{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.876857306709765,
  "eval_steps": 1000,
  "global_step": 9000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 8.544243812561035,
      "learning_rate": 1.0427528675703859e-07,
      "loss": 1.8298,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 7.121495723724365,
      "learning_rate": 2.0855057351407717e-07,
      "loss": 1.8251,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.022141933441162,
      "learning_rate": 3.128258602711158e-07,
      "loss": 1.8055,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.6211042404174805,
      "learning_rate": 4.1710114702815434e-07,
      "loss": 1.8076,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.393233299255371,
      "learning_rate": 5.213764337851929e-07,
      "loss": 1.7492,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.075599193572998,
      "learning_rate": 6.256517205422316e-07,
      "loss": 1.7474,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.686915636062622,
      "learning_rate": 7.299270072992701e-07,
      "loss": 1.7052,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.735376358032227,
      "learning_rate": 8.342022940563087e-07,
      "loss": 1.6499,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.812263011932373,
      "learning_rate": 9.384775808133473e-07,
      "loss": 1.6299,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.213949680328369,
      "learning_rate": 1.0427528675703859e-06,
      "loss": 1.6052,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.180210590362549,
      "learning_rate": 1.1470281543274244e-06,
      "loss": 1.5974,
      "step": 110
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.1223864555358887,
      "learning_rate": 1.2513034410844632e-06,
      "loss": 1.5442,
      "step": 120
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.524357557296753,
      "learning_rate": 1.3555787278415016e-06,
      "loss": 1.5487,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7285054922103882,
      "learning_rate": 1.4598540145985402e-06,
      "loss": 1.5652,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.6190853118896484,
      "learning_rate": 1.5641293013555788e-06,
      "loss": 1.5206,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7167280912399292,
      "learning_rate": 1.6684045881126174e-06,
      "loss": 1.5195,
      "step": 160
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7101649045944214,
      "learning_rate": 1.7726798748696562e-06,
      "loss": 1.5388,
      "step": 170
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7588235139846802,
      "learning_rate": 1.8769551616266945e-06,
      "loss": 1.5155,
      "step": 180
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.6155401468276978,
      "learning_rate": 1.981230448383733e-06,
      "loss": 1.5149,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.809321403503418,
      "learning_rate": 2.0855057351407717e-06,
      "loss": 1.5159,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.6688252687454224,
      "learning_rate": 2.1897810218978103e-06,
      "loss": 1.4955,
      "step": 210
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.6307278871536255,
      "learning_rate": 2.294056308654849e-06,
      "loss": 1.5238,
      "step": 220
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.4807227849960327,
      "learning_rate": 2.398331595411888e-06,
      "loss": 1.5086,
      "step": 230
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.4992481470108032,
      "learning_rate": 2.5026068821689265e-06,
      "loss": 1.469,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.8425967693328857,
      "learning_rate": 2.606882168925965e-06,
      "loss": 1.5177,
      "step": 250
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.5325536727905273,
      "learning_rate": 2.7111574556830032e-06,
      "loss": 1.4757,
      "step": 260
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.5502701997756958,
      "learning_rate": 2.815432742440042e-06,
      "loss": 1.4808,
      "step": 270
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.5416339635849,
      "learning_rate": 2.9197080291970804e-06,
      "loss": 1.493,
      "step": 280
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.7341797351837158,
      "learning_rate": 3.0239833159541194e-06,
      "loss": 1.4425,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.4314051866531372,
      "learning_rate": 3.1282586027111576e-06,
      "loss": 1.4498,
      "step": 300
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.6512941122055054,
      "learning_rate": 3.232533889468196e-06,
      "loss": 1.4687,
      "step": 310
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.5534512996673584,
      "learning_rate": 3.3368091762252348e-06,
      "loss": 1.4615,
      "step": 320
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.5043481588363647,
      "learning_rate": 3.4410844629822738e-06,
      "loss": 1.4729,
      "step": 330
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.6013394594192505,
      "learning_rate": 3.5453597497393123e-06,
      "loss": 1.4477,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.5246102809906006,
      "learning_rate": 3.6496350364963505e-06,
      "loss": 1.4602,
      "step": 350
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5013917684555054,
      "learning_rate": 3.753910323253389e-06,
