      "grad_norm": 3.2143821716308594,
+      "learning_rate": 3.621375781694145e-05,
+      "loss": 5.5284,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2960199004975124,
+      "grad_norm": 3.2884023189544678,
+      "learning_rate": 3.619954519613417e-05,
+      "loss": 5.3533,
+      "step": 5355
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 2.707595109939575,
+      "learning_rate": 3.618533257532689e-05,
+      "loss": 5.4062,
+      "step": 5360
+    },
+    {
+      "epoch": 0.29657269209508014,
+      "grad_norm": 2.758535146713257,
+      "learning_rate": 3.6171119954519614e-05,
+      "loss": 5.3409,
+      "step": 5365
+    },
+    {
+      "epoch": 0.296849087893864,
+      "grad_norm": 3.2588820457458496,
+      "learning_rate": 3.615690733371234e-05,
+      "loss": 5.4564,
+      "step": 5370
+    },
+    {
+      "epoch": 0.29712548369264785,
+      "grad_norm": 3.819500207901001,
+      "learning_rate": 3.614269471290506e-05,
+      "loss": 5.5978,
+      "step": 5375
+    },
+    {
+      "epoch": 0.2974018794914317,
+      "grad_norm": 3.1476078033447266,
+      "learning_rate": 3.6128482092097786e-05,
+      "loss": 5.1632,
+      "step": 5380
+    },
+    {
+      "epoch": 0.29767827529021557,
+      "grad_norm": 2.8561973571777344,
+      "learning_rate": 3.611426947129051e-05,
+      "loss": 5.5271,
+      "step": 5385
+    },
+    {
+      "epoch": 0.2979546710889994,
+      "grad_norm": 2.6981911659240723,
+      "learning_rate": 3.6100056850483234e-05,
+      "loss": 5.5108,
+      "step": 5390
+    },
+    {
+      "epoch": 0.2982310668877833,
+      "grad_norm": 3.0033581256866455,
+      "learning_rate": 3.608584422967595e-05,
+      "loss": 5.2921,
+      "step": 5395
+    },
+    {
+      "epoch": 0.29850746268656714,
+      "grad_norm": 3.7791874408721924,
+      "learning_rate": 3.6071631608868675e-05,
+      "loss": 5.5129,
+      "step": 5400
+    },
+    {
+      "epoch": 0.29878385848535105,
+      "grad_norm": 2.5647666454315186,
+      "learning_rate": 3.60574189880614e-05,
+      "loss": 5.7754,
+      "step": 5405
+    },
+    {
+      "epoch": 0.2990602542841349,
+      "grad_norm": 2.5552051067352295,
+      "learning_rate": 3.604320636725412e-05,
+      "loss": 5.2537,
+      "step": 5410
+    },
+    {
+      "epoch": 0.29933665008291876,
+      "grad_norm": 3.173316478729248,
+      "learning_rate": 3.602899374644685e-05,
+      "loss": 5.7515,
+      "step": 5415
+    },
+    {
+      "epoch": 0.2996130458817026,
+      "grad_norm": 2.4736528396606445,
+      "learning_rate": 3.601478112563957e-05,
+      "loss": 5.0059,
+      "step": 5420
+    },
+    {
+      "epoch": 0.2998894416804865,
+      "grad_norm": 3.348816394805908,
+      "learning_rate": 3.6000568504832295e-05,
+      "loss": 5.477,
+      "step": 5425
+    },
+    {
+      "epoch": 0.30016583747927034,
+      "grad_norm": 2.6805813312530518,
+      "learning_rate": 3.598635588402501e-05,
+      "loss": 5.5837,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3004422332780542,
+      "grad_norm": 3.264838218688965,
+      "learning_rate": 3.5972143263217736e-05,
+      "loss": 5.6984,
+      "step": 5435
+    },
+    {
+      "epoch": 0.30071862907683805,
+      "grad_norm": 2.8391873836517334,
+      "learning_rate": 3.595793064241047e-05,
+      "loss": 5.2333,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3009950248756219,
+      "grad_norm": 3.601405382156372,
+      "learning_rate": 3.5943718021603184e-05,
+      "loss": 5.2084,
+      "step": 5445
+    },
+    {
+      "epoch": 0.30127142067440577,
+      "grad_norm": 3.3208744525909424,
+      "learning_rate": 3.592950540079591e-05,
+      "loss": 5.18,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3015478164731896,
+      "grad_norm": 3.2721526622772217,
+      "learning_rate": 3.591529277998863e-05,
+      "loss": 5.6478,
+      "step": 5455
+    },
+    {
+      "epoch": 0.3018242122719735,
+      "grad_norm": 3.912829637527466,
+      "learning_rate": 3.5901080159181356e-05,
+      "loss": 5.4232,
+      "step": 5460
+    },
+    {
+      "epoch": 0.30210060807075734,
+      "grad_norm": 3.12005615234375,
+      "learning_rate": 3.588686753837408e-05,
+      "loss": 5.6941,
+      "step": 5465
+    },
+    {
+      "epoch": 0.3023770038695412,
+      "grad_norm": 3.007559061050415,
+      "learning_rate": 3.58726549175668e-05,
+      "loss": 5.3598,
+      "step": 5470
+    },
+    {
+      "epoch": 0.30265339966832505,
+      "grad_norm": 3.6492421627044678,
+      "learning_rate": 3.585844229675953e-05,
+      "loss": 5.6721,
+      "step": 5475
+    },
+    {
+      "epoch": 0.3029297954671089,
+      "grad_norm": 3.9073736667633057,
+      "learning_rate": 3.5844229675952245e-05,
+      "loss": 5.4005,
+      "step": 5480
+    },
+    {
+      "epoch": 0.30320619126589277,
+      "grad_norm": 3.031067371368408,
+      "learning_rate": 3.583001705514497e-05,
+      "loss": 5.3484,
+      "step": 5485
+    },
+    {
+      "epoch": 0.3034825870646766,
+      "grad_norm": 3.599224090576172,
+      "learning_rate": 3.581580443433769e-05,
+      "loss": 5.4078,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3037589828634605,
+      "grad_norm": 3.183856964111328,
+      "learning_rate": 3.580159181353042e-05,
+      "loss": 5.5047,
+      "step": 5495
+    },
+    {
+      "epoch": 0.30403537866224434,
+      "grad_norm": 3.2303097248077393,
+      "learning_rate": 3.578737919272314e-05,
+      "loss": 5.5218,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3043117744610282,
+      "grad_norm": 3.168322801589966,
+      "learning_rate": 3.577316657191586e-05,
+      "loss": 5.4699,
+      "step": 5505
+    },
+    {
+      "epoch": 0.30458817025981205,
+      "grad_norm": 2.8198087215423584,
+      "learning_rate": 3.575895395110859e-05,
+      "loss": 4.9456,
+      "step": 5510
+    },
+    {
+      "epoch": 0.3048645660585959,
+      "grad_norm": 3.3285229206085205,
+      "learning_rate": 3.5744741330301307e-05,
+      "loss": 5.5881,
+      "step": 5515
+    },
+    {
+      "epoch": 0.30514096185737977,
+      "grad_norm": 3.381110668182373,
+      "learning_rate": 3.573052870949403e-05,
+      "loss": 5.8306,
+      "step": 5520
+    },
+    {
+      "epoch": 0.3054173576561636,
+      "grad_norm": 2.8767781257629395,
+      "learning_rate": 3.5716316088686754e-05,
+      "loss": 5.2034,
+      "step": 5525
+    },
+    {
+      "epoch": 0.3056937534549475,
+      "grad_norm": 4.404655456542969,
+      "learning_rate": 3.570210346787948e-05,
+      "loss": 5.4832,
+      "step": 5530
+    },
+    {
+      "epoch": 0.30597014925373134,
+      "grad_norm": 2.4534425735473633,
+      "learning_rate": 3.56878908470722e-05,
+      "loss": 5.1591,
+      "step": 5535
+    },
+    {
+      "epoch": 0.3062465450525152,
+      "grad_norm": 2.891896963119507,
+      "learning_rate": 3.567367822626492e-05,
+      "loss": 5.253,
+      "step": 5540
+    },
+    {
+      "epoch": 0.30652294085129905,
+      "grad_norm": 2.3647806644439697,
+      "learning_rate": 3.565946560545765e-05,
+      "loss": 5.4199,
+      "step": 5545
+    },
+    {
+      "epoch": 0.3067993366500829,
+      "grad_norm": 3.0437514781951904,
+      "learning_rate": 3.5645252984650374e-05,
+      "loss": 5.4336,
+      "step": 5550
+    },
+    {
+      "epoch": 0.30707573244886677,
+      "grad_norm": 3.7891740798950195,
+      "learning_rate": 3.563104036384309e-05,
+      "loss": 5.5751,
+      "step": 5555
+    },
+    {
+      "epoch": 0.3073521282476506,
+      "grad_norm": 3.7321300506591797,
+      "learning_rate": 3.5616827743035816e-05,
+      "loss": 5.319,
+      "step": 5560
+    },
+    {
+      "epoch": 0.3076285240464345,
+      "grad_norm": 3.2207529544830322,
+      "learning_rate": 3.560261512222854e-05,
+      "loss": 5.74,
+      "step": 5565
+    },
+    {
+      "epoch": 0.30790491984521834,
+      "grad_norm": 3.2805068492889404,
+      "learning_rate": 3.5588402501421263e-05,
+      "loss": 5.2347,
+      "step": 5570
+    },
+    {
+      "epoch": 0.3081813156440022,
+      "grad_norm": 2.671035051345825,
+      "learning_rate": 3.557418988061399e-05,
+      "loss": 5.4821,
+      "step": 5575
+    },
+    {
+      "epoch": 0.30845771144278605,
+      "grad_norm": 2.748237133026123,
+      "learning_rate": 3.555997725980671e-05,
+      "loss": 5.2953,
+      "step": 5580
+    },
+    {
+      "epoch": 0.3087341072415699,
+      "grad_norm": 3.34366512298584,
+      "learning_rate": 3.5545764638999435e-05,
+      "loss": 5.336,
+      "step": 5585
+    },
+    {
+      "epoch": 0.30901050304035377,
+      "grad_norm": 3.0358309745788574,
+      "learning_rate": 3.553155201819215e-05,
+      "loss": 5.6294,
+      "step": 5590
+    },
+    {
+      "epoch": 0.3092868988391376,
+      "grad_norm": 2.842928171157837,
+      "learning_rate": 3.5517339397384883e-05,
+      "loss": 5.2573,
+      "step": 5595
+    },
+    {
+      "epoch": 0.3095632946379215,
+      "grad_norm": 3.203237533569336,
+      "learning_rate": 3.55031267765776e-05,
+      "loss": 5.4097,
+      "step": 5600
+    },
+    {
+      "epoch": 0.30983969043670534,
+      "grad_norm": 3.44087290763855,
+      "learning_rate": 3.5488914155770325e-05,
+      "loss": 5.252,
+      "step": 5605
+    },
+    {
+      "epoch": 0.3101160862354892,
+      "grad_norm": 2.9737708568573,
+      "learning_rate": 3.547470153496305e-05,
+      "loss": 5.3748,
+      "step": 5610
+    },
+    {
+      "epoch": 0.31039248203427305,
+      "grad_norm": 2.6989824771881104,
+      "learning_rate": 3.546048891415577e-05,
+      "loss": 5.612,
+      "step": 5615
+    },
+    {
+      "epoch": 0.3106688778330569,
+      "grad_norm": 3.0308315753936768,
+      "learning_rate": 3.5446276293348497e-05,
+      "loss": 5.5512,
+      "step": 5620
+    },
+    {
+      "epoch": 0.31094527363184077,
+      "grad_norm": 3.383033275604248,
+      "learning_rate": 3.5432063672541214e-05,
+      "loss": 5.4498,
+      "step": 5625
+    },
+    {
+      "epoch": 0.3112216694306247,
+      "grad_norm": 2.5991249084472656,
+      "learning_rate": 3.5417851051733944e-05,
+      "loss": 5.3198,
+      "step": 5630
+    },
+    {
+      "epoch": 0.31149806522940854,
+      "grad_norm": 3.038034439086914,
+      "learning_rate": 3.540363843092667e-05,
+      "loss": 5.6306,
+      "step": 5635
+    },
+    {
+      "epoch": 0.3117744610281924,
+      "grad_norm": 3.4985435009002686,
+      "learning_rate": 3.5389425810119386e-05,
+      "loss": 5.7529,
+      "step": 5640
+    },
+    {
+      "epoch": 0.31205085682697625,
+      "grad_norm": 3.8740835189819336,
+      "learning_rate": 3.537521318931211e-05,
+      "loss": 5.5961,
+      "step": 5645
+    },
+    {
+      "epoch": 0.3123272526257601,
+      "grad_norm": 2.630774974822998,
+      "learning_rate": 3.5361000568504834e-05,
+      "loss": 5.4196,
+      "step": 5650
+    },
+    {
+      "epoch": 0.31260364842454397,
+      "grad_norm": 2.864548683166504,
+      "learning_rate": 3.534678794769756e-05,
+      "loss": 5.4804,
+      "step": 5655
+    },
+    {
+      "epoch": 0.3128800442233278,
+      "grad_norm": 2.944843053817749,
+      "learning_rate": 3.533257532689028e-05,
+      "loss": 5.6182,
+      "step": 5660
+    },
+    {
+      "epoch": 0.3131564400221117,
+      "grad_norm": 2.843414545059204,
+      "learning_rate": 3.5318362706083006e-05,
+      "loss": 5.3153,
+      "step": 5665
+    },
+    {
+      "epoch": 0.31343283582089554,
+      "grad_norm": 2.6151304244995117,
+      "learning_rate": 3.530415008527573e-05,
+      "loss": 5.2085,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3137092316196794,
+      "grad_norm": 3.2855122089385986,
+      "learning_rate": 3.528993746446845e-05,
+      "loss": 5.1338,
+      "step": 5675
+    },
+    {
+      "epoch": 0.31398562741846325,
+      "grad_norm": 2.8010687828063965,
+      "learning_rate": 3.527572484366117e-05,
+      "loss": 5.644,
+      "step": 5680
+    },
+    {
+      "epoch": 0.3142620232172471,
+      "grad_norm": 3.428339958190918,
+      "learning_rate": 3.52615122228539e-05,
+      "loss": 5.4467,
+      "step": 5685
+    },
+    {
+      "epoch": 0.31453841901603097,
+      "grad_norm": 2.4719932079315186,
+      "learning_rate": 3.524729960204662e-05,
+      "loss": 5.5976,
+      "step": 5690
+    },
+    {
+      "epoch": 0.3148148148148148,
+      "grad_norm": 3.2966175079345703,
+      "learning_rate": 3.523308698123934e-05,
+      "loss": 5.6092,
+      "step": 5695
+    },
+    {
+      "epoch": 0.3150912106135987,
+      "grad_norm": 3.152592658996582,
+      "learning_rate": 3.521887436043207e-05,
+      "loss": 5.3276,
+      "step": 5700
+    },
+    {
+      "epoch": 0.31536760641238254,
+      "grad_norm": 2.5936498641967773,
+      "learning_rate": 3.520466173962479e-05,
+      "loss": 5.5664,
+      "step": 5705
+    },
+    {
+      "epoch": 0.3156440022111664,
+      "grad_norm": 3.2857308387756348,
+      "learning_rate": 3.519044911881751e-05,
+      "loss": 5.0564,
+      "step": 5710
+    },
+    {
+      "epoch": 0.31592039800995025,
+      "grad_norm": 3.022956371307373,
+      "learning_rate": 3.517623649801023e-05,
+      "loss": 5.3535,
+      "step": 5715
+    },
+    {
+      "epoch": 0.3161967938087341,
+      "grad_norm": 2.88362455368042,
+      "learning_rate": 3.516202387720296e-05,
+      "loss": 5.6187,
+      "step": 5720
+    },
+    {
+      "epoch": 0.31647318960751797,
+      "grad_norm": 3.379072427749634,
+      "learning_rate": 3.514781125639568e-05,
+      "loss": 5.6233,
+      "step": 5725
+    },
+    {
+      "epoch": 0.3167495854063018,
+      "grad_norm": 2.9026386737823486,
+      "learning_rate": 3.5133598635588404e-05,
+      "loss": 5.5723,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3170259812050857,
+      "grad_norm": 3.1735050678253174,
+      "learning_rate": 3.511938601478113e-05,
+      "loss": 5.3733,
+      "step": 5735
+    },
+    {
+      "epoch": 0.31730237700386954,
+      "grad_norm": 2.6387147903442383,
+      "learning_rate": 3.510517339397385e-05,
+      "loss": 5.5433,
+      "step": 5740
+    },
+    {
+      "epoch": 0.3175787728026534,
+      "grad_norm": 2.5088794231414795,
+      "learning_rate": 3.5090960773166576e-05,
+      "loss": 5.3205,
+      "step": 5745
+    },
+    {
+      "epoch": 0.31785516860143725,
+      "grad_norm": 3.163351535797119,
+      "learning_rate": 3.507674815235929e-05,
+      "loss": 5.4311,
+      "step": 5750
+    },
+    {
+      "epoch": 0.3181315644002211,
+      "grad_norm": 3.1320881843566895,
+      "learning_rate": 3.5062535531552024e-05,
+      "loss": 5.3945,
+      "step": 5755
+    },
+    {
+      "epoch": 0.31840796019900497,
+      "grad_norm": 2.6302387714385986,
+      "learning_rate": 3.504832291074474e-05,
+      "loss": 5.2206,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3186843559977888,
+      "grad_norm": 2.8344478607177734,
+      "learning_rate": 3.5034110289937465e-05,
+      "loss": 5.2453,
+      "step": 5765
+    },
+    {
+      "epoch": 0.3189607517965727,
+      "grad_norm": 3.076878786087036,
+      "learning_rate": 3.501989766913019e-05,
+      "loss": 5.6607,
+      "step": 5770
+    },
+    {
+      "epoch": 0.31923714759535654,
+      "grad_norm": 2.7953310012817383,
+      "learning_rate": 3.500568504832291e-05,
+      "loss": 5.4197,
+      "step": 5775
+    },
+    {
+      "epoch": 0.3195135433941404,
+      "grad_norm": 3.1712965965270996,
+      "learning_rate": 3.499147242751564e-05,
+      "loss": 5.5598,
+      "step": 5780
+    },
+    {
+      "epoch": 0.31978993919292426,
+      "grad_norm": 4.416378021240234,
+      "learning_rate": 3.4977259806708354e-05,
+      "loss": 5.5493,
+      "step": 5785
+    },
+    {
+      "epoch": 0.3200663349917081,
+      "grad_norm": 3.5662577152252197,
+      "learning_rate": 3.4963047185901085e-05,
+      "loss": 5.9117,
+      "step": 5790
+    },
+    {
+      "epoch": 0.32034273079049197,
+      "grad_norm": 3.769775390625,
+      "learning_rate": 3.494883456509381e-05,
+      "loss": 5.3641,
+      "step": 5795
+    },
+    {
+      "epoch": 0.3206191265892758,
+      "grad_norm": 3.322216749191284,
+      "learning_rate": 3.4934621944286526e-05,
+      "loss": 5.5554,
+      "step": 5800
+    },
+    {
+      "epoch": 0.3208955223880597,
+      "grad_norm": 2.9367949962615967,
+      "learning_rate": 3.492040932347925e-05,
+      "loss": 5.403,
+      "step": 5805
+    },
+    {
+      "epoch": 0.32117191818684354,
+      "grad_norm": 2.9416840076446533,
+      "learning_rate": 3.4906196702671974e-05,
+      "loss": 5.3599,
+      "step": 5810
+    },
+    {
+      "epoch": 0.3214483139856274,
+      "grad_norm": 3.0525002479553223,
+      "learning_rate": 3.48919840818647e-05,
+      "loss": 5.3231,
+      "step": 5815
+    },
+    {
+      "epoch": 0.32172470978441126,
+      "grad_norm": 3.840785503387451,
+      "learning_rate": 3.4877771461057415e-05,
+      "loss": 5.6919,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3220011055831951,
+      "grad_norm": 2.738666296005249,
+      "learning_rate": 3.4863558840250146e-05,
+      "loss": 5.6685,
+      "step": 5825
+    },
+    {
+      "epoch": 0.32227750138197897,
+      "grad_norm": 4.050598621368408,
+      "learning_rate": 3.484934621944287e-05,
+      "loss": 5.4941,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3225538971807628,
+      "grad_norm": 3.0634477138519287,
+      "learning_rate": 3.483513359863559e-05,
+      "loss": 5.3485,
+      "step": 5835
+    },
+    {
+      "epoch": 0.3228302929795467,
+      "grad_norm": 3.11482572555542,
+      "learning_rate": 3.482092097782831e-05,
+      "loss": 5.1452,
+      "step": 5840
+    },
+    {
+      "epoch": 0.32310668877833054,
+      "grad_norm": 2.8689663410186768,
+      "learning_rate": 3.4806708357021035e-05,
+      "loss": 5.397,
+      "step": 5845
+    },
+    {
+      "epoch": 0.32338308457711445,
+      "grad_norm": 3.4078924655914307,
+      "learning_rate": 3.479249573621376e-05,
+      "loss": 5.6103,
+      "step": 5850
+    },
+    {
+      "epoch": 0.3236594803758983,
+      "grad_norm": 3.1890652179718018,
+      "learning_rate": 3.477828311540648e-05,
+      "loss": 5.482,
+      "step": 5855
+    },
+    {
+      "epoch": 0.32393587617468217,
+      "grad_norm": 3.211745023727417,
+      "learning_rate": 3.476407049459921e-05,
+      "loss": 5.2123,
+      "step": 5860
+    },
+    {
+      "epoch": 0.324212271973466,
+      "grad_norm": 3.373166799545288,
+      "learning_rate": 3.474985787379193e-05,
+      "loss": 5.4259,
+      "step": 5865
+    },
+    {
+      "epoch": 0.3244886677722499,
+      "grad_norm": 3.0592551231384277,
+      "learning_rate": 3.473564525298465e-05,
+      "loss": 5.4406,
+      "step": 5870
+    },
+    {
+      "epoch": 0.32476506357103374,
+      "grad_norm": 3.850008010864258,
+      "learning_rate": 3.472143263217737e-05,
+      "loss": 5.4733,
+      "step": 5875
+    },
+    {
+      "epoch": 0.3250414593698176,
+      "grad_norm": 3.4215028285980225,
+      "learning_rate": 3.47072200113701e-05,
+      "loss": 5.6001,
+      "step": 5880
+    },
+    {
+      "epoch": 0.32531785516860146,
+      "grad_norm": 2.7632102966308594,
+      "learning_rate": 3.469300739056282e-05,
+      "loss": 5.3779,
+      "step": 5885
+    },
+    {
+      "epoch": 0.3255942509673853,
+      "grad_norm": 2.7702176570892334,
+      "learning_rate": 3.4678794769755544e-05,
+      "loss": 5.3826,
+      "step": 5890
+    },
+    {
+      "epoch": 0.32587064676616917,
+      "grad_norm": 2.870290517807007,
+      "learning_rate": 3.466458214894827e-05,
+      "loss": 5.4052,
+      "step": 5895
+    },
+    {
+      "epoch": 0.326147042564953,
+      "grad_norm": 3.039081573486328,
+      "learning_rate": 3.465036952814099e-05,
+      "loss": 5.2297,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3264234383637369,
+      "grad_norm": 3.192931890487671,
+      "learning_rate": 3.4636156907333716e-05,
+      "loss": 5.424,
+      "step": 5905
+    },
+    {
+      "epoch": 0.32669983416252074,
+      "grad_norm": 2.7073121070861816,
+      "learning_rate": 3.462194428652644e-05,
+      "loss": 5.4958,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3269762299613046,
+      "grad_norm": 3.2365293502807617,
+      "learning_rate": 3.4607731665719164e-05,
+      "loss": 5.099,
+      "step": 5915
+    },
+    {
+      "epoch": 0.32725262576008846,
+      "grad_norm": 3.3595356941223145,
+      "learning_rate": 3.459351904491188e-05,
+      "loss": 5.5425,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3275290215588723,
+      "grad_norm": 2.8691320419311523,
+      "learning_rate": 3.4579306424104605e-05,
+      "loss": 5.4369,
+      "step": 5925
+    },
+    {
+      "epoch": 0.32780541735765617,
+      "grad_norm": 3.031155586242676,
+      "learning_rate": 3.456509380329733e-05,
+      "loss": 5.4964,
+      "step": 5930
+    },
+    {
+      "epoch": 0.32808181315644,
+      "grad_norm": 2.736032247543335,
+      "learning_rate": 3.455088118249005e-05,
+      "loss": 5.3402,
+      "step": 5935
+    },
+    {
+      "epoch": 0.3283582089552239,
+      "grad_norm": 3.4750332832336426,
+      "learning_rate": 3.453666856168278e-05,
+      "loss": 5.2481,
+      "step": 5940
+    },
+    {
+      "epoch": 0.32863460475400774,
+      "grad_norm": 2.969907522201538,
+      "learning_rate": 3.45224559408755e-05,
+      "loss": 5.2518,
+      "step": 5945
+    },
+    {
+      "epoch": 0.3289110005527916,
+      "grad_norm": 3.4692533016204834,
+      "learning_rate": 3.4508243320068225e-05,
+      "loss": 5.7448,
+      "step": 5950
+    },
+    {
+      "epoch": 0.32918739635157546,
+      "grad_norm": 2.769881010055542,
+      "learning_rate": 3.449403069926094e-05,
+      "loss": 5.5949,
+      "step": 5955
+    },
+    {
+      "epoch": 0.3294637921503593,
+      "grad_norm": 2.9238779544830322,
+      "learning_rate": 3.4479818078453666e-05,
+      "loss": 5.495,
+      "step": 5960
+    },
+    {
+      "epoch": 0.32974018794914317,
+      "grad_norm": 3.0299391746520996,
+      "learning_rate": 3.44656054576464e-05,
+      "loss": 5.2859,
+      "step": 5965
+    },
+    {
+      "epoch": 0.33001658374792703,
+      "grad_norm": 3.4507875442504883,
+      "learning_rate": 3.4451392836839114e-05,
+      "loss": 5.1926,
+      "step": 5970
+    },
+    {
+      "epoch": 0.3302929795467109,
+      "grad_norm": 3.147733688354492,
+      "learning_rate": 3.443718021603184e-05,
+      "loss": 5.4821,
+      "step": 5975
+    },
+    {
+      "epoch": 0.33056937534549474,
+      "grad_norm": 2.713547945022583,
+      "learning_rate": 3.442296759522456e-05,
+      "loss": 5.3562,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3308457711442786,
+      "grad_norm": 3.38211727142334,
+      "learning_rate": 3.4408754974417286e-05,
+      "loss": 5.69,
+      "step": 5985
+    },
+    {
+      "epoch": 0.33112216694306246,
+      "grad_norm": 3.258375644683838,
+      "learning_rate": 3.439454235361001e-05,
+      "loss": 5.45,
+      "step": 5990
+    },
+    {
+      "epoch": 0.3313985627418463,
+      "grad_norm": 3.048330307006836,
+      "learning_rate": 3.438032973280273e-05,
+      "loss": 5.2408,
+      "step": 5995
+    },
+    {
+      "epoch": 0.33167495854063017,
+      "grad_norm": 3.6864914894104004,
+      "learning_rate": 3.436611711199546e-05,
+      "loss": 5.5333,
+      "step": 6000
+    },
+    {
+      "epoch": 0.33195135433941403,
+      "grad_norm": 3.327286958694458,
+      "learning_rate": 3.4351904491188175e-05,
+      "loss": 5.4176,
+      "step": 6005
+    },
+    {
+      "epoch": 0.3322277501381979,
+      "grad_norm": 2.82293438911438,
+      "learning_rate": 3.43376918703809e-05,
+      "loss": 5.7049,
+      "step": 6010
+    },
+    {
+      "epoch": 0.33250414593698174,
+      "grad_norm": 2.9626595973968506,
+      "learning_rate": 3.432347924957362e-05,
+      "loss": 5.7763,
+      "step": 6015
+    },
+    {
+      "epoch": 0.3327805417357656,
+      "grad_norm": 3.8101632595062256,
+      "learning_rate": 3.430926662876635e-05,
+      "loss": 5.1365,
+      "step": 6020
+    },
+    {
+      "epoch": 0.33305693753454946,
+      "grad_norm": 3.6021718978881836,
+      "learning_rate": 3.429505400795907e-05,
+      "loss": 5.3274,
+      "step": 6025
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 2.758277654647827,
+      "learning_rate": 3.428084138715179e-05,
+      "loss": 5.4596,
+      "step": 6030
+    },
+    {
+      "epoch": 0.33360972913211717,
+      "grad_norm": 3.2654664516448975,
+      "learning_rate": 3.426662876634452e-05,
+      "loss": 5.4182,
+      "step": 6035
+    },
+    {
+      "epoch": 0.33388612493090103,
+      "grad_norm": 3.1895666122436523,
+      "learning_rate": 3.4252416145537236e-05,
+      "loss": 5.4077,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3341625207296849,
+      "grad_norm": 2.5739054679870605,
+      "learning_rate": 3.423820352472996e-05,
+      "loss": 5.4277,
+      "step": 6045
+    },
+    {
+      "epoch": 0.33443891652846874,
+      "grad_norm": 2.6572439670562744,
+      "learning_rate": 3.4223990903922684e-05,
+      "loss": 5.7119,
+      "step": 6050
+    },
+    {
+      "epoch": 0.3347153123272526,
+      "grad_norm": 3.490938663482666,
+      "learning_rate": 3.420977828311541e-05,
+      "loss": 5.3248,
+      "step": 6055
+    },
+    {
+      "epoch": 0.33499170812603646,
+      "grad_norm": 3.159533977508545,
+      "learning_rate": 3.419556566230813e-05,
+      "loss": 5.3683,
+      "step": 6060
+    },
+    {
+      "epoch": 0.3352681039248203,
+      "grad_norm": 2.4630796909332275,
+      "learning_rate": 3.418135304150085e-05,
+      "loss": 5.6767,
+      "step": 6065
+    },
+    {
+      "epoch": 0.33554449972360423,
+      "grad_norm": 2.582796812057495,
+      "learning_rate": 3.416714042069358e-05,
+      "loss": 5.3356,
+      "step": 6070
+    },
+    {
+      "epoch": 0.3358208955223881,
+      "grad_norm": 2.7957217693328857,
+      "learning_rate": 3.4152927799886304e-05,
+      "loss": 5.3454,
+      "step": 6075
+    },
+    {
+      "epoch": 0.33609729132117194,
+      "grad_norm": 2.7028167247772217,
+      "learning_rate": 3.413871517907902e-05,
+      "loss": 5.0287,
+      "step": 6080
+    },
+    {
+      "epoch": 0.3363736871199558,
+      "grad_norm": 2.7556872367858887,
+      "learning_rate": 3.4124502558271745e-05,
+      "loss": 5.5228,
+      "step": 6085
+    },
+    {
+      "epoch": 0.33665008291873966,
+      "grad_norm": 3.9431374073028564,
+      "learning_rate": 3.411028993746447e-05,
+      "loss": 5.4459,
+      "step": 6090
+    },
+    {
+      "epoch": 0.3369264787175235,
+      "grad_norm": 3.6302506923675537,
+      "learning_rate": 3.409607731665719e-05,
+      "loss": 5.7821,
+      "step": 6095
+    },
+    {
+      "epoch": 0.33720287451630737,
+      "grad_norm": 2.845989227294922,
+      "learning_rate": 3.408186469584992e-05,
+      "loss": 5.214,
+      "step": 6100
+    },
+    {
+      "epoch": 0.33747927031509123,
+      "grad_norm": 2.877218008041382,
+      "learning_rate": 3.406765207504264e-05,
+      "loss": 5.2843,
+      "step": 6105
+    },
+    {
+      "epoch": 0.3377556661138751,
+      "grad_norm": 2.90212082862854,
+      "learning_rate": 3.4053439454235365e-05,
+      "loss": 5.1521,
+      "step": 6110
+    },
+    {
+      "epoch": 0.33803206191265894,
+      "grad_norm": 3.701557159423828,
+      "learning_rate": 3.403922683342808e-05,
+      "loss": 5.4906,
+      "step": 6115
+    },
+    {
+      "epoch": 0.3383084577114428,
+      "grad_norm": 2.650724411010742,
+      "learning_rate": 3.4025014212620806e-05,
+      "loss": 5.339,
+      "step": 6120
+    },
+    {
+      "epoch": 0.33858485351022666,
+      "grad_norm": 3.82738995552063,
+      "learning_rate": 3.401080159181354e-05,
+      "loss": 5.489,
+      "step": 6125
+    },
+    {
+      "epoch": 0.3388612493090105,
+      "grad_norm": 3.1155989170074463,
+      "learning_rate": 3.3996588971006254e-05,
+      "loss": 4.9967,
+      "step": 6130
+    },
+    {
+      "epoch": 0.33913764510779437,
+      "grad_norm": 2.9814980030059814,
+      "learning_rate": 3.398237635019898e-05,
+      "loss": 5.3042,
+      "step": 6135
+    },
+    {
+      "epoch": 0.33941404090657823,
+      "grad_norm": 4.761054515838623,
+      "learning_rate": 3.39681637293917e-05,
+      "loss": 5.4005,
+      "step": 6140
+    },
+    {
+      "epoch": 0.3396904367053621,
+      "grad_norm": 2.9018607139587402,
+      "learning_rate": 3.3953951108584426e-05,
+      "loss": 5.2394,
+      "step": 6145
+    },
+    {
+      "epoch": 0.33996683250414594,
+      "grad_norm": 2.923269748687744,
+      "learning_rate": 3.3939738487777143e-05,
+      "loss": 5.3934,
+      "step": 6150
+    },
+    {
+      "epoch": 0.3402432283029298,
+      "grad_norm": 3.082612991333008,
+      "learning_rate": 3.392552586696987e-05,
+      "loss": 5.3588,
+      "step": 6155
+    },
+    {
+      "epoch": 0.34051962410171366,
+      "grad_norm": 2.4649646282196045,
+      "learning_rate": 3.39113132461626e-05,
+      "loss": 4.7427,
+      "step": 6160
+    },
+    {
+      "epoch": 0.3407960199004975,
+      "grad_norm": 4.229916095733643,
+      "learning_rate": 3.3897100625355315e-05,
+      "loss": 5.5868,
+      "step": 6165
+    },
+    {
+      "epoch": 0.3410724156992814,
+      "grad_norm": 2.851255416870117,
+      "learning_rate": 3.388288800454804e-05,
+      "loss": 5.097,
+      "step": 6170
+    },
+    {
+      "epoch": 0.34134881149806523,
+      "grad_norm": 3.1447596549987793,
+      "learning_rate": 3.386867538374076e-05,
+      "loss": 5.5086,
+      "step": 6175
+    },
+    {
+      "epoch": 0.3416252072968491,
+      "grad_norm": 3.191239833831787,
+      "learning_rate": 3.385446276293349e-05,
+      "loss": 5.7917,
+      "step": 6180
+    },
+    {