      "loss": 1.4512,
      "step": 360
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.8450820446014404,
      "learning_rate": 3.858185610010428e-06,
      "loss": 1.4804,
      "step": 370
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5352061986923218,
      "learning_rate": 3.962460896767466e-06,
      "loss": 1.4408,
      "step": 380
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5943782329559326,
      "learning_rate": 4.066736183524505e-06,
      "loss": 1.4231,
      "step": 390
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.7596439123153687,
      "learning_rate": 4.1710114702815434e-06,
      "loss": 1.446,
      "step": 400
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.470418095588684,
      "learning_rate": 4.275286757038582e-06,
      "loss": 1.4482,
      "step": 410
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.4170681238174438,
      "learning_rate": 4.379562043795621e-06,
      "loss": 1.4475,
      "step": 420
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.409445881843567,
      "learning_rate": 4.483837330552659e-06,
      "loss": 1.4355,
      "step": 430
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.552442193031311,
      "learning_rate": 4.588112617309698e-06,
      "loss": 1.4281,
      "step": 440
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.527559518814087,
      "learning_rate": 4.692387904066736e-06,
      "loss": 1.4482,
      "step": 450
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4792981147766113,
      "learning_rate": 4.796663190823776e-06,
      "loss": 1.422,
      "step": 460
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5031851530075073,
      "learning_rate": 4.9009384775808135e-06,
      "loss": 1.415,
      "step": 470
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4678523540496826,
      "learning_rate": 5.005213764337853e-06,
      "loss": 1.4362,
      "step": 480
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4980840682983398,
      "learning_rate": 5.1094890510948916e-06,
      "loss": 1.4098,
      "step": 490
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.487581491470337,
      "learning_rate": 5.21376433785193e-06,
      "loss": 1.4337,
      "step": 500
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5128278732299805,
      "learning_rate": 5.318039624608968e-06,
      "loss": 1.3967,
      "step": 510
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.463364839553833,
      "learning_rate": 5.4223149113660065e-06,
      "loss": 1.4004,
      "step": 520
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5302982330322266,
      "learning_rate": 5.526590198123045e-06,
      "loss": 1.4289,
      "step": 530
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.3981858491897583,
      "learning_rate": 5.630865484880084e-06,
      "loss": 1.4046,
      "step": 540
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.553695797920227,
      "learning_rate": 5.735140771637122e-06,
      "loss": 1.4155,
      "step": 550
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.411907434463501,
      "learning_rate": 5.839416058394161e-06,
      "loss": 1.4248,
      "step": 560
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.7118998765945435,
      "learning_rate": 5.9436913451512e-06,
      "loss": 1.3852,
      "step": 570
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.4115082025527954,
      "learning_rate": 6.047966631908239e-06,
      "loss": 1.4028,
      "step": 580
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.365872859954834,
      "learning_rate": 6.152241918665277e-06,
      "loss": 1.407,
      "step": 590
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4455780982971191,
      "learning_rate": 6.256517205422315e-06,
      "loss": 1.4141,
      "step": 600
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.3968677520751953,
      "learning_rate": 6.360792492179354e-06,
      "loss": 1.3952,
      "step": 610
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.5289356708526611,
      "learning_rate": 6.465067778936392e-06,
      "loss": 1.3892,
      "step": 620
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7357593774795532,
      "learning_rate": 6.569343065693431e-06,
      "loss": 1.4117,
      "step": 630
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.5576961040496826,
      "learning_rate": 6.6736183524504695e-06,
      "loss": 1.3937,
      "step": 640
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3234304189682007,
      "learning_rate": 6.777893639207508e-06,
      "loss": 1.3851,
      "step": 650
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4000426530838013,
      "learning_rate": 6.8821689259645475e-06,
      "loss": 1.398,
      "step": 660
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.5590957403182983,
      "learning_rate": 6.986444212721586e-06,
      "loss": 1.3891,
      "step": 670
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4518722295761108,
      "learning_rate": 7.090719499478625e-06,
      "loss": 1.404,
      "step": 680
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.5161577463150024,
      "learning_rate": 7.1949947862356624e-06,
      "loss": 1.4004,
      "step": 690
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.3116180896759033,
      "learning_rate": 7.299270072992701e-06,
      "loss": 1.3831,
      "step": 700
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5861413478851318,
      "learning_rate": 7.40354535974974e-06,
      "loss": 1.3939,
      "step": 710
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6353439092636108,
      "learning_rate": 7.507820646506778e-06,
      "loss": 1.4047,
      "step": 720
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.4873729944229126,
      "learning_rate": 7.612095933263817e-06,
      "loss": 1.4024,
      "step": 730
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.4138253927230835,
      "learning_rate": 7.716371220020855e-06,
      "loss": 1.4055,
      "step": 740
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4762160778045654,
      "learning_rate": 7.820646506777895e-06,
      "loss": 1.3882,
      "step": 750
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3965837955474854,
      "learning_rate": 7.924921793534933e-06,
      "loss": 1.3633,
      "step": 760
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4030544757843018,
      "learning_rate": 8.029197080291972e-06,
      "loss": 1.3644,
      "step": 770
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4032145738601685,
      "learning_rate": 8.13347236704901e-06,
      "loss": 1.387,
      "step": 780
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.5707696676254272,
      "learning_rate": 8.237747653806049e-06,
      "loss": 1.3528,
      "step": 790
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4654338359832764,
      "learning_rate": 8.342022940563087e-06,
      "loss": 1.4175,
      "step": 800
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.5189216136932373,
      "learning_rate": 8.446298227320126e-06,
      "loss": 1.365,
      "step": 810
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4133602380752563,
      "learning_rate": 8.550573514077164e-06,
      "loss": 1.3441,
      "step": 820
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.5661669969558716,
      "learning_rate": 8.654848800834203e-06,
      "loss": 1.3698,
      "step": 830
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.5398190021514893,
      "learning_rate": 8.759124087591241e-06,
      "loss": 1.3708,
      "step": 840
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.4412083625793457,
      "learning_rate": 8.86339937434828e-06,
      "loss": 1.3779,
      "step": 850
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3615409135818481,
      "learning_rate": 8.967674661105318e-06,
      "loss": 1.3742,
      "step": 860
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.4569330215454102,
      "learning_rate": 9.071949947862358e-06,
      "loss": 1.3732,
      "step": 870
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.5276408195495605,
      "learning_rate": 9.176225234619396e-06,
      "loss": 1.3875,
      "step": 880
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3974804878234863,
      "learning_rate": 9.280500521376435e-06,
      "loss": 1.3681,
      "step": 890
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.4797160625457764,
      "learning_rate": 9.384775808133473e-06,
      "loss": 1.3544,
      "step": 900
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.4647297859191895,
      "learning_rate": 9.48905109489051e-06,
      "loss": 1.3627,
      "step": 910
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.4966884851455688,
      "learning_rate": 9.593326381647552e-06,
      "loss": 1.3628,
      "step": 920
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.4639487266540527,
      "learning_rate": 9.69760166840459e-06,
      "loss": 1.3687,
      "step": 930
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4451621770858765,
      "learning_rate": 9.801876955161627e-06,
      "loss": 1.3855,
      "step": 940
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.382853388786316,
      "learning_rate": 9.906152241918667e-06,
      "loss": 1.3615,
      "step": 950
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2870492935180664,
      "learning_rate": 9.9988413857027e-06,
      "loss": 1.3495,
      "step": 960
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3984755277633667,
      "learning_rate": 9.987255242729697e-06,
      "loss": 1.3597,
      "step": 970
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3395270109176636,
      "learning_rate": 9.975669099756692e-06,
      "loss": 1.3536,
      "step": 980
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.303995132446289,
      "learning_rate": 9.964082956783688e-06,
      "loss": 1.3537,
      "step": 990
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.629632830619812,
      "learning_rate": 9.952496813810684e-06,
      "loss": 1.3498,
      "step": 1000
    },
    {
      "epoch": 0.21,
      "eval_loss": 1.3506057262420654,
      "eval_runtime": 60.4268,
      "eval_samples_per_second": 102.603,
      "eval_steps_per_second": 12.825,
      "step": 1000
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3876349925994873,
      "learning_rate": 9.940910670837679e-06,
      "loss": 1.3521,
      "step": 1010
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4891064167022705,
      "learning_rate": 9.929324527864674e-06,
      "loss": 1.3506,
      "step": 1020
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3278287649154663,
      "learning_rate": 9.91773838489167e-06,
      "loss": 1.3458,
      "step": 1030
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5127308368682861,