+      "epoch": 0.34190160309563294,
+      "grad_norm": 3.4441893100738525,
+      "learning_rate": 3.384025014212621e-05,
+      "loss": 5.2703,
+      "step": 6185
+    },
+    {
+      "epoch": 0.3421779988944168,
+      "grad_norm": 2.9748077392578125,
+      "learning_rate": 3.382603752131893e-05,
+      "loss": 5.4914,
+      "step": 6190
+    },
+    {
+      "epoch": 0.34245439469320066,
+      "grad_norm": 2.859182834625244,
+      "learning_rate": 3.381182490051166e-05,
+      "loss": 5.1884,
+      "step": 6195
+    },
+    {
+      "epoch": 0.3427307904919845,
+      "grad_norm": 3.529243230819702,
+      "learning_rate": 3.3797612279704376e-05,
+      "loss": 5.2572,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3430071862907684,
+      "grad_norm": 3.0292012691497803,
+      "learning_rate": 3.37833996588971e-05,
+      "loss": 5.4534,
+      "step": 6205
+    },
+    {
+      "epoch": 0.34328358208955223,
+      "grad_norm": 2.9836554527282715,
+      "learning_rate": 3.376918703808983e-05,
+      "loss": 5.3841,
+      "step": 6210
+    },
+    {
+      "epoch": 0.3435599778883361,
+      "grad_norm": 2.9530041217803955,
+      "learning_rate": 3.375497441728255e-05,
+      "loss": 5.4718,
+      "step": 6215
+    },
+    {
+      "epoch": 0.34383637368711994,
+      "grad_norm": 4.38737678527832,
+      "learning_rate": 3.374076179647527e-05,
+      "loss": 5.5249,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3441127694859038,
+      "grad_norm": 2.9150431156158447,
+      "learning_rate": 3.3726549175667996e-05,
+      "loss": 5.4543,
+      "step": 6225
+    },
+    {
+      "epoch": 0.34438916528468766,
+      "grad_norm": 2.641899824142456,
+      "learning_rate": 3.371233655486072e-05,
+      "loss": 5.3053,
+      "step": 6230
+    },
+    {
+      "epoch": 0.3446655610834715,
+      "grad_norm": 2.8919003009796143,
+      "learning_rate": 3.3698123934053444e-05,
+      "loss": 5.5736,
+      "step": 6235
+    },
+    {
+      "epoch": 0.3449419568822554,
+      "grad_norm": 4.31245756149292,
+      "learning_rate": 3.368391131324616e-05,
+      "loss": 5.4375,
+      "step": 6240
+    },
+    {
+      "epoch": 0.34521835268103923,
+      "grad_norm": 2.8349204063415527,
+      "learning_rate": 3.366969869243889e-05,
+      "loss": 5.4986,
+      "step": 6245
+    },
+    {
+      "epoch": 0.3454947484798231,
+      "grad_norm": 3.073392629623413,
+      "learning_rate": 3.365548607163161e-05,
+      "loss": 5.345,
+      "step": 6250
+    },
+    {
+      "epoch": 0.34577114427860695,
+      "grad_norm": 3.0366547107696533,
+      "learning_rate": 3.3641273450824333e-05,
+      "loss": 5.484,
+      "step": 6255
+    },
+    {
+      "epoch": 0.3460475400773908,
+      "grad_norm": 2.6022896766662598,
+      "learning_rate": 3.362706083001706e-05,
+      "loss": 5.1542,
+      "step": 6260
+    },
+    {
+      "epoch": 0.34632393587617466,
+      "grad_norm": 2.9307637214660645,
+      "learning_rate": 3.361284820920978e-05,
+      "loss": 5.3493,
+      "step": 6265
+    },
+    {
+      "epoch": 0.3466003316749585,
+      "grad_norm": 3.0827174186706543,
+      "learning_rate": 3.3598635588402505e-05,
+      "loss": 5.039,
+      "step": 6270
+    },
+    {
+      "epoch": 0.3468767274737424,
+      "grad_norm": 3.8074021339416504,
+      "learning_rate": 3.358442296759522e-05,
+      "loss": 5.382,
+      "step": 6275
+    },
+    {
+      "epoch": 0.34715312327252623,
+      "grad_norm": 3.806523561477661,
+      "learning_rate": 3.357021034678795e-05,
+      "loss": 5.5725,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3474295190713101,
+      "grad_norm": 2.643564224243164,
+      "learning_rate": 3.355599772598067e-05,
+      "loss": 5.5118,
+      "step": 6285
+    },
+    {
+      "epoch": 0.347705914870094,
+      "grad_norm": 3.194028377532959,
+      "learning_rate": 3.3541785105173395e-05,
+      "loss": 5.5817,
+      "step": 6290
+    },
+    {
+      "epoch": 0.34798231066887786,
+      "grad_norm": 2.6427788734436035,
+      "learning_rate": 3.352757248436612e-05,
+      "loss": 5.3618,
+      "step": 6295
+    },
+    {
+      "epoch": 0.3482587064676617,
+      "grad_norm": 2.87471604347229,
+      "learning_rate": 3.351335986355884e-05,
+      "loss": 5.3261,
+      "step": 6300
+    },
+    {
+      "epoch": 0.3485351022664456,
+      "grad_norm": 3.181290864944458,
+      "learning_rate": 3.3499147242751566e-05,
+      "loss": 5.2144,
+      "step": 6305
+    },
+    {
+      "epoch": 0.34881149806522943,
+      "grad_norm": 3.502073049545288,
+      "learning_rate": 3.3484934621944284e-05,
+      "loss": 5.4402,
+      "step": 6310
+    },
+    {
+      "epoch": 0.3490878938640133,
+      "grad_norm": 3.2654974460601807,
+      "learning_rate": 3.3470722001137014e-05,
+      "loss": 5.5757,
+      "step": 6315
+    },
+    {
+      "epoch": 0.34936428966279715,
+      "grad_norm": 2.6679444313049316,
+      "learning_rate": 3.345650938032974e-05,
+      "loss": 5.2295,
+      "step": 6320
+    },
+    {
+      "epoch": 0.349640685461581,
+      "grad_norm": 3.1147098541259766,
+      "learning_rate": 3.3442296759522456e-05,
+      "loss": 5.3529,
+      "step": 6325
+    },
+    {
+      "epoch": 0.34991708126036486,
+      "grad_norm": 2.5592527389526367,
+      "learning_rate": 3.342808413871518e-05,
+      "loss": 5.1588,
+      "step": 6330
+    },
+    {
+      "epoch": 0.3501934770591487,
+      "grad_norm": 2.633533000946045,
+      "learning_rate": 3.3413871517907904e-05,
+      "loss": 5.4131,
+      "step": 6335
+    },
+    {
+      "epoch": 0.3504698728579326,
+      "grad_norm": 3.3299381732940674,
+      "learning_rate": 3.339965889710063e-05,
+      "loss": 5.1725,
+      "step": 6340
+    },
+    {
+      "epoch": 0.35074626865671643,
+      "grad_norm": 2.768136978149414,
+      "learning_rate": 3.338544627629335e-05,
+      "loss": 5.6988,
+      "step": 6345
+    },
+    {
+      "epoch": 0.3510226644555003,
+      "grad_norm": 3.701432228088379,
+      "learning_rate": 3.3371233655486076e-05,
+      "loss": 5.5261,
+      "step": 6350
+    },
+    {
+      "epoch": 0.35129906025428415,
+      "grad_norm": 4.025837421417236,
+      "learning_rate": 3.33570210346788e-05,
+      "loss": 5.1812,
+      "step": 6355
+    },
+    {
+      "epoch": 0.351575456053068,
+      "grad_norm": 2.9328114986419678,
+      "learning_rate": 3.334280841387152e-05,
+      "loss": 5.3862,
+      "step": 6360
+    },
+    {
+      "epoch": 0.35185185185185186,
+      "grad_norm": 2.7843148708343506,
+      "learning_rate": 3.332859579306424e-05,
+      "loss": 5.6493,
+      "step": 6365
+    },
+    {
+      "epoch": 0.3521282476506357,
+      "grad_norm": 3.647606134414673,
+      "learning_rate": 3.3314383172256965e-05,
+      "loss": 5.7016,
+      "step": 6370
+    },
+    {
+      "epoch": 0.3524046434494196,
+      "grad_norm": 2.85952091217041,
+      "learning_rate": 3.330017055144969e-05,
+      "loss": 5.287,
+      "step": 6375
+    },
+    {
+      "epoch": 0.35268103924820343,
+      "grad_norm": 2.931126356124878,
+      "learning_rate": 3.328595793064241e-05,
+      "loss": 5.2523,
+      "step": 6380
+    },
+    {
+      "epoch": 0.3529574350469873,
+      "grad_norm": 3.174170970916748,
+      "learning_rate": 3.3271745309835137e-05,
+      "loss": 5.1911,
+      "step": 6385
+    },
+    {
+      "epoch": 0.35323383084577115,
+      "grad_norm": 3.573742389678955,
+      "learning_rate": 3.325753268902786e-05,
+      "loss": 4.888,
+      "step": 6390
+    },
+    {
+      "epoch": 0.353510226644555,
+      "grad_norm": 2.854358673095703,
+      "learning_rate": 3.324332006822058e-05,
+      "loss": 5.3531,
+      "step": 6395
+    },
+    {
+      "epoch": 0.35378662244333886,
+      "grad_norm": 3.4788529872894287,
+      "learning_rate": 3.32291074474133e-05,
+      "loss": 5.5047,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3540630182421227,
+      "grad_norm": 3.1905970573425293,
+      "learning_rate": 3.321489482660603e-05,
+      "loss": 5.311,
+      "step": 6405
+    },
+    {
+      "epoch": 0.3543394140409066,
+      "grad_norm": 3.261735439300537,
+      "learning_rate": 3.320068220579875e-05,
+      "loss": 5.3405,
+      "step": 6410
+    },
+    {
+      "epoch": 0.35461580983969043,
+      "grad_norm": 3.806363344192505,
+      "learning_rate": 3.3186469584991474e-05,
+      "loss": 5.2041,
+      "step": 6415
+    },
+    {
+      "epoch": 0.3548922056384743,
+      "grad_norm": 3.2285566329956055,
+      "learning_rate": 3.31722569641842e-05,
+      "loss": 5.2239,
+      "step": 6420
+    },
+    {
+      "epoch": 0.35516860143725815,
+      "grad_norm": 3.194967269897461,
+      "learning_rate": 3.315804434337692e-05,
+      "loss": 5.4333,
+      "step": 6425
+    },
+    {
+      "epoch": 0.355444997236042,
+      "grad_norm": 3.9801723957061768,
+      "learning_rate": 3.3143831722569646e-05,
+      "loss": 5.4375,
+      "step": 6430
+    },
+    {
+      "epoch": 0.35572139303482586,
+      "grad_norm": 2.690793514251709,
+      "learning_rate": 3.312961910176236e-05,
+      "loss": 5.3818,
+      "step": 6435
+    },
+    {
+      "epoch": 0.3559977888336097,
+      "grad_norm": 3.140059232711792,
+      "learning_rate": 3.3115406480955094e-05,
+      "loss": 5.2608,
+      "step": 6440
+    },
+    {
+      "epoch": 0.3562741846323936,
+      "grad_norm": 3.5246164798736572,
+      "learning_rate": 3.310119386014781e-05,
+      "loss": 5.4467,
+      "step": 6445
+    },
+    {
+      "epoch": 0.35655058043117743,
+      "grad_norm": 3.004757881164551,
+      "learning_rate": 3.3086981239340535e-05,
+      "loss": 5.6051,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3568269762299613,
+      "grad_norm": 2.6584270000457764,
+      "learning_rate": 3.307276861853326e-05,
+      "loss": 5.4181,
+      "step": 6455
+    },
+    {
+      "epoch": 0.35710337202874515,
+      "grad_norm": 2.8345117568969727,
+      "learning_rate": 3.305855599772598e-05,
+      "loss": 5.5933,
+      "step": 6460
+    },
+    {
+      "epoch": 0.357379767827529,
+      "grad_norm": 2.8853676319122314,
+      "learning_rate": 3.304434337691871e-05,
+      "loss": 5.1924,
+      "step": 6465
+    },
+    {
+      "epoch": 0.35765616362631286,
+      "grad_norm": 2.9934706687927246,
+      "learning_rate": 3.3030130756111424e-05,
+      "loss": 5.5229,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3579325594250967,
+      "grad_norm": 3.472885847091675,
+      "learning_rate": 3.3015918135304155e-05,
+      "loss": 5.6357,
+      "step": 6475
+    },
+    {
+      "epoch": 0.3582089552238806,
+      "grad_norm": 3.2652275562286377,
+      "learning_rate": 3.300170551449687e-05,
+      "loss": 5.2829,
+      "step": 6480
+    },
+    {
+      "epoch": 0.35848535102266443,
+      "grad_norm": 3.120035171508789,
+      "learning_rate": 3.2987492893689596e-05,
+      "loss": 5.3109,
+      "step": 6485
+    },
+    {
+      "epoch": 0.3587617468214483,
+      "grad_norm": 2.817810297012329,
+      "learning_rate": 3.297328027288232e-05,
+      "loss": 5.243,
+      "step": 6490
+    },
+    {
+      "epoch": 0.35903814262023215,
+      "grad_norm": 2.679710626602173,
+      "learning_rate": 3.2959067652075044e-05,
+      "loss": 5.6024,
+      "step": 6495
+    },
+    {
+      "epoch": 0.359314538419016,
+      "grad_norm": 3.362010955810547,
+      "learning_rate": 3.294485503126777e-05,
+      "loss": 5.3039,
+      "step": 6500
+    },
+    {
+      "epoch": 0.35959093421779986,
+      "grad_norm": 3.8571290969848633,
+      "learning_rate": 3.2930642410460485e-05,
+      "loss": 5.4973,
+      "step": 6505
+    },
+    {
+      "epoch": 0.3598673300165838,
+      "grad_norm": 3.4675230979919434,
+      "learning_rate": 3.2916429789653216e-05,
+      "loss": 5.2216,
+      "step": 6510
+    },
+    {
+      "epoch": 0.36014372581536763,
+      "grad_norm": 2.4890525341033936,
+      "learning_rate": 3.290221716884594e-05,
+      "loss": 5.2977,
+      "step": 6515
+    },
+    {
+      "epoch": 0.3604201216141515,
+      "grad_norm": 4.0375518798828125,
+      "learning_rate": 3.288800454803866e-05,
+      "loss": 5.5202,
+      "step": 6520
+    },
+    {
+      "epoch": 0.36069651741293535,
+      "grad_norm": 3.2754342555999756,
+      "learning_rate": 3.287379192723139e-05,
+      "loss": 5.082,
+      "step": 6525
+    },
+    {
+      "epoch": 0.3609729132117192,
+      "grad_norm": 2.7093911170959473,
+      "learning_rate": 3.2859579306424105e-05,
+      "loss": 5.4033,
+      "step": 6530
+    },
+    {
+      "epoch": 0.36124930901050306,
+      "grad_norm": 4.006176948547363,
+      "learning_rate": 3.284536668561683e-05,
+      "loss": 5.5329,
+      "step": 6535
+    },
+    {
+      "epoch": 0.3615257048092869,
+      "grad_norm": 4.503550052642822,
+      "learning_rate": 3.283115406480955e-05,
+      "loss": 5.3618,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3618021006080708,
+      "grad_norm": 3.1193487644195557,
+      "learning_rate": 3.281694144400228e-05,
+      "loss": 5.3315,
+      "step": 6545
+    },
+    {
+      "epoch": 0.36207849640685463,
+      "grad_norm": 2.99080228805542,
+      "learning_rate": 3.2802728823195e-05,
+      "loss": 5.4393,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3623548922056385,
+      "grad_norm": 3.051684617996216,
+      "learning_rate": 3.278851620238772e-05,
+      "loss": 5.2381,
+      "step": 6555
+    },
+    {
+      "epoch": 0.36263128800442235,
+      "grad_norm": 3.2539477348327637,
+      "learning_rate": 3.277430358158045e-05,
+      "loss": 5.1091,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3629076838032062,
+      "grad_norm": 2.8935561180114746,
+      "learning_rate": 3.2760090960773166e-05,
+      "loss": 5.3882,
+      "step": 6565
+    },
+    {
+      "epoch": 0.36318407960199006,
+      "grad_norm": 3.101651430130005,
+      "learning_rate": 3.274587833996589e-05,
+      "loss": 5.5792,
+      "step": 6570
+    },
+    {
+      "epoch": 0.3634604754007739,
+      "grad_norm": 2.6292476654052734,
+      "learning_rate": 3.2731665719158614e-05,
+      "loss": 5.1631,
+      "step": 6575
+    },
+    {
+      "epoch": 0.3637368711995578,
+      "grad_norm": 3.1118204593658447,
+      "learning_rate": 3.271745309835134e-05,
+      "loss": 5.7077,
+      "step": 6580
+    },
+    {
+      "epoch": 0.36401326699834163,
+      "grad_norm": 3.469219207763672,
+      "learning_rate": 3.270324047754406e-05,
+      "loss": 5.4348,
+      "step": 6585
+    },
+    {
+      "epoch": 0.3642896627971255,
+      "grad_norm": 3.5681636333465576,
+      "learning_rate": 3.268902785673678e-05,
+      "loss": 5.1626,
+      "step": 6590
+    },
+    {
+      "epoch": 0.36456605859590935,
+      "grad_norm": 3.1905977725982666,
+      "learning_rate": 3.267481523592951e-05,
+      "loss": 5.0887,
+      "step": 6595
+    },
+    {
+      "epoch": 0.3648424543946932,
+      "grad_norm": 3.0895135402679443,
+      "learning_rate": 3.2660602615122234e-05,
+      "loss": 5.3645,
+      "step": 6600
+    },
+    {
+      "epoch": 0.36511885019347706,
+      "grad_norm": 2.88983154296875,
+      "learning_rate": 3.264638999431495e-05,
+      "loss": 5.0051,
+      "step": 6605
+    },
+    {
+      "epoch": 0.3653952459922609,
+      "grad_norm": 2.7491235733032227,
+      "learning_rate": 3.2632177373507675e-05,
+      "loss": 5.4856,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3656716417910448,
+      "grad_norm": 2.5615084171295166,
+      "learning_rate": 3.26179647527004e-05,
+      "loss": 5.5509,
+      "step": 6615
+    },
+    {
+      "epoch": 0.36594803758982863,
+      "grad_norm": 3.6004726886749268,
+      "learning_rate": 3.260375213189312e-05,
+      "loss": 4.955,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3662244333886125,
+      "grad_norm": 3.0466911792755127,
+      "learning_rate": 3.258953951108585e-05,
+      "loss": 5.3448,
+      "step": 6625
+    },
+    {
+      "epoch": 0.36650082918739635,
+      "grad_norm": 3.489597797393799,
+      "learning_rate": 3.257532689027857e-05,
+      "loss": 5.239,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3667772249861802,
+      "grad_norm": 3.6642837524414062,
+      "learning_rate": 3.2561114269471295e-05,
+      "loss": 5.7073,
+      "step": 6635
+    },
+    {
+      "epoch": 0.36705362078496406,
+      "grad_norm": 3.3257694244384766,
+      "learning_rate": 3.254690164866401e-05,
+      "loss": 5.5338,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3673300165837479,
+      "grad_norm": 2.889758825302124,
+      "learning_rate": 3.2532689027856736e-05,
+      "loss": 5.4763,
+      "step": 6645
+    },
+    {
+      "epoch": 0.3676064123825318,
+      "grad_norm": 3.0709822177886963,
+      "learning_rate": 3.251847640704947e-05,
+      "loss": 5.4343,
+      "step": 6650
+    },
+    {
+      "epoch": 0.36788280818131563,
+      "grad_norm": 3.2317843437194824,
+      "learning_rate": 3.2504263786242184e-05,
+      "loss": 5.48,
+      "step": 6655
+    },
+    {
+      "epoch": 0.3681592039800995,
+      "grad_norm": 2.843003034591675,
+      "learning_rate": 3.249005116543491e-05,
+      "loss": 5.3191,
+      "step": 6660
+    },
+    {
+      "epoch": 0.36843559977888335,
+      "grad_norm": 2.82918119430542,
+      "learning_rate": 3.247583854462763e-05,
+      "loss": 5.2936,
+      "step": 6665
+    },
+    {
+      "epoch": 0.3687119955776672,
+      "grad_norm": 3.3583202362060547,
+      "learning_rate": 3.2461625923820356e-05,
+      "loss": 5.5645,
+      "step": 6670
+    },
+    {
+      "epoch": 0.36898839137645106,
+      "grad_norm": 3.859200954437256,
+      "learning_rate": 3.244741330301307e-05,
+      "loss": 5.3669,
+      "step": 6675
+    },
+    {
+      "epoch": 0.3692647871752349,
+      "grad_norm": 2.8152244091033936,
+      "learning_rate": 3.24332006822058e-05,
+      "loss": 5.3654,
+      "step": 6680
+    },
+    {
+      "epoch": 0.3695411829740188,
+      "grad_norm": 2.6449549198150635,
+      "learning_rate": 3.241898806139853e-05,
+      "loss": 5.3025,
+      "step": 6685
+    },
+    {
+      "epoch": 0.36981757877280264,
+      "grad_norm": 2.791011095046997,
+      "learning_rate": 3.2404775440591245e-05,
+      "loss": 5.2969,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3700939745715865,
+      "grad_norm": 3.3841540813446045,
+      "learning_rate": 3.239056281978397e-05,
+      "loss": 5.2648,
+      "step": 6695
+    },
+    {
+      "epoch": 0.37037037037037035,
+      "grad_norm": 2.6592512130737305,
+      "learning_rate": 3.237635019897669e-05,
+      "loss": 5.0054,
+      "step": 6700
+    },
+    {
+      "epoch": 0.3706467661691542,
+      "grad_norm": 3.7593088150024414,
+      "learning_rate": 3.236213757816942e-05,
+      "loss": 5.3497,
+      "step": 6705
+    },
+    {
+      "epoch": 0.37092316196793806,
+      "grad_norm": 3.4678800106048584,
+      "learning_rate": 3.234792495736214e-05,
+      "loss": 5.5969,
+      "step": 6710
+    },
+    {
+      "epoch": 0.3711995577667219,
+      "grad_norm": 2.172253370285034,
+      "learning_rate": 3.233371233655486e-05,
+      "loss": 5.2902,
+      "step": 6715
+    },
+    {
+      "epoch": 0.3714759535655058,
+      "grad_norm": 3.2627487182617188,
+      "learning_rate": 3.231949971574759e-05,
+      "loss": 5.5915,
+      "step": 6720
+    },
+    {
+      "epoch": 0.37175234936428964,
+      "grad_norm": 3.7664973735809326,
+      "learning_rate": 3.2305287094940306e-05,
+      "loss": 5.2726,
+      "step": 6725
+    },
+    {
+      "epoch": 0.37202874516307355,
+      "grad_norm": 4.006364345550537,
+      "learning_rate": 3.229107447413303e-05,
+      "loss": 5.1299,
+      "step": 6730
+    },
+    {
+      "epoch": 0.3723051409618574,
+      "grad_norm": 3.697758436203003,
+      "learning_rate": 3.2276861853325754e-05,
+      "loss": 5.4359,
+      "step": 6735
+    },
+    {
+      "epoch": 0.37258153676064126,
+      "grad_norm": 3.6661410331726074,
+      "learning_rate": 3.226264923251848e-05,
+      "loss": 5.3954,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3728579325594251,
+      "grad_norm": 2.565558910369873,
+      "learning_rate": 3.22484366117112e-05,
+      "loss": 5.2579,
+      "step": 6745
+    },
+    {
+      "epoch": 0.373134328358209,
+      "grad_norm": 4.1350483894348145,
+      "learning_rate": 3.223422399090392e-05,
+      "loss": 4.9195,
+      "step": 6750
+    },
+    {
+      "epoch": 0.37341072415699283,
+      "grad_norm": 2.775804042816162,
+      "learning_rate": 3.222001137009665e-05,
+      "loss": 5.2634,
+      "step": 6755
+    },
+    {
+      "epoch": 0.3736871199557767,
+      "grad_norm": 3.0823562145233154,
+      "learning_rate": 3.2205798749289374e-05,
+      "loss": 5.2957,
+      "step": 6760
+    },
+    {
+      "epoch": 0.37396351575456055,
+      "grad_norm": 2.7218778133392334,
+      "learning_rate": 3.219158612848209e-05,
+      "loss": 5.2254,
+      "step": 6765
+    },
+    {
+      "epoch": 0.3742399115533444,
+      "grad_norm": 2.774827718734741,
+      "learning_rate": 3.2177373507674815e-05,
+      "loss": 5.5636,
+      "step": 6770
+    },
+    {
+      "epoch": 0.37451630735212826,
+      "grad_norm": 3.3417789936065674,
+      "learning_rate": 3.216316088686754e-05,
+      "loss": 5.0876,
+      "step": 6775
+    },
+    {
+      "epoch": 0.3747927031509121,
+      "grad_norm": 2.745445489883423,
+      "learning_rate": 3.214894826606026e-05,
+      "loss": 5.5064,
+      "step": 6780
+    },
+    {
+      "epoch": 0.375069098949696,
+      "grad_norm": 3.0469281673431396,
+      "learning_rate": 3.213473564525298e-05,
+      "loss": 5.238,
+      "step": 6785
+    },
+    {
+      "epoch": 0.37534549474847984,
+      "grad_norm": 3.1263833045959473,
+      "learning_rate": 3.212052302444571e-05,
+      "loss": 5.1806,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3756218905472637,
+      "grad_norm": 3.453270435333252,
+      "learning_rate": 3.2106310403638435e-05,
+      "loss": 5.2468,
+      "step": 6795
+    },
+    {
+      "epoch": 0.37589828634604755,
+      "grad_norm": 4.632005214691162,
+      "learning_rate": 3.209209778283115e-05,
+      "loss": 5.4471,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3761746821448314,
+      "grad_norm": 3.877882242202759,
+      "learning_rate": 3.2077885162023876e-05,
+      "loss": 5.5113,
+      "step": 6805
+    },
+    {
+      "epoch": 0.37645107794361526,
+      "grad_norm": 3.9937760829925537,
+      "learning_rate": 3.20636725412166e-05,
+      "loss": 5.4494,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3767274737423991,
+      "grad_norm": 2.7250454425811768,
+      "learning_rate": 3.2049459920409324e-05,
+      "loss": 5.1192,
+      "step": 6815
+    },
+    {
+      "epoch": 0.377003869541183,
+      "grad_norm": 2.984262466430664,
+      "learning_rate": 3.203524729960205e-05,
+      "loss": 5.1072,
+      "step": 6820
+    },
+    {
+      "epoch": 0.37728026533996684,
+      "grad_norm": 3.4100584983825684,
+      "learning_rate": 3.202103467879477e-05,
+      "loss": 5.2542,
+      "step": 6825
+    },
+    {
+      "epoch": 0.3775566611387507,
+      "grad_norm": 3.297354221343994,
+      "learning_rate": 3.2006822057987496e-05,
+      "loss": 5.2157,
+      "step": 6830
+    },
+    {
+      "epoch": 0.37783305693753455,
+      "grad_norm": 3.0777430534362793,
+      "learning_rate": 3.1992609437180213e-05,
+      "loss": 5.516,
+      "step": 6835
+    },
+    {
+      "epoch": 0.3781094527363184,
+      "grad_norm": 3.4484028816223145,
+      "learning_rate": 3.1978396816372944e-05,
+      "loss": 5.5581,
+      "step": 6840
+    },
+    {
+      "epoch": 0.37838584853510226,
+      "grad_norm": 3.6601834297180176,
+      "learning_rate": 3.196418419556567e-05,
+      "loss": 5.2055,
+      "step": 6845
+    },
+    {
+      "epoch": 0.3786622443338861,
+      "grad_norm": 3.5895004272460938,
+      "learning_rate": 3.1949971574758385e-05,
+      "loss": 5.1609,
+      "step": 6850
+    },
+    {
+      "epoch": 0.37893864013267,
+      "grad_norm": 2.961439371109009,
+      "learning_rate": 3.193575895395111e-05,
+      "loss": 5.4183,
+      "step": 6855
+    },
+    {
+      "epoch": 0.37921503593145384,
+      "grad_norm": 4.192780494689941,
+      "learning_rate": 3.192154633314383e-05,
+      "loss": 5.4031,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3794914317302377,
+      "grad_norm": 3.4532811641693115,
+      "learning_rate": 3.190733371233656e-05,
+      "loss": 5.3026,
+      "step": 6865
+    },
+    {
+      "epoch": 0.37976782752902155,
+      "grad_norm": 3.2710492610931396,
+      "learning_rate": 3.189312109152928e-05,
+      "loss": 5.061,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3800442233278054,
+      "grad_norm": 3.0584661960601807,
+      "learning_rate": 3.1878908470722005e-05,
+      "loss": 5.2326,
+      "step": 6875
+    },
+    {
+      "epoch": 0.38032061912658927,
+      "grad_norm": 2.271395206451416,
+      "learning_rate": 3.186469584991473e-05,
+      "loss": 4.9213,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3805970149253731,
+      "grad_norm": 3.3883402347564697,
+      "learning_rate": 3.1850483229107446e-05,
+      "loss": 5.1073,
+      "step": 6885
+    },
+    {
+      "epoch": 0.380873410724157,
+      "grad_norm": 3.4375452995300293,
+      "learning_rate": 3.183627060830017e-05,
+      "loss": 5.5934,
+      "step": 6890
+    },
+    {
+      "epoch": 0.38114980652294084,
+      "grad_norm": 2.793851137161255,
+      "learning_rate": 3.1822057987492894e-05,
+      "loss": 5.0503,
+      "step": 6895
+    },
+    {
+      "epoch": 0.3814262023217247,
+      "grad_norm": 3.7272047996520996,
+      "learning_rate": 3.180784536668562e-05,
+      "loss": 5.3093,
+      "step": 6900
+    },
+    {
+      "epoch": 0.38170259812050855,
+      "grad_norm": 3.7055723667144775,
+      "learning_rate": 3.179363274587834e-05,
+      "loss": 5.3679,
+      "step": 6905
+    },
+    {
+      "epoch": 0.3819789939192924,
+      "grad_norm": 5.639072418212891,
+      "learning_rate": 3.1779420125071066e-05,
+      "loss": 5.4524,
+      "step": 6910
+    },
+    {
+      "epoch": 0.38225538971807627,
+      "grad_norm": 2.6650073528289795,
+      "learning_rate": 3.176520750426379e-05,
+      "loss": 5.022,
+      "step": 6915
+    },
+    {
+      "epoch": 0.3825317855168601,
+      "grad_norm": 3.388301372528076,
+      "learning_rate": 3.175099488345651e-05,
+      "loss": 5.1516,
+      "step": 6920
+    },
+    {
+      "epoch": 0.382808181315644,
+      "grad_norm": 4.096010208129883,
+      "learning_rate": 3.173678226264923e-05,
+      "loss": 5.127,
+      "step": 6925
+    },
+    {
+      "epoch": 0.38308457711442784,
+      "grad_norm": 3.0948524475097656,
+      "learning_rate": 3.172256964184196e-05,
+      "loss": 5.192,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3833609729132117,
+      "grad_norm": 3.4886627197265625,
+      "learning_rate": 3.170835702103468e-05,
+      "loss": 5.19,
+      "step": 6935
+    },
+    {
+      "epoch": 0.38363736871199555,
+      "grad_norm": 3.2240042686462402,
+      "learning_rate": 3.1694144400227403e-05,
+      "loss": 5.6154,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3839137645107794,
+      "grad_norm": 2.617563247680664,
+      "learning_rate": 3.167993177942013e-05,
+      "loss": 5.3793,
+      "step": 6945
+    },
+    {
+      "epoch": 0.3841901603095633,
+      "grad_norm": 3.2528016567230225,
+      "learning_rate": 3.166571915861285e-05,
+      "loss": 5.3784,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3844665561083472,
+      "grad_norm": 3.294482707977295,
+      "learning_rate": 3.1651506537805575e-05,
+      "loss": 5.2618,
+      "step": 6955
+    },
+    {
+      "epoch": 0.38474295190713104,
+      "grad_norm": 2.874786615371704,
+      "learning_rate": 3.163729391699829e-05,
+      "loss": 5.4889,
+      "step": 6960
+    },
+    {
+      "epoch": 0.3850193477059149,
+      "grad_norm": 2.737448215484619,
+      "learning_rate": 3.162308129619102e-05,
+      "loss": 5.2753,
+      "step": 6965
+    },
+    {
+      "epoch": 0.38529574350469875,
+      "grad_norm": 3.391822338104248,
+      "learning_rate": 3.160886867538374e-05,
+      "loss": 5.3451,
+      "step": 6970
+    },
+    {
+      "epoch": 0.3855721393034826,
+      "grad_norm": 3.012768268585205,
+      "learning_rate": 3.1594656054576464e-05,
+      "loss": 5.1552,
+      "step": 6975
+    },
+    {
+      "epoch": 0.38584853510226647,
+      "grad_norm": 2.8616573810577393,
+      "learning_rate": 3.158044343376919e-05,
+      "loss": 5.3288,
+      "step": 6980
+    },
+    {
+      "epoch": 0.3861249309010503,
+      "grad_norm": 2.7854886054992676,
+      "learning_rate": 3.156623081296191e-05,
+      "loss": 5.2121,
+      "step": 6985
+    },
+    {
+      "epoch": 0.3864013266998342,
+      "grad_norm": 2.713104009628296,
+      "learning_rate": 3.1552018192154636e-05,
+      "loss": 5.1097,
+      "step": 6990
+    },
+    {
+      "epoch": 0.38667772249861804,
+      "grad_norm": 3.1725711822509766,
+      "learning_rate": 3.1537805571347354e-05,
+      "loss": 5.2764,
+      "step": 6995
+    },
+    {
+      "epoch": 0.3869541182974019,
+      "grad_norm": 2.950223684310913,
+      "learning_rate": 3.1523592950540084e-05,
+      "loss": 5.3748,
+      "step": 7000
+    },
+    {
+      "epoch": 0.38723051409618575,
+      "grad_norm": 3.557173252105713,
+      "learning_rate": 3.15093803297328e-05,
+      "loss": 5.7142,
+      "step": 7005
+    },
+    {
+      "epoch": 0.3875069098949696,
+      "grad_norm": 3.303514242172241,
+      "learning_rate": 3.1495167708925526e-05,
+      "loss": 5.4981,
+      "step": 7010
+    },
+    {
+      "epoch": 0.38778330569375347,
+      "grad_norm": 3.2154033184051514,
+      "learning_rate": 3.148095508811825e-05,
+      "loss": 5.3371,