      "learning_rate": 9.906152241918667e-06,
      "loss": 1.3477,
      "step": 1040
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3162195682525635,
      "learning_rate": 9.894566098945661e-06,
      "loss": 1.3444,
      "step": 1050
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4123969078063965,
      "learning_rate": 9.882979955972658e-06,
      "loss": 1.3568,
      "step": 1060
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.531692624092102,
      "learning_rate": 9.871393812999654e-06,
      "loss": 1.3464,
      "step": 1070
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4226391315460205,
      "learning_rate": 9.859807670026649e-06,
      "loss": 1.3327,
      "step": 1080
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4860808849334717,
      "learning_rate": 9.848221527053643e-06,
      "loss": 1.3826,
      "step": 1090
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4818434715270996,
      "learning_rate": 9.836635384080641e-06,
      "loss": 1.3327,
      "step": 1100
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.58639395236969,
      "learning_rate": 9.825049241107636e-06,
      "loss": 1.3461,
      "step": 1110
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4077178239822388,
      "learning_rate": 9.81346309813463e-06,
      "loss": 1.309,
      "step": 1120
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.3896112442016602,
      "learning_rate": 9.801876955161627e-06,
      "loss": 1.3266,
      "step": 1130
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.508527159690857,
      "learning_rate": 9.790290812188623e-06,
      "loss": 1.3764,
      "step": 1140
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.311788558959961,
      "learning_rate": 9.778704669215618e-06,
      "loss": 1.3445,
      "step": 1150
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4742310047149658,
      "learning_rate": 9.767118526242615e-06,
      "loss": 1.3338,
      "step": 1160
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.3195024728775024,
      "learning_rate": 9.755532383269611e-06,
      "loss": 1.3444,
      "step": 1170
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4474618434906006,
      "learning_rate": 9.743946240296606e-06,
      "loss": 1.3332,
      "step": 1180
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3813029527664185,
      "learning_rate": 9.732360097323602e-06,
      "loss": 1.3745,
      "step": 1190
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.370364785194397,
      "learning_rate": 9.720773954350598e-06,
      "loss": 1.3124,
      "step": 1200
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4230958223342896,
      "learning_rate": 9.709187811377593e-06,
      "loss": 1.3224,
      "step": 1210
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.5277326107025146,
      "learning_rate": 9.69760166840459e-06,
      "loss": 1.3368,
      "step": 1220
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.424777626991272,
      "learning_rate": 9.686015525431584e-06,
      "loss": 1.3202,
      "step": 1230
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5502588748931885,
      "learning_rate": 9.67442938245858e-06,
      "loss": 1.3127,
      "step": 1240
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5083343982696533,
      "learning_rate": 9.662843239485577e-06,
      "loss": 1.3234,
      "step": 1250
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.3840395212173462,
      "learning_rate": 9.651257096512571e-06,
      "loss": 1.3241,
      "step": 1260
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.3413887023925781,
      "learning_rate": 9.639670953539568e-06,
      "loss": 1.3247,
      "step": 1270
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3408432006835938,
      "learning_rate": 9.628084810566564e-06,
      "loss": 1.3439,
      "step": 1280
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.302042007446289,
      "learning_rate": 9.616498667593559e-06,
      "loss": 1.3261,
      "step": 1290
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3523017168045044,
      "learning_rate": 9.604912524620554e-06,
      "loss": 1.3412,
      "step": 1300
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3820520639419556,
      "learning_rate": 9.593326381647552e-06,
      "loss": 1.3457,
      "step": 1310
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5620774030685425,
      "learning_rate": 9.581740238674546e-06,
      "loss": 1.3086,
      "step": 1320
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.440212607383728,
      "learning_rate": 9.570154095701541e-06,
      "loss": 1.3197,
      "step": 1330
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.2081576585769653,
      "learning_rate": 9.558567952728537e-06,
      "loss": 1.314,
      "step": 1340
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5691925287246704,
      "learning_rate": 9.546981809755534e-06,
      "loss": 1.319,
      "step": 1350
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5542395114898682,
      "learning_rate": 9.535395666782528e-06,
      "loss": 1.293,
      "step": 1360
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3244450092315674,
      "learning_rate": 9.523809523809525e-06,
      "loss": 1.3092,
      "step": 1370
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.5976145267486572,
      "learning_rate": 9.512223380836521e-06,
      "loss": 1.3379,
      "step": 1380
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3805755376815796,
      "learning_rate": 9.500637237863516e-06,
      "loss": 1.3325,
      "step": 1390
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.6162505149841309,
      "learning_rate": 9.48905109489051e-06,
      "loss": 1.3056,
      "step": 1400
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.4099946022033691,
      "learning_rate": 9.477464951917509e-06,
      "loss": 1.3308,
      "step": 1410
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3236525058746338,
      "learning_rate": 9.465878808944503e-06,
      "loss": 1.3299,
      "step": 1420
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3962891101837158,
      "learning_rate": 9.454292665971498e-06,
      "loss": 1.3196,
      "step": 1430
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3816227912902832,
      "learning_rate": 9.442706522998494e-06,
      "loss": 1.3104,
      "step": 1440
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3028792142868042,
      "learning_rate": 9.43112038002549e-06,
      "loss": 1.3022,
      "step": 1450
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.361894965171814,
      "learning_rate": 9.419534237052485e-06,
      "loss": 1.3456,
      "step": 1460
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.392577886581421,
      "learning_rate": 9.407948094079482e-06,
      "loss": 1.3036,
      "step": 1470
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4495409727096558,
      "learning_rate": 9.396361951106478e-06,
      "loss": 1.3233,
      "step": 1480
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.3723502159118652,
      "learning_rate": 9.384775808133473e-06,
      "loss": 1.3269,
      "step": 1490
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.2673066854476929,
      "learning_rate": 9.373189665160469e-06,
      "loss": 1.2977,
      "step": 1500
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.491845965385437,
      "learning_rate": 9.361603522187464e-06,
      "loss": 1.3207,
      "step": 1510
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.456560730934143,
      "learning_rate": 9.35001737921446e-06,
      "loss": 1.3064,
      "step": 1520
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.399476170539856,
      "learning_rate": 9.338431236241457e-06,
      "loss": 1.3232,
      "step": 1530
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3272968530654907,
      "learning_rate": 9.326845093268451e-06,
      "loss": 1.3271,
      "step": 1540
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3097448348999023,
      "learning_rate": 9.315258950295448e-06,
      "loss": 1.3015,
      "step": 1550
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.3764145374298096,
      "learning_rate": 9.303672807322444e-06,
      "loss": 1.3044,
      "step": 1560
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4176369905471802,
      "learning_rate": 9.292086664349439e-06,
      "loss": 1.3069,
      "step": 1570
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.3119934797286987,
      "learning_rate": 9.280500521376435e-06,
      "loss": 1.3008,
      "step": 1580
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.2886179685592651,
      "learning_rate": 9.268914378403431e-06,
      "loss": 1.2901,
      "step": 1590
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.2770109176635742,
      "learning_rate": 9.257328235430426e-06,
      "loss": 1.2761,
      "step": 1600
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4204245805740356,
      "learning_rate": 9.24574209245742e-06,
      "loss": 1.2895,
      "step": 1610
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4266266822814941,
      "learning_rate": 9.234155949484417e-06,
      "loss": 1.2987,
      "step": 1620
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3138302564620972,
      "learning_rate": 9.222569806511413e-06,
      "loss": 1.3048,
      "step": 1630
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3465932607650757,
      "learning_rate": 9.210983663538408e-06,
      "loss": 1.3133,
      "step": 1640
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4104775190353394,
      "learning_rate": 9.199397520565405e-06,
      "loss": 1.3001,
      "step": 1650
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3274825811386108,
      "learning_rate": 9.187811377592401e-06,
      "loss": 1.3009,
      "step": 1660
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4182919263839722,
      "learning_rate": 9.176225234619396e-06,
      "loss": 1.3104,
      "step": 1670
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.34902024269104,
      "learning_rate": 9.164639091646392e-06,
      "loss": 1.281,
      "step": 1680
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.463240146636963,
      "learning_rate": 9.153052948673388e-06,
      "loss": 1.3048,
      "step": 1690
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3565565347671509,
      "learning_rate": 9.141466805700383e-06,
      "loss": 1.2868,
      "step": 1700
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.2957673072814941,
      "learning_rate": 9.129880662727378e-06,
      "loss": 1.2985,
      "step": 1710
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.392398476600647,
      "learning_rate": 9.118294519754374e-06,
      "loss": 1.2876,
      "step": 1720
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.4857003688812256,