+      "step": 7015
+    },
+    {
+      "epoch": 0.3880597014925373,
+      "grad_norm": 3.158167600631714,
+      "learning_rate": 3.1466742467310974e-05,
+      "loss": 5.2136,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3883360972913212,
+      "grad_norm": 2.4815750122070312,
+      "learning_rate": 3.14525298465037e-05,
+      "loss": 5.3322,
+      "step": 7025
+    },
+    {
+      "epoch": 0.38861249309010504,
+      "grad_norm": 2.9701693058013916,
+      "learning_rate": 3.1438317225696415e-05,
+      "loss": 5.0189,
+      "step": 7030
+    },
+    {
+      "epoch": 0.3888888888888889,
+      "grad_norm": 2.927236318588257,
+      "learning_rate": 3.1424104604889145e-05,
+      "loss": 5.224,
+      "step": 7035
+    },
+    {
+      "epoch": 0.38916528468767275,
+      "grad_norm": 3.0851492881774902,
+      "learning_rate": 3.140989198408187e-05,
+      "loss": 5.1077,
+      "step": 7040
+    },
+    {
+      "epoch": 0.3894416804864566,
+      "grad_norm": 3.0460946559906006,
+      "learning_rate": 3.139567936327459e-05,
+      "loss": 4.9981,
+      "step": 7045
+    },
+    {
+      "epoch": 0.38971807628524047,
+      "grad_norm": 3.0376229286193848,
+      "learning_rate": 3.138146674246731e-05,
+      "loss": 5.4228,
+      "step": 7050
+    },
+    {
+      "epoch": 0.3899944720840243,
+      "grad_norm": 3.2579638957977295,
+      "learning_rate": 3.1367254121660035e-05,
+      "loss": 5.3469,
+      "step": 7055
+    },
+    {
+      "epoch": 0.3902708678828082,
+      "grad_norm": 3.601778984069824,
+      "learning_rate": 3.135304150085276e-05,
+      "loss": 5.3353,
+      "step": 7060
+    },
+    {
+      "epoch": 0.39054726368159204,
+      "grad_norm": 3.496253252029419,
+      "learning_rate": 3.133882888004548e-05,
+      "loss": 5.2927,
+      "step": 7065
+    },
+    {
+      "epoch": 0.3908236594803759,
+      "grad_norm": 2.954893112182617,
+      "learning_rate": 3.1324616259238207e-05,
+      "loss": 5.1752,
+      "step": 7070
+    },
+    {
+      "epoch": 0.39110005527915975,
+      "grad_norm": 3.126713752746582,
+      "learning_rate": 3.131040363843093e-05,
+      "loss": 5.1443,
+      "step": 7075
+    },
+    {
+      "epoch": 0.3913764510779436,
+      "grad_norm": 3.489055633544922,
+      "learning_rate": 3.129619101762365e-05,
+      "loss": 5.1166,
+      "step": 7080
+    },
+    {
+      "epoch": 0.39165284687672747,
+      "grad_norm": 2.989750623703003,
+      "learning_rate": 3.128197839681637e-05,
+      "loss": 5.2128,
+      "step": 7085
+    },
+    {
+      "epoch": 0.3919292426755113,
+      "grad_norm": 3.5772876739501953,
+      "learning_rate": 3.12677657760091e-05,
+      "loss": 5.371,
+      "step": 7090
+    },
+    {
+      "epoch": 0.3922056384742952,
+      "grad_norm": 3.1399166584014893,
+      "learning_rate": 3.125355315520182e-05,
+      "loss": 5.0302,
+      "step": 7095
+    },
+    {
+      "epoch": 0.39248203427307904,
+      "grad_norm": 3.717059373855591,
+      "learning_rate": 3.1239340534394544e-05,
+      "loss": 5.1814,
+      "step": 7100
+    },
+    {
+      "epoch": 0.3927584300718629,
+      "grad_norm": 3.603480815887451,
+      "learning_rate": 3.122512791358727e-05,
+      "loss": 5.2066,
+      "step": 7105
+    },
+    {
+      "epoch": 0.39303482587064675,
+      "grad_norm": 2.86629319190979,
+      "learning_rate": 3.121091529277999e-05,
+      "loss": 5.0585,
+      "step": 7110
+    },
+    {
+      "epoch": 0.3933112216694306,
+      "grad_norm": 3.81514573097229,
+      "learning_rate": 3.119670267197271e-05,
+      "loss": 5.4419,
+      "step": 7115
+    },
+    {
+      "epoch": 0.39358761746821447,
+      "grad_norm": 3.7256438732147217,
+      "learning_rate": 3.118249005116543e-05,
+      "loss": 5.4916,
+      "step": 7120
+    },
+    {
+      "epoch": 0.3938640132669983,
+      "grad_norm": 2.9998602867126465,
+      "learning_rate": 3.1168277430358164e-05,
+      "loss": 5.3949,
+      "step": 7125
+    },
+    {
+      "epoch": 0.3941404090657822,
+      "grad_norm": 3.518113613128662,
+      "learning_rate": 3.115406480955088e-05,
+      "loss": 5.4088,
+      "step": 7130
+    },
+    {
+      "epoch": 0.39441680486456604,
+      "grad_norm": 2.6121208667755127,
+      "learning_rate": 3.1139852188743605e-05,
+      "loss": 5.4199,
+      "step": 7135
+    },
+    {
+      "epoch": 0.3946932006633499,
+      "grad_norm": 3.279484510421753,
+      "learning_rate": 3.112563956793633e-05,
+      "loss": 5.1075,
+      "step": 7140
+    },
+    {
+      "epoch": 0.39496959646213375,
+      "grad_norm": 3.5185775756835938,
+      "learning_rate": 3.111142694712905e-05,
+      "loss": 5.2746,
+      "step": 7145
+    },
+    {
+      "epoch": 0.3952459922609176,
+      "grad_norm": 3.7516870498657227,
+      "learning_rate": 3.109721432632178e-05,
+      "loss": 5.1902,
+      "step": 7150
+    },
+    {
+      "epoch": 0.39552238805970147,
+      "grad_norm": 3.0849173069000244,
+      "learning_rate": 3.10830017055145e-05,
+      "loss": 5.1215,
+      "step": 7155
+    },
+    {
+      "epoch": 0.3957987838584853,
+      "grad_norm": 2.995088577270508,
+      "learning_rate": 3.1068789084707225e-05,
+      "loss": 5.2936,
+      "step": 7160
+    },
+    {
+      "epoch": 0.3960751796572692,
+      "grad_norm": 3.2057669162750244,
+      "learning_rate": 3.105457646389994e-05,
+      "loss": 5.5044,
+      "step": 7165
+    },
+    {
+      "epoch": 0.3963515754560531,
+      "grad_norm": 4.289910793304443,
+      "learning_rate": 3.1040363843092666e-05,
+      "loss": 5.522,
+      "step": 7170
+    },
+    {
+      "epoch": 0.39662797125483695,
+      "grad_norm": 2.816082000732422,
+      "learning_rate": 3.1026151222285397e-05,
+      "loss": 5.1474,
+      "step": 7175
+    },
+    {
+      "epoch": 0.3969043670536208,
+      "grad_norm": 3.6746461391448975,
+      "learning_rate": 3.1011938601478114e-05,
+      "loss": 5.5855,
+      "step": 7180
+    },
+    {
+      "epoch": 0.39718076285240467,
+      "grad_norm": 3.0824356079101562,
+      "learning_rate": 3.099772598067084e-05,
+      "loss": 5.1413,
+      "step": 7185
+    },
+    {
+      "epoch": 0.3974571586511885,
+      "grad_norm": 3.6951475143432617,
+      "learning_rate": 3.098351335986356e-05,
+      "loss": 5.3582,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3977335544499724,
+      "grad_norm": 3.4342403411865234,
+      "learning_rate": 3.0969300739056286e-05,
+      "loss": 5.4512,
+      "step": 7195
+    },
+    {
+      "epoch": 0.39800995024875624,
+      "grad_norm": 2.806812047958374,
+      "learning_rate": 3.095508811824901e-05,
+      "loss": 4.7637,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3982863460475401,
+      "grad_norm": 3.490265130996704,
+      "learning_rate": 3.094087549744173e-05,
+      "loss": 5.6499,
+      "step": 7205
+    },
+    {
+      "epoch": 0.39856274184632395,
+      "grad_norm": 2.978156566619873,
+      "learning_rate": 3.092666287663446e-05,
+      "loss": 5.4028,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3988391376451078,
+      "grad_norm": 2.9881980419158936,
+      "learning_rate": 3.0912450255827175e-05,
+      "loss": 5.4023,
+      "step": 7215
+    },
+    {
+      "epoch": 0.39911553344389167,
+      "grad_norm": 3.111189365386963,
+      "learning_rate": 3.08982376350199e-05,
+      "loss": 5.2614,
+      "step": 7220
+    },
+    {
+      "epoch": 0.3993919292426755,
+      "grad_norm": 2.9220035076141357,
+      "learning_rate": 3.088402501421262e-05,
+      "loss": 5.1271,
+      "step": 7225
+    },
+    {
+      "epoch": 0.3996683250414594,
+      "grad_norm": 3.135206460952759,
+      "learning_rate": 3.086981239340535e-05,
+      "loss": 5.0486,
+      "step": 7230
+    },
+    {
+      "epoch": 0.39994472084024324,
+      "grad_norm": 3.383234739303589,
+      "learning_rate": 3.085559977259807e-05,
+      "loss": 4.99,
+      "step": 7235
+    },
+    {
+      "epoch": 0.4002211166390271,
+      "grad_norm": 3.352858066558838,
+      "learning_rate": 3.084138715179079e-05,
+      "loss": 5.206,
+      "step": 7240
+    },
+    {
+      "epoch": 0.40049751243781095,
+      "grad_norm": 2.6485846042633057,
+      "learning_rate": 3.082717453098352e-05,
+      "loss": 5.3789,
+      "step": 7245
+    },
+    {
+      "epoch": 0.4007739082365948,
+      "grad_norm": 2.586437702178955,
+      "learning_rate": 3.0812961910176236e-05,
+      "loss": 5.2531,
+      "step": 7250
+    },
+    {
+      "epoch": 0.40105030403537867,
+      "grad_norm": 3.3769021034240723,
+      "learning_rate": 3.079874928936896e-05,
+      "loss": 5.3068,
+      "step": 7255
+    },
+    {
+      "epoch": 0.4013266998341625,
+      "grad_norm": 2.9882419109344482,
+      "learning_rate": 3.0784536668561684e-05,
+      "loss": 5.1777,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4016030956329464,
+      "grad_norm": 3.543286085128784,
+      "learning_rate": 3.077032404775441e-05,
+      "loss": 5.4205,
+      "step": 7265
+    },
+    {
+      "epoch": 0.40187949143173024,
+      "grad_norm": 3.5167641639709473,
+      "learning_rate": 3.075611142694713e-05,
+      "loss": 5.2147,
+      "step": 7270
+    },
+    {
+      "epoch": 0.4021558872305141,
+      "grad_norm": 4.086061477661133,
+      "learning_rate": 3.074189880613985e-05,
+      "loss": 5.3399,
+      "step": 7275
+    },
+    {
+      "epoch": 0.40243228302929795,
+      "grad_norm": 2.9162538051605225,
+      "learning_rate": 3.072768618533258e-05,
+      "loss": 5.2173,
+      "step": 7280
+    },
+    {
+      "epoch": 0.4027086788280818,
+      "grad_norm": 3.2344117164611816,
+      "learning_rate": 3.0713473564525304e-05,
+      "loss": 5.446,
+      "step": 7285
+    },
+    {
+      "epoch": 0.40298507462686567,
+      "grad_norm": 2.547792673110962,
+      "learning_rate": 3.069926094371802e-05,
+      "loss": 5.2879,
+      "step": 7290
+    },
+    {
+      "epoch": 0.4032614704256495,
+      "grad_norm": 2.705897092819214,
+      "learning_rate": 3.0685048322910745e-05,
+      "loss": 5.243,
+      "step": 7295
+    },
+    {
+      "epoch": 0.4035378662244334,
+      "grad_norm": 3.052686929702759,
+      "learning_rate": 3.067083570210347e-05,
+      "loss": 5.5547,
+      "step": 7300
+    },
+    {
+      "epoch": 0.40381426202321724,
+      "grad_norm": 2.7619709968566895,
+      "learning_rate": 3.065662308129619e-05,
+      "loss": 5.1307,
+      "step": 7305
+    },
+    {
+      "epoch": 0.4040906578220011,
+      "grad_norm": 2.8688414096832275,
+      "learning_rate": 3.064241046048892e-05,
+      "loss": 5.1096,
+      "step": 7310
+    },
+    {
+      "epoch": 0.40436705362078496,
+      "grad_norm": 3.4232749938964844,
+      "learning_rate": 3.062819783968164e-05,
+      "loss": 5.318,
+      "step": 7315
+    },
+    {
+      "epoch": 0.4046434494195688,
+      "grad_norm": 4.028477191925049,
+      "learning_rate": 3.0613985218874365e-05,
+      "loss": 5.4557,
+      "step": 7320
+    },
+    {
+      "epoch": 0.40491984521835267,
+      "grad_norm": 3.430436134338379,
+      "learning_rate": 3.059977259806708e-05,
+      "loss": 5.3102,
+      "step": 7325
+    },
+    {
+      "epoch": 0.4051962410171365,
+      "grad_norm": 2.725626230239868,
+      "learning_rate": 3.0585559977259806e-05,
+      "loss": 5.3974,
+      "step": 7330
+    },
+    {
+      "epoch": 0.4054726368159204,
+      "grad_norm": 3.0082714557647705,
+      "learning_rate": 3.057134735645253e-05,
+      "loss": 5.2993,
+      "step": 7335
+    },
+    {
+      "epoch": 0.40574903261470424,
+      "grad_norm": 3.552365303039551,
+      "learning_rate": 3.0557134735645254e-05,
+      "loss": 5.2335,
+      "step": 7340
+    },
+    {
+      "epoch": 0.4060254284134881,
+      "grad_norm": 3.0008647441864014,
+      "learning_rate": 3.054292211483798e-05,
+      "loss": 5.179,
+      "step": 7345
+    },
+    {
+      "epoch": 0.40630182421227196,
+      "grad_norm": 3.4368932247161865,
+      "learning_rate": 3.05287094940307e-05,
+      "loss": 5.1877,
+      "step": 7350
+    },
+    {
+      "epoch": 0.4065782200110558,
+      "grad_norm": 2.775163173675537,
+      "learning_rate": 3.0514496873223426e-05,
+      "loss": 5.3432,
+      "step": 7355
+    },
+    {
+      "epoch": 0.40685461580983967,
+      "grad_norm": 3.331177234649658,
+      "learning_rate": 3.0500284252416143e-05,
+      "loss": 5.6045,
+      "step": 7360
+    },
+    {
+      "epoch": 0.4071310116086235,
+      "grad_norm": 3.2628567218780518,
+      "learning_rate": 3.048607163160887e-05,
+      "loss": 5.3768,
+      "step": 7365
+    },
+    {
+      "epoch": 0.4074074074074074,
+      "grad_norm": 3.5824832916259766,
+      "learning_rate": 3.0471859010801594e-05,
+      "loss": 5.5101,
+      "step": 7370
+    },
+    {
+      "epoch": 0.40768380320619124,
+      "grad_norm": 2.661316394805908,
+      "learning_rate": 3.0457646389994315e-05,
+      "loss": 5.4241,
+      "step": 7375
+    },
+    {
+      "epoch": 0.4079601990049751,
+      "grad_norm": 3.2753899097442627,
+      "learning_rate": 3.0443433769187042e-05,
+      "loss": 5.1645,
+      "step": 7380
+    },
+    {
+      "epoch": 0.40823659480375896,
+      "grad_norm": 3.1966748237609863,
+      "learning_rate": 3.042922114837976e-05,
+      "loss": 5.3537,
+      "step": 7385
+    },
+    {
+      "epoch": 0.4085129906025428,
+      "grad_norm": 3.2661285400390625,
+      "learning_rate": 3.0415008527572487e-05,
+      "loss": 5.3202,
+      "step": 7390
+    },
+    {
+      "epoch": 0.4087893864013267,
+      "grad_norm": 2.994544506072998,
+      "learning_rate": 3.040079590676521e-05,
+      "loss": 5.3407,
+      "step": 7395
+    },
+    {
+      "epoch": 0.4090657822001106,
+      "grad_norm": 3.363358974456787,
+      "learning_rate": 3.038658328595793e-05,
+      "loss": 5.3274,
+      "step": 7400
+    },
+    {
+      "epoch": 0.40934217799889444,
+      "grad_norm": 3.417901039123535,
+      "learning_rate": 3.0372370665150656e-05,
+      "loss": 5.2245,
+      "step": 7405
+    },
+    {
+      "epoch": 0.4096185737976783,
+      "grad_norm": 3.499995470046997,
+      "learning_rate": 3.0358158044343376e-05,
+      "loss": 5.043,
+      "step": 7410
+    },
+    {
+      "epoch": 0.40989496959646216,
+      "grad_norm": 3.142343044281006,
+      "learning_rate": 3.0343945423536104e-05,
+      "loss": 5.2181,
+      "step": 7415
+    },
+    {
+      "epoch": 0.410171365395246,
+      "grad_norm": 2.7809529304504395,
+      "learning_rate": 3.032973280272882e-05,
+      "loss": 5.3378,
+      "step": 7420
+    },
+    {
+      "epoch": 0.41044776119402987,
+      "grad_norm": 3.567267656326294,
+      "learning_rate": 3.0315520181921548e-05,
+      "loss": 5.3781,
+      "step": 7425
+    },
+    {
+      "epoch": 0.4107241569928137,
+      "grad_norm": 4.007451057434082,
+      "learning_rate": 3.0301307561114272e-05,
+      "loss": 5.1938,
+      "step": 7430
+    },
+    {
+      "epoch": 0.4110005527915976,
+      "grad_norm": 3.4437944889068604,
+      "learning_rate": 3.0287094940306993e-05,
+      "loss": 5.0837,
+      "step": 7435
+    },
+    {
+      "epoch": 0.41127694859038144,
+      "grad_norm": 3.051156997680664,
+      "learning_rate": 3.0272882319499717e-05,
+      "loss": 5.3,
+      "step": 7440
+    },
+    {
+      "epoch": 0.4115533443891653,
+      "grad_norm": 3.0758109092712402,
+      "learning_rate": 3.0258669698692437e-05,
+      "loss": 5.0707,
+      "step": 7445
+    },
+    {
+      "epoch": 0.41182974018794916,
+      "grad_norm": 3.035916328430176,
+      "learning_rate": 3.0244457077885165e-05,
+      "loss": 5.1578,
+      "step": 7450
+    },
+    {
+      "epoch": 0.412106135986733,
+      "grad_norm": 3.541250228881836,
+      "learning_rate": 3.023024445707789e-05,
+      "loss": 4.8507,
+      "step": 7455
+    },
+    {
+      "epoch": 0.41238253178551687,
+      "grad_norm": 4.51204252243042,
+      "learning_rate": 3.021603183627061e-05,
+      "loss": 5.1498,
+      "step": 7460
+    },
+    {
+      "epoch": 0.4126589275843007,
+      "grad_norm": 2.9354135990142822,
+      "learning_rate": 3.0201819215463333e-05,
+      "loss": 5.0576,
+      "step": 7465
+    },
+    {
+      "epoch": 0.4129353233830846,
+      "grad_norm": 2.9570817947387695,
+      "learning_rate": 3.0187606594656054e-05,
+      "loss": 5.3592,
+      "step": 7470
+    },
+    {
+      "epoch": 0.41321171918186844,
+      "grad_norm": 2.823472499847412,
+      "learning_rate": 3.0173393973848778e-05,
+      "loss": 4.9364,
+      "step": 7475
+    },
+    {
+      "epoch": 0.4134881149806523,
+      "grad_norm": 2.96844744682312,
+      "learning_rate": 3.0159181353041505e-05,
+      "loss": 5.3399,
+      "step": 7480
+    },
+    {
+      "epoch": 0.41376451077943616,
+      "grad_norm": 3.269542694091797,
+      "learning_rate": 3.0144968732234226e-05,
+      "loss": 5.2783,
+      "step": 7485
+    },
+    {
+      "epoch": 0.41404090657822,
+      "grad_norm": 2.990311622619629,
+      "learning_rate": 3.013075611142695e-05,
+      "loss": 5.4059,
+      "step": 7490
+    },
+    {
+      "epoch": 0.41431730237700387,
+      "grad_norm": 2.818763494491577,
+      "learning_rate": 3.011654349061967e-05,
+      "loss": 5.3341,
+      "step": 7495
+    },
+    {
+      "epoch": 0.41459369817578773,
+      "grad_norm": 4.055943489074707,
+      "learning_rate": 3.0102330869812394e-05,
+      "loss": 5.2108,
+      "step": 7500
+    },
+    {
+      "epoch": 0.4148700939745716,
+      "grad_norm": 3.448615789413452,
+      "learning_rate": 3.008811824900512e-05,
+      "loss": 5.0803,
+      "step": 7505
+    },
+    {
+      "epoch": 0.41514648977335544,
+      "grad_norm": 2.7099623680114746,
+      "learning_rate": 3.007390562819784e-05,
+      "loss": 5.1597,
+      "step": 7510
+    },
+    {
+      "epoch": 0.4154228855721393,
+      "grad_norm": 2.860887050628662,
+      "learning_rate": 3.0059693007390566e-05,
+      "loss": 5.0554,
+      "step": 7515
+    },
+    {
+      "epoch": 0.41569928137092316,
+      "grad_norm": 3.508277177810669,
+      "learning_rate": 3.0045480386583287e-05,
+      "loss": 5.2466,
+      "step": 7520
+    },
+    {
+      "epoch": 0.415975677169707,
+      "grad_norm": 3.1834499835968018,
+      "learning_rate": 3.003126776577601e-05,
+      "loss": 5.4212,
+      "step": 7525
+    },
+    {
+      "epoch": 0.41625207296849087,
+      "grad_norm": 3.1005613803863525,
+      "learning_rate": 3.001705514496873e-05,
+      "loss": 5.4608,
+      "step": 7530
+    },
+    {
+      "epoch": 0.41652846876727473,
+      "grad_norm": 3.1550991535186768,
+      "learning_rate": 3.0002842524161455e-05,
+      "loss": 5.2565,
+      "step": 7535
+    },
+    {
+      "epoch": 0.4168048645660586,
+      "grad_norm": 3.293140172958374,
+      "learning_rate": 2.9988629903354183e-05,
+      "loss": 5.5235,
+      "step": 7540
+    },
+    {
+      "epoch": 0.41708126036484244,
+      "grad_norm": 3.3371574878692627,
+      "learning_rate": 2.9974417282546903e-05,
+      "loss": 5.3766,
+      "step": 7545
+    },
+    {
+      "epoch": 0.4173576561636263,
+      "grad_norm": 3.0962579250335693,
+      "learning_rate": 2.9960204661739627e-05,
+      "loss": 5.1429,
+      "step": 7550
+    },
+    {
+      "epoch": 0.41763405196241016,
+      "grad_norm": 2.632319927215576,
+      "learning_rate": 2.9945992040932348e-05,
+      "loss": 5.2072,
+      "step": 7555
+    },
+    {
+      "epoch": 0.417910447761194,
+      "grad_norm": 3.5730085372924805,
+      "learning_rate": 2.9931779420125072e-05,
+      "loss": 5.3899,
+      "step": 7560
+    },
+    {
+      "epoch": 0.41818684355997787,
+      "grad_norm": 3.1113953590393066,
+      "learning_rate": 2.99175667993178e-05,
+      "loss": 5.2587,
+      "step": 7565
+    },
+    {
+      "epoch": 0.41846323935876173,
+      "grad_norm": 2.9837839603424072,
+      "learning_rate": 2.9903354178510516e-05,
+      "loss": 5.1953,
+      "step": 7570
+    },
+    {
+      "epoch": 0.4187396351575456,
+      "grad_norm": 3.156015634536743,
+      "learning_rate": 2.9889141557703244e-05,
+      "loss": 5.0675,
+      "step": 7575
+    },
+    {
+      "epoch": 0.41901603095632944,
+      "grad_norm": 2.8620293140411377,
+      "learning_rate": 2.9874928936895964e-05,
+      "loss": 5.2232,
+      "step": 7580
+    },
+    {
+      "epoch": 0.4192924267551133,
+      "grad_norm": 2.7657828330993652,
+      "learning_rate": 2.9860716316088688e-05,
+      "loss": 5.0537,
+      "step": 7585
+    },
+    {
+      "epoch": 0.41956882255389716,
+      "grad_norm": 2.991236448287964,
+      "learning_rate": 2.9846503695281412e-05,
+      "loss": 5.3795,
+      "step": 7590
+    },
+    {
+      "epoch": 0.419845218352681,
+      "grad_norm": 2.935781717300415,
+      "learning_rate": 2.9832291074474133e-05,
+      "loss": 5.0797,
+      "step": 7595
+    },
+    {
+      "epoch": 0.4201216141514649,
+      "grad_norm": 3.5586657524108887,
+      "learning_rate": 2.981807845366686e-05,
+      "loss": 5.3075,
+      "step": 7600
+    },
+    {
+      "epoch": 0.42039800995024873,
+      "grad_norm": 3.6937458515167236,
+      "learning_rate": 2.9803865832859577e-05,
+      "loss": 5.2942,
+      "step": 7605
+    },
+    {
+      "epoch": 0.4206744057490326,
+      "grad_norm": 3.0351436138153076,
+      "learning_rate": 2.9789653212052305e-05,
+      "loss": 5.1602,
+      "step": 7610
+    },
+    {
+      "epoch": 0.4209508015478165,
+      "grad_norm": 2.95098614692688,
+      "learning_rate": 2.977544059124503e-05,
+      "loss": 5.2933,
+      "step": 7615
+    },
+    {
+      "epoch": 0.42122719734660036,
+      "grad_norm": 2.8165371417999268,
+      "learning_rate": 2.976122797043775e-05,
+      "loss": 5.1234,
+      "step": 7620
+    },
+    {
+      "epoch": 0.4215035931453842,
+      "grad_norm": 3.436764717102051,
+      "learning_rate": 2.9747015349630473e-05,
+      "loss": 5.2743,
+      "step": 7625
+    },
+    {
+      "epoch": 0.42177998894416807,
+      "grad_norm": 3.4942984580993652,
+      "learning_rate": 2.9732802728823194e-05,
+      "loss": 5.5733,
+      "step": 7630
+    },
+    {
+      "epoch": 0.42205638474295193,
+      "grad_norm": 3.76875901222229,
+      "learning_rate": 2.971859010801592e-05,
+      "loss": 5.3345,
+      "step": 7635
+    },
+    {
+      "epoch": 0.4223327805417358,
+      "grad_norm": 3.2860209941864014,
+      "learning_rate": 2.970437748720864e-05,
+      "loss": 5.3476,
+      "step": 7640
+    },
+    {
+      "epoch": 0.42260917634051964,
+      "grad_norm": 3.403754949569702,
+      "learning_rate": 2.9690164866401366e-05,
+      "loss": 5.1001,
+      "step": 7645
+    },
+    {
+      "epoch": 0.4228855721393035,
+      "grad_norm": 3.613391876220703,
+      "learning_rate": 2.967595224559409e-05,
+      "loss": 5.0587,
+      "step": 7650
+    },
+    {
+      "epoch": 0.42316196793808736,
+      "grad_norm": 3.401045083999634,
+      "learning_rate": 2.966173962478681e-05,
+      "loss": 4.985,
+      "step": 7655
+    },
+    {
+      "epoch": 0.4234383637368712,
+      "grad_norm": 3.0673391819000244,
+      "learning_rate": 2.9647527003979534e-05,
+      "loss": 5.115,
+      "step": 7660
+    },
+    {
+      "epoch": 0.42371475953565507,
+      "grad_norm": 3.054534673690796,
+      "learning_rate": 2.9633314383172255e-05,
+      "loss": 5.1765,
+      "step": 7665
+    },
+    {
+      "epoch": 0.42399115533443893,
+      "grad_norm": 4.545713424682617,
+      "learning_rate": 2.9619101762364982e-05,
+      "loss": 5.0729,
+      "step": 7670
+    },
+    {
+      "epoch": 0.4242675511332228,
+      "grad_norm": 3.014993190765381,
+      "learning_rate": 2.9604889141557706e-05,
+      "loss": 5.4135,
+      "step": 7675
+    },
+    {
+      "epoch": 0.42454394693200664,
+      "grad_norm": 2.8630824089050293,
+      "learning_rate": 2.9590676520750427e-05,
+      "loss": 5.1434,
+      "step": 7680
+    },
+    {
+      "epoch": 0.4248203427307905,
+      "grad_norm": 3.143050193786621,
+      "learning_rate": 2.957646389994315e-05,
+      "loss": 5.2194,
+      "step": 7685
+    },
+    {
+      "epoch": 0.42509673852957436,
+      "grad_norm": 3.1831743717193604,
+      "learning_rate": 2.956225127913587e-05,
+      "loss": 5.4145,
+      "step": 7690
+    },
+    {
+      "epoch": 0.4253731343283582,
+      "grad_norm": 3.281905174255371,
+      "learning_rate": 2.95480386583286e-05,
+      "loss": 5.2458,
+      "step": 7695
+    },
+    {
+      "epoch": 0.4256495301271421,
+      "grad_norm": 3.2459661960601807,
+      "learning_rate": 2.9533826037521323e-05,
+      "loss": 5.1643,
+      "step": 7700
+    },
+    {
+      "epoch": 0.42592592592592593,
+      "grad_norm": 3.014573097229004,
+      "learning_rate": 2.9519613416714043e-05,
+      "loss": 5.3188,
+      "step": 7705
+    },
+    {
+      "epoch": 0.4262023217247098,
+      "grad_norm": 3.489968776702881,
+      "learning_rate": 2.9505400795906767e-05,
+      "loss": 5.359,
+      "step": 7710
+    },
+    {
+      "epoch": 0.42647871752349364,
+      "grad_norm": 2.6818461418151855,
+      "learning_rate": 2.9491188175099488e-05,
+      "loss": 5.2443,
+      "step": 7715
+    },
+    {
+      "epoch": 0.4267551133222775,
+      "grad_norm": 3.4301328659057617,
+      "learning_rate": 2.9476975554292212e-05,
+      "loss": 5.3404,
+      "step": 7720
+    },
+    {
+      "epoch": 0.42703150912106136,
+      "grad_norm": 2.975738525390625,
+      "learning_rate": 2.946276293348494e-05,
+      "loss": 5.5118,
+      "step": 7725
+    },
+    {
+      "epoch": 0.4273079049198452,
+      "grad_norm": 3.66871976852417,
+      "learning_rate": 2.944855031267766e-05,
+      "loss": 5.2737,
+      "step": 7730
+    },
+    {
+      "epoch": 0.4275843007186291,
+      "grad_norm": 2.812068462371826,
+      "learning_rate": 2.9434337691870384e-05,
+      "loss": 5.0211,
+      "step": 7735
+    },
+    {
+      "epoch": 0.42786069651741293,
+      "grad_norm": 3.8144164085388184,
+      "learning_rate": 2.9420125071063105e-05,
+      "loss": 5.3493,
+      "step": 7740
+    },
+    {
+      "epoch": 0.4281370923161968,
+      "grad_norm": 3.173715353012085,
+      "learning_rate": 2.940591245025583e-05,
+      "loss": 5.144,
+      "step": 7745
+    },
+    {
+      "epoch": 0.42841348811498065,
+      "grad_norm": 3.548175573348999,
+      "learning_rate": 2.939169982944855e-05,
+      "loss": 5.4388,
+      "step": 7750
+    },
+    {
+      "epoch": 0.4286898839137645,
+      "grad_norm": 2.4336702823638916,
+      "learning_rate": 2.9377487208641273e-05,
+      "loss": 5.1597,
+      "step": 7755
+    },
+    {
+      "epoch": 0.42896627971254836,
+      "grad_norm": 3.0814414024353027,
+      "learning_rate": 2.9363274587834e-05,
+      "loss": 5.2932,
+      "step": 7760
+    },
+    {
+      "epoch": 0.4292426755113322,
+      "grad_norm": 2.701829671859741,
+      "learning_rate": 2.934906196702672e-05,
+      "loss": 5.2772,
+      "step": 7765
+    },
+    {
+      "epoch": 0.4295190713101161,
+      "grad_norm": 2.623934268951416,
+      "learning_rate": 2.9334849346219445e-05,
+      "loss": 4.9666,
+      "step": 7770
+    },
+    {
+      "epoch": 0.42979546710889993,
+      "grad_norm": 3.162205696105957,
+      "learning_rate": 2.9320636725412166e-05,
+      "loss": 5.5104,
+      "step": 7775
+    },
+    {
+      "epoch": 0.4300718629076838,
+      "grad_norm": 3.1001057624816895,
+      "learning_rate": 2.930642410460489e-05,
+      "loss": 5.2675,
+      "step": 7780
+    },
+    {
+      "epoch": 0.43034825870646765,
+      "grad_norm": 3.4800355434417725,
+      "learning_rate": 2.9292211483797617e-05,
+      "loss": 5.246,
+      "step": 7785
+    },
+    {
+      "epoch": 0.4306246545052515,
+      "grad_norm": 3.1043055057525635,
+      "learning_rate": 2.9277998862990334e-05,
+      "loss": 4.9439,
+      "step": 7790
+    },
+    {
+      "epoch": 0.43090105030403536,
+      "grad_norm": 2.9988179206848145,
+      "learning_rate": 2.926378624218306e-05,
+      "loss": 5.1472,
+      "step": 7795
+    },
+    {
+      "epoch": 0.4311774461028192,
+      "grad_norm": 3.2718756198883057,
+      "learning_rate": 2.9249573621375782e-05,
+      "loss": 5.3382,
+      "step": 7800
+    },
+    {
+      "epoch": 0.4314538419016031,
+      "grad_norm": 2.6709251403808594,
+      "learning_rate": 2.9235361000568506e-05,
+      "loss": 5.1973,
+      "step": 7805
+    },
+    {
+      "epoch": 0.43173023770038693,
+      "grad_norm": 2.908034086227417,
+      "learning_rate": 2.922114837976123e-05,
+      "loss": 5.0482,
+      "step": 7810
+    },
+    {
+      "epoch": 0.4320066334991708,
+      "grad_norm": 2.8936550617218018,
+      "learning_rate": 2.920693575895395e-05,
+      "loss": 5.223,
+      "step": 7815
+    },
+    {
+      "epoch": 0.43228302929795465,
+      "grad_norm": 3.31170654296875,
+      "learning_rate": 2.9192723138146678e-05,
+      "loss": 5.218,
+      "step": 7820
+    },
+    {
+      "epoch": 0.4325594250967385,
+      "grad_norm": 3.3530709743499756,
+      "learning_rate": 2.9178510517339395e-05,
+      "loss": 5.1359,
+      "step": 7825
+    },
+    {
+      "epoch": 0.43283582089552236,
+      "grad_norm": 2.950207233428955,
+      "learning_rate": 2.9164297896532123e-05,
+      "loss": 5.538,
+      "step": 7830
+    },
+    {
+      "epoch": 0.4331122166943063,
+      "grad_norm": 3.7131307125091553,
+      "learning_rate": 2.9150085275724847e-05,
+      "loss": 5.2753,
+      "step": 7835
+    },
+    {
+      "epoch": 0.43338861249309013,
+      "grad_norm": 3.750091075897217,
+      "learning_rate": 2.9135872654917567e-05,
+      "loss": 5.0678,
+      "step": 7840
+    },
+    {
+      "epoch": 0.433665008291874,
+      "grad_norm": 2.7476789951324463,
+      "learning_rate": 2.9121660034110295e-05,
+      "loss": 5.3974,
+      "step": 7845
+    },
+    {
+      "epoch": 0.43394140409065785,