      "learning_rate": 9.10670837678137e-06,
      "loss": 1.3095,
      "step": 1730
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.3010202646255493,
      "learning_rate": 9.095122233808365e-06,
      "loss": 1.2914,
      "step": 1740
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.4109833240509033,
      "learning_rate": 9.083536090835361e-06,
      "loss": 1.2908,
      "step": 1750
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.4451119899749756,
      "learning_rate": 9.071949947862358e-06,
      "loss": 1.301,
      "step": 1760
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.3937684297561646,
      "learning_rate": 9.060363804889353e-06,
      "loss": 1.3257,
      "step": 1770
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.2829424142837524,
      "learning_rate": 9.048777661916349e-06,
      "loss": 1.2825,
      "step": 1780
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.3617833852767944,
      "learning_rate": 9.037191518943345e-06,
      "loss": 1.2924,
      "step": 1790
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.2105286121368408,
      "learning_rate": 9.02560537597034e-06,
      "loss": 1.312,
      "step": 1800
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.2995182275772095,
      "learning_rate": 9.014019232997336e-06,
      "loss": 1.2883,
      "step": 1810
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.3444055318832397,
      "learning_rate": 9.002433090024331e-06,
      "loss": 1.2948,
      "step": 1820
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.4586807489395142,
      "learning_rate": 8.990846947051327e-06,
      "loss": 1.3089,
      "step": 1830
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.3522496223449707,
      "learning_rate": 8.979260804078324e-06,
      "loss": 1.309,
      "step": 1840
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.3606278896331787,
      "learning_rate": 8.967674661105318e-06,
      "loss": 1.284,
      "step": 1850
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.3653103113174438,
      "learning_rate": 8.956088518132315e-06,
      "loss": 1.2998,
      "step": 1860
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.3912575244903564,
      "learning_rate": 8.944502375159311e-06,
      "loss": 1.292,
      "step": 1870
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.2404125928878784,
      "learning_rate": 8.932916232186306e-06,
      "loss": 1.3031,
      "step": 1880
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4737727642059326,
      "learning_rate": 8.921330089213302e-06,
      "loss": 1.274,
      "step": 1890
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5364305973052979,
      "learning_rate": 8.909743946240297e-06,
      "loss": 1.2855,
      "step": 1900
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.2872428894042969,
      "learning_rate": 8.898157803267293e-06,
      "loss": 1.2933,
      "step": 1910
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.2924387454986572,
      "learning_rate": 8.886571660294288e-06,
      "loss": 1.2801,
      "step": 1920
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4313395023345947,
      "learning_rate": 8.874985517321284e-06,
      "loss": 1.2686,
      "step": 1930
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.3312171697616577,
      "learning_rate": 8.86339937434828e-06,
      "loss": 1.2913,
      "step": 1940
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3304767608642578,
      "learning_rate": 8.851813231375275e-06,
      "loss": 1.3036,
      "step": 1950
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3129816055297852,
      "learning_rate": 8.840227088402272e-06,
      "loss": 1.3102,
      "step": 1960
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4595379829406738,
      "learning_rate": 8.828640945429268e-06,
      "loss": 1.2758,
      "step": 1970
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3031460046768188,
      "learning_rate": 8.817054802456263e-06,
      "loss": 1.2598,
      "step": 1980
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4015743732452393,
      "learning_rate": 8.805468659483259e-06,
      "loss": 1.2897,
      "step": 1990
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.456390380859375,
      "learning_rate": 8.793882516510255e-06,
      "loss": 1.2952,
      "step": 2000
    },
    {
      "epoch": 0.42,
      "eval_loss": 1.2834819555282593,
      "eval_runtime": 60.3193,
      "eval_samples_per_second": 102.786,
      "eval_steps_per_second": 12.848,
      "step": 2000
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4190367460250854,
      "learning_rate": 8.78229637353725e-06,
      "loss": 1.2863,
      "step": 2010
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.3781564235687256,
      "learning_rate": 8.770710230564245e-06,
      "loss": 1.2898,
      "step": 2020
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4007055759429932,
      "learning_rate": 8.759124087591241e-06,
      "loss": 1.3049,
      "step": 2030
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.46671462059021,
      "learning_rate": 8.747537944618238e-06,
      "loss": 1.2931,
      "step": 2040
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.3067024946212769,
      "learning_rate": 8.735951801645232e-06,
      "loss": 1.2758,
      "step": 2050
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.3156636953353882,
      "learning_rate": 8.724365658672229e-06,
      "loss": 1.2912,
      "step": 2060
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4328694343566895,
      "learning_rate": 8.712779515699225e-06,
      "loss": 1.2687,
      "step": 2070
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.3482651710510254,
      "learning_rate": 8.70119337272622e-06,
      "loss": 1.2815,
      "step": 2080
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.3155994415283203,
      "learning_rate": 8.689607229753216e-06,
      "loss": 1.2646,
      "step": 2090
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.3547475337982178,
      "learning_rate": 8.678021086780212e-06,
      "loss": 1.2726,
      "step": 2100
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.439499020576477,
      "learning_rate": 8.666434943807207e-06,
      "loss": 1.2896,
      "step": 2110
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.302492618560791,
      "learning_rate": 8.654848800834203e-06,
      "loss": 1.263,
      "step": 2120
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.2857979536056519,
      "learning_rate": 8.643262657861198e-06,
      "loss": 1.2636,
      "step": 2130
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.362838864326477,
      "learning_rate": 8.631676514888195e-06,
      "loss": 1.2554,
      "step": 2140
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4692515134811401,
      "learning_rate": 8.62009037191519e-06,
      "loss": 1.2589,
      "step": 2150
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.3445467948913574,
      "learning_rate": 8.608504228942186e-06,
      "loss": 1.2829,
      "step": 2160
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.2820574045181274,
      "learning_rate": 8.596918085969182e-06,
      "loss": 1.266,
      "step": 2170
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4762849807739258,
      "learning_rate": 8.585331942996177e-06,
      "loss": 1.2924,
      "step": 2180
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.3395150899887085,
      "learning_rate": 8.573745800023173e-06,
      "loss": 1.2724,
      "step": 2190
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4047057628631592,
      "learning_rate": 8.56215965705017e-06,
      "loss": 1.2877,
      "step": 2200
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.3061277866363525,
      "learning_rate": 8.550573514077164e-06,
      "loss": 1.2805,
      "step": 2210
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.300742268562317,
      "learning_rate": 8.53898737110416e-06,
      "loss": 1.2851,
      "step": 2220
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.3588734865188599,
      "learning_rate": 8.527401228131155e-06,
      "loss": 1.2537,
      "step": 2230
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.4370704889297485,
      "learning_rate": 8.515815085158151e-06,
      "loss": 1.2779,
      "step": 2240
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.3714896440505981,
      "learning_rate": 8.504228942185148e-06,
      "loss": 1.2707,
      "step": 2250
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.4633715152740479,
      "learning_rate": 8.492642799212143e-06,
      "loss": 1.2882,
      "step": 2260
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.459731101989746,
      "learning_rate": 8.481056656239139e-06,
      "loss": 1.2896,
      "step": 2270
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.3027414083480835,
      "learning_rate": 8.469470513266135e-06,
      "loss": 1.2703,
      "step": 2280
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.460031509399414,
      "learning_rate": 8.45788437029313e-06,
      "loss": 1.2871,
      "step": 2290
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.3085455894470215,
      "learning_rate": 8.446298227320126e-06,
      "loss": 1.2552,
      "step": 2300
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.2865551710128784,
      "learning_rate": 8.434712084347123e-06,
      "loss": 1.2706,
      "step": 2310
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.357158899307251,
      "learning_rate": 8.423125941374117e-06,
      "loss": 1.2862,
      "step": 2320
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4073933362960815,
      "learning_rate": 8.411539798401112e-06,
      "loss": 1.2753,
      "step": 2330
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.3302937746047974,
      "learning_rate": 8.399953655428108e-06,
      "loss": 1.2641,
      "step": 2340
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.2936241626739502,
      "learning_rate": 8.388367512455105e-06,
      "loss": 1.2823,
      "step": 2350
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.339048147201538,
      "learning_rate": 8.3767813694821e-06,
      "loss": 1.2477,
      "step": 2360
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.3758487701416016,
      "learning_rate": 8.365195226509096e-06,
      "loss": 1.2625,
      "step": 2370
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.2881660461425781,
      "learning_rate": 8.353609083536092e-06,
      "loss": 1.2886,
      "step": 2380
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.3712249994277954,
      "learning_rate": 8.342022940563087e-06,
      "loss": 1.2708,
      "step": 2390
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.2772799730300903,
      "learning_rate": 8.330436797590082e-06,