+      "grad_norm": 3.338824510574341,
+      "learning_rate": 2.9107447413303012e-05,
+      "loss": 4.9142,
+      "step": 7850
+    },
+    {
+      "epoch": 0.4342177998894417,
+      "grad_norm": 3.1479055881500244,
+      "learning_rate": 2.909323479249574e-05,
+      "loss": 5.0689,
+      "step": 7855
+    },
+    {
+      "epoch": 0.43449419568822556,
+      "grad_norm": 2.765441656112671,
+      "learning_rate": 2.907902217168846e-05,
+      "loss": 5.2719,
+      "step": 7860
+    },
+    {
+      "epoch": 0.4347705914870094,
+      "grad_norm": 3.1556508541107178,
+      "learning_rate": 2.9064809550881184e-05,
+      "loss": 5.391,
+      "step": 7865
+    },
+    {
+      "epoch": 0.4350469872857933,
+      "grad_norm": 3.2448055744171143,
+      "learning_rate": 2.9050596930073908e-05,
+      "loss": 5.3468,
+      "step": 7870
+    },
+    {
+      "epoch": 0.43532338308457713,
+      "grad_norm": 3.5034596920013428,
+      "learning_rate": 2.9036384309266628e-05,
+      "loss": 5.1951,
+      "step": 7875
+    },
+    {
+      "epoch": 0.435599778883361,
+      "grad_norm": 3.018533945083618,
+      "learning_rate": 2.9022171688459356e-05,
+      "loss": 5.1914,
+      "step": 7880
+    },
+    {
+      "epoch": 0.43587617468214485,
+      "grad_norm": 2.8897950649261475,
+      "learning_rate": 2.9007959067652073e-05,
+      "loss": 5.6935,
+      "step": 7885
+    },
+    {
+      "epoch": 0.4361525704809287,
+      "grad_norm": 2.643296957015991,
+      "learning_rate": 2.89937464468448e-05,
+      "loss": 5.1013,
+      "step": 7890
+    },
+    {
+      "epoch": 0.43642896627971256,
+      "grad_norm": 3.041177988052368,
+      "learning_rate": 2.8979533826037524e-05,
+      "loss": 5.3812,
+      "step": 7895
+    },
+    {
+      "epoch": 0.4367053620784964,
+      "grad_norm": 3.4055657386779785,
+      "learning_rate": 2.8965321205230245e-05,
+      "loss": 5.5113,
+      "step": 7900
+    },
+    {
+      "epoch": 0.4369817578772803,
+      "grad_norm": 4.0776495933532715,
+      "learning_rate": 2.895110858442297e-05,
+      "loss": 5.4321,
+      "step": 7905
+    },
+    {
+      "epoch": 0.43725815367606413,
+      "grad_norm": 3.4030396938323975,
+      "learning_rate": 2.893689596361569e-05,
+      "loss": 5.091,
+      "step": 7910
+    },
+    {
+      "epoch": 0.437534549474848,
+      "grad_norm": 3.5989456176757812,
+      "learning_rate": 2.8922683342808417e-05,
+      "loss": 5.48,
+      "step": 7915
+    },
+    {
+      "epoch": 0.43781094527363185,
+      "grad_norm": 2.3618738651275635,
+      "learning_rate": 2.890847072200114e-05,
+      "loss": 5.0508,
+      "step": 7920
+    },
+    {
+      "epoch": 0.4380873410724157,
+      "grad_norm": 3.1228749752044678,
+      "learning_rate": 2.889425810119386e-05,
+      "loss": 5.3904,
+      "step": 7925
+    },
+    {
+      "epoch": 0.43836373687119956,
+      "grad_norm": 3.0725607872009277,
+      "learning_rate": 2.8880045480386585e-05,
+      "loss": 5.4847,
+      "step": 7930
+    },
+    {
+      "epoch": 0.4386401326699834,
+      "grad_norm": 2.9387130737304688,
+      "learning_rate": 2.8865832859579306e-05,
+      "loss": 4.9448,
+      "step": 7935
+    },
+    {
+      "epoch": 0.4389165284687673,
+      "grad_norm": 3.3646342754364014,
+      "learning_rate": 2.885162023877203e-05,
+      "loss": 5.3713,
+      "step": 7940
+    },
+    {
+      "epoch": 0.43919292426755113,
+      "grad_norm": 3.699636697769165,
+      "learning_rate": 2.8837407617964757e-05,
+      "loss": 5.2954,
+      "step": 7945
+    },
+    {
+      "epoch": 0.439469320066335,
+      "grad_norm": 3.304562568664551,
+      "learning_rate": 2.8823194997157478e-05,
+      "loss": 5.1658,
+      "step": 7950
+    },
+    {
+      "epoch": 0.43974571586511885,
+      "grad_norm": 2.908881902694702,
+      "learning_rate": 2.8808982376350202e-05,
+      "loss": 5.0667,
+      "step": 7955
+    },
+    {
+      "epoch": 0.4400221116639027,
+      "grad_norm": 3.2574126720428467,
+      "learning_rate": 2.8794769755542922e-05,
+      "loss": 5.1662,
+      "step": 7960
+    },
+    {
+      "epoch": 0.44029850746268656,
+      "grad_norm": 3.021467924118042,
+      "learning_rate": 2.8780557134735646e-05,
+      "loss": 5.0305,
+      "step": 7965
+    },
+    {
+      "epoch": 0.4405749032614704,
+      "grad_norm": 2.6084797382354736,
+      "learning_rate": 2.8766344513928367e-05,
+      "loss": 5.3558,
+      "step": 7970
+    },
+    {
+      "epoch": 0.4408512990602543,
+      "grad_norm": 3.0199179649353027,
+      "learning_rate": 2.875213189312109e-05,
+      "loss": 5.2878,
+      "step": 7975
+    },
+    {
+      "epoch": 0.44112769485903813,
+      "grad_norm": 3.2974507808685303,
+      "learning_rate": 2.8737919272313818e-05,
+      "loss": 5.3241,
+      "step": 7980
+    },
+    {
+      "epoch": 0.441404090657822,
+      "grad_norm": 3.1707582473754883,
+      "learning_rate": 2.872370665150654e-05,
+      "loss": 4.94,
+      "step": 7985
+    },
+    {
+      "epoch": 0.44168048645660585,
+      "grad_norm": 2.6704931259155273,
+      "learning_rate": 2.8709494030699263e-05,
+      "loss": 5.1259,
+      "step": 7990
+    },
+    {
+      "epoch": 0.4419568822553897,
+      "grad_norm": 4.149257659912109,
+      "learning_rate": 2.8695281409891983e-05,
+      "loss": 5.4461,
+      "step": 7995
+    },
+    {
+      "epoch": 0.44223327805417356,
+      "grad_norm": 3.305680274963379,
+      "learning_rate": 2.8681068789084707e-05,
+      "loss": 5.1553,
+      "step": 8000
+    },
+    {
+      "epoch": 0.4425096738529574,
+      "grad_norm": 3.288146734237671,
+      "learning_rate": 2.8666856168277435e-05,
+      "loss": 5.2207,
+      "step": 8005
+    },
+    {
+      "epoch": 0.4427860696517413,
+      "grad_norm": 4.230424880981445,
+      "learning_rate": 2.8652643547470155e-05,
+      "loss": 5.3263,
+      "step": 8010
+    },
+    {
+      "epoch": 0.44306246545052513,
+      "grad_norm": 3.5503792762756348,
+      "learning_rate": 2.863843092666288e-05,
+      "loss": 5.2507,
+      "step": 8015
+    },
+    {
+      "epoch": 0.443338861249309,
+      "grad_norm": 2.99174427986145,
+      "learning_rate": 2.86242183058556e-05,
+      "loss": 5.2627,
+      "step": 8020
+    },
+    {
+      "epoch": 0.44361525704809285,
+      "grad_norm": 3.6940457820892334,
+      "learning_rate": 2.8610005685048324e-05,
+      "loss": 4.9232,
+      "step": 8025
+    },
+    {
+      "epoch": 0.4438916528468767,
+      "grad_norm": 3.4294097423553467,
+      "learning_rate": 2.859579306424105e-05,
+      "loss": 5.2177,
+      "step": 8030
+    },
+    {
+      "epoch": 0.44416804864566056,
+      "grad_norm": 3.6956839561462402,
+      "learning_rate": 2.858158044343377e-05,
+      "loss": 5.3431,
+      "step": 8035
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 3.1307976245880127,
+      "learning_rate": 2.8567367822626496e-05,
+      "loss": 5.3812,
+      "step": 8040
+    },
+    {
+      "epoch": 0.4447208402432283,
+      "grad_norm": 2.7979776859283447,
+      "learning_rate": 2.8553155201819216e-05,
+      "loss": 5.6822,
+      "step": 8045
+    },
+    {
+      "epoch": 0.44499723604201213,
+      "grad_norm": 2.8679885864257812,
+      "learning_rate": 2.853894258101194e-05,
+      "loss": 5.278,
+      "step": 8050
+    },
+    {
+      "epoch": 0.44527363184079605,
+      "grad_norm": 2.877314805984497,
+      "learning_rate": 2.8524729960204664e-05,
+      "loss": 5.2747,
+      "step": 8055
+    },
+    {
+      "epoch": 0.4455500276395799,
+      "grad_norm": 3.238851308822632,
+      "learning_rate": 2.8510517339397385e-05,
+      "loss": 5.3333,
+      "step": 8060
+    },
+    {
+      "epoch": 0.44582642343836376,
+      "grad_norm": 2.7757999897003174,
+      "learning_rate": 2.8496304718590112e-05,
+      "loss": 5.0878,
+      "step": 8065
+    },
+    {
+      "epoch": 0.4461028192371476,
+      "grad_norm": 3.3367576599121094,
+      "learning_rate": 2.848209209778283e-05,
+      "loss": 5.2714,
+      "step": 8070
+    },
+    {
+      "epoch": 0.4463792150359315,
+      "grad_norm": 2.983865976333618,
+      "learning_rate": 2.8467879476975557e-05,
+      "loss": 5.4693,
+      "step": 8075
+    },
+    {
+      "epoch": 0.44665561083471533,
+      "grad_norm": 3.211333751678467,
+      "learning_rate": 2.8453666856168278e-05,
+      "loss": 5.1891,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4469320066334992,
+      "grad_norm": 3.608457565307617,
+      "learning_rate": 2.8439454235361e-05,
+      "loss": 4.9361,
+      "step": 8085
+    },
+    {
+      "epoch": 0.44720840243228305,
+      "grad_norm": 3.1287999153137207,
+      "learning_rate": 2.8425241614553725e-05,
+      "loss": 5.5907,
+      "step": 8090
+    },
+    {
+      "epoch": 0.4474847982310669,
+      "grad_norm": 3.5823168754577637,
+      "learning_rate": 2.8411028993746446e-05,
+      "loss": 5.1325,
+      "step": 8095
+    },
+    {
+      "epoch": 0.44776119402985076,
+      "grad_norm": 4.0038299560546875,
+      "learning_rate": 2.8396816372939173e-05,
+      "loss": 5.0001,
+      "step": 8100
+    },
+    {
+      "epoch": 0.4480375898286346,
+      "grad_norm": 3.3478615283966064,
+      "learning_rate": 2.838260375213189e-05,
+      "loss": 5.1217,
+      "step": 8105
+    },
+    {
+      "epoch": 0.4483139856274185,
+      "grad_norm": 3.304795503616333,
+      "learning_rate": 2.8368391131324618e-05,
+      "loss": 5.5211,
+      "step": 8110
+    },
+    {
+      "epoch": 0.44859038142620233,
+      "grad_norm": 3.8344953060150146,
+      "learning_rate": 2.8354178510517342e-05,
+      "loss": 5.2864,
+      "step": 8115
+    },
+    {
+      "epoch": 0.4488667772249862,
+      "grad_norm": 3.359548807144165,
+      "learning_rate": 2.8339965889710063e-05,
+      "loss": 5.5551,
+      "step": 8120
+    },
+    {
+      "epoch": 0.44914317302377005,
+      "grad_norm": 3.1302855014801025,
+      "learning_rate": 2.8325753268902787e-05,
+      "loss": 5.1885,
+      "step": 8125
+    },
+    {
+      "epoch": 0.4494195688225539,
+      "grad_norm": 3.1038334369659424,
+      "learning_rate": 2.8311540648095507e-05,
+      "loss": 5.1696,
+      "step": 8130
+    },
+    {
+      "epoch": 0.44969596462133776,
+      "grad_norm": 4.253241062164307,
+      "learning_rate": 2.8297328027288235e-05,
+      "loss": 5.5296,
+      "step": 8135
+    },
+    {
+      "epoch": 0.4499723604201216,
+      "grad_norm": 3.2857799530029297,
+      "learning_rate": 2.828311540648096e-05,
+      "loss": 5.2782,
+      "step": 8140
+    },
+    {
+      "epoch": 0.4502487562189055,
+      "grad_norm": 3.255506753921509,
+      "learning_rate": 2.826890278567368e-05,
+      "loss": 5.2445,
+      "step": 8145
+    },
+    {
+      "epoch": 0.45052515201768933,
+      "grad_norm": 4.212364673614502,
+      "learning_rate": 2.8254690164866403e-05,
+      "loss": 5.1744,
+      "step": 8150
+    },
+    {
+      "epoch": 0.4508015478164732,
+      "grad_norm": 3.6351518630981445,
+      "learning_rate": 2.8240477544059124e-05,
+      "loss": 5.3138,
+      "step": 8155
+    },
+    {
+      "epoch": 0.45107794361525705,
+      "grad_norm": 2.9749603271484375,
+      "learning_rate": 2.822626492325185e-05,
+      "loss": 5.2675,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4513543394140409,
+      "grad_norm": 3.6853575706481934,
+      "learning_rate": 2.8212052302444575e-05,
+      "loss": 5.2495,
+      "step": 8165
+    },
+    {
+      "epoch": 0.45163073521282476,
+      "grad_norm": 3.170837640762329,
+      "learning_rate": 2.8197839681637296e-05,
+      "loss": 5.2689,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4519071310116086,
+      "grad_norm": 3.9865872859954834,
+      "learning_rate": 2.818362706083002e-05,
+      "loss": 5.0987,
+      "step": 8175
+    },
+    {
+      "epoch": 0.4521835268103925,
+      "grad_norm": 3.1320040225982666,
+      "learning_rate": 2.816941444002274e-05,
+      "loss": 5.2081,
+      "step": 8180
+    },
+    {
+      "epoch": 0.45245992260917633,
+      "grad_norm": 3.1708438396453857,
+      "learning_rate": 2.8155201819215464e-05,
+      "loss": 5.1956,
+      "step": 8185
+    },
+    {
+      "epoch": 0.4527363184079602,
+      "grad_norm": 2.9454898834228516,
+      "learning_rate": 2.8140989198408185e-05,
+      "loss": 5.2115,
+      "step": 8190
+    },
+    {
+      "epoch": 0.45301271420674405,
+      "grad_norm": 3.4514904022216797,
+      "learning_rate": 2.8126776577600912e-05,
+      "loss": 5.1788,
+      "step": 8195
+    },
+    {
+      "epoch": 0.4532891100055279,
+      "grad_norm": 3.5502679347991943,
+      "learning_rate": 2.8112563956793636e-05,
+      "loss": 5.1326,
+      "step": 8200
+    },
+    {
+      "epoch": 0.45356550580431176,
+      "grad_norm": 3.267252206802368,
+      "learning_rate": 2.8098351335986357e-05,
+      "loss": 5.2808,
+      "step": 8205
+    },
+    {
+      "epoch": 0.4538419016030956,
+      "grad_norm": 3.928968667984009,
+      "learning_rate": 2.808413871517908e-05,
+      "loss": 5.262,
+      "step": 8210
+    },
+    {
+      "epoch": 0.4541182974018795,
+      "grad_norm": 3.202829360961914,
+      "learning_rate": 2.80699260943718e-05,
+      "loss": 5.2817,
+      "step": 8215
+    },
+    {
+      "epoch": 0.45439469320066334,
+      "grad_norm": 3.040146589279175,
+      "learning_rate": 2.8055713473564525e-05,
+      "loss": 5.3738,
+      "step": 8220
+    },
+    {
+      "epoch": 0.4546710889994472,
+      "grad_norm": 2.829195499420166,
+      "learning_rate": 2.8041500852757253e-05,
+      "loss": 4.9243,
+      "step": 8225
+    },
+    {
+      "epoch": 0.45494748479823105,
+      "grad_norm": 2.964588165283203,
+      "learning_rate": 2.8027288231949973e-05,
+      "loss": 5.2799,
+      "step": 8230
+    },
+    {
+      "epoch": 0.4552238805970149,
+      "grad_norm": 3.4491324424743652,
+      "learning_rate": 2.8013075611142697e-05,
+      "loss": 5.3802,
+      "step": 8235
+    },
+    {
+      "epoch": 0.45550027639579876,
+      "grad_norm": 3.446174383163452,
+      "learning_rate": 2.7998862990335418e-05,
+      "loss": 5.2512,
+      "step": 8240
+    },
+    {
+      "epoch": 0.4557766721945826,
+      "grad_norm": 3.0602688789367676,
+      "learning_rate": 2.7984650369528142e-05,
+      "loss": 5.281,
+      "step": 8245
+    },
+    {
+      "epoch": 0.4560530679933665,
+      "grad_norm": 2.8950977325439453,
+      "learning_rate": 2.797043774872087e-05,
+      "loss": 4.9311,
+      "step": 8250
+    },
+    {
+      "epoch": 0.45632946379215034,
+      "grad_norm": 3.147683620452881,
+      "learning_rate": 2.7956225127913586e-05,
+      "loss": 5.1642,
+      "step": 8255
+    },
+    {
+      "epoch": 0.4566058595909342,
+      "grad_norm": 3.45888090133667,
+      "learning_rate": 2.7942012507106314e-05,
+      "loss": 5.1064,
+      "step": 8260
+    },
+    {
+      "epoch": 0.45688225538971805,
+      "grad_norm": 3.435093641281128,
+      "learning_rate": 2.7927799886299034e-05,
+      "loss": 5.4801,
+      "step": 8265
+    },
+    {
+      "epoch": 0.4571586511885019,
+      "grad_norm": 3.746147394180298,
+      "learning_rate": 2.7913587265491758e-05,
+      "loss": 5.3686,
+      "step": 8270
+    },
+    {
+      "epoch": 0.4574350469872858,
+      "grad_norm": 3.0401806831359863,
+      "learning_rate": 2.7899374644684482e-05,
+      "loss": 5.0805,
+      "step": 8275
+    },
+    {
+      "epoch": 0.4577114427860697,
+      "grad_norm": 4.183225154876709,
+      "learning_rate": 2.7885162023877203e-05,
+      "loss": 5.5761,
+      "step": 8280
+    },
+    {
+      "epoch": 0.45798783858485353,
+      "grad_norm": 3.05330491065979,
+      "learning_rate": 2.787094940306993e-05,
+      "loss": 5.195,
+      "step": 8285
+    },
+    {
+      "epoch": 0.4582642343836374,
+      "grad_norm": 3.5704240798950195,
+      "learning_rate": 2.7856736782262647e-05,
+      "loss": 5.01,
+      "step": 8290
+    },
+    {
+      "epoch": 0.45854063018242125,
+      "grad_norm": 3.817938804626465,
+      "learning_rate": 2.7842524161455375e-05,
+      "loss": 5.2507,
+      "step": 8295
+    },
+    {
+      "epoch": 0.4588170259812051,
+      "grad_norm": 2.9435267448425293,
+      "learning_rate": 2.7828311540648095e-05,
+      "loss": 5.3353,
+      "step": 8300
+    },
+    {
+      "epoch": 0.45909342177998896,
+      "grad_norm": 3.9453604221343994,
+      "learning_rate": 2.781409891984082e-05,
+      "loss": 5.2214,
+      "step": 8305
+    },
+    {
+      "epoch": 0.4593698175787728,
+      "grad_norm": 3.7097527980804443,
+      "learning_rate": 2.7799886299033547e-05,
+      "loss": 4.9059,
+      "step": 8310
+    },
+    {
+      "epoch": 0.4596462133775567,
+      "grad_norm": 2.8728573322296143,
+      "learning_rate": 2.7785673678226264e-05,
+      "loss": 5.2303,
+      "step": 8315
+    },
+    {
+      "epoch": 0.45992260917634054,
+      "grad_norm": 2.946532726287842,
+      "learning_rate": 2.777146105741899e-05,
+      "loss": 5.1568,
+      "step": 8320
+    },
+    {
+      "epoch": 0.4601990049751244,
+      "grad_norm": 2.8694427013397217,
+      "learning_rate": 2.7757248436611712e-05,
+      "loss": 5.2556,
+      "step": 8325
+    },
+    {
+      "epoch": 0.46047540077390825,
+      "grad_norm": 2.947298049926758,
+      "learning_rate": 2.7743035815804436e-05,
+      "loss": 5.269,
+      "step": 8330
+    },
+    {
+      "epoch": 0.4607517965726921,
+      "grad_norm": 2.9417827129364014,
+      "learning_rate": 2.772882319499716e-05,
+      "loss": 5.0295,
+      "step": 8335
+    },
+    {
+      "epoch": 0.46102819237147596,
+      "grad_norm": 3.7710001468658447,
+      "learning_rate": 2.771461057418988e-05,
+      "loss": 5.403,
+      "step": 8340
+    },
+    {
+      "epoch": 0.4613045881702598,
+      "grad_norm": 3.4150478839874268,
+      "learning_rate": 2.7700397953382608e-05,
+      "loss": 5.2718,
+      "step": 8345
+    },
+    {
+      "epoch": 0.4615809839690437,
+      "grad_norm": 3.6512367725372314,
+      "learning_rate": 2.7686185332575325e-05,
+      "loss": 5.1832,
+      "step": 8350
+    },
+    {
+      "epoch": 0.46185737976782754,
+      "grad_norm": 3.080050468444824,
+      "learning_rate": 2.7671972711768052e-05,
+      "loss": 5.4398,
+      "step": 8355
+    },
+    {
+      "epoch": 0.4621337755666114,
+      "grad_norm": 3.351170778274536,
+      "learning_rate": 2.7657760090960776e-05,
+      "loss": 5.3517,
+      "step": 8360
+    },
+    {
+      "epoch": 0.46241017136539525,
+      "grad_norm": 3.467395782470703,
+      "learning_rate": 2.7643547470153497e-05,
+      "loss": 5.3345,
+      "step": 8365
+    },
+    {
+      "epoch": 0.4626865671641791,
+      "grad_norm": 3.781153440475464,
+      "learning_rate": 2.762933484934622e-05,
+      "loss": 5.2006,
+      "step": 8370
+    },
+    {
+      "epoch": 0.46296296296296297,
+      "grad_norm": 3.0200061798095703,
+      "learning_rate": 2.761512222853894e-05,
+      "loss": 4.8137,
+      "step": 8375
+    },
+    {
+      "epoch": 0.4632393587617468,
+      "grad_norm": 3.3280367851257324,
+      "learning_rate": 2.760090960773167e-05,
+      "loss": 5.0456,
+      "step": 8380
+    },
+    {
+      "epoch": 0.4635157545605307,
+      "grad_norm": 3.1769330501556396,
+      "learning_rate": 2.7586696986924386e-05,
+      "loss": 5.3049,
+      "step": 8385
+    },
+    {
+      "epoch": 0.46379215035931454,
+      "grad_norm": 4.280707836151123,
+      "learning_rate": 2.7572484366117113e-05,
+      "loss": 5.2953,
+      "step": 8390
+    },
+    {
+      "epoch": 0.4640685461580984,
+      "grad_norm": 2.720076322555542,
+      "learning_rate": 2.7558271745309837e-05,
+      "loss": 5.269,
+      "step": 8395
+    },
+    {
+      "epoch": 0.46434494195688225,
+      "grad_norm": 3.4472744464874268,
+      "learning_rate": 2.7544059124502558e-05,
+      "loss": 5.1512,
+      "step": 8400
+    },
+    {
+      "epoch": 0.4646213377556661,
+      "grad_norm": 3.2112371921539307,
+      "learning_rate": 2.7529846503695282e-05,
+      "loss": 5.3101,
+      "step": 8405
+    },
+    {
+      "epoch": 0.46489773355444997,
+      "grad_norm": 3.637249708175659,
+      "learning_rate": 2.7515633882888003e-05,
+      "loss": 5.1246,
+      "step": 8410
+    },
+    {
+      "epoch": 0.4651741293532338,
+      "grad_norm": 2.8876161575317383,
+      "learning_rate": 2.750142126208073e-05,
+      "loss": 5.4245,
+      "step": 8415
+    },
+    {
+      "epoch": 0.4654505251520177,
+      "grad_norm": 3.1725854873657227,
+      "learning_rate": 2.7487208641273454e-05,
+      "loss": 5.0848,
+      "step": 8420
+    },
+    {
+      "epoch": 0.46572692095080154,
+      "grad_norm": 2.7039942741394043,
+      "learning_rate": 2.7472996020466175e-05,
+      "loss": 5.2807,
+      "step": 8425
+    },
+    {
+      "epoch": 0.4660033167495854,
+      "grad_norm": 2.8240089416503906,
+      "learning_rate": 2.74587833996589e-05,
+      "loss": 5.4056,
+      "step": 8430
+    },
+    {
+      "epoch": 0.46627971254836925,
+      "grad_norm": 3.1696321964263916,
+      "learning_rate": 2.744457077885162e-05,
+      "loss": 5.4693,
+      "step": 8435
+    },
+    {
+      "epoch": 0.4665561083471531,
+      "grad_norm": 2.855938196182251,
+      "learning_rate": 2.7430358158044343e-05,
+      "loss": 5.1644,
+      "step": 8440
+    },
+    {
+      "epoch": 0.46683250414593697,
+      "grad_norm": 3.3975577354431152,
+      "learning_rate": 2.741614553723707e-05,
+      "loss": 5.1207,
+      "step": 8445
+    },
+    {
+      "epoch": 0.4671088999447208,
+      "grad_norm": 3.4103002548217773,
+      "learning_rate": 2.740193291642979e-05,
+      "loss": 5.2251,
+      "step": 8450
+    },
+    {
+      "epoch": 0.4673852957435047,
+      "grad_norm": 3.4105570316314697,
+      "learning_rate": 2.7387720295622515e-05,
+      "loss": 5.1505,
+      "step": 8455
+    },
+    {
+      "epoch": 0.46766169154228854,
+      "grad_norm": 3.4730947017669678,
+      "learning_rate": 2.7373507674815236e-05,
+      "loss": 5.056,
+      "step": 8460
+    },
+    {
+      "epoch": 0.4679380873410724,
+      "grad_norm": 3.272139549255371,
+      "learning_rate": 2.735929505400796e-05,
+      "loss": 5.3615,
+      "step": 8465
+    },
+    {
+      "epoch": 0.46821448313985625,
+      "grad_norm": 2.538783073425293,
+      "learning_rate": 2.7345082433200687e-05,
+      "loss": 5.3061,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4684908789386401,
+      "grad_norm": 2.9536359310150146,
+      "learning_rate": 2.7330869812393408e-05,
+      "loss": 5.2621,
+      "step": 8475
+    },
+    {
+      "epoch": 0.46876727473742397,
+      "grad_norm": 4.758022308349609,
+      "learning_rate": 2.731665719158613e-05,
+      "loss": 5.0966,
+      "step": 8480
+    },
+    {
+      "epoch": 0.4690436705362078,
+      "grad_norm": 3.3707218170166016,
+      "learning_rate": 2.7302444570778852e-05,
+      "loss": 5.4979,
+      "step": 8485
+    },
+    {
+      "epoch": 0.4693200663349917,
+      "grad_norm": 2.765289783477783,
+      "learning_rate": 2.7288231949971576e-05,
+      "loss": 5.4462,
+      "step": 8490
+    },
+    {
+      "epoch": 0.4695964621337756,
+      "grad_norm": 2.88114070892334,
+      "learning_rate": 2.7274019329164297e-05,
+      "loss": 5.7333,
+      "step": 8495
+    },
+    {
+      "epoch": 0.46987285793255945,
+      "grad_norm": 3.3201422691345215,
+      "learning_rate": 2.725980670835702e-05,
+      "loss": 5.2429,
+      "step": 8500
+    },
+    {
+      "epoch": 0.4701492537313433,
+      "grad_norm": 3.3583285808563232,
+      "learning_rate": 2.7245594087549748e-05,
+      "loss": 5.0037,
+      "step": 8505
+    },
+    {
+      "epoch": 0.47042564953012717,
+      "grad_norm": 3.485041618347168,
+      "learning_rate": 2.723138146674247e-05,
+      "loss": 5.1081,
+      "step": 8510
+    },
+    {
+      "epoch": 0.470702045328911,
+      "grad_norm": 3.287743330001831,
+      "learning_rate": 2.7217168845935193e-05,
+      "loss": 5.1691,
+      "step": 8515
+    },
+    {
+      "epoch": 0.4709784411276949,
+      "grad_norm": 3.2845332622528076,
+      "learning_rate": 2.7202956225127913e-05,
+      "loss": 5.0228,
+      "step": 8520
+    },
+    {
+      "epoch": 0.47125483692647874,
+      "grad_norm": 3.746253728866577,
+      "learning_rate": 2.7188743604320637e-05,
+      "loss": 4.901,
+      "step": 8525
+    },
+    {
+      "epoch": 0.4715312327252626,
+      "grad_norm": 3.2223970890045166,
+      "learning_rate": 2.7174530983513365e-05,
+      "loss": 4.915,
+      "step": 8530
+    },
+    {
+      "epoch": 0.47180762852404645,
+      "grad_norm": 3.4350712299346924,
+      "learning_rate": 2.7160318362706082e-05,
+      "loss": 5.2032,
+      "step": 8535
+    },
+    {
+      "epoch": 0.4720840243228303,
+      "grad_norm": 3.5283148288726807,
+      "learning_rate": 2.714610574189881e-05,
+      "loss": 5.0011,
+      "step": 8540
+    },
+    {
+      "epoch": 0.47236042012161417,
+      "grad_norm": 3.2964229583740234,
+      "learning_rate": 2.713189312109153e-05,
+      "loss": 5.1857,
+      "step": 8545
+    },
+    {
+      "epoch": 0.472636815920398,
+      "grad_norm": 2.962416172027588,
+      "learning_rate": 2.7117680500284254e-05,
+      "loss": 4.9636,
+      "step": 8550
+    },
+    {
+      "epoch": 0.4729132117191819,
+      "grad_norm": 3.4900479316711426,
+      "learning_rate": 2.7103467879476978e-05,
+      "loss": 4.8722,
+      "step": 8555
+    },
+    {
+      "epoch": 0.47318960751796574,
+      "grad_norm": 3.9601972103118896,
+      "learning_rate": 2.7089255258669698e-05,
+      "loss": 5.3924,
+      "step": 8560
+    },
+    {
+      "epoch": 0.4734660033167496,
+      "grad_norm": 3.298952102661133,
+      "learning_rate": 2.7075042637862426e-05,
+      "loss": 4.971,
+      "step": 8565
+    },
+    {
+      "epoch": 0.47374239911553345,
+      "grad_norm": 3.38555645942688,
+      "learning_rate": 2.7060830017055143e-05,
+      "loss": 5.0353,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4740187949143173,
+      "grad_norm": 2.793022632598877,
+      "learning_rate": 2.704661739624787e-05,
+      "loss": 5.0542,
+      "step": 8575
+    },
+    {
+      "epoch": 0.47429519071310117,
+      "grad_norm": 2.7033803462982178,
+      "learning_rate": 2.7032404775440594e-05,
+      "loss": 5.1854,
+      "step": 8580
+    },
+    {
+      "epoch": 0.474571586511885,
+      "grad_norm": 3.3331148624420166,
+      "learning_rate": 2.7018192154633315e-05,
+      "loss": 5.3229,
+      "step": 8585
+    },
+    {
+      "epoch": 0.4748479823106689,
+      "grad_norm": 3.3170993328094482,
+      "learning_rate": 2.700397953382604e-05,
+      "loss": 5.2236,
+      "step": 8590
+    },
+    {
+      "epoch": 0.47512437810945274,
+      "grad_norm": 3.055724859237671,
+      "learning_rate": 2.698976691301876e-05,
+      "loss": 5.447,
+      "step": 8595
+    },
+    {
+      "epoch": 0.4754007739082366,
+      "grad_norm": 3.705082654953003,
+      "learning_rate": 2.6975554292211487e-05,
+      "loss": 5.2754,
+      "step": 8600
+    },
+    {
+      "epoch": 0.47567716970702045,
+      "grad_norm": 3.6504838466644287,
+      "learning_rate": 2.6961341671404204e-05,
+      "loss": 5.1527,
+      "step": 8605
+    },
+    {
+      "epoch": 0.4759535655058043,
+      "grad_norm": 3.557499647140503,
+      "learning_rate": 2.694712905059693e-05,
+      "loss": 5.1805,
+      "step": 8610
+    },
+    {
+      "epoch": 0.47622996130458817,
+      "grad_norm": 3.1619584560394287,
+      "learning_rate": 2.6932916429789655e-05,
+      "loss": 5.0161,
+      "step": 8615
+    },
+    {
+      "epoch": 0.476506357103372,
+      "grad_norm": 3.1936593055725098,
+      "learning_rate": 2.6918703808982376e-05,
+      "loss": 5.0251,
+      "step": 8620
+    },
+    {
+      "epoch": 0.4767827529021559,
+      "grad_norm": 3.5965256690979004,
+      "learning_rate": 2.6904491188175103e-05,
+      "loss": 4.8984,
+      "step": 8625
+    },
+    {
+      "epoch": 0.47705914870093974,
+      "grad_norm": 3.2955245971679688,
+      "learning_rate": 2.689027856736782e-05,
+      "loss": 5.1065,
+      "step": 8630
+    },
+    {
+      "epoch": 0.4773355444997236,
+      "grad_norm": 2.86582350730896,
+      "learning_rate": 2.6876065946560548e-05,
+      "loss": 5.3203,
+      "step": 8635
+    },
+    {
+      "epoch": 0.47761194029850745,
+      "grad_norm": 2.8077163696289062,
+      "learning_rate": 2.6861853325753272e-05,
+      "loss": 5.1535,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4778883360972913,
+      "grad_norm": 3.2445852756500244,
+      "learning_rate": 2.6847640704945992e-05,
+      "loss": 5.3598,
+      "step": 8645
+    },
+    {
+      "epoch": 0.47816473189607517,
+      "grad_norm": 2.902517557144165,
+      "learning_rate": 2.6833428084138716e-05,
+      "loss": 5.1782,
+      "step": 8650
+    },
+    {
+      "epoch": 0.478441127694859,
+      "grad_norm": 3.0056138038635254,
+      "learning_rate": 2.6819215463331437e-05,
+      "loss": 5.2646,
+      "step": 8655
+    },
+    {
+      "epoch": 0.4787175234936429,
+      "grad_norm": 3.432762861251831,
+      "learning_rate": 2.6805002842524164e-05,
+      "loss": 5.3692,
+      "step": 8660
+    },
+    {
+      "epoch": 0.47899391929242674,
+      "grad_norm": 2.9771177768707275,
+      "learning_rate": 2.6790790221716888e-05,
+      "loss": 5.2922,
+      "step": 8665
+    },
+    {
+      "epoch": 0.4792703150912106,
+      "grad_norm": 2.8784945011138916,
+      "learning_rate": 2.677657760090961e-05,
+      "loss": 4.8865,
+      "step": 8670
+    },
+    {
+      "epoch": 0.47954671088999445,
+      "grad_norm": 3.0475375652313232,
+      "learning_rate": 2.6762364980102333e-05,
+      "loss": 5.0122,
+      "step": 8675
+    },
+    {
+      "epoch": 0.4798231066887783,
+      "grad_norm": 3.4154276847839355,
+      "learning_rate": 2.6748152359295053e-05,
+      "loss": 5.1384,
+      "step": 8680
+    },
+    {
+      "epoch": 0.48009950248756217,
+      "grad_norm": 2.841428279876709,
+      "learning_rate": 2.6733939738487777e-05,
+      "loss": 5.3496,
+      "step": 8685
+    },
+    {
+      "epoch": 0.480375898286346,
+      "grad_norm": 3.4203035831451416,
+      "learning_rate": 2.6719727117680505e-05,
+      "loss": 5.2612,
+      "step": 8690
+    },
+    {
+      "epoch": 0.4806522940851299,
+      "grad_norm": 2.9312570095062256,
+      "learning_rate": 2.6705514496873225e-05,
+      "loss": 5.065,
+      "step": 8695
+    },
+    {
+      "epoch": 0.48092868988391374,
+      "grad_norm": 4.5229668617248535,
+      "learning_rate": 2.669130187606595e-05,
+      "loss": 5.2577,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4812050856826976,
+      "grad_norm": 4.566796779632568,
+      "learning_rate": 2.667708925525867e-05,
+      "loss": 5.4462,
+      "step": 8705
+    },
+    {
+      "epoch": 0.48148148148148145,
+      "grad_norm": 3.907799005508423,
+      "learning_rate": 2.6662876634451394e-05,
+      "loss": 5.3402,
+      "step": 8710
+    },
+    {
+      "epoch": 0.48175787728026537,
+      "grad_norm": 4.1701836585998535,
+      "learning_rate": 2.6648664013644114e-05,
+      "loss": 5.3027,
+      "step": 8715
+    },
+    {
+      "epoch": 0.4820342730790492,
+      "grad_norm": 2.956747055053711,
+      "learning_rate": 2.663445139283684e-05,
+      "loss": 5.2394,
+      "step": 8720
+    },
+    {
+      "epoch": 0.4823106688778331,
+      "grad_norm": 3.6254959106445312,
+      "learning_rate": 2.6620238772029566e-05,
+      "loss": 5.1001,
+      "step": 8725
+    },
+    {
+      "epoch": 0.48258706467661694,
+      "grad_norm": 3.541015386581421,
+      "learning_rate": 2.6606026151222286e-05,
+      "loss": 5.2681,
+      "step": 8730
+    },
+    {
+      "epoch": 0.4828634604754008,
+      "grad_norm": 2.8555665016174316,
+      "learning_rate": 2.659181353041501e-05,
+      "loss": 5.1359,
+      "step": 8735
+    },
+    {
+      "epoch": 0.48313985627418465,
+      "grad_norm": 3.2647249698638916,
+      "learning_rate": 2.657760090960773e-05,
+      "loss": 5.1848,
+      "step": 8740
+    },
+    {
+      "epoch": 0.4834162520729685,
+      "grad_norm": 3.473285675048828,
+      "learning_rate": 2.6563388288800455e-05,
+      "loss": 5.1448,
+      "step": 8745
+    },
+    {
+      "epoch": 0.48369264787175237,
+      "grad_norm": 4.414984226226807,
+      "learning_rate": 2.6549175667993182e-05,
+      "loss": 4.8777,
+      "step": 8750
+    },
+    {
+      "epoch": 0.4839690436705362,
+      "grad_norm": 3.5118448734283447,
+      "learning_rate": 2.65349630471859e-05,
+      "loss": 4.973,
+      "step": 8755
+    },
+    {
+      "epoch": 0.4842454394693201,
+      "grad_norm": 3.8873729705810547,
+      "learning_rate": 2.6520750426378627e-05,
+      "loss": 5.1884,
+      "step": 8760
+    },
+    {
+      "epoch": 0.48452183526810394,
+      "grad_norm": 2.8551294803619385,
+      "learning_rate": 2.6506537805571347e-05,
+      "loss": 5.2593,
+      "step": 8765
+    },
+    {
+      "epoch": 0.4847982310668878,
+      "grad_norm": 3.043487310409546,
+      "learning_rate": 2.649232518476407e-05,
+      "loss": 5.3543,
+      "step": 8770
+    },
+    {
+      "epoch": 0.48507462686567165,
+      "grad_norm": 2.915309429168701,
+      "learning_rate": 2.64781125639568e-05,
+      "loss": 5.1256,
+      "step": 8775
+    },
+    {
+      "epoch": 0.4853510226644555,
+      "grad_norm": 3.0859193801879883,
+      "learning_rate": 2.6463899943149516e-05,
+      "loss": 5.1491,
+      "step": 8780
+    },
+    {
+      "epoch": 0.48562741846323937,
+      "grad_norm": 3.597959518432617,
+      "learning_rate": 2.6449687322342243e-05,
+      "loss": 5.2776,
+      "step": 8785
+    },
+    {
+      "epoch": 0.4859038142620232,
+      "grad_norm": 3.311375617980957,
+      "learning_rate": 2.6435474701534964e-05,
+      "loss": 5.0004,
+      "step": 8790
+    },
+    {
+      "epoch": 0.4861802100608071,
+      "grad_norm": 2.9994888305664062,
+      "learning_rate": 2.6421262080727688e-05,
+      "loss": 5.2205,
+      "step": 8795
+    },
+    {
+      "epoch": 0.48645660585959094,
+      "grad_norm": 3.1175289154052734,
+      "learning_rate": 2.6407049459920412e-05,
+      "loss": 4.9917,
+      "step": 8800
+    },
+    {
+      "epoch": 0.4867330016583748,
+      "grad_norm": 3.356494426727295,
+      "learning_rate": 2.6392836839113133e-05,
+      "loss": 5.119,
+      "step": 8805
+    },
+    {
+      "epoch": 0.48700939745715865,
+      "grad_norm": 2.93296480178833,
+      "learning_rate": 2.637862421830586e-05,
+      "loss": 5.1587,
+      "step": 8810
+    },
+    {
+      "epoch": 0.4872857932559425,
+      "grad_norm": 3.3215854167938232,
+      "learning_rate": 2.6364411597498577e-05,
+      "loss": 5.3971,
+      "step": 8815
+    },
+    {
+      "epoch": 0.48756218905472637,
+      "grad_norm": 2.9889779090881348,
+      "learning_rate": 2.6350198976691304e-05,
+      "loss": 5.023,
+      "step": 8820
+    },
+    {
+      "epoch": 0.4878385848535102,
+      "grad_norm": 4.003651142120361,
+      "learning_rate": 2.6335986355884025e-05,
+      "loss": 5.3191,
+      "step": 8825
+    },
+    {
+      "epoch": 0.4881149806522941,
+      "grad_norm": 4.0643744468688965,
+      "learning_rate": 2.632177373507675e-05,
+      "loss": 5.4204,
+      "step": 8830
+    },
+    {
+      "epoch": 0.48839137645107794,
+      "grad_norm": 3.0099825859069824,
+      "learning_rate": 2.6307561114269473e-05,
+      "loss": 5.0147,
+      "step": 8835
+    },
+    {
+      "epoch": 0.4886677722498618,
+      "grad_norm": 2.961827516555786,
+      "learning_rate": 2.6293348493462194e-05,
+      "loss": 4.8997,
+      "step": 8840
+    },
+    {
+      "epoch": 0.48894416804864566,
+      "grad_norm": 2.8388969898223877,
+      "learning_rate": 2.627913587265492e-05,
+      "loss": 5.0859,
+      "step": 8845
+    },
+    {
+      "epoch": 0.4892205638474295,
+      "grad_norm": 3.230729341506958,
+      "learning_rate": 2.6264923251847638e-05,
+      "loss": 5.2606,
+      "step": 8850
+    },
+    {
+      "epoch": 0.48949695964621337,
+      "grad_norm": 2.873793363571167,
+      "learning_rate": 2.6250710631040366e-05,
+      "loss": 5.5914,
+      "step": 8855
+    },
+    {
+      "epoch": 0.4897733554449972,
+      "grad_norm": 3.9639649391174316,
+      "learning_rate": 2.623649801023309e-05,
+      "loss": 4.8028,
+      "step": 8860
+    },
+    {
+      "epoch": 0.4900497512437811,
+      "grad_norm": 3.220949411392212,
+      "learning_rate": 2.622228538942581e-05,
+      "loss": 5.2824,
+      "step": 8865
+    },
+    {
+      "epoch": 0.49032614704256494,
+      "grad_norm": 4.145904541015625,
+      "learning_rate": 2.6208072768618534e-05,
+      "loss": 5.3472,
+      "step": 8870
+    },
+    {
+      "epoch": 0.4906025428413488,
+      "grad_norm": 3.1538095474243164,
+      "learning_rate": 2.6193860147811255e-05,
+      "loss": 5.2069,
+      "step": 8875
+    },
+    {
+      "epoch": 0.49087893864013266,
+      "grad_norm": 3.5244972705841064,
+      "learning_rate": 2.6179647527003982e-05,
+      "loss": 5.048,
+      "step": 8880
+    },
+    {
+      "epoch": 0.4911553344389165,
+      "grad_norm": 3.0251150131225586,
+      "learning_rate": 2.6165434906196706e-05,
+      "loss": 5.1505,
+      "step": 8885
+    },
+    {
+      "epoch": 0.49143173023770037,
+      "grad_norm": 2.746638298034668,
+      "learning_rate": 2.6151222285389427e-05,
+      "loss": 5.2146,
+      "step": 8890
+    },
+    {
+      "epoch": 0.4917081260364842,
+      "grad_norm": 2.9804468154907227,
+      "learning_rate": 2.613700966458215e-05,
+      "loss": 5.1231,
+      "step": 8895
+    },
+    {
+      "epoch": 0.4919845218352681,
+      "grad_norm": 3.2078375816345215,
+      "learning_rate": 2.612279704377487e-05,
+      "loss": 4.9723,
+      "step": 8900
+    },
+    {
+      "epoch": 0.49226091763405194,
+      "grad_norm": 3.1018552780151367,
+      "learning_rate": 2.6108584422967595e-05,
+      "loss": 5.1348,
+      "step": 8905
+    },
+    {
+      "epoch": 0.4925373134328358,
+      "grad_norm": 2.991896390914917,
+      "learning_rate": 2.6094371802160323e-05,
+      "loss": 5.1945,
+      "step": 8910
+    },
+    {
+      "epoch": 0.49281370923161966,
+      "grad_norm": 3.790259599685669,
+      "learning_rate": 2.6080159181353043e-05,
+      "loss": 5.2229,
+      "step": 8915
+    },
+    {
+      "epoch": 0.4930901050304035,
+      "grad_norm": 2.9007620811462402,
+      "learning_rate": 2.6065946560545767e-05,
+      "loss": 4.7525,
+      "step": 8920
+    },
+    {
+      "epoch": 0.49336650082918737,
+      "grad_norm": 3.5712127685546875,
+      "learning_rate": 2.6051733939738488e-05,
+      "loss": 5.2762,
+      "step": 8925
+    },
+    {
+      "epoch": 0.49364289662797123,
+      "grad_norm": 3.451338291168213,
+      "learning_rate": 2.6037521318931212e-05,
+      "loss": 4.9248,
+      "step": 8930
+    },
+    {
+      "epoch": 0.49391929242675514,
+      "grad_norm": 3.268249988555908,
+      "learning_rate": 2.6023308698123932e-05,
+      "loss": 5.2842,
+      "step": 8935
+    },
+    {
+      "epoch": 0.494195688225539,
+      "grad_norm": 2.8075647354125977,
+      "learning_rate": 2.600909607731666e-05,
+      "loss": 5.4364,
+      "step": 8940
+    },
+    {
+      "epoch": 0.49447208402432286,
+      "grad_norm": 3.8568859100341797,
+      "learning_rate": 2.5994883456509384e-05,
+      "loss": 5.1617,
+      "step": 8945
+    },
+    {
+      "epoch": 0.4947484798231067,
+      "grad_norm": 4.422865867614746,
+      "learning_rate": 2.5980670835702104e-05,
+      "loss": 5.2414,
+      "step": 8950
+    },
+    {
+      "epoch": 0.49502487562189057,
+      "grad_norm": 3.7072761058807373,
+      "learning_rate": 2.5966458214894828e-05,
+      "loss": 4.9748,
+      "step": 8955
+    },
+    {
+      "epoch": 0.4953012714206744,
+      "grad_norm": 3.6415674686431885,
+      "learning_rate": 2.595224559408755e-05,
+      "loss": 5.3649,
+      "step": 8960
+    },
+    {
+      "epoch": 0.4955776672194583,
+      "grad_norm": 3.360414505004883,
+      "learning_rate": 2.5938032973280273e-05,
+      "loss": 4.9524,
+      "step": 8965
+    },
+    {
+      "epoch": 0.49585406301824214,
+      "grad_norm": 3.060049057006836,
+      "learning_rate": 2.5923820352473e-05,
+      "loss": 5.1118,
+      "step": 8970
+    },
+    {
+      "epoch": 0.496130458817026,
+      "grad_norm": 4.066228866577148,
+      "learning_rate": 2.590960773166572e-05,
+      "loss": 5.4221,
+      "step": 8975
+    },
+    {
+      "epoch": 0.49640685461580986,
+      "grad_norm": 3.416016101837158,
+      "learning_rate": 2.5895395110858445e-05,
+      "loss": 5.1152,
+      "step": 8980
+    },
+    {
+      "epoch": 0.4966832504145937,
+      "grad_norm": 3.874455451965332,
+      "learning_rate": 2.5881182490051165e-05,
+      "loss": 5.1813,
+      "step": 8985
+    },
+    {
+      "epoch": 0.49695964621337757,
+      "grad_norm": 3.0348269939422607,
+      "learning_rate": 2.586696986924389e-05,
+      "loss": 5.1617,
+      "step": 8990
+    },
+    {
+      "epoch": 0.4972360420121614,
+      "grad_norm": 3.266130208969116,
+      "learning_rate": 2.5852757248436617e-05,
+      "loss": 5.2492,
+      "step": 8995
+    },
+    {
+      "epoch": 0.4975124378109453,
+      "grad_norm": 3.8973846435546875,
+      "learning_rate": 2.5838544627629334e-05,
+      "loss": 5.1134,
+      "step": 9000
+    },
+    {
+      "epoch": 0.49778883360972914,
+      "grad_norm": 3.322026491165161,
+      "learning_rate": 2.582433200682206e-05,
+      "loss": 5.2699,
+      "step": 9005
+    },
+    {
+      "epoch": 0.498065229408513,
+      "grad_norm": 3.2580060958862305,
+      "learning_rate": 2.5810119386014782e-05,
+      "loss": 5.0649,
+      "step": 9010
+    },
+    {
+      "epoch": 0.49834162520729686,
+      "grad_norm": 3.0472872257232666,
+      "learning_rate": 2.5795906765207506e-05,
+      "loss": 5.1669,
+      "step": 9015
+    },
+    {
+      "epoch": 0.4986180210060807,
+      "grad_norm": 3.3452460765838623,
+      "learning_rate": 2.578169414440023e-05,
+      "loss": 5.4016,
+      "step": 9020
+    },
+    {
+      "epoch": 0.49889441680486457,
+      "grad_norm": 3.5364160537719727,
+      "learning_rate": 2.576748152359295e-05,
+      "loss": 5.2059,
+      "step": 9025
+    },
+    {
+      "epoch": 0.49917081260364843,
+      "grad_norm": 3.439129114151001,
+      "learning_rate": 2.5753268902785678e-05,
+      "loss": 5.4018,
+      "step": 9030
+    },
+    {
+      "epoch": 0.4994472084024323,
+      "grad_norm": 4.166476249694824,
+      "learning_rate": 2.5739056281978395e-05,
+      "loss": 4.9097,
+      "step": 9035
+    },
+    {
+      "epoch": 0.49972360420121614,
+      "grad_norm": 2.5692834854125977,
+      "learning_rate": 2.5724843661171122e-05,
+      "loss": 5.2335,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3796117305755615,
+      "learning_rate": 2.5710631040363843e-05,
+      "loss": 5.1816,
+      "step": 9045
+    },
+    {
+      "epoch": 0.5002763957987839,
+      "grad_norm": 3.1776998043060303,
+      "learning_rate": 2.5696418419556567e-05,
+      "loss": 5.1573,
+      "step": 9050
+    },
+    {
+      "epoch": 0.5005527915975677,
+      "grad_norm": 3.153939723968506,
+      "learning_rate": 2.5682205798749294e-05,
+      "loss": 5.2645,
+      "step": 9055
+    },
+    {
+      "epoch": 0.5008291873963516,
+      "grad_norm": 2.713528633117676,
+      "learning_rate": 2.566799317794201e-05,
+      "loss": 4.9227,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5011055831951354,
+      "grad_norm": 3.3882882595062256,
+      "learning_rate": 2.565378055713474e-05,
+      "loss": 5.1723,
+      "step": 9065
+    },
+    {
+      "epoch": 0.5013819789939193,
+      "grad_norm": 3.6385247707366943,
+      "learning_rate": 2.5639567936327456e-05,
+      "loss": 5.0634,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5016583747927031,
+      "grad_norm": 3.6225831508636475,
+      "learning_rate": 2.5625355315520183e-05,
+      "loss": 4.8257,
+      "step": 9075
+    },
+    {
+      "epoch": 0.5019347705914871,
+      "grad_norm": 3.2693662643432617,
+      "learning_rate": 2.5611142694712907e-05,
+      "loss": 5.0228,
+      "step": 9080
+    },
+    {
+      "epoch": 0.5022111663902709,
+      "grad_norm": 3.6240193843841553,
+      "learning_rate": 2.5596930073905628e-05,
+      "loss": 5.0373,
+      "step": 9085
+    },
+    {
+      "epoch": 0.5024875621890548,
+      "grad_norm": 3.114393472671509,
+      "learning_rate": 2.5582717453098355e-05,
+      "loss": 5.4028,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5027639579878386,
+      "grad_norm": 3.7577624320983887,
+      "learning_rate": 2.5568504832291073e-05,
+      "loss": 5.0717,
+      "step": 9095
+    },
+    {
+      "epoch": 0.5030403537866225,
+      "grad_norm": 3.0983474254608154,
+      "learning_rate": 2.55542922114838e-05,
+      "loss": 5.5069,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5033167495854063,
+      "grad_norm": 3.490330696105957,
+      "learning_rate": 2.5540079590676524e-05,
+      "loss": 5.1826,
+      "step": 9105
+    },
+    {
+      "epoch": 0.5035931453841902,
+      "grad_norm": 4.066411018371582,
+      "learning_rate": 2.5525866969869244e-05,
+      "loss": 5.2115,
+      "step": 9110
+    },
+    {
+      "epoch": 0.503869541182974,
+      "grad_norm": 3.8119757175445557,
+      "learning_rate": 2.551165434906197e-05,
+      "loss": 5.2514,
+      "step": 9115
+    },
+    {
+      "epoch": 0.5041459369817579,
+      "grad_norm": 3.7238693237304688,
+      "learning_rate": 2.549744172825469e-05,
+      "loss": 4.7289,
+      "step": 9120
+    },
+    {
+      "epoch": 0.5044223327805417,
+      "grad_norm": 3.154292583465576,
+      "learning_rate": 2.5483229107447416e-05,
+      "loss": 5.1676,
+      "step": 9125
+    },
+    {
+      "epoch": 0.5046987285793256,
+      "grad_norm": 3.4549925327301025,
+      "learning_rate": 2.546901648664014e-05,
+      "loss": 5.1312,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5049751243781094,
+      "grad_norm": 3.0969717502593994,
+      "learning_rate": 2.545480386583286e-05,
+      "loss": 5.2531,
+      "step": 9135
+    },
+    {
+      "epoch": 0.5052515201768933,
+      "grad_norm": 6.122801780700684,
+      "learning_rate": 2.5440591245025585e-05,
+      "loss": 5.0781,
+      "step": 9140
+    },
+    {
+      "epoch": 0.5055279159756771,
+      "grad_norm": 3.613739013671875,
+      "learning_rate": 2.5426378624218306e-05,
+      "loss": 5.2013,
+      "step": 9145
+    },
+    {
+      "epoch": 0.5058043117744611,
+      "grad_norm": 3.0882883071899414,
+      "learning_rate": 2.541216600341103e-05,
+      "loss": 5.1607,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5060807075732449,
+      "grad_norm": 3.7881696224212646,
+      "learning_rate": 2.539795338260375e-05,
+      "loss": 5.0951,
+      "step": 9155
+    },
+    {
+      "epoch": 0.5063571033720288,
+      "grad_norm": 3.161926507949829,
+      "learning_rate": 2.5383740761796477e-05,
+      "loss": 5.1998,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5066334991708126,
+      "grad_norm": 3.223527193069458,
+      "learning_rate": 2.53695281409892e-05,
+      "loss": 5.2554,
+      "step": 9165
+    },
+    {
+      "epoch": 0.5069098949695965,
+      "grad_norm": 3.8171050548553467,
+      "learning_rate": 2.5355315520181922e-05,
+      "loss": 5.2001,
+      "step": 9170
+    },
+    {
+      "epoch": 0.5071862907683803,
+      "grad_norm": 2.8175442218780518,
+      "learning_rate": 2.5341102899374646e-05,
+      "loss": 5.1805,
+      "step": 9175
+    },
+    {
+      "epoch": 0.5074626865671642,
+      "grad_norm": 3.0957937240600586,
+      "learning_rate": 2.5326890278567367e-05,
+      "loss": 5.4076,
+      "step": 9180
+    },
+    {
+      "epoch": 0.507739082365948,
+      "grad_norm": 2.975939989089966,
+      "learning_rate": 2.531267765776009e-05,
+      "loss": 5.2884,
+      "step": 9185
+    },
+    {
+      "epoch": 0.5080154781647319,
+      "grad_norm": 4.01318359375,
+      "learning_rate": 2.5298465036952818e-05,
+      "loss": 5.4881,
+      "step": 9190
+    },
+    {
+      "epoch": 0.5082918739635157,
+      "grad_norm": 2.6178340911865234,
+      "learning_rate": 2.528425241614554e-05,
+      "loss": 5.0587,
+      "step": 9195
+    },
+    {
+      "epoch": 0.5085682697622996,
+      "grad_norm": 2.9565234184265137,
+      "learning_rate": 2.5270039795338263e-05,
+      "loss": 5.16,
+      "step": 9200
+    },
+    {
+      "epoch": 0.5088446655610834,
+      "grad_norm": 3.489684820175171,
+      "learning_rate": 2.5255827174530983e-05,
+      "loss": 5.0378,
+      "step": 9205
+    },
+    {
+      "epoch": 0.5091210613598673,
+      "grad_norm": 2.891176700592041,
+      "learning_rate": 2.5241614553723707e-05,
+      "loss": 5.0495,
+      "step": 9210
+    },
+    {
+      "epoch": 0.5093974571586511,
+      "grad_norm": 3.31199312210083,
+      "learning_rate": 2.5227401932916434e-05,
+      "loss": 5.2217,
+      "step": 9215
+    },
+    {
+      "epoch": 0.5096738529574351,
+      "grad_norm": 2.859729290008545,
+      "learning_rate": 2.521318931210915e-05,
+      "loss": 5.2588,
+      "step": 9220
+    },
+    {
+      "epoch": 0.5099502487562189,
+      "grad_norm": 3.0680506229400635,
+      "learning_rate": 2.519897669130188e-05,
+      "loss": 4.9947,
+      "step": 9225
+    },
+    {
+      "epoch": 0.5102266445550028,
+      "grad_norm": 3.4097602367401123,
+      "learning_rate": 2.51847640704946e-05,
+      "loss": 5.156,
+      "step": 9230
+    },
+    {
+      "epoch": 0.5105030403537866,
+      "grad_norm": 3.063750743865967,
+      "learning_rate": 2.5170551449687324e-05,
+      "loss": 5.1035,
+      "step": 9235
+    },
+    {
+      "epoch": 0.5107794361525705,
+      "grad_norm": 3.8064587116241455,
+      "learning_rate": 2.515633882888005e-05,
+      "loss": 5.3786,
+      "step": 9240
+    },
+    {
+      "epoch": 0.5110558319513543,
+      "grad_norm": 2.8018996715545654,
+      "learning_rate": 2.5142126208072768e-05,
+      "loss": 5.0211,
+      "step": 9245
+    },
+    {
+      "epoch": 0.5113322277501382,
+      "grad_norm": 2.8001601696014404,
+      "learning_rate": 2.5127913587265496e-05,
+      "loss": 5.3029,
+      "step": 9250
+    },
+    {
+      "epoch": 0.511608623548922,
+      "grad_norm": 3.171142339706421,
+      "learning_rate": 2.5113700966458216e-05,
+      "loss": 4.9662,
+      "step": 9255
+    },
+    {
+      "epoch": 0.5118850193477059,
+      "grad_norm": 2.9701507091522217,
+      "learning_rate": 2.509948834565094e-05,
+      "loss": 4.8072,
+      "step": 9260
+    },
+    {
+      "epoch": 0.5121614151464897,
+      "grad_norm": 3.0788140296936035,
+      "learning_rate": 2.508527572484366e-05,
+      "loss": 4.9462,
+      "step": 9265
+    },
+    {
+      "epoch": 0.5124378109452736,
+      "grad_norm": 3.2449519634246826,
+      "learning_rate": 2.5071063104036385e-05,
+      "loss": 5.1457,
+      "step": 9270
+    },
+    {
+      "epoch": 0.5127142067440575,
+      "grad_norm": 3.2045812606811523,
+      "learning_rate": 2.5056850483229112e-05,
+      "loss": 5.3124,
+      "step": 9275
+    },
+    {
+      "epoch": 0.5129906025428413,
+      "grad_norm": 3.2076756954193115,
+      "learning_rate": 2.504263786242183e-05,
+      "loss": 5.187,
+      "step": 9280
+    },
+    {
+      "epoch": 0.5132669983416253,
+      "grad_norm": 3.471508264541626,
+      "learning_rate": 2.5028425241614557e-05,
+      "loss": 5.3885,
+      "step": 9285
+    },
+    {
+      "epoch": 0.5135433941404091,
+      "grad_norm": 3.8236048221588135,
+      "learning_rate": 2.5014212620807277e-05,
+      "loss": 5.1868,
+      "step": 9290
+    },
+    {
+      "epoch": 0.513819789939193,
+      "grad_norm": 3.278804302215576,
+      "learning_rate": 2.5e-05,
+      "loss": 5.2836,
+      "step": 9295
+    },
+    {
+      "epoch": 0.5140961857379768,
+      "grad_norm": 3.5111825466156006,
+      "learning_rate": 2.4985787379192725e-05,
+      "loss": 5.3669,
+      "step": 9300
+    },
+    {
+      "epoch": 0.5143725815367607,
+      "grad_norm": 3.7894766330718994,
+      "learning_rate": 2.4971574758385446e-05,
+      "loss": 4.9516,
+      "step": 9305
+    },
+    {
+      "epoch": 0.5146489773355445,
+      "grad_norm": 2.5481996536254883,
+      "learning_rate": 2.4957362137578173e-05,
+      "loss": 5.1002,
+      "step": 9310
+    },
+    {
+      "epoch": 0.5149253731343284,
+      "grad_norm": 3.0527219772338867,
+      "learning_rate": 2.4943149516770894e-05,
+      "loss": 5.3666,
+      "step": 9315
+    },
+    {
+      "epoch": 0.5152017689331122,
+      "grad_norm": 3.6825625896453857,
+      "learning_rate": 2.4928936895963618e-05,
+      "loss": 5.0798,
+      "step": 9320
+    },
+    {
+      "epoch": 0.5154781647318961,
+      "grad_norm": 3.366027593612671,
+      "learning_rate": 2.4914724275156338e-05,
+      "loss": 5.2866,
+      "step": 9325
+    },
+    {
+      "epoch": 0.5157545605306799,
+      "grad_norm": 3.307328462600708,
+      "learning_rate": 2.4900511654349062e-05,
+      "loss": 4.9818,
+      "step": 9330
+    },
+    {
+      "epoch": 0.5160309563294638,
+      "grad_norm": 3.386768102645874,
+      "learning_rate": 2.4886299033541786e-05,
+      "loss": 5.1491,
+      "step": 9335
+    },
+    {
+      "epoch": 0.5163073521282476,
+      "grad_norm": 3.7182559967041016,
+      "learning_rate": 2.487208641273451e-05,
+      "loss": 5.2406,
+      "step": 9340
+    },
+    {
+      "epoch": 0.5165837479270315,
+      "grad_norm": 4.657354354858398,
+      "learning_rate": 2.4857873791927234e-05,
+      "loss": 4.9486,
+      "step": 9345
+    },
+    {
+      "epoch": 0.5168601437258153,
+      "grad_norm": 3.5175914764404297,
+      "learning_rate": 2.4843661171119955e-05,
+      "loss": 5.1466,
+      "step": 9350
+    },
+    {
+      "epoch": 0.5171365395245993,
+      "grad_norm": 3.5786592960357666,
+      "learning_rate": 2.482944855031268e-05,
+      "loss": 5.1679,
+      "step": 9355
+    },
+    {
+      "epoch": 0.5174129353233831,
+      "grad_norm": 2.8684964179992676,
+      "learning_rate": 2.48152359295054e-05,
+      "loss": 4.9781,
+      "step": 9360
+    },
+    {
+      "epoch": 0.517689331122167,
+      "grad_norm": 3.130319356918335,
+      "learning_rate": 2.4801023308698127e-05,
+      "loss": 5.1694,
+      "step": 9365
+    },
+    {
+      "epoch": 0.5179657269209508,
+      "grad_norm": 3.6741483211517334,
+      "learning_rate": 2.478681068789085e-05,
+      "loss": 4.9329,
+      "step": 9370
+    },
+    {
+      "epoch": 0.5182421227197347,
+      "grad_norm": 3.3555543422698975,
+      "learning_rate": 2.477259806708357e-05,
+      "loss": 5.0482,
+      "step": 9375
+    },
+    {
+      "epoch": 0.5185185185185185,
+      "grad_norm": 3.1766357421875,
+      "learning_rate": 2.4758385446276295e-05,
+      "loss": 5.0693,
+      "step": 9380
+    },
+    {
+      "epoch": 0.5187949143173024,
+      "grad_norm": 3.1514878273010254,
+      "learning_rate": 2.4744172825469016e-05,
+      "loss": 4.7144,
+      "step": 9385
+    },
+    {
+      "epoch": 0.5190713101160862,
+      "grad_norm": 3.2456977367401123,
+      "learning_rate": 2.472996020466174e-05,
+      "loss": 4.9498,
+      "step": 9390
+    },
+    {
+      "epoch": 0.5193477059148701,
+      "grad_norm": 3.266662836074829,
+      "learning_rate": 2.4715747583854464e-05,
+      "loss": 5.3611,
+      "step": 9395
+    },
+    {
+      "epoch": 0.5196241017136539,
+      "grad_norm": 2.9359843730926514,
+      "learning_rate": 2.4701534963047188e-05,
+      "loss": 5.0735,
+      "step": 9400
+    },
+    {
+      "epoch": 0.5199004975124378,
+      "grad_norm": 2.956838369369507,
+      "learning_rate": 2.4687322342239912e-05,
+      "loss": 5.2266,
+      "step": 9405
+    },
+    {
+      "epoch": 0.5201768933112216,
+      "grad_norm": 3.2221481800079346,
+      "learning_rate": 2.4673109721432632e-05,
+      "loss": 4.8549,
+      "step": 9410
+    },
+    {
+      "epoch": 0.5204532891100055,
+      "grad_norm": 4.0131635665893555,
+      "learning_rate": 2.4658897100625356e-05,
+      "loss": 5.0946,
+      "step": 9415
+    },
+    {
+      "epoch": 0.5207296849087893,
+      "grad_norm": 3.1099162101745605,
+      "learning_rate": 2.464468447981808e-05,
+      "loss": 5.4147,
+      "step": 9420
+    },
+    {
+      "epoch": 0.5210060807075733,
+      "grad_norm": 2.945998430252075,
+      "learning_rate": 2.4630471859010804e-05,
+      "loss": 5.0275,
+      "step": 9425
+    },
+    {
+      "epoch": 0.5212824765063571,
+      "grad_norm": 3.6736671924591064,
+      "learning_rate": 2.4616259238203525e-05,
+      "loss": 5.0205,
+      "step": 9430
+    },
+    {
+      "epoch": 0.521558872305141,
+      "grad_norm": 3.3827338218688965,
+      "learning_rate": 2.460204661739625e-05,
+      "loss": 5.2464,
+      "step": 9435
+    },
+    {
+      "epoch": 0.5218352681039248,
+      "grad_norm": 3.2636919021606445,
+      "learning_rate": 2.4587833996588973e-05,
+      "loss": 4.7875,
+      "step": 9440
+    },
+    {
+      "epoch": 0.5221116639027087,
+      "grad_norm": 2.948478937149048,
+      "learning_rate": 2.4573621375781693e-05,
+      "loss": 5.0072,
+      "step": 9445
+    },
+    {
+      "epoch": 0.5223880597014925,
+      "grad_norm": 3.093000888824463,
+      "learning_rate": 2.455940875497442e-05,
+      "loss": 4.9569,
+      "step": 9450
+    },
+    {
+      "epoch": 0.5226644555002764,
+      "grad_norm": 3.29803466796875,
+      "learning_rate": 2.454519613416714e-05,
+      "loss": 5.0124,
+      "step": 9455
+    },
+    {
+      "epoch": 0.5229408512990602,
+      "grad_norm": 2.948936700820923,
+      "learning_rate": 2.4530983513359865e-05,
+      "loss": 5.0278,
+      "step": 9460
+    },
+    {
+      "epoch": 0.5232172470978441,
+      "grad_norm": 3.211668014526367,
+      "learning_rate": 2.4516770892552586e-05,
+      "loss": 5.1416,
+      "step": 9465
+    },
+    {
+      "epoch": 0.5234936428966279,
+      "grad_norm": 4.215364456176758,
+      "learning_rate": 2.450255827174531e-05,
+      "loss": 4.8566,
+      "step": 9470
+    },
+    {
+      "epoch": 0.5237700386954118,
+      "grad_norm": 3.8499324321746826,
+      "learning_rate": 2.4488345650938034e-05,
+      "loss": 4.9764,
+      "step": 9475
+    },
+    {
+      "epoch": 0.5240464344941956,
+      "grad_norm": 3.159186840057373,
+      "learning_rate": 2.4474133030130758e-05,
+      "loss": 5.1786,
+      "step": 9480
+    },
+    {
+      "epoch": 0.5243228302929795,
+      "grad_norm": 3.4853217601776123,
+      "learning_rate": 2.4459920409323482e-05,
+      "loss": 5.2136,
+      "step": 9485
+    },
+    {
+      "epoch": 0.5245992260917635,
+      "grad_norm": 3.228022813796997,
+      "learning_rate": 2.4445707788516203e-05,
+      "loss": 5.498,
+      "step": 9490
+    },
+    {
+      "epoch": 0.5248756218905473,
+      "grad_norm": 2.8631765842437744,
+      "learning_rate": 2.4431495167708926e-05,
+      "loss": 5.1078,
+      "step": 9495
+    },
+    {
+      "epoch": 0.5251520176893312,
+      "grad_norm": 3.351837396621704,
+      "learning_rate": 2.4417282546901647e-05,
+      "loss": 5.2389,
+      "step": 9500
+    },
+    {
+      "epoch": 0.525428413488115,
+      "grad_norm": 4.084488391876221,
+      "learning_rate": 2.4403069926094374e-05,
+      "loss": 5.1804,
+      "step": 9505
+    },
+    {
+      "epoch": 0.5257048092868989,
+      "grad_norm": 4.037084579467773,
+      "learning_rate": 2.4388857305287095e-05,
+      "loss": 5.162,
+      "step": 9510
+    },
+    {
+      "epoch": 0.5259812050856827,
+      "grad_norm": 3.0944149494171143,
+      "learning_rate": 2.437464468447982e-05,
+      "loss": 5.0158,
+      "step": 9515
+    },
+    {
+      "epoch": 0.5262576008844666,
+      "grad_norm": 3.6755268573760986,
+      "learning_rate": 2.4360432063672543e-05,
+      "loss": 5.1988,
+      "step": 9520
+    },
+    {
+      "epoch": 0.5265339966832504,
+      "grad_norm": 3.344731330871582,
+      "learning_rate": 2.4346219442865264e-05,
+      "loss": 5.2125,
+      "step": 9525
+    },
+    {
+      "epoch": 0.5268103924820343,
+      "grad_norm": 4.068390369415283,
+      "learning_rate": 2.433200682205799e-05,
+      "loss": 5.4192,
+      "step": 9530
+    },
+    {
+      "epoch": 0.5270867882808181,
+      "grad_norm": 2.7342958450317383,
+      "learning_rate": 2.431779420125071e-05,
+      "loss": 5.1935,
+      "step": 9535
+    },
+    {
+      "epoch": 0.527363184079602,
+      "grad_norm": 3.347498893737793,
+      "learning_rate": 2.4303581580443436e-05,
+      "loss": 4.9082,
+      "step": 9540
+    },
+    {
+      "epoch": 0.5276395798783858,
+      "grad_norm": 3.082366466522217,
+      "learning_rate": 2.428936895963616e-05,
+      "loss": 5.1855,
+      "step": 9545
+    },
+    {
+      "epoch": 0.5279159756771697,
+      "grad_norm": 3.0448951721191406,
+      "learning_rate": 2.427515633882888e-05,
+      "loss": 5.3123,
+      "step": 9550
+    },
+    {
+      "epoch": 0.5281923714759535,
+      "grad_norm": 3.1284632682800293,
+      "learning_rate": 2.4260943718021604e-05,
+      "loss": 5.0266,
+      "step": 9555
+    },
+    {
+      "epoch": 0.5284687672747375,
+      "grad_norm": 3.3635149002075195,
+      "learning_rate": 2.4246731097214328e-05,
+      "loss": 5.1604,
+      "step": 9560
+    },
+    {
+      "epoch": 0.5287451630735213,
+      "grad_norm": 4.114374160766602,
+      "learning_rate": 2.4232518476407052e-05,
+      "loss": 5.027,
+      "step": 9565
+    },
+    {
+      "epoch": 0.5290215588723052,
+      "grad_norm": 2.970109701156616,
+      "learning_rate": 2.4218305855599773e-05,
+      "loss": 5.0342,
+      "step": 9570
+    },
+    {
+      "epoch": 0.529297954671089,
+      "grad_norm": 3.081437110900879,
+      "learning_rate": 2.4204093234792497e-05,
+      "loss": 4.9838,
+      "step": 9575
+    },
+    {
+      "epoch": 0.5295743504698729,
+      "grad_norm": 3.3371689319610596,
+      "learning_rate": 2.418988061398522e-05,
+      "loss": 5.0638,
+      "step": 9580
+    },
+    {
+      "epoch": 0.5298507462686567,
+      "grad_norm": 3.4338042736053467,
+      "learning_rate": 2.4175667993177945e-05,
+      "loss": 5.2537,
+      "step": 9585
+    },
+    {
+      "epoch": 0.5301271420674406,
+      "grad_norm": 2.9125192165374756,
+      "learning_rate": 2.416145537237067e-05,
+      "loss": 5.1985,
+      "step": 9590
+    },
+    {
+      "epoch": 0.5304035378662244,
+      "grad_norm": 2.703503131866455,
+      "learning_rate": 2.414724275156339e-05,
+      "loss": 5.0267,
+      "step": 9595
+    },
+    {
+      "epoch": 0.5306799336650083,
+      "grad_norm": 2.609586238861084,
+      "learning_rate": 2.4133030130756113e-05,
+      "loss": 5.181,
+      "step": 9600
+    },
+    {
+      "epoch": 0.5309563294637921,
+      "grad_norm": 3.443605661392212,
+      "learning_rate": 2.4118817509948834e-05,
+      "loss": 5.0051,
+      "step": 9605
+    },
+    {
+      "epoch": 0.531232725262576,
+      "grad_norm": 3.419793128967285,
+      "learning_rate": 2.4104604889141558e-05,
+      "loss": 5.0689,
+      "step": 9610
+    },
+    {
+      "epoch": 0.5315091210613598,
+      "grad_norm": 3.239157199859619,
+      "learning_rate": 2.409039226833428e-05,
+      "loss": 5.023,
+      "step": 9615
+    },
+    {
+      "epoch": 0.5317855168601437,
+      "grad_norm": 3.928612470626831,
+      "learning_rate": 2.4076179647527006e-05,
+      "loss": 5.0672,
+      "step": 9620
+    },
+    {
+      "epoch": 0.5320619126589275,
+      "grad_norm": 3.0533933639526367,
+      "learning_rate": 2.406196702671973e-05,
+      "loss": 4.8787,
+      "step": 9625
+    },
+    {
+      "epoch": 0.5323383084577115,
+      "grad_norm": 2.9441888332366943,
+      "learning_rate": 2.404775440591245e-05,
+      "loss": 4.7492,
+      "step": 9630
+    },
+    {
+      "epoch": 0.5326147042564953,
+      "grad_norm": 3.513918399810791,
+      "learning_rate": 2.4033541785105174e-05,
+      "loss": 5.1696,
+      "step": 9635
+    },
+    {
+      "epoch": 0.5328911000552792,
+      "grad_norm": 3.191592216491699,
+      "learning_rate": 2.4019329164297898e-05,
+      "loss": 5.0818,
+      "step": 9640
+    },
+    {
+      "epoch": 0.533167495854063,
+      "grad_norm": 3.9577038288116455,
+      "learning_rate": 2.4005116543490622e-05,
+      "loss": 5.0411,
+      "step": 9645
+    },
+    {
+      "epoch": 0.5334438916528469,
+      "grad_norm": 3.297288656234741,
+      "learning_rate": 2.3990903922683343e-05,
+      "loss": 5.009,
+      "step": 9650
+    },
+    {
+      "epoch": 0.5337202874516307,
+      "grad_norm": 2.9676246643066406,
+      "learning_rate": 2.3976691301876067e-05,
+      "loss": 5.2229,
+      "step": 9655
+    },
+    {
+      "epoch": 0.5339966832504146,
+      "grad_norm": 3.918351650238037,
+      "learning_rate": 2.396247868106879e-05,
+      "loss": 5.3025,
+      "step": 9660
+    },
+    {
+      "epoch": 0.5342730790491984,
+      "grad_norm": 3.9034459590911865,
+      "learning_rate": 2.394826606026151e-05,
+      "loss": 4.985,
+      "step": 9665
+    },
+    {
+      "epoch": 0.5345494748479823,
+      "grad_norm": 3.6231179237365723,
+      "learning_rate": 2.393405343945424e-05,
+      "loss": 4.8507,
+      "step": 9670
+    },
+    {
+      "epoch": 0.5348258706467661,
+      "grad_norm": 2.965517044067383,
+      "learning_rate": 2.391984081864696e-05,
+      "loss": 5.2807,
+      "step": 9675
+    },
+    {
+      "epoch": 0.53510226644555,
+      "grad_norm": 3.329669713973999,
+      "learning_rate": 2.3905628197839683e-05,
+      "loss": 5.1108,
+      "step": 9680
+    },
+    {
+      "epoch": 0.5353786622443338,
+      "grad_norm": 3.134333372116089,
+      "learning_rate": 2.3891415577032407e-05,
+      "loss": 5.1821,
+      "step": 9685
+    },
+    {
+      "epoch": 0.5356550580431177,
+      "grad_norm": 2.873486042022705,
+      "learning_rate": 2.3877202956225128e-05,
+      "loss": 5.0137,
+      "step": 9690
+    },
+    {
+      "epoch": 0.5359314538419016,
+      "grad_norm": 3.7581984996795654,
+      "learning_rate": 2.3862990335417855e-05,
+      "loss": 5.1051,
+      "step": 9695
+    },
+    {
+      "epoch": 0.5362078496406855,
+      "grad_norm": 4.077391147613525,
+      "learning_rate": 2.3848777714610576e-05,
+      "loss": 5.2823,
+      "step": 9700
+    },
+    {
+      "epoch": 0.5364842454394693,
+      "grad_norm": 3.292609930038452,
+      "learning_rate": 2.38345650938033e-05,
+      "loss": 5.0104,
+      "step": 9705
+    },
+    {
+      "epoch": 0.5367606412382532,
+      "grad_norm": 3.4207916259765625,
+      "learning_rate": 2.382035247299602e-05,
+      "loss": 5.2349,
+      "step": 9710
+    },
+    {
+      "epoch": 0.5370370370370371,
+      "grad_norm": 3.6559712886810303,
+      "learning_rate": 2.3806139852188744e-05,
+      "loss": 5.3262,
+      "step": 9715
+    },
+    {
+      "epoch": 0.5373134328358209,
+      "grad_norm": 2.987576484680176,
+      "learning_rate": 2.3791927231381468e-05,
+      "loss": 5.1957,
+      "step": 9720
+    },
+    {
+      "epoch": 0.5375898286346048,
+      "grad_norm": 3.834031105041504,
+      "learning_rate": 2.3777714610574192e-05,
+      "loss": 5.2105,
+      "step": 9725
+    },
+    {
+      "epoch": 0.5378662244333886,
+      "grad_norm": 2.903601884841919,
+      "learning_rate": 2.3763501989766916e-05,
+      "loss": 5.0302,
+      "step": 9730
+    },
+    {
+      "epoch": 0.5381426202321725,
+      "grad_norm": 4.139296531677246,
+      "learning_rate": 2.3749289368959637e-05,
+      "loss": 5.1185,
+      "step": 9735
+    },
+    {
+      "epoch": 0.5384190160309563,
+      "grad_norm": 3.0780575275421143,
+      "learning_rate": 2.373507674815236e-05,
+      "loss": 5.0262,
+      "step": 9740
+    },
+    {
+      "epoch": 0.5386954118297402,
+      "grad_norm": 2.722825288772583,
+      "learning_rate": 2.372086412734508e-05,
+      "loss": 4.9768,
+      "step": 9745
+    },
+    {
+      "epoch": 0.538971807628524,
+      "grad_norm": 2.6656720638275146,
+      "learning_rate": 2.370665150653781e-05,
+      "loss": 5.0315,
+      "step": 9750
+    },
+    {
+      "epoch": 0.539248203427308,
+      "grad_norm": 3.2058663368225098,
+      "learning_rate": 2.369243888573053e-05,
+      "loss": 4.9424,
+      "step": 9755
+    },
+    {
+      "epoch": 0.5395245992260918,
+      "grad_norm": 3.5652551651000977,
+      "learning_rate": 2.3678226264923253e-05,
+      "loss": 4.9202,
+      "step": 9760
+    },
+    {
+      "epoch": 0.5398009950248757,
+      "grad_norm": 4.415959358215332,
+      "learning_rate": 2.3664013644115977e-05,
+      "loss": 5.3065,
+      "step": 9765
+    },
+    {
+      "epoch": 0.5400773908236595,
+      "grad_norm": 3.7161080837249756,
+      "learning_rate": 2.3649801023308698e-05,
+      "loss": 5.3898,
+      "step": 9770
+    },
+    {
+      "epoch": 0.5403537866224434,
+      "grad_norm": 4.07873010635376,
+      "learning_rate": 2.3635588402501422e-05,
+      "loss": 5.0935,
+      "step": 9775
+    },
+    {
+      "epoch": 0.5406301824212272,
+      "grad_norm": 4.138256549835205,
+      "learning_rate": 2.3621375781694146e-05,
+      "loss": 4.8865,
+      "step": 9780
+    },
+    {
+      "epoch": 0.5409065782200111,
+      "grad_norm": 3.804762601852417,
+      "learning_rate": 2.360716316088687e-05,
+      "loss": 5.1764,
+      "step": 9785
+    },
+    {
+      "epoch": 0.5411829740187949,
+      "grad_norm": 3.2528018951416016,
+      "learning_rate": 2.359295054007959e-05,
+      "loss": 5.3817,
+      "step": 9790
+    },
+    {
+      "epoch": 0.5414593698175788,
+      "grad_norm": 3.966082811355591,
+      "learning_rate": 2.3578737919272314e-05,
+      "loss": 5.1415,
+      "step": 9795
+    },
+    {
+      "epoch": 0.5417357656163626,
+      "grad_norm": 3.4930150508880615,
+      "learning_rate": 2.356452529846504e-05,
+      "loss": 5.2255,
+      "step": 9800
+    },
+    {
+      "epoch": 0.5420121614151465,
+      "grad_norm": 3.7764456272125244,
+      "learning_rate": 2.355031267765776e-05,
+      "loss": 5.3015,
+      "step": 9805
+    },
+    {
+      "epoch": 0.5422885572139303,
+      "grad_norm": 3.298708200454712,
+      "learning_rate": 2.3536100056850486e-05,
+      "loss": 4.8978,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5425649530127142,
+      "grad_norm": 3.262017011642456,
+      "learning_rate": 2.3521887436043207e-05,
+      "loss": 5.2939,
+      "step": 9815
+    },
+    {
+      "epoch": 0.542841348811498,
+      "grad_norm": 3.5957589149475098,
+      "learning_rate": 2.350767481523593e-05,
+      "loss": 5.3993,
+      "step": 9820
+    },
+    {
+      "epoch": 0.543117744610282,
+      "grad_norm": 3.324526071548462,
+      "learning_rate": 2.349346219442865e-05,
+      "loss": 4.8801,
+      "step": 9825
+    },
+    {
+      "epoch": 0.5433941404090658,
+      "grad_norm": 2.6901962757110596,
+      "learning_rate": 2.3479249573621375e-05,
+      "loss": 4.8502,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5436705362078497,
+      "grad_norm": 3.8433539867401123,
+      "learning_rate": 2.3465036952814103e-05,
+      "loss": 5.1071,
+      "step": 9835
+    },
+    {
+      "epoch": 0.5439469320066335,
+      "grad_norm": 3.26355242729187,
+      "learning_rate": 2.3450824332006823e-05,
+      "loss": 5.0674,
+      "step": 9840
+    },
+    {
+      "epoch": 0.5442233278054174,
+      "grad_norm": 3.1263980865478516,
+      "learning_rate": 2.3436611711199547e-05,
+      "loss": 5.2538,
+      "step": 9845
+    },
+    {
+      "epoch": 0.5444997236042012,
+      "grad_norm": 3.164491891860962,
+      "learning_rate": 2.3422399090392268e-05,
+      "loss": 5.1808,
+      "step": 9850
+    },
+    {
+      "epoch": 0.5447761194029851,
+      "grad_norm": 2.887119770050049,
+      "learning_rate": 2.3408186469584992e-05,
+      "loss": 4.8469,
+      "step": 9855
+    },
+    {
+      "epoch": 0.5450525152017689,
+      "grad_norm": 3.320695161819458,
+      "learning_rate": 2.3393973848777716e-05,
+      "loss": 4.9163,
+      "step": 9860
+    },
+    {
+      "epoch": 0.5453289110005528,
+      "grad_norm": 3.467114210128784,
+      "learning_rate": 2.337976122797044e-05,
+      "loss": 5.0402,
+      "step": 9865
+    },
+    {
+      "epoch": 0.5456053067993366,
+      "grad_norm": 3.7807321548461914,
+      "learning_rate": 2.3365548607163164e-05,
+      "loss": 5.4195,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5458817025981205,
+      "grad_norm": 3.7920944690704346,
+      "learning_rate": 2.3351335986355885e-05,
+      "loss": 4.8416,
+      "step": 9875
+    },
+    {
+      "epoch": 0.5461580983969043,
+      "grad_norm": 3.7439687252044678,
+      "learning_rate": 2.333712336554861e-05,
+      "loss": 5.0364,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5464344941956882,
+      "grad_norm": 3.955336093902588,
+      "learning_rate": 2.332291074474133e-05,
+      "loss": 5.5306,
+      "step": 9885
+    },
+    {
+      "epoch": 0.546710889994472,
+      "grad_norm": 3.5154666900634766,
+      "learning_rate": 2.3308698123934056e-05,
+      "loss": 5.1119,
+      "step": 9890
+    },
+    {
+      "epoch": 0.546987285793256,
+      "grad_norm": 3.323625326156616,
+      "learning_rate": 2.3294485503126777e-05,
+      "loss": 5.1608,
+      "step": 9895
+    },
+    {
+      "epoch": 0.5472636815920398,
+      "grad_norm": 2.7888081073760986,
+      "learning_rate": 2.32802728823195e-05,
+      "loss": 5.173,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5475400773908237,
+      "grad_norm": 3.7210378646850586,
+      "learning_rate": 2.3266060261512225e-05,
+      "loss": 5.3936,
+      "step": 9905
+    },
+    {
+      "epoch": 0.5478164731896075,
+      "grad_norm": 3.4708786010742188,
+      "learning_rate": 2.3251847640704946e-05,
+      "loss": 5.1886,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5480928689883914,
+      "grad_norm": 3.851804494857788,
+      "learning_rate": 2.323763501989767e-05,
+      "loss": 5.0378,
+      "step": 9915
+    },
+    {
+      "epoch": 0.5483692647871752,
+      "grad_norm": 3.7334494590759277,
+      "learning_rate": 2.3223422399090394e-05,
+      "loss": 4.8619,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5486456605859591,
+      "grad_norm": 3.100353479385376,
+      "learning_rate": 2.3209209778283118e-05,
+      "loss": 4.8678,
+      "step": 9925
+    },
+    {
+      "epoch": 0.548922056384743,
+      "grad_norm": 3.700317621231079,
+      "learning_rate": 2.3194997157475838e-05,
+      "loss": 5.2288,
+      "step": 9930
+    },
+    {
+      "epoch": 0.5491984521835268,
+      "grad_norm": 2.94370174407959,
+      "learning_rate": 2.3180784536668562e-05,
+      "loss": 5.097,
+      "step": 9935
+    },
+    {
+      "epoch": 0.5494748479823107,
+      "grad_norm": 2.9802255630493164,
+      "learning_rate": 2.3166571915861286e-05,
+      "loss": 5.1708,
+      "step": 9940
+    },
+    {
+      "epoch": 0.5497512437810945,
+      "grad_norm": 3.3415024280548096,
+      "learning_rate": 2.315235929505401e-05,
+      "loss": 5.2635,
+      "step": 9945
+    },
+    {
+      "epoch": 0.5500276395798784,
+      "grad_norm": 2.9417765140533447,
+      "learning_rate": 2.3138146674246734e-05,
+      "loss": 5.2387,
+      "step": 9950
+    },
+    {
+      "epoch": 0.5503040353786622,
+      "grad_norm": 3.112800359725952,
+      "learning_rate": 2.3123934053439455e-05,
+      "loss": 5.0443,
+      "step": 9955
+    },
+    {
+      "epoch": 0.5505804311774462,
+      "grad_norm": 2.728961944580078,
+      "learning_rate": 2.310972143263218e-05,
+      "loss": 5.0209,
+      "step": 9960
+    },
+    {
+      "epoch": 0.55085682697623,
+      "grad_norm": 3.2710981369018555,
+      "learning_rate": 2.30955088118249e-05,
+      "loss": 5.2521,
+      "step": 9965
+    },
+    {
+      "epoch": 0.5511332227750139,
+      "grad_norm": 3.748812675476074,
+      "learning_rate": 2.3081296191017623e-05,
+      "loss": 5.3927,
+      "step": 9970
+    },
+    {
+      "epoch": 0.5514096185737977,
+      "grad_norm": 3.3356282711029053,
+      "learning_rate": 2.3067083570210347e-05,
+      "loss": 5.1706,
+      "step": 9975
+    },
+    {
+      "epoch": 0.5516860143725816,
+      "grad_norm": 2.5249369144439697,
+      "learning_rate": 2.305287094940307e-05,
+      "loss": 5.0159,
+      "step": 9980
+    },
+    {
+      "epoch": 0.5519624101713654,
+      "grad_norm": 3.26520037651062,
+      "learning_rate": 2.3038658328595795e-05,
+      "loss": 5.1818,
+      "step": 9985
+    },
+    {
+      "epoch": 0.5522388059701493,
+      "grad_norm": 3.965653896331787,
+      "learning_rate": 2.3024445707788516e-05,
+      "loss": 4.8768,
+      "step": 9990
+    },
+    {
+      "epoch": 0.5525152017689331,
+      "grad_norm": 4.425838947296143,
+      "learning_rate": 2.301023308698124e-05,
+      "loss": 5.2812,
+      "step": 9995
+    },
+    {
+      "epoch": 0.552791597567717,
+      "grad_norm": 3.3867270946502686,
+      "learning_rate": 2.2996020466173964e-05,
+      "loss": 5.1826,
+      "step": 10000
+    },
+    {
+      "epoch": 0.5530679933665008,
+      "grad_norm": 3.590095043182373,
+      "learning_rate": 2.2981807845366688e-05,
+      "loss": 5.2498,
+      "step": 10005
+    },
+    {
+      "epoch": 0.5533443891652847,
+      "grad_norm": 3.230135440826416,
+      "learning_rate": 2.296759522455941e-05,
+      "loss": 5.0662,
+      "step": 10010
+    },
+    {
+      "epoch": 0.5536207849640685,
+      "grad_norm": 3.142805337905884,
+      "learning_rate": 2.2953382603752132e-05,
+      "loss": 5.1993,
+      "step": 10015
+    },
+    {
+      "epoch": 0.5538971807628524,
+      "grad_norm": 4.192418098449707,
+      "learning_rate": 2.2939169982944856e-05,
+      "loss": 5.1139,
+      "step": 10020
+    },
+    {
+      "epoch": 0.5541735765616362,
+      "grad_norm": 3.831026315689087,
+      "learning_rate": 2.2924957362137577e-05,
+      "loss": 5.1316,
+      "step": 10025
+    },
+    {
+      "epoch": 0.5544499723604202,
+      "grad_norm": 3.54022216796875,
+      "learning_rate": 2.2910744741330304e-05,
+      "loss": 5.0901,
+      "step": 10030
+    },
+    {
+      "epoch": 0.554726368159204,
+      "grad_norm": 4.166489601135254,
+      "learning_rate": 2.2896532120523025e-05,
+      "loss": 5.1592,
+      "step": 10035
+    },
+    {
+      "epoch": 0.5550027639579879,
+      "grad_norm": 3.5303165912628174,
+      "learning_rate": 2.288231949971575e-05,
+      "loss": 4.9975,
+      "step": 10040
+    },
+    {
+      "epoch": 0.5552791597567717,
+      "grad_norm": 3.065279245376587,
+      "learning_rate": 2.2868106878908473e-05,
+      "loss": 5.1,
+      "step": 10045
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 4.913681507110596,
+      "learning_rate": 2.2853894258101193e-05,
+      "loss": 5.207,
+      "step": 10050
+    },
+    {
+      "epoch": 0.5558319513543394,
+      "grad_norm": 3.3215198516845703,
+      "learning_rate": 2.283968163729392e-05,
+      "loss": 5.0969,
+      "step": 10055
+    },
+    {
+      "epoch": 0.5561083471531233,
+      "grad_norm": 3.600184440612793,
+      "learning_rate": 2.282546901648664e-05,
+      "loss": 4.8922,
+      "step": 10060
+    },
+    {
+      "epoch": 0.5563847429519071,
+      "grad_norm": 3.041132926940918,
+      "learning_rate": 2.2811256395679365e-05,
+      "loss": 5.2081,
+      "step": 10065
+    },
+    {
+      "epoch": 0.556661138750691,
+      "grad_norm": 3.7934274673461914,
+      "learning_rate": 2.2797043774872086e-05,
+      "loss": 5.1229,
+      "step": 10070
+    },
+    {
+      "epoch": 0.5569375345494748,
+      "grad_norm": 2.82381010055542,
+      "learning_rate": 2.278283115406481e-05,
+      "loss": 5.2129,
+      "step": 10075
+    },
+    {
+      "epoch": 0.5572139303482587,
+      "grad_norm": 3.184391736984253,
+      "learning_rate": 2.2768618533257534e-05,
+      "loss": 5.199,
+      "step": 10080
+    },
+    {
+      "epoch": 0.5574903261470425,
+      "grad_norm": 2.923619031906128,
+      "learning_rate": 2.2754405912450258e-05,
+      "loss": 5.2113,
+      "step": 10085
+    },
+    {
+      "epoch": 0.5577667219458264,
+      "grad_norm": 3.289689540863037,
+      "learning_rate": 2.2740193291642982e-05,
+      "loss": 5.3625,
+      "step": 10090
+    },
+    {
+      "epoch": 0.5580431177446102,
+      "grad_norm": 3.128100633621216,
+      "learning_rate": 2.2725980670835702e-05,
+      "loss": 4.9404,
+      "step": 10095
+    },
+    {
+      "epoch": 0.5583195135433942,
+      "grad_norm": 3.116595983505249,
+      "learning_rate": 2.2711768050028426e-05,
+      "loss": 4.6508,
+      "step": 10100
+    },
+    {
+      "epoch": 0.558595909342178,
+      "grad_norm": 3.509528636932373,
+      "learning_rate": 2.2697555429221147e-05,
+      "loss": 5.3815,
+      "step": 10105
+    },
+    {
+      "epoch": 0.5588723051409619,
+      "grad_norm": 2.830974578857422,
+      "learning_rate": 2.2683342808413874e-05,
+      "loss": 5.0021,
+      "step": 10110
+    },
+    {
+      "epoch": 0.5591487009397457,
+      "grad_norm": 3.5547916889190674,
+      "learning_rate": 2.2669130187606595e-05,
+      "loss": 5.1721,
+      "step": 10115
+    },
+    {
+      "epoch": 0.5594250967385296,
+      "grad_norm": 3.2966182231903076,
+      "learning_rate": 2.265491756679932e-05,
+      "loss": 5.1488,
+      "step": 10120
+    },
+    {
+      "epoch": 0.5597014925373134,
+      "grad_norm": 3.2352070808410645,
+      "learning_rate": 2.2640704945992043e-05,
+      "loss": 4.8284,
+      "step": 10125
+    },
+    {
+      "epoch": 0.5599778883360973,
+      "grad_norm": 3.2427544593811035,
+      "learning_rate": 2.2626492325184763e-05,
+      "loss": 5.1183,
+      "step": 10130
+    },
+    {
+      "epoch": 0.5602542841348811,
+      "grad_norm": 2.5661115646362305,
+      "learning_rate": 2.2612279704377487e-05,
+      "loss": 5.1111,
+      "step": 10135
+    },
+    {
+      "epoch": 0.560530679933665,
+      "grad_norm": 3.292429208755493,
+      "learning_rate": 2.259806708357021e-05,
+      "loss": 5.2935,
+      "step": 10140
+    },
+    {
+      "epoch": 0.5608070757324488,
+      "grad_norm": 3.2697348594665527,
+      "learning_rate": 2.2583854462762935e-05,
+      "loss": 5.3207,
+      "step": 10145
+    },
+    {
+      "epoch": 0.5610834715312327,
+      "grad_norm": 2.9878225326538086,
+      "learning_rate": 2.256964184195566e-05,
+      "loss": 4.8464,
+      "step": 10150
+    },
+    {
+      "epoch": 0.5613598673300166,
+      "grad_norm": 3.4964377880096436,
+      "learning_rate": 2.255542922114838e-05,
+      "loss": 5.2375,
+      "step": 10155
+    },
+    {
+      "epoch": 0.5616362631288004,
+      "grad_norm": 3.2263262271881104,
+      "learning_rate": 2.2541216600341104e-05,
+      "loss": 5.1451,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5619126589275844,
+      "grad_norm": 2.8740015029907227,
+      "learning_rate": 2.2527003979533828e-05,
+      "loss": 4.9799,
+      "step": 10165
+    },
+    {
+      "epoch": 0.5621890547263682,
+      "grad_norm": 3.3548688888549805,
+      "learning_rate": 2.2512791358726552e-05,
+      "loss": 5.0609,
+      "step": 10170
+    },
+    {
+      "epoch": 0.5624654505251521,
+      "grad_norm": 3.348917245864868,
+      "learning_rate": 2.2498578737919272e-05,
+      "loss": 5.0375,
+      "step": 10175
+    },
+    {
+      "epoch": 0.5627418463239359,
+      "grad_norm": 2.9906487464904785,
+      "learning_rate": 2.2484366117111996e-05,
+      "loss": 5.109,
+      "step": 10180
+    },
+    {
+      "epoch": 0.5630182421227198,
+      "grad_norm": 3.7559046745300293,
+      "learning_rate": 2.247015349630472e-05,
+      "loss": 4.9875,
+      "step": 10185
+    },
+    {
+      "epoch": 0.5632946379215036,
+      "grad_norm": 4.169631004333496,
+      "learning_rate": 2.245594087549744e-05,
+      "loss": 5.0137,
+      "step": 10190
+    },
+    {
+      "epoch": 0.5635710337202875,
+      "grad_norm": 3.692657709121704,
+      "learning_rate": 2.244172825469017e-05,
+      "loss": 5.1208,
+      "step": 10195
+    },
+    {
+      "epoch": 0.5638474295190713,
+      "grad_norm": 4.431186199188232,
+      "learning_rate": 2.242751563388289e-05,
+      "loss": 5.0208,
+      "step": 10200
+    },
+    {
+      "epoch": 0.5641238253178552,
+      "grad_norm": 3.045299530029297,
+      "learning_rate": 2.2413303013075613e-05,
+      "loss": 4.8395,
+      "step": 10205
+    },
+    {
+      "epoch": 0.564400221116639,
+      "grad_norm": 2.9593913555145264,
+      "learning_rate": 2.2399090392268334e-05,
+      "loss": 5.0236,
+      "step": 10210
+    },
+    {
+      "epoch": 0.5646766169154229,
+      "grad_norm": 3.3709757328033447,
+      "learning_rate": 2.2384877771461058e-05,
+      "loss": 5.3689,
+      "step": 10215
+    },
+    {
+      "epoch": 0.5649530127142067,
+      "grad_norm": 3.7486398220062256,
+      "learning_rate": 2.237066515065378e-05,
+      "loss": 5.1189,
+      "step": 10220
+    },
+    {
+      "epoch": 0.5652294085129906,
+      "grad_norm": 2.766789674758911,
+      "learning_rate": 2.2356452529846505e-05,
+      "loss": 5.1984,
+      "step": 10225
+    },
+    {
+      "epoch": 0.5655058043117744,
+      "grad_norm": 3.7790632247924805,
+      "learning_rate": 2.234223990903923e-05,
+      "loss": 5.0615,
+      "step": 10230
+    },
+    {
+      "epoch": 0.5657822001105584,
+      "grad_norm": 3.1158392429351807,
+      "learning_rate": 2.232802728823195e-05,
+      "loss": 5.3859,
+      "step": 10235
+    },
+    {
+      "epoch": 0.5660585959093422,
+      "grad_norm": 3.7767012119293213,
+      "learning_rate": 2.2313814667424674e-05,
+      "loss": 5.1952,
+      "step": 10240
+    },
+    {
+      "epoch": 0.5663349917081261,
+      "grad_norm": 3.6504740715026855,
+      "learning_rate": 2.2299602046617395e-05,
+      "loss": 4.9556,
+      "step": 10245
+    },
+    {
+      "epoch": 0.5666113875069099,
+      "grad_norm": 3.328760862350464,
+      "learning_rate": 2.2285389425810122e-05,
+      "loss": 5.1943,
+      "step": 10250
+    },
+    {
+      "epoch": 0.5668877833056938,
+      "grad_norm": 2.9243085384368896,
+      "learning_rate": 2.2271176805002843e-05,
+      "loss": 4.821,
+      "step": 10255
+    },
+    {
+      "epoch": 0.5671641791044776,
+      "grad_norm": 3.004983425140381,
+      "learning_rate": 2.2256964184195567e-05,
+      "loss": 5.1676,
+      "step": 10260
+    },
+    {
+      "epoch": 0.5674405749032615,
+      "grad_norm": 3.309654474258423,
+      "learning_rate": 2.224275156338829e-05,
+      "loss": 4.9507,
+      "step": 10265
+    },
+    {
+      "epoch": 0.5677169707020453,
+      "grad_norm": 2.664963960647583,
+      "learning_rate": 2.222853894258101e-05,
+      "loss": 4.8282,
+      "step": 10270
+    },
+    {
+      "epoch": 0.5679933665008292,
+      "grad_norm": 3.0774271488189697,
+      "learning_rate": 2.221432632177374e-05,
+      "loss": 4.9935,
+      "step": 10275
+    },
+    {
+      "epoch": 0.568269762299613,
+      "grad_norm": 3.3895280361175537,
+      "learning_rate": 2.220011370096646e-05,
+      "loss": 5.2864,
+      "step": 10280
+    },
+    {
+      "epoch": 0.5685461580983969,
+      "grad_norm": 4.060614109039307,
+      "learning_rate": 2.2185901080159183e-05,
+      "loss": 5.4247,
+      "step": 10285
+    },
+    {
+      "epoch": 0.5688225538971807,
+      "grad_norm": 4.643860816955566,
+      "learning_rate": 2.2171688459351904e-05,
+      "loss": 5.0794,
+      "step": 10290
+    },
+    {
+      "epoch": 0.5690989496959646,
+      "grad_norm": 3.1240062713623047,
+      "learning_rate": 2.2157475838544628e-05,
+      "loss": 5.1919,
+      "step": 10295
+    },
+    {
+      "epoch": 0.5693753454947484,
+      "grad_norm": 2.9779624938964844,
+      "learning_rate": 2.214326321773735e-05,
+      "loss": 5.066,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5696517412935324,
+      "grad_norm": 3.8186936378479004,
+      "learning_rate": 2.2129050596930076e-05,
+      "loss": 5.2885,
+      "step": 10305
+    },
+    {
+      "epoch": 0.5699281370923162,
+      "grad_norm": 3.409640073776245,
+      "learning_rate": 2.21148379761228e-05,
+      "loss": 4.9745,
+      "step": 10310
+    },
+    {
+      "epoch": 0.5702045328911001,
+      "grad_norm": 4.826499938964844,
+      "learning_rate": 2.210062535531552e-05,
+      "loss": 5.1617,
+      "step": 10315
+    },
+    {
+      "epoch": 0.5704809286898839,
+      "grad_norm": 3.2550175189971924,
+      "learning_rate": 2.2086412734508244e-05,
+      "loss": 5.0947,
+      "step": 10320
+    },
+    {
+      "epoch": 0.5707573244886678,
+      "grad_norm": 3.287470579147339,
+      "learning_rate": 2.2072200113700968e-05,
+      "loss": 4.8948,
+      "step": 10325
+    },
+    {
+      "epoch": 0.5710337202874516,
+      "grad_norm": 2.8862740993499756,
+      "learning_rate": 2.2057987492893692e-05,
+      "loss": 4.8842,
+      "step": 10330
+    },
+    {
+      "epoch": 0.5713101160862355,
+      "grad_norm": 3.2922773361206055,
+      "learning_rate": 2.2043774872086416e-05,
+      "loss": 5.13,
+      "step": 10335
+    },
+    {
+      "epoch": 0.5715865118850193,
+      "grad_norm": 3.6037707328796387,
+      "learning_rate": 2.2029562251279137e-05,
+      "loss": 4.9092,
+      "step": 10340
+    },
+    {
+      "epoch": 0.5718629076838032,
+      "grad_norm": 3.365394115447998,
+      "learning_rate": 2.201534963047186e-05,
+      "loss": 5.3031,
+      "step": 10345
+    },
+    {
+      "epoch": 0.572139303482587,
+      "grad_norm": 4.060079574584961,
+      "learning_rate": 2.200113700966458e-05,
+      "loss": 5.4657,
+      "step": 10350
+    },
+    {
+      "epoch": 0.5724156992813709,
+      "grad_norm": 3.4800689220428467,
+      "learning_rate": 2.1986924388857305e-05,
+      "loss": 5.0092,
+      "step": 10355
+    },
+    {
+      "epoch": 0.5726920950801547,
+      "grad_norm": 3.4629600048065186,
+      "learning_rate": 2.197271176805003e-05,
+      "loss": 5.0127,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5729684908789386,
+      "grad_norm": 3.4687135219573975,
+      "learning_rate": 2.1958499147242753e-05,
+      "loss": 5.2457,
+      "step": 10365
+    },
+    {
+      "epoch": 0.5732448866777226,
+      "grad_norm": 3.8498566150665283,
+      "learning_rate": 2.1944286526435477e-05,
+      "loss": 5.4415,
+      "step": 10370
+    },
+    {
+      "epoch": 0.5735212824765064,
+      "grad_norm": 3.0555355548858643,
+      "learning_rate": 2.1930073905628198e-05,
+      "loss": 4.9743,
+      "step": 10375
+    },
+    {
+      "epoch": 0.5737976782752903,
+      "grad_norm": 3.5073928833007812,
+      "learning_rate": 2.1915861284820922e-05,
+      "loss": 5.0117,
+      "step": 10380
+    },
+    {
+      "epoch": 0.5740740740740741,
+      "grad_norm": 3.9393508434295654,
+      "learning_rate": 2.1901648664013646e-05,
+      "loss": 5.0825,
+      "step": 10385
+    },
+    {
+      "epoch": 0.574350469872858,
+      "grad_norm": 3.139087438583374,
+      "learning_rate": 2.188743604320637e-05,
+      "loss": 4.9374,
+      "step": 10390
+    },
+    {
+      "epoch": 0.5746268656716418,
+      "grad_norm": 3.968113660812378,
+      "learning_rate": 2.187322342239909e-05,
+      "loss": 5.1305,
+      "step": 10395
+    },
+    {
+      "epoch": 0.5749032614704257,
+      "grad_norm": 3.1973788738250732,
+      "learning_rate": 2.1859010801591814e-05,
+      "loss": 4.7387,
+      "step": 10400
+    },
+    {
+      "epoch": 0.5751796572692095,
+      "grad_norm": 3.473482847213745,
+      "learning_rate": 2.1844798180784538e-05,
+      "loss": 4.8958,
+      "step": 10405
+    },
+    {
+      "epoch": 0.5754560530679934,
+      "grad_norm": 3.418353796005249,
+      "learning_rate": 2.183058555997726e-05,
+      "loss": 4.8574,
+      "step": 10410
+    },
+    {
+      "epoch": 0.5757324488667772,
+      "grad_norm": 2.979325532913208,
+      "learning_rate": 2.1816372939169986e-05,
+      "loss": 5.1611,
+      "step": 10415
+    },
+    {
+      "epoch": 0.5760088446655611,
+      "grad_norm": 3.108954668045044,
+      "learning_rate": 2.1802160318362707e-05,
+      "loss": 5.1672,
+      "step": 10420
+    },
+    {
+      "epoch": 0.5762852404643449,
+      "grad_norm": 2.963421106338501,
+      "learning_rate": 2.178794769755543e-05,
+      "loss": 4.4638,
+      "step": 10425
+    },
+    {
+      "epoch": 0.5765616362631288,
+      "grad_norm": 4.1467461585998535,
+      "learning_rate": 2.177373507674815e-05,
+      "loss": 4.8556,
+      "step": 10430
+    },
+    {
+      "epoch": 0.5768380320619126,
+      "grad_norm": 4.003355503082275,
+      "learning_rate": 2.1759522455940875e-05,
+      "loss": 5.0152,
+      "step": 10435
+    },
+    {
+      "epoch": 0.5771144278606966,
+      "grad_norm": 3.502562999725342,
+      "learning_rate": 2.17453098351336e-05,
+      "loss": 5.0242,
+      "step": 10440
+    },
+    {
+      "epoch": 0.5773908236594804,
+      "grad_norm": 3.727094888687134,
+      "learning_rate": 2.1731097214326323e-05,
+      "loss": 5.2192,
+      "step": 10445
+    },
+    {
+      "epoch": 0.5776672194582643,
+      "grad_norm": 3.4509377479553223,
+      "learning_rate": 2.1716884593519047e-05,
+      "loss": 5.3475,
+      "step": 10450
+    },
+    {
+      "epoch": 0.5779436152570481,
+      "grad_norm": 3.1824676990509033,
+      "learning_rate": 2.1702671972711768e-05,
+      "loss": 4.8505,
+      "step": 10455
+    },
+    {
+      "epoch": 0.578220011055832,
+      "grad_norm": 4.357178211212158,
+      "learning_rate": 2.1688459351904492e-05,
+      "loss": 5.2004,
+      "step": 10460
+    },
+    {
+      "epoch": 0.5784964068546158,
+      "grad_norm": 4.844539642333984,
+      "learning_rate": 2.1674246731097216e-05,
+      "loss": 5.1357,
+      "step": 10465
+    },
+    {
+      "epoch": 0.5787728026533997,
+      "grad_norm": 3.694406509399414,
+      "learning_rate": 2.166003411028994e-05,
+      "loss": 4.9931,
+      "step": 10470
+    },
+    {
+      "epoch": 0.5790491984521835,
+      "grad_norm": 3.6863090991973877,
+      "learning_rate": 2.1645821489482664e-05,
+      "loss": 5.4604,
+      "step": 10475
+    },
+    {
+      "epoch": 0.5793255942509674,
+      "grad_norm": 3.620305299758911,
+      "learning_rate": 2.1631608868675384e-05,
+      "loss": 4.958,
+      "step": 10480
+    },
+    {
+      "epoch": 0.5796019900497512,
+      "grad_norm": 3.349443197250366,
+      "learning_rate": 2.161739624786811e-05,
+      "loss": 5.0294,
+      "step": 10485
+    },
+    {
+      "epoch": 0.5798783858485351,
+      "grad_norm": 3.268993854522705,
+      "learning_rate": 2.160318362706083e-05,
+      "loss": 5.0303,
+      "step": 10490
+    },
+    {
+      "epoch": 0.5801547816473189,
+      "grad_norm": 3.0426461696624756,
+      "learning_rate": 2.1588971006253556e-05,
+      "loss": 5.0315,
+      "step": 10495
+    },
+    {
+      "epoch": 0.5804311774461028,
+      "grad_norm": 3.8381333351135254,
+      "learning_rate": 2.1574758385446277e-05,
+      "loss": 5.1518,
+      "step": 10500
+    },
+    {
+      "epoch": 0.5807075732448866,
+      "grad_norm": 4.0485687255859375,
+      "learning_rate": 2.1560545764639e-05,
+      "loss": 5.1779,
+      "step": 10505
+    },
+    {
+      "epoch": 0.5809839690436706,
+      "grad_norm": 3.4191527366638184,
+      "learning_rate": 2.1546333143831725e-05,
+      "loss": 5.2983,
+      "step": 10510
+    },
+    {
+      "epoch": 0.5812603648424544,
+      "grad_norm": 3.35133695602417,
+      "learning_rate": 2.1532120523024445e-05,
+      "loss": 5.1839,
+      "step": 10515
+    },
+    {
+      "epoch": 0.5815367606412383,
+      "grad_norm": 2.7311925888061523,
+      "learning_rate": 2.151790790221717e-05,
+      "loss": 4.8678,
+      "step": 10520
+    },
+    {
+      "epoch": 0.5818131564400221,
+      "grad_norm": 4.017518043518066,
+      "learning_rate": 2.1503695281409893e-05,
+      "loss": 4.8128,
+      "step": 10525
+    },
+    {
+      "epoch": 0.582089552238806,
+      "grad_norm": 3.112291097640991,
+      "learning_rate": 2.1489482660602617e-05,
+      "loss": 5.3122,
+      "step": 10530
+    },
+    {
+      "epoch": 0.5823659480375898,
+      "grad_norm": 3.437387704849243,
+      "learning_rate": 2.1475270039795338e-05,
+      "loss": 4.9359,
+      "step": 10535
+    },
+    {
+      "epoch": 0.5826423438363737,
+      "grad_norm": 3.4737443923950195,
+      "learning_rate": 2.1461057418988062e-05,
+      "loss": 5.0828,
+      "step": 10540
+    },
+    {
+      "epoch": 0.5829187396351575,
+      "grad_norm": 4.825445175170898,
+      "learning_rate": 2.1446844798180786e-05,
+      "loss": 5.0617,
+      "step": 10545
+    },
+    {
+      "epoch": 0.5831951354339414,
+      "grad_norm": 2.9588332176208496,
+      "learning_rate": 2.143263217737351e-05,
+      "loss": 5.124,
+      "step": 10550
+    },
+    {
+      "epoch": 0.5834715312327252,
+      "grad_norm": 3.551666736602783,
+      "learning_rate": 2.1418419556566234e-05,
+      "loss": 5.1687,
+      "step": 10555
+    },
+    {
+      "epoch": 0.5837479270315091,
+      "grad_norm": 3.1426239013671875,
+      "learning_rate": 2.1404206935758954e-05,
+      "loss": 4.7921,
+      "step": 10560
+    },
+    {
+      "epoch": 0.5840243228302929,
+      "grad_norm": 3.088836193084717,
+      "learning_rate": 2.138999431495168e-05,
+      "loss": 5.3397,
+      "step": 10565
+    },
+    {
+      "epoch": 0.5843007186290768,
+      "grad_norm": 3.6306614875793457,
+      "learning_rate": 2.13757816941444e-05,
+      "loss": 5.3874,
+      "step": 10570
+    },
+    {
+      "epoch": 0.5845771144278606,
+      "grad_norm": 4.261897563934326,
+      "learning_rate": 2.1361569073337123e-05,
+      "loss": 5.0643,
+      "step": 10575
+    },
+    {
+      "epoch": 0.5848535102266446,
+      "grad_norm": 3.7982075214385986,
+      "learning_rate": 2.1347356452529847e-05,
+      "loss": 4.9988,
+      "step": 10580
+    },
+    {
+      "epoch": 0.5851299060254284,
+      "grad_norm": 3.575862407684326,
+      "learning_rate": 2.133314383172257e-05,
+      "loss": 5.3502,
+      "step": 10585
+    },
+    {
+      "epoch": 0.5854063018242123,
+      "grad_norm": 3.2091081142425537,
+      "learning_rate": 2.1318931210915295e-05,
+      "loss": 4.929,
+      "step": 10590
+    },
+    {
+      "epoch": 0.5856826976229962,
+      "grad_norm": 3.6131644248962402,
+      "learning_rate": 2.1304718590108016e-05,
+      "loss": 4.9555,
+      "step": 10595
+    },
+    {
+      "epoch": 0.58595909342178,
+      "grad_norm": 3.6742172241210938,
+      "learning_rate": 2.129050596930074e-05,
+      "loss": 5.0453,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5862354892205639,
+      "grad_norm": 3.404371738433838,
+      "learning_rate": 2.1276293348493464e-05,
+      "loss": 4.9043,
+      "step": 10605
+    },
+    {
+      "epoch": 0.5865118850193477,
+      "grad_norm": 3.1416611671447754,
+      "learning_rate": 2.1262080727686187e-05,
+      "loss": 5.0373,
+      "step": 10610
+    },
+    {
+      "epoch": 0.5867882808181316,
+      "grad_norm": 3.7025134563446045,
+      "learning_rate": 2.124786810687891e-05,
+      "loss": 4.9599,
+      "step": 10615
+    },
+    {
+      "epoch": 0.5870646766169154,
+      "grad_norm": 3.171035051345825,
+      "learning_rate": 2.1233655486071632e-05,
+      "loss": 5.4027,
+      "step": 10620
+    },
+    {
+      "epoch": 0.5873410724156993,
+      "grad_norm": 3.402039051055908,
+      "learning_rate": 2.1219442865264356e-05,
+      "loss": 5.0835,
+      "step": 10625
+    },
+    {
+      "epoch": 0.5876174682144831,
+      "grad_norm": 3.2764732837677,
+      "learning_rate": 2.1205230244457077e-05,
+      "loss": 4.7431,
+      "step": 10630
+    },
+    {
+      "epoch": 0.587893864013267,
+      "grad_norm": 3.389270305633545,
+      "learning_rate": 2.1191017623649804e-05,
+      "loss": 5.0233,
+      "step": 10635
+    },
+    {
+      "epoch": 0.5881702598120508,
+      "grad_norm": 3.1218948364257812,
+      "learning_rate": 2.1176805002842525e-05,
+      "loss": 5.0945,
+      "step": 10640
+    },
+    {
+      "epoch": 0.5884466556108348,
+      "grad_norm": 3.973499298095703,
+      "learning_rate": 2.116259238203525e-05,
+      "loss": 4.7494,
+      "step": 10645
+    },
+    {
+      "epoch": 0.5887230514096186,
+      "grad_norm": 3.017624616622925,
+      "learning_rate": 2.1148379761227973e-05,
+      "loss": 4.8494,
+      "step": 10650
+    },
+    {
+      "epoch": 0.5889994472084025,
+      "grad_norm": 4.036647796630859,
+      "learning_rate": 2.1134167140420693e-05,
+      "loss": 5.053,
+      "step": 10655
+    },
+    {
+      "epoch": 0.5892758430071863,
+      "grad_norm": 3.0895462036132812,
+      "learning_rate": 2.111995451961342e-05,
+      "loss": 4.7882,
+      "step": 10660
+    },
+    {
+      "epoch": 0.5895522388059702,
+      "grad_norm": 3.1549055576324463,
+      "learning_rate": 2.110574189880614e-05,
+      "loss": 4.8596,
+      "step": 10665
+    },
+    {
+      "epoch": 0.589828634604754,
+      "grad_norm": 3.222073793411255,
+      "learning_rate": 2.1091529277998865e-05,
+      "loss": 5.1663,
+      "step": 10670
+    },
+    {
+      "epoch": 0.5901050304035379,
+      "grad_norm": 3.191425323486328,
+      "learning_rate": 2.1077316657191586e-05,
+      "loss": 5.1247,
+      "step": 10675
+    },
+    {
+      "epoch": 0.5903814262023217,
+      "grad_norm": 3.3451499938964844,
+      "learning_rate": 2.106310403638431e-05,
+      "loss": 5.0468,
+      "step": 10680
+    },
+    {
+      "epoch": 0.5906578220011056,
+      "grad_norm": 3.7087557315826416,
+      "learning_rate": 2.1048891415577034e-05,
+      "loss": 5.2902,
+      "step": 10685
+    },
+    {
+      "epoch": 0.5909342177998894,
+      "grad_norm": 3.4653618335723877,
+      "learning_rate": 2.1034678794769758e-05,
+      "loss": 4.8757,
+      "step": 10690
+    },
+    {
+      "epoch": 0.5912106135986733,
+      "grad_norm": 3.368715763092041,
+      "learning_rate": 2.102046617396248e-05,
+      "loss": 5.0481,
+      "step": 10695
+    },
+    {
+      "epoch": 0.5914870093974571,
+      "grad_norm": 3.7154407501220703,
+      "learning_rate": 2.1006253553155202e-05,
+      "loss": 4.8707,
+      "step": 10700
+    },
+    {
+      "epoch": 0.591763405196241,
+      "grad_norm": 2.9506330490112305,
+      "learning_rate": 2.0992040932347926e-05,
+      "loss": 4.9962,
+      "step": 10705
+    },
+    {
+      "epoch": 0.5920398009950248,
+      "grad_norm": 2.9487926959991455,
+      "learning_rate": 2.0977828311540647e-05,
+      "loss": 4.864,
+      "step": 10710
+    },
+    {
+      "epoch": 0.5923161967938088,
+      "grad_norm": 2.97090482711792,
+      "learning_rate": 2.0963615690733374e-05,
+      "loss": 4.7697,
+      "step": 10715
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 3.8541948795318604,
+      "learning_rate": 2.0949403069926095e-05,
+      "loss": 5.1119,
+      "step": 10720
+    },
+    {
+      "epoch": 0.5928689883913765,
+      "grad_norm": 3.0364882946014404,
+      "learning_rate": 2.093519044911882e-05,
+      "loss": 5.2197,
+      "step": 10725
+    },
+    {
+      "epoch": 0.5931453841901603,
+      "grad_norm": 3.167233943939209,
+      "learning_rate": 2.0920977828311543e-05,
+      "loss": 5.1334,
+      "step": 10730
+    },
+    {
+      "epoch": 0.5934217799889442,
+      "grad_norm": 2.8649866580963135,
+      "learning_rate": 2.0906765207504263e-05,
+      "loss": 4.886,
+      "step": 10735
+    },
+    {
+      "epoch": 0.593698175787728,
+      "grad_norm": 4.133547782897949,
+      "learning_rate": 2.0892552586696987e-05,
+      "loss": 5.1874,
+      "step": 10740
+    },
+    {
+      "epoch": 0.5939745715865119,
+      "grad_norm": 3.462965488433838,
+      "learning_rate": 2.087833996588971e-05,
+      "loss": 5.0722,
+      "step": 10745
+    },
+    {
+      "epoch": 0.5942509673852957,
+      "grad_norm": 3.6210412979125977,
+      "learning_rate": 2.0864127345082435e-05,
+      "loss": 4.7343,
+      "step": 10750
+    },
+    {
+      "epoch": 0.5945273631840796,
+      "grad_norm": 3.1893310546875,
+      "learning_rate": 2.0849914724275156e-05,
+      "loss": 4.9851,
+      "step": 10755
+    },
+    {
+      "epoch": 0.5948037589828634,
+      "grad_norm": 2.8404297828674316,
+      "learning_rate": 2.083570210346788e-05,
+      "loss": 5.1774,
+      "step": 10760
+    },
+    {
+      "epoch": 0.5950801547816473,
+      "grad_norm": 3.4504430294036865,
+      "learning_rate": 2.0821489482660604e-05,
+      "loss": 5.1506,
+      "step": 10765
+    },
+    {
+      "epoch": 0.5953565505804311,
+      "grad_norm": 3.756342649459839,
+      "learning_rate": 2.0807276861853324e-05,
+      "loss": 5.1785,
+      "step": 10770
+    },
+    {
+      "epoch": 0.595632946379215,
+      "grad_norm": 2.698963165283203,
+      "learning_rate": 2.0793064241046052e-05,
+      "loss": 5.1673,
+      "step": 10775
+    },
+    {
+      "epoch": 0.5959093421779988,
+      "grad_norm": 3.2271690368652344,
+      "learning_rate": 2.0778851620238772e-05,
+      "loss": 5.037,
+      "step": 10780
+    },
+    {
+      "epoch": 0.5961857379767828,
+      "grad_norm": 3.4815831184387207,
+      "learning_rate": 2.0764638999431496e-05,
+      "loss": 4.9544,
+      "step": 10785
+    },
+    {
+      "epoch": 0.5964621337755666,
+      "grad_norm": 3.6066627502441406,
+      "learning_rate": 2.075042637862422e-05,
+      "loss": 5.0425,
+      "step": 10790
+    },
+    {
+      "epoch": 0.5967385295743505,
+      "grad_norm": 3.649170160293579,
+      "learning_rate": 2.073621375781694e-05,
+      "loss": 5.303,
+      "step": 10795
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 3.320713996887207,
+      "learning_rate": 2.0722001137009668e-05,
+      "loss": 4.8067,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5972913211719182,
+      "grad_norm": 2.9411003589630127,
+      "learning_rate": 2.070778851620239e-05,
+      "loss": 5.1317,
+      "step": 10805
+    },
+    {
+      "epoch": 0.5975677169707021,
+      "grad_norm": 3.370591878890991,
+      "learning_rate": 2.0693575895395113e-05,
+      "loss": 5.191,
+      "step": 10810
+    },
+    {
+      "epoch": 0.5978441127694859,
+      "grad_norm": 3.282804250717163,
+      "learning_rate": 2.0679363274587833e-05,
+      "loss": 5.2755,
+      "step": 10815
+    },
+    {
+      "epoch": 0.5981205085682698,
+      "grad_norm": 3.2764978408813477,
+      "learning_rate": 2.0665150653780557e-05,
+      "loss": 5.0115,
+      "step": 10820
+    },
+    {
+      "epoch": 0.5983969043670536,
+      "grad_norm": 3.380523920059204,
+      "learning_rate": 2.065093803297328e-05,
+      "loss": 4.9544,
+      "step": 10825
+    },
+    {
+      "epoch": 0.5986733001658375,
+      "grad_norm": 3.9343714714050293,
+      "learning_rate": 2.0636725412166005e-05,
+      "loss": 4.836,
+      "step": 10830
+    },
+    {
+      "epoch": 0.5989496959646213,
+      "grad_norm": 2.9544639587402344,
+      "learning_rate": 2.062251279135873e-05,
+      "loss": 4.939,
+      "step": 10835
+    },
+    {
+      "epoch": 0.5992260917634052,
+      "grad_norm": 3.6150293350219727,
+      "learning_rate": 2.060830017055145e-05,
+      "loss": 5.26,
+      "step": 10840
+    },
+    {
+      "epoch": 0.599502487562189,
+      "grad_norm": 3.480836868286133,
+      "learning_rate": 2.0594087549744174e-05,
+      "loss": 5.2313,
+      "step": 10845
+    },
+    {
+      "epoch": 0.599778883360973,
+      "grad_norm": 3.0056374073028564,
+      "learning_rate": 2.0579874928936894e-05,
+      "loss": 4.9642,
+      "step": 10850
+    },
+    {
+      "epoch": 0.6000552791597568,
+      "grad_norm": 3.1670284271240234,
+      "learning_rate": 2.0565662308129622e-05,
+      "loss": 5.4859,
+      "step": 10855
+    },
+    {
+      "epoch": 0.6003316749585407,
+      "grad_norm": 3.991672992706299,
+      "learning_rate": 2.0551449687322342e-05,
+      "loss": 5.167,
+      "step": 10860
+    },
+    {
+      "epoch": 0.6006080707573245,
+      "grad_norm": 3.690068244934082,
+      "learning_rate": 2.0537237066515066e-05,
+      "loss": 5.0009,
+      "step": 10865
+    },
+    {
+      "epoch": 0.6008844665561084,
+      "grad_norm": 3.5538177490234375,
+      "learning_rate": 2.052302444570779e-05,
+      "loss": 4.9969,
+      "step": 10870
+    },
+    {
+      "epoch": 0.6011608623548922,
+      "grad_norm": 3.6092166900634766,
+      "learning_rate": 2.050881182490051e-05,
+      "loss": 5.2605,
+      "step": 10875
+    },
+    {
+      "epoch": 0.6014372581536761,
+      "grad_norm": 3.576063871383667,
+      "learning_rate": 2.0494599204093235e-05,
+      "loss": 5.2605,
+      "step": 10880
+    },
+    {
+      "epoch": 0.6017136539524599,
+      "grad_norm": 3.104238271713257,
+      "learning_rate": 2.048038658328596e-05,
+      "loss": 4.956,
+      "step": 10885
+    },
+    {
+      "epoch": 0.6019900497512438,
+      "grad_norm": 2.788604974746704,
+      "learning_rate": 2.0466173962478683e-05,
+      "loss": 5.0503,
+      "step": 10890
+    },
+    {
+      "epoch": 0.6022664455500276,
+      "grad_norm": 3.3958215713500977,
+      "learning_rate": 2.0451961341671403e-05,
+      "loss": 5.4058,
+      "step": 10895
+    },
+    {
+      "epoch": 0.6025428413488115,
+      "grad_norm": 3.638075113296509,
+      "learning_rate": 2.0437748720864127e-05,
+      "loss": 5.0203,
+      "step": 10900
+    },
+    {
+      "epoch": 0.6028192371475953,
+      "grad_norm": 3.6704213619232178,
+      "learning_rate": 2.042353610005685e-05,
+      "loss": 5.0472,
+      "step": 10905
+    },
+    {
+      "epoch": 0.6030956329463792,
+      "grad_norm": 4.042377471923828,
+      "learning_rate": 2.0409323479249575e-05,
+      "loss": 4.8711,
+      "step": 10910
+    },
+    {
+      "epoch": 0.603372028745163,
+      "grad_norm": 3.991793632507324,
+      "learning_rate": 2.03951108584423e-05,
+      "loss": 5.3127,
+      "step": 10915
+    },
+    {
+      "epoch": 0.603648424543947,
+      "grad_norm": 3.0052876472473145,
+      "learning_rate": 2.038089823763502e-05,
+      "loss": 5.3358,
+      "step": 10920
+    },
+    {
+      "epoch": 0.6039248203427308,
+      "grad_norm": 3.848688840866089,
+      "learning_rate": 2.0366685616827744e-05,
+      "loss": 5.1195,
+      "step": 10925
+    },
+    {
+      "epoch": 0.6042012161415147,
+      "grad_norm": 3.1843132972717285,
+      "learning_rate": 2.0352472996020468e-05,
+      "loss": 5.0706,
+      "step": 10930
+    },
+    {
+      "epoch": 0.6044776119402985,
+      "grad_norm": 3.1511127948760986,
+      "learning_rate": 2.033826037521319e-05,
+      "loss": 5.0114,
+      "step": 10935
+    },
+    {
+      "epoch": 0.6047540077390824,
+      "grad_norm": 3.199019432067871,
+      "learning_rate": 2.0324047754405916e-05,
+      "loss": 4.9726,
+      "step": 10940
+    },
+    {
+      "epoch": 0.6050304035378662,
+      "grad_norm": 3.4878625869750977,
+      "learning_rate": 2.0309835133598637e-05,
+      "loss": 5.249,
+      "step": 10945
+    },
+    {
+      "epoch": 0.6053067993366501,
+      "grad_norm": 2.9809420108795166,
+      "learning_rate": 2.029562251279136e-05,
+      "loss": 5.2818,
+      "step": 10950
+    },
+    {
+      "epoch": 0.6055831951354339,
+      "grad_norm": 3.647505521774292,
+      "learning_rate": 2.028140989198408e-05,
+      "loss": 5.1207,
+      "step": 10955
+    },
+    {
+      "epoch": 0.6058595909342178,
+      "grad_norm": 3.185030698776245,
+      "learning_rate": 2.0267197271176805e-05,
+      "loss": 5.0107,
+      "step": 10960
+    },
+    {
+      "epoch": 0.6061359867330016,
+      "grad_norm": 4.0943803787231445,
+      "learning_rate": 2.025298465036953e-05,
+      "loss": 5.0815,
+      "step": 10965
+    },
+    {
+      "epoch": 0.6064123825317855,
+      "grad_norm": 2.887913703918457,
+      "learning_rate": 2.0238772029562253e-05,
+      "loss": 4.8405,
+      "step": 10970
+    },
+    {
+      "epoch": 0.6066887783305693,
+      "grad_norm": 3.262972831726074,
+      "learning_rate": 2.0224559408754977e-05,
+      "loss": 4.8914,
+      "step": 10975
+    },
+    {
+      "epoch": 0.6069651741293532,
+      "grad_norm": 4.215073585510254,
+      "learning_rate": 2.0210346787947698e-05,
+      "loss": 4.9772,
+      "step": 10980
+    },
+    {
+      "epoch": 0.607241569928137,
+      "grad_norm": 3.0118231773376465,
+      "learning_rate": 2.019613416714042e-05,
+      "loss": 4.9951,
+      "step": 10985
+    },
+    {
+      "epoch": 0.607517965726921,
+      "grad_norm": 4.327718257904053,
+      "learning_rate": 2.0181921546333142e-05,
+      "loss": 5.2855,
+      "step": 10990
+    },
+    {
+      "epoch": 0.6077943615257048,
+      "grad_norm": 3.5684874057769775,
+      "learning_rate": 2.016770892552587e-05,
+      "loss": 5.302,
+      "step": 10995
+    },
+    {
+      "epoch": 0.6080707573244887,
+      "grad_norm": 3.2276368141174316,
+      "learning_rate": 2.015349630471859e-05,
+      "loss": 5.1272,
+      "step": 11000
+    },
+    {
+      "epoch": 0.6083471531232725,
+      "grad_norm": 3.3742012977600098,
+      "learning_rate": 2.0139283683911314e-05,
+      "loss": 5.1438,
+      "step": 11005
+    },
+    {
+      "epoch": 0.6086235489220564,
+      "grad_norm": 3.5114636421203613,
+      "learning_rate": 2.0125071063104038e-05,
+      "loss": 4.9084,
+      "step": 11010
+    },
+    {
+      "epoch": 0.6088999447208402,
+      "grad_norm": 3.3797311782836914,
+      "learning_rate": 2.011085844229676e-05,
+      "loss": 4.9402,
+      "step": 11015
+    },
+    {
+      "epoch": 0.6091763405196241,
+      "grad_norm": 3.398404121398926,
+      "learning_rate": 2.0096645821489486e-05,
+      "loss": 4.8323,
+      "step": 11020
+    },
+    {
+      "epoch": 0.6094527363184079,
+      "grad_norm": 3.7108492851257324,
+      "learning_rate": 2.0082433200682207e-05,
+      "loss": 4.7783,
+      "step": 11025
+    },
+    {
+      "epoch": 0.6097291321171918,
+      "grad_norm": 3.121354341506958,
+      "learning_rate": 2.006822057987493e-05,
+      "loss": 5.045,
+      "step": 11030
+    },
+    {
+      "epoch": 0.6100055279159757,
+      "grad_norm": 4.430671691894531,
+      "learning_rate": 2.005400795906765e-05,
+      "loss": 5.0018,
+      "step": 11035
+    },
+    {
+      "epoch": 0.6102819237147595,
+      "grad_norm": 3.9316344261169434,
+      "learning_rate": 2.0039795338260375e-05,
+      "loss": 5.2804,
+      "step": 11040
+    },
+    {
+      "epoch": 0.6105583195135434,
+      "grad_norm": 3.610534906387329,
+      "learning_rate": 2.00255827174531e-05,
+      "loss": 5.0137,
+      "step": 11045
+    },
+    {
+      "epoch": 0.6108347153123272,
+      "grad_norm": 3.248753786087036,
+      "learning_rate": 2.0011370096645823e-05,
+      "loss": 4.9726,
+      "step": 11050
+    },
+    {
+      "epoch": 0.6111111111111112,
+      "grad_norm": 4.037750244140625,
+      "learning_rate": 1.9997157475838547e-05,
+      "loss": 5.3961,
+      "step": 11055
+    },
+    {
+      "epoch": 0.611387506909895,
+      "grad_norm": 3.1049892902374268,
+      "learning_rate": 1.9982944855031268e-05,
+      "loss": 5.026,
+      "step": 11060
+    },
+    {
+      "epoch": 0.6116639027086789,
+      "grad_norm": 3.5560545921325684,
+      "learning_rate": 1.996873223422399e-05,
+      "loss": 5.3866,
+      "step": 11065
+    },
+    {
+      "epoch": 0.6119402985074627,
+      "grad_norm": 3.1912312507629395,
+      "learning_rate": 1.9954519613416712e-05,
+      "loss": 5.1017,
+      "step": 11070
+    },
+    {
+      "epoch": 0.6122166943062466,
+      "grad_norm": 3.402047634124756,
+      "learning_rate": 1.994030699260944e-05,
+      "loss": 5.0437,
+      "step": 11075
+    },
+    {
+      "epoch": 0.6124930901050304,
+      "grad_norm": 4.292905807495117,
+      "learning_rate": 1.9926094371802164e-05,
+      "loss": 5.2194,
+      "step": 11080
+    },
+    {
+      "epoch": 0.6127694859038143,
+      "grad_norm": 3.1720829010009766,
+      "learning_rate": 1.9911881750994884e-05,
+      "loss": 4.989,
+      "step": 11085
+    },
+    {
+      "epoch": 0.6130458817025981,
+      "grad_norm": 3.152600049972534,
+      "learning_rate": 1.9897669130187608e-05,
+      "loss": 5.4581,
+      "step": 11090
+    },
+    {
+      "epoch": 0.613322277501382,
+      "grad_norm": 3.7568509578704834,
+      "learning_rate": 1.988345650938033e-05,
+      "loss": 5.183,
+      "step": 11095
+    },
+    {
+      "epoch": 0.6135986733001658,
+      "grad_norm": 3.13639497756958,
+      "learning_rate": 1.9869243888573053e-05,
+      "loss": 5.2017,
+      "step": 11100
+    },
+    {
+      "epoch": 0.6138750690989497,
+      "grad_norm": 3.1426844596862793,
+      "learning_rate": 1.9855031267765777e-05,
+      "loss": 4.9344,
+      "step": 11105
+    },
+    {
+      "epoch": 0.6141514648977335,
+      "grad_norm": 3.382380723953247,
+      "learning_rate": 1.98408186469585e-05,
+      "loss": 5.0299,
+      "step": 11110
+    },
+    {
+      "epoch": 0.6144278606965174,
+      "grad_norm": 3.4279749393463135,
+      "learning_rate": 1.9826606026151225e-05,
+      "loss": 5.0376,
+      "step": 11115
+    },
+    {
+      "epoch": 0.6147042564953012,
+      "grad_norm": 2.9934186935424805,
+      "learning_rate": 1.9812393405343945e-05,
+      "loss": 4.8845,
+      "step": 11120
+    },
+    {
+      "epoch": 0.6149806522940852,
+      "grad_norm": 3.1613216400146484,
+      "learning_rate": 1.979818078453667e-05,
+      "loss": 5.2844,
+      "step": 11125
+    },
+    {
+      "epoch": 0.615257048092869,
+      "grad_norm": 2.9240121841430664,
+      "learning_rate": 1.9783968163729393e-05,
+      "loss": 5.0547,
+      "step": 11130
+    },
+    {
+      "epoch": 0.6155334438916529,
+      "grad_norm": 3.2536087036132812,
+      "learning_rate": 1.9769755542922117e-05,
+      "loss": 4.9044,
+      "step": 11135
+    },
+    {
+      "epoch": 0.6158098396904367,
+      "grad_norm": 3.3475849628448486,
+      "learning_rate": 1.9755542922114838e-05,
+      "loss": 5.041,
+      "step": 11140
+    },
+    {
+      "epoch": 0.6160862354892206,
+      "grad_norm": 3.322903871536255,
+      "learning_rate": 1.9741330301307562e-05,
+      "loss": 4.9652,
+      "step": 11145
+    },
+    {
+      "epoch": 0.6163626312880044,
+      "grad_norm": 3.3216490745544434,
+      "learning_rate": 1.9727117680500286e-05,
+      "loss": 4.9637,
+      "step": 11150
+    },
+    {
+      "epoch": 0.6166390270867883,
+      "grad_norm": 3.5304057598114014,
+      "learning_rate": 1.9712905059693006e-05,
+      "loss": 5.2,
+      "step": 11155
+    },
+    {
+      "epoch": 0.6169154228855721,
+      "grad_norm": 3.0232765674591064,
+      "learning_rate": 1.9698692438885734e-05,
+      "loss": 4.9661,
+      "step": 11160
+    },
+    {
+      "epoch": 0.617191818684356,
+      "grad_norm": 3.3707964420318604,
+      "learning_rate": 1.9684479818078454e-05,
+      "loss": 5.2281,
+      "step": 11165
+    },
+    {
+      "epoch": 0.6174682144831398,
+      "grad_norm": 3.333315134048462,
+      "learning_rate": 1.9670267197271178e-05,
+      "loss": 5.0546,
+      "step": 11170
+    },
+    {
+      "epoch": 0.6177446102819237,
+      "grad_norm": 3.7123754024505615,
+      "learning_rate": 1.96560545764639e-05,
+      "loss": 4.9796,
+      "step": 11175
+    },
+    {
+      "epoch": 0.6180210060807075,
+      "grad_norm": 3.348001480102539,
+      "learning_rate": 1.9641841955656623e-05,
+      "loss": 5.06,
+      "step": 11180
+    },
+    {
+      "epoch": 0.6182974018794914,
+      "grad_norm": 3.643882989883423,
+      "learning_rate": 1.9627629334849347e-05,
+      "loss": 4.9581,
+      "step": 11185
+    },
+    {
+      "epoch": 0.6185737976782753,
+      "grad_norm": 4.164022922515869,
+      "learning_rate": 1.961341671404207e-05,
+      "loss": 5.1829,
+      "step": 11190
+    },
+    {
+      "epoch": 0.6188501934770592,
+      "grad_norm": 2.8525750637054443,
+      "learning_rate": 1.9599204093234795e-05,
+      "loss": 5.1341,
+      "step": 11195
+    },
+    {
+      "epoch": 0.619126589275843,
+      "grad_norm": 3.13280987739563,
+      "learning_rate": 1.9584991472427515e-05,
+      "loss": 4.6817,
+      "step": 11200
+    },
+    {
+      "epoch": 0.6194029850746269,
+      "grad_norm": 3.3129093647003174,
+      "learning_rate": 1.957077885162024e-05,
+      "loss": 4.7727,
+      "step": 11205
+    },
+    {
+      "epoch": 0.6196793808734107,
+      "grad_norm": 3.110025405883789,
+      "learning_rate": 1.955656623081296e-05,
+      "loss": 5.0595,
+      "step": 11210
+    },
+    {
+      "epoch": 0.6199557766721946,
+      "grad_norm": 3.4862754344940186,
+      "learning_rate": 1.9542353610005687e-05,
+      "loss": 4.9615,
+      "step": 11215
+    },
+    {
+      "epoch": 0.6202321724709784,
+      "grad_norm": 2.9289820194244385,
+      "learning_rate": 1.9528140989198408e-05,
+      "loss": 5.0799,
+      "step": 11220
+    },
+    {
+      "epoch": 0.6205085682697623,
+      "grad_norm": 3.868337631225586,
+      "learning_rate": 1.9513928368391132e-05,
+      "loss": 5.0724,
+      "step": 11225
+    },
+    {
+      "epoch": 0.6207849640685461,
+      "grad_norm": 3.828287124633789,
+      "learning_rate": 1.9499715747583856e-05,
+      "loss": 4.9354,
+      "step": 11230
+    },
+    {
+      "epoch": 0.62106135986733,
+      "grad_norm": 3.591191291809082,
+      "learning_rate": 1.9485503126776576e-05,
+      "loss": 5.0352,
+      "step": 11235
+    },
+    {
+      "epoch": 0.6213377556661138,
+      "grad_norm": 3.2487423419952393,
+      "learning_rate": 1.9471290505969304e-05,
+      "loss": 4.6987,
+      "step": 11240
+    },
+    {
+      "epoch": 0.6216141514648977,
+      "grad_norm": 3.1650702953338623,
+      "learning_rate": 1.9457077885162024e-05,
+      "loss": 5.0245,
+      "step": 11245
+    },
+    {
+      "epoch": 0.6218905472636815,
+      "grad_norm": 3.4682087898254395,
+      "learning_rate": 1.944286526435475e-05,
+      "loss": 4.9166,
+      "step": 11250
+    },
+    {
+      "epoch": 0.6221669430624654,
+      "grad_norm": 3.3193416595458984,
+      "learning_rate": 1.9428652643547472e-05,
+      "loss": 5.1592,
+      "step": 11255
+    },
+    {
+      "epoch": 0.6224433388612494,
+      "grad_norm": 3.075951337814331,
+      "learning_rate": 1.9414440022740193e-05,
+      "loss": 5.0866,
+      "step": 11260
+    },
+    {
+      "epoch": 0.6227197346600332,
+      "grad_norm": 3.391075372695923,
+      "learning_rate": 1.9400227401932917e-05,
+      "loss": 5.1489,
+      "step": 11265
+    },
+    {
+      "epoch": 0.6229961304588171,
+      "grad_norm": 3.094604015350342,
+      "learning_rate": 1.938601478112564e-05,
+      "loss": 5.0992,
+      "step": 11270
+    },
+    {
+      "epoch": 0.6232725262576009,
+      "grad_norm": 3.3130276203155518,
+      "learning_rate": 1.9371802160318365e-05,
+      "loss": 4.941,
+      "step": 11275
+    },
+    {
+      "epoch": 0.6235489220563848,
+      "grad_norm": 4.256862640380859,
+      "learning_rate": 1.9357589539511086e-05,
+      "loss": 5.0641,
+      "step": 11280
+    },
+    {
+      "epoch": 0.6238253178551686,
+      "grad_norm": 3.174084186553955,
+      "learning_rate": 1.934337691870381e-05,
+      "loss": 5.2618,
+      "step": 11285
+    },
+    {
+      "epoch": 0.6241017136539525,
+      "grad_norm": 4.607797145843506,
+      "learning_rate": 1.9329164297896533e-05,
+      "loss": 4.6887,
+      "step": 11290
+    },
+    {
+      "epoch": 0.6243781094527363,
+      "grad_norm": 3.27817964553833,
+      "learning_rate": 1.9314951677089257e-05,
+      "loss": 5.25,
+      "step": 11295
+    },
+    {
+      "epoch": 0.6246545052515202,
+      "grad_norm": 3.901153564453125,
+      "learning_rate": 1.930073905628198e-05,
+      "loss": 5.1067,
+      "step": 11300
+    },
+    {
+      "epoch": 0.624930901050304,
+      "grad_norm": 4.034906387329102,
+      "learning_rate": 1.9286526435474702e-05,
+      "loss": 5.0124,
+      "step": 11305
+    },
+    {
+      "epoch": 0.6252072968490879,
+      "grad_norm": 3.6741175651550293,
+      "learning_rate": 1.9272313814667426e-05,
+      "loss": 5.1274,
+      "step": 11310
+    },
+    {
+      "epoch": 0.6254836926478717,
+      "grad_norm": 3.136146068572998,
+      "learning_rate": 1.9258101193860147e-05,
+      "loss": 4.9906,
+      "step": 11315
+    },
+    {
+      "epoch": 0.6257600884466556,
+      "grad_norm": 2.70194673538208,
+      "learning_rate": 1.924388857305287e-05,
+      "loss": 4.8283,
+      "step": 11320
+    },
+    {
+      "epoch": 0.6260364842454395,
+      "grad_norm": 3.145467758178711,
+      "learning_rate": 1.9229675952245595e-05,
+      "loss": 5.2228,
+      "step": 11325
+    },
+    {
+      "epoch": 0.6263128800442234,
+      "grad_norm": 3.2286715507507324,
+      "learning_rate": 1.921546333143832e-05,
+      "loss": 4.9371,
+      "step": 11330
+    },
+    {
+      "epoch": 0.6265892758430072,
+      "grad_norm": 3.4829323291778564,
+      "learning_rate": 1.9201250710631043e-05,
+      "loss": 4.8272,
+      "step": 11335
+    },
+    {
+      "epoch": 0.6268656716417911,
+      "grad_norm": 3.874124050140381,
+      "learning_rate": 1.9187038089823763e-05,
+      "loss": 5.1033,
+      "step": 11340
+    },
+    {
+      "epoch": 0.6271420674405749,
+      "grad_norm": 3.66743540763855,
+      "learning_rate": 1.9172825469016487e-05,
+      "loss": 5.012,
+      "step": 11345
+    },
+    {
+      "epoch": 0.6274184632393588,
+      "grad_norm": 3.3614590167999268,
+      "learning_rate": 1.915861284820921e-05,
+      "loss": 5.2151,
+      "step": 11350
+    },
+    {
+      "epoch": 0.6276948590381426,
+      "grad_norm": 3.108370780944824,
+      "learning_rate": 1.9144400227401935e-05,
+      "loss": 4.9544,
+      "step": 11355
+    },
+    {
+      "epoch": 0.6279712548369265,
+      "grad_norm": 4.2630391120910645,
+      "learning_rate": 1.9130187606594656e-05,
+      "loss": 4.6829,
+      "step": 11360
+    },
+    {
+      "epoch": 0.6282476506357103,
+      "grad_norm": 4.369514465332031,
+      "learning_rate": 1.911597498578738e-05,
+      "loss": 5.0551,
+      "step": 11365
+    },
+    {
+      "epoch": 0.6285240464344942,
+      "grad_norm": 3.091458559036255,
+      "learning_rate": 1.9101762364980104e-05,
+      "loss": 5.2453,
+      "step": 11370
+    },
+    {
+      "epoch": 0.628800442233278,
+      "grad_norm": 3.507791757583618,
+      "learning_rate": 1.9087549744172824e-05,
+      "loss": 4.803,
+      "step": 11375
+    },
+    {
+      "epoch": 0.6290768380320619,
+      "grad_norm": 3.9560585021972656,
+      "learning_rate": 1.907333712336555e-05,
+      "loss": 4.9198,
+      "step": 11380
+    },
+    {
+      "epoch": 0.6293532338308457,
+      "grad_norm": 3.2745492458343506,
+      "learning_rate": 1.9059124502558272e-05,
+      "loss": 5.143,
+      "step": 11385
+    },
+    {
+      "epoch": 0.6296296296296297,
+      "grad_norm": 3.749288558959961,
+      "learning_rate": 1.9044911881750996e-05,
+      "loss": 5.0508,
+      "step": 11390
+    },
+    {
+      "epoch": 0.6299060254284135,
+      "grad_norm": 3.2741405963897705,
+      "learning_rate": 1.903069926094372e-05,
+      "loss": 4.849,
+      "step": 11395
+    },
+    {
+      "epoch": 0.6301824212271974,
+      "grad_norm": 3.3491995334625244,
+      "learning_rate": 1.901648664013644e-05,
+      "loss": 5.0678,
+      "step": 11400
+    },
+    {
+      "epoch": 0.6304588170259812,
+      "grad_norm": 3.6836395263671875,
+      "learning_rate": 1.9002274019329168e-05,
+      "loss": 5.2173,
+      "step": 11405
+    },
+    {
+      "epoch": 0.6307352128247651,
+      "grad_norm": 3.4828968048095703,
+      "learning_rate": 1.898806139852189e-05,
+      "loss": 4.8904,
+      "step": 11410
+    },
+    {
+      "epoch": 0.6310116086235489,
+      "grad_norm": 3.472628355026245,
+      "learning_rate": 1.8973848777714613e-05,
+      "loss": 5.1827,
+      "step": 11415
+    },
+    {
+      "epoch": 0.6312880044223328,
+      "grad_norm": 3.0321156978607178,
+      "learning_rate": 1.8959636156907333e-05,
+      "loss": 4.8185,
+      "step": 11420
+    },
+    {
+      "epoch": 0.6315644002211166,
+      "grad_norm": 3.027168035507202,
+      "learning_rate": 1.8945423536100057e-05,
+      "loss": 5.1646,
+      "step": 11425
+    },
+    {
+      "epoch": 0.6318407960199005,
+      "grad_norm": 3.633697748184204,
+      "learning_rate": 1.893121091529278e-05,
+      "loss": 4.8653,
+      "step": 11430
+    },
+    {
+      "epoch": 0.6321171918186843,
+      "grad_norm": 3.8917782306671143,
+      "learning_rate": 1.8916998294485505e-05,
+      "loss": 5.1636,
+      "step": 11435
+    },
+    {
+      "epoch": 0.6323935876174682,
+      "grad_norm": 3.511700391769409,
+      "learning_rate": 1.890278567367823e-05,
+      "loss": 5.0657,
+      "step": 11440
+    },
+    {
+      "epoch": 0.632669983416252,
+      "grad_norm": 3.8616366386413574,
+      "learning_rate": 1.888857305287095e-05,
+      "loss": 5.085,
+      "step": 11445
+    },
+    {
+      "epoch": 0.6329463792150359,
+      "grad_norm": 3.9748528003692627,
+      "learning_rate": 1.8874360432063674e-05,
+      "loss": 5.2986,
+      "step": 11450
+    },
+    {
+      "epoch": 0.6332227750138197,
+      "grad_norm": 3.7701029777526855,
+      "learning_rate": 1.8860147811256394e-05,
+      "loss": 5.0137,
+      "step": 11455
+    },
+    {
+      "epoch": 0.6334991708126037,
+      "grad_norm": 3.903155565261841,
+      "learning_rate": 1.884593519044912e-05,
+      "loss": 5.0449,
+      "step": 11460
+    },
+    {
+      "epoch": 0.6337755666113875,
+      "grad_norm": 4.124213218688965,
+      "learning_rate": 1.8831722569641842e-05,
+      "loss": 4.9746,
+      "step": 11465
+    },
+    {
+      "epoch": 0.6340519624101714,
+      "grad_norm": 3.3113105297088623,
+      "learning_rate": 1.8817509948834566e-05,
+      "loss": 5.3676,
+      "step": 11470
+    },
+    {
+      "epoch": 0.6343283582089553,
+      "grad_norm": 3.800441265106201,
+      "learning_rate": 1.880329732802729e-05,
+      "loss": 4.9518,
+      "step": 11475
+    },
+    {
+      "epoch": 0.6346047540077391,
+      "grad_norm": 3.8035666942596436,
+      "learning_rate": 1.878908470722001e-05,
+      "loss": 5.1185,
+      "step": 11480
+    },
+    {
+      "epoch": 0.634881149806523,
+      "grad_norm": 3.5756587982177734,
+      "learning_rate": 1.8774872086412735e-05,
+      "loss": 5.027,
+      "step": 11485
+    },
+    {
+      "epoch": 0.6351575456053068,
+      "grad_norm": 4.0546875,
+      "learning_rate": 1.876065946560546e-05,
+      "loss": 5.163,
+      "step": 11490
+    },
+    {
+      "epoch": 0.6354339414040907,
+      "grad_norm": 3.088939666748047,
+      "learning_rate": 1.8746446844798183e-05,
+      "loss": 5.0475,
+      "step": 11495
+    },
+    {
+      "epoch": 0.6357103372028745,
+      "grad_norm": 3.1749298572540283,
+      "learning_rate": 1.8732234223990903e-05,
+      "loss": 4.7649,
+      "step": 11500
+    },
+    {
+      "epoch": 0.6359867330016584,
+      "grad_norm": 3.7194294929504395,
+      "learning_rate": 1.8718021603183627e-05,
+      "loss": 4.9005,
+      "step": 11505
+    },
+    {
+      "epoch": 0.6362631288004422,
+      "grad_norm": 3.6377458572387695,
+      "learning_rate": 1.870380898237635e-05,
+      "loss": 4.8569,
+      "step": 11510
+    },
+    {
+      "epoch": 0.6365395245992261,
+      "grad_norm": 3.6304149627685547,
+      "learning_rate": 1.8689596361569075e-05,
+      "loss": 5.1913,
+      "step": 11515
+    },
+    {
+      "epoch": 0.6368159203980099,
+      "grad_norm": 3.7646942138671875,
+      "learning_rate": 1.86753837407618e-05,
+      "loss": 4.9044,
+      "step": 11520
+    },
+    {
+      "epoch": 0.6370923161967939,
+      "grad_norm": 2.9954702854156494,
+      "learning_rate": 1.866117111995452e-05,
+      "loss": 4.9666,
+      "step": 11525
+    },
+    {
+      "epoch": 0.6373687119955777,
+      "grad_norm": 3.282142162322998,
+      "learning_rate": 1.8646958499147244e-05,
+      "loss": 4.8544,
+      "step": 11530
+    },
+    {
+      "epoch": 0.6376451077943616,
+      "grad_norm": 3.545342206954956,
+      "learning_rate": 1.8632745878339964e-05,
+      "loss": 4.9769,
+      "step": 11535
+    },
+    {
+      "epoch": 0.6379215035931454,
+      "grad_norm": 3.873244285583496,
+      "learning_rate": 1.861853325753269e-05,
+      "loss": 5.1524,
+      "step": 11540
+    },
+    {
+      "epoch": 0.6381978993919293,
+      "grad_norm": 3.273556709289551,
+      "learning_rate": 1.8604320636725416e-05,
+      "loss": 4.9264,
+      "step": 11545
+    },
+    {
+      "epoch": 0.6384742951907131,
+      "grad_norm": 4.054984092712402,
+      "learning_rate": 1.8590108015918136e-05,
+      "loss": 4.9509,
+      "step": 11550
+    },
+    {
+      "epoch": 0.638750690989497,
+      "grad_norm": 4.351573467254639,
+      "learning_rate": 1.857589539511086e-05,
+      "loss": 5.0652,
+      "step": 11555
+    },
+    {
+      "epoch": 0.6390270867882808,
+      "grad_norm": 3.1132781505584717,
+      "learning_rate": 1.856168277430358e-05,
+      "loss": 5.0784,
+      "step": 11560
+    },
+    {
+      "epoch": 0.6393034825870647,
+      "grad_norm": 3.2859065532684326,
+      "learning_rate": 1.8547470153496305e-05,
+      "loss": 5.099,
+      "step": 11565
+    },
+    {
+      "epoch": 0.6395798783858485,
+      "grad_norm": 3.499131441116333,
+      "learning_rate": 1.853325753268903e-05,
+      "loss": 5.0266,
+      "step": 11570
+    },
+    {
+      "epoch": 0.6398562741846324,
+      "grad_norm": 3.202064037322998,
+      "learning_rate": 1.8519044911881753e-05,
+      "loss": 4.8385,
+      "step": 11575
+    },
+    {
+      "epoch": 0.6401326699834162,
+      "grad_norm": 3.816030502319336,
+      "learning_rate": 1.8504832291074477e-05,
+      "loss": 4.8693,
+      "step": 11580
+    },
+    {
+      "epoch": 0.6404090657822001,
+      "grad_norm": 3.5522069931030273,
+      "learning_rate": 1.8490619670267197e-05,
+      "loss": 5.2284,
+      "step": 11585
+    },
+    {
+      "epoch": 0.6406854615809839,
+      "grad_norm": 2.9385523796081543,
+      "learning_rate": 1.847640704945992e-05,
+      "loss": 4.9548,
+      "step": 11590
+    },
+    {
+      "epoch": 0.6409618573797679,
+      "grad_norm": 3.6049933433532715,
+      "learning_rate": 1.8462194428652642e-05,
+      "loss": 4.9881,
+      "step": 11595
+    },
+    {
+      "epoch": 0.6412382531785517,
+      "grad_norm": 2.8920090198516846,
+      "learning_rate": 1.844798180784537e-05,
+      "loss": 4.9215,
+      "step": 11600
+    },
+    {
+      "epoch": 0.6415146489773356,
+      "grad_norm": 3.008737564086914,
+      "learning_rate": 1.843376918703809e-05,
+      "loss": 4.9579,
+      "step": 11605
+    },
+    {
+      "epoch": 0.6417910447761194,
+      "grad_norm": 3.340202808380127,
+      "learning_rate": 1.8419556566230814e-05,
+      "loss": 4.9546,
+      "step": 11610
+    },
+    {
+      "epoch": 0.6420674405749033,
+      "grad_norm": 4.0580973625183105,
+      "learning_rate": 1.8405343945423538e-05,
+      "loss": 5.3337,
+      "step": 11615
+    },
+    {
+      "epoch": 0.6423438363736871,
+      "grad_norm": 3.040929079055786,
+      "learning_rate": 1.839113132461626e-05,
+      "loss": 4.8372,
+      "step": 11620
+    },
+    {
+      "epoch": 0.642620232172471,
+      "grad_norm": 3.111201524734497,
+      "learning_rate": 1.8376918703808986e-05,
+      "loss": 5.1435,
+      "step": 11625
+    },
+    {
+      "epoch": 0.6428966279712548,
+      "grad_norm": 3.6376893520355225,
+      "learning_rate": 1.8362706083001706e-05,
+      "loss": 4.9107,
+      "step": 11630
+    },
+    {
+      "epoch": 0.6431730237700387,
+      "grad_norm": 3.81954288482666,
+      "learning_rate": 1.834849346219443e-05,
+      "loss": 5.0791,
+      "step": 11635
+    },
+    {
+      "epoch": 0.6434494195688225,
+      "grad_norm": 2.9065468311309814,
+      "learning_rate": 1.833428084138715e-05,
+      "loss": 5.0171,
+      "step": 11640
+    },
+    {
+      "epoch": 0.6437258153676064,
+      "grad_norm": 3.3919105529785156,
+      "learning_rate": 1.8320068220579875e-05,
+      "loss": 4.9347,
+      "step": 11645
+    },
+    {
+      "epoch": 0.6440022111663902,
+      "grad_norm": 2.9932496547698975,
+      "learning_rate": 1.83058555997726e-05,
+      "loss": 5.207,
+      "step": 11650
+    },
+    {
+      "epoch": 0.6442786069651741,
+      "grad_norm": 3.7456681728363037,
+      "learning_rate": 1.8291642978965323e-05,
+      "loss": 4.9296,
+      "step": 11655
+    },
+    {
+      "epoch": 0.6445550027639579,
+      "grad_norm": 3.1489369869232178,
+      "learning_rate": 1.8277430358158047e-05,
+      "loss": 4.8545,
+      "step": 11660
+    },
+    {
+      "epoch": 0.6448313985627419,
+      "grad_norm": 3.770603895187378,
+      "learning_rate": 1.8263217737350768e-05,
+      "loss": 4.9182,
+      "step": 11665
+    },
+    {
+      "epoch": 0.6451077943615257,
+      "grad_norm": 3.3885293006896973,
+      "learning_rate": 1.824900511654349e-05,
+      "loss": 5.0724,
+      "step": 11670
+    },
+    {
+      "epoch": 0.6453841901603096,
+      "grad_norm": 3.6072707176208496,
+      "learning_rate": 1.8234792495736212e-05,
+      "loss": 4.9864,
+      "step": 11675
+    },
+    {
+      "epoch": 0.6456605859590934,
+      "grad_norm": 3.027808904647827,
+      "learning_rate": 1.822057987492894e-05,
+      "loss": 5.3012,
+      "step": 11680
+    },
+    {
+      "epoch": 0.6459369817578773,
+      "grad_norm": 3.8706979751586914,
+      "learning_rate": 1.8206367254121663e-05,
+      "loss": 5.0479,
+      "step": 11685
+    },
+    {
+      "epoch": 0.6462133775566611,
+      "grad_norm": 2.9442152976989746,
+      "learning_rate": 1.8192154633314384e-05,
+      "loss": 5.153,
+      "step": 11690
+    },
+    {
+      "epoch": 0.646489773355445,
+      "grad_norm": 3.645040512084961,
+      "learning_rate": 1.8177942012507108e-05,
+      "loss": 5.0203,
+      "step": 11695
+    },
+    {
+      "epoch": 0.6467661691542289,
+      "grad_norm": 3.7537355422973633,
+      "learning_rate": 1.816372939169983e-05,
+      "loss": 5.155,
+      "step": 11700
+    },
+    {
+      "epoch": 0.6470425649530127,
+      "grad_norm": 3.36226749420166,
+      "learning_rate": 1.8149516770892553e-05,
+      "loss": 5.1796,
+      "step": 11705
+    },
+    {
+      "epoch": 0.6473189607517966,
+      "grad_norm": 3.477741241455078,
+      "learning_rate": 1.8135304150085277e-05,
+      "loss": 4.8361,
+      "step": 11710
+    },
+    {
+      "epoch": 0.6475953565505804,
+      "grad_norm": 3.8094239234924316,
+      "learning_rate": 1.8121091529278e-05,
+      "loss": 4.9998,
+      "step": 11715
+    },
+    {
+      "epoch": 0.6478717523493643,
+      "grad_norm": 3.418806791305542,
+      "learning_rate": 1.8106878908470725e-05,
+      "loss": 5.1284,
+      "step": 11720
+    },
+    {
+      "epoch": 0.6481481481481481,
+      "grad_norm": 3.108595132827759,
+      "learning_rate": 1.8092666287663445e-05,
+      "loss": 5.0816,
+      "step": 11725
+    },
+    {
+      "epoch": 0.648424543946932,
+      "grad_norm": 3.583040475845337,
+      "learning_rate": 1.807845366685617e-05,
+      "loss": 5.0151,
+      "step": 11730
+    },
+    {
+      "epoch": 0.6487009397457159,
+      "grad_norm": 2.922956705093384,
+      "learning_rate": 1.8064241046048893e-05,
+      "loss": 5.2425,
+      "step": 11735
+    },
+    {
+      "epoch": 0.6489773355444998,
+      "grad_norm": 3.604804039001465,
+      "learning_rate": 1.8050028425241617e-05,
+      "loss": 4.9848,
+      "step": 11740
+    },
+    {
+      "epoch": 0.6492537313432836,
+      "grad_norm": 3.4737608432769775,
+      "learning_rate": 1.8035815804434338e-05,
+      "loss": 5.0245,
+      "step": 11745
+    },
+    {
+      "epoch": 0.6495301271420675,
+      "grad_norm": 3.2414677143096924,
+      "learning_rate": 1.802160318362706e-05,
+      "loss": 4.8838,
+      "step": 11750
+    },
+    {
+      "epoch": 0.6498065229408513,
+      "grad_norm": 3.7715156078338623,
+      "learning_rate": 1.8007390562819786e-05,
+      "loss": 4.8937,
+      "step": 11755
+    },
+    {
+      "epoch": 0.6500829187396352,
+      "grad_norm": 3.0882837772369385,
+      "learning_rate": 1.7993177942012506e-05,
+      "loss": 5.4637,
+      "step": 11760
+    },
+    {
+      "epoch": 0.650359314538419,
+      "grad_norm": 3.7343342304229736,
+      "learning_rate": 1.7978965321205234e-05,
+      "loss": 4.9524,
+      "step": 11765
+    },
+    {
+      "epoch": 0.6506357103372029,
+      "grad_norm": 3.9432291984558105,
+      "learning_rate": 1.7964752700397954e-05,
+      "loss": 5.1691,
+      "step": 11770
+    },
+    {
+      "epoch": 0.6509121061359867,
+      "grad_norm": 3.1902172565460205,
+      "learning_rate": 1.7950540079590678e-05,
+      "loss": 5.2231,
+      "step": 11775
+    },
+    {
+      "epoch": 0.6511885019347706,
+      "grad_norm": 3.733781099319458,
+      "learning_rate": 1.79363274587834e-05,
+      "loss": 5.1506,
+      "step": 11780
+    },
+    {
+      "epoch": 0.6514648977335544,
+      "grad_norm": 3.6218392848968506,
+      "learning_rate": 1.7922114837976123e-05,
+      "loss": 5.1997,
+      "step": 11785
+    },
+    {
+      "epoch": 0.6517412935323383,
+      "grad_norm": 4.095831394195557,
+      "learning_rate": 1.7907902217168847e-05,
+      "loss": 4.934,
+      "step": 11790
+    },
+    {
+      "epoch": 0.6520176893311221,
+      "grad_norm": 2.9039041996002197,
+      "learning_rate": 1.789368959636157e-05,
+      "loss": 4.8885,
+      "step": 11795
+    },
+    {
+      "epoch": 0.652294085129906,
+      "grad_norm": 2.837620496749878,
+      "learning_rate": 1.7879476975554295e-05,
+      "loss": 4.7736,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6525704809286899,
+      "grad_norm": 4.38239049911499,
+      "learning_rate": 1.7865264354747015e-05,
+      "loss": 5.1993,
+      "step": 11805
+    },
+    {
+      "epoch": 0.6528468767274738,
+      "grad_norm": 3.1232762336730957,
+      "learning_rate": 1.785105173393974e-05,
+      "loss": 4.8217,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6531232725262576,
+      "grad_norm": 3.8106727600097656,
+      "learning_rate": 1.783683911313246e-05,
+      "loss": 5.0121,
+      "step": 11815
+    },
+    {
+      "epoch": 0.6533996683250415,
+      "grad_norm": 3.4995906352996826,
+      "learning_rate": 1.7822626492325187e-05,
+      "loss": 5.0138,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6536760641238253,
+      "grad_norm": 3.8078784942626953,
+      "learning_rate": 1.7808413871517908e-05,
+      "loss": 5.102,
+      "step": 11825
+    },
+    {
+      "epoch": 0.6539524599226092,
+      "grad_norm": 3.3027868270874023,
+      "learning_rate": 1.7794201250710632e-05,
+      "loss": 4.8747,
+      "step": 11830
+    },
+    {
+      "epoch": 0.654228855721393,
+      "grad_norm": 3.113663673400879,
+      "learning_rate": 1.7779988629903356e-05,
+      "loss": 4.9301,
+      "step": 11835
+    },
+    {
+      "epoch": 0.6545052515201769,
+      "grad_norm": 3.6533520221710205,
+      "learning_rate": 1.7765776009096076e-05,
+      "loss": 4.8014,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6547816473189607,
+      "grad_norm": 3.131390333175659,
+      "learning_rate": 1.77515633882888e-05,
+      "loss": 4.8878,
+      "step": 11845
+    },
+    {
+      "epoch": 0.6550580431177446,
+      "grad_norm": 3.265279769897461,
+      "learning_rate": 1.7737350767481524e-05,
+      "loss": 5.1533,
+      "step": 11850
+    },
+    {
+      "epoch": 0.6553344389165284,
+      "grad_norm": 3.4372310638427734,
+      "learning_rate": 1.7723138146674248e-05,
+      "loss": 4.766,
+      "step": 11855
+    },
+    {
+      "epoch": 0.6556108347153123,
+      "grad_norm": 3.4847805500030518,
+      "learning_rate": 1.7708925525866972e-05,
+      "loss": 4.963,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6558872305140961,
+      "grad_norm": 4.266146659851074,
+      "learning_rate": 1.7694712905059693e-05,
+      "loss": 4.8469,
+      "step": 11865
+    },
+    {
+      "epoch": 0.65616362631288,
+      "grad_norm": 3.8087639808654785,
+      "learning_rate": 1.7680500284252417e-05,
+      "loss": 5.1229,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6564400221116639,
+      "grad_norm": 3.209784507751465,
+      "learning_rate": 1.766628766344514e-05,
+      "loss": 5.0318,
+      "step": 11875
+    },
+    {
+      "epoch": 0.6567164179104478,
+      "grad_norm": 3.074418306350708,
+      "learning_rate": 1.7652075042637865e-05,
+      "loss": 4.9064,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6569928137092316,
+      "grad_norm": 3.3850672245025635,
+      "learning_rate": 1.7637862421830585e-05,
+      "loss": 4.882,
+      "step": 11885
+    },
+    {
+      "epoch": 0.6572692095080155,
+      "grad_norm": 2.94960618019104,
+      "learning_rate": 1.762364980102331e-05,
+      "loss": 5.1099,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6575456053067993,
+      "grad_norm": 3.0878140926361084,
+      "learning_rate": 1.7609437180216033e-05,
+      "loss": 5.1642,
+      "step": 11895
+    },
+    {
+      "epoch": 0.6578220011055832,
+      "grad_norm": 3.899074077606201,
+      "learning_rate": 1.7595224559408754e-05,
+      "loss": 5.0711,
+      "step": 11900
+    },
+    {
+      "epoch": 0.658098396904367,
+      "grad_norm": 3.6305460929870605,
+      "learning_rate": 1.758101193860148e-05,
+      "loss": 4.9042,
+      "step": 11905
+    },
+    {
+      "epoch": 0.6583747927031509,
+      "grad_norm": 3.7980971336364746,
+      "learning_rate": 1.7566799317794202e-05,
+      "loss": 5.0719,
+      "step": 11910
+    },
+    {
+      "epoch": 0.6586511885019348,
+      "grad_norm": 3.5552990436553955,
+      "learning_rate": 1.7552586696986926e-05,
+      "loss": 5.0205,
+      "step": 11915
+    },
+    {
+      "epoch": 0.6589275843007186,
+      "grad_norm": 3.524707794189453,
+      "learning_rate": 1.7538374076179646e-05,
+      "loss": 4.978,
+      "step": 11920
+    },
+    {
+      "epoch": 0.6592039800995025,
+      "grad_norm": 3.4494073390960693,
+      "learning_rate": 1.752416145537237e-05,
+      "loss": 4.9419,
+      "step": 11925
+    },
+    {
+      "epoch": 0.6594803758982863,
+      "grad_norm": 2.977220058441162,
+      "learning_rate": 1.7509948834565094e-05,
+      "loss": 5.1477,
+      "step": 11930
+    },
+    {
+      "epoch": 0.6597567716970703,
+      "grad_norm": 4.73137903213501,
+      "learning_rate": 1.749573621375782e-05,
+      "loss": 4.9816,
+      "step": 11935
+    },
+    {
+      "epoch": 0.6600331674958541,
+      "grad_norm": 3.988065719604492,
+      "learning_rate": 1.7481523592950542e-05,
+      "loss": 4.8524,
+      "step": 11940
+    },
+    {
+      "epoch": 0.660309563294638,
+      "grad_norm": 3.581176996231079,
+      "learning_rate": 1.7467310972143263e-05,
+      "loss": 5.3324,
+      "step": 11945
+    },
+    {
+      "epoch": 0.6605859590934218,
+      "grad_norm": 3.1877663135528564,
+      "learning_rate": 1.7453098351335987e-05,
+      "loss": 4.9964,
+      "step": 11950
+    },
+    {
+      "epoch": 0.6608623548922057,
+      "grad_norm": 3.8775217533111572,
+      "learning_rate": 1.7438885730528708e-05,
+      "loss": 4.9952,
+      "step": 11955
+    },
+    {
+      "epoch": 0.6611387506909895,
+      "grad_norm": 3.670485496520996,
+      "learning_rate": 1.7424673109721435e-05,
+      "loss": 4.9151,
+      "step": 11960
+    },
+    {
+      "epoch": 0.6614151464897734,
+      "grad_norm": 5.177221298217773,
+      "learning_rate": 1.7410460488914155e-05,
+      "loss": 5.3336,
+      "step": 11965
+    },
+    {
+      "epoch": 0.6616915422885572,
+      "grad_norm": 3.854691982269287,
+      "learning_rate": 1.739624786810688e-05,
+      "loss": 5.2463,
+      "step": 11970
+    },
+    {
+      "epoch": 0.6619679380873411,
+      "grad_norm": 3.330227851867676,
+      "learning_rate": 1.7382035247299603e-05,
+      "loss": 4.4221,
+      "step": 11975
+    },
+    {
+      "epoch": 0.6622443338861249,
+      "grad_norm": 3.3024230003356934,
+      "learning_rate": 1.7367822626492324e-05,
+      "loss": 4.9856,
+      "step": 11980
+    },
+    {
+      "epoch": 0.6625207296849088,
+      "grad_norm": 3.446526527404785,
+      "learning_rate": 1.735361000568505e-05,
+      "loss": 4.9725,
+      "step": 11985
+    },
+    {
+      "epoch": 0.6627971254836926,
+      "grad_norm": 3.300661087036133,
+      "learning_rate": 1.7339397384877772e-05,
+      "loss": 4.965,
+      "step": 11990
+    },
+    {
+      "epoch": 0.6630735212824765,
+      "grad_norm": 3.3495724201202393,
+      "learning_rate": 1.7325184764070496e-05,
+      "loss": 5.1418,
+      "step": 11995
+    },
+    {
+      "epoch": 0.6633499170812603,
+      "grad_norm": 3.493957281112671,
+      "learning_rate": 1.731097214326322e-05,
+      "loss": 4.8989,
+      "step": 12000
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 18090,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1151619342811136e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}