      "loss": 1.2669,
      "step": 2400
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.2048228979110718,
      "learning_rate": 8.31885065461708e-06,
      "loss": 1.2543,
      "step": 2410
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.4433040618896484,
      "learning_rate": 8.307264511644074e-06,
      "loss": 1.2656,
      "step": 2420
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.366408109664917,
      "learning_rate": 8.295678368671069e-06,
      "loss": 1.2384,
      "step": 2430
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.311207890510559,
      "learning_rate": 8.284092225698065e-06,
      "loss": 1.2826,
      "step": 2440
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.3012316226959229,
      "learning_rate": 8.272506082725062e-06,
      "loss": 1.2813,
      "step": 2450
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.3766871690750122,
      "learning_rate": 8.260919939752056e-06,
      "loss": 1.2777,
      "step": 2460
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.328155279159546,
      "learning_rate": 8.249333796779053e-06,
      "loss": 1.2778,
      "step": 2470
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.354087471961975,
      "learning_rate": 8.237747653806049e-06,
      "loss": 1.2766,
      "step": 2480
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.2256219387054443,
      "learning_rate": 8.226161510833044e-06,
      "loss": 1.2542,
      "step": 2490
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.453528881072998,
      "learning_rate": 8.21457536786004e-06,
      "loss": 1.2778,
      "step": 2500
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4232090711593628,
      "learning_rate": 8.202989224887037e-06,
      "loss": 1.2963,
      "step": 2510
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.3284271955490112,
      "learning_rate": 8.191403081914031e-06,
      "loss": 1.27,
      "step": 2520
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4262343645095825,
      "learning_rate": 8.179816938941028e-06,
      "loss": 1.2728,
      "step": 2530
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.3023653030395508,
      "learning_rate": 8.168230795968022e-06,
      "loss": 1.2658,
      "step": 2540
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.3244940042495728,
      "learning_rate": 8.156644652995019e-06,
      "loss": 1.2631,
      "step": 2550
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.3604815006256104,
      "learning_rate": 8.145058510022015e-06,
      "loss": 1.2793,
      "step": 2560
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.3496869802474976,
      "learning_rate": 8.13347236704901e-06,
      "loss": 1.2839,
      "step": 2570
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.476453423500061,
      "learning_rate": 8.121886224076006e-06,
      "loss": 1.2556,
      "step": 2580
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4949761629104614,
      "learning_rate": 8.110300081103002e-06,
      "loss": 1.2769,
      "step": 2590
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.2385618686676025,
      "learning_rate": 8.098713938129997e-06,
      "loss": 1.2714,
      "step": 2600
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.3485174179077148,
      "learning_rate": 8.087127795156992e-06,
      "loss": 1.2749,
      "step": 2610
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4038894176483154,
      "learning_rate": 8.07554165218399e-06,
      "loss": 1.2733,
      "step": 2620
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.3258107900619507,
      "learning_rate": 8.063955509210985e-06,
      "loss": 1.2767,
      "step": 2630
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.3624732494354248,
      "learning_rate": 8.05236936623798e-06,
      "loss": 1.2651,
      "step": 2640
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4360556602478027,
      "learning_rate": 8.040783223264976e-06,
      "loss": 1.2519,
      "step": 2650
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4374834299087524,
      "learning_rate": 8.029197080291972e-06,
      "loss": 1.2664,
      "step": 2660
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4204131364822388,
      "learning_rate": 8.017610937318967e-06,
      "loss": 1.271,
      "step": 2670
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.3522765636444092,
      "learning_rate": 8.006024794345963e-06,
      "loss": 1.2499,
      "step": 2680
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.373672604560852,
      "learning_rate": 7.99443865137296e-06,
      "loss": 1.2821,
      "step": 2690
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.474179983139038,
      "learning_rate": 7.982852508399954e-06,
      "loss": 1.2627,
      "step": 2700
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.4560604095458984,
      "learning_rate": 7.971266365426949e-06,
      "loss": 1.2479,
      "step": 2710
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.3803871870040894,
      "learning_rate": 7.959680222453947e-06,
      "loss": 1.2693,
      "step": 2720
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.2998815774917603,
      "learning_rate": 7.948094079480941e-06,
      "loss": 1.2477,
      "step": 2730
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.3741968870162964,
      "learning_rate": 7.936507936507936e-06,
      "loss": 1.25,
      "step": 2740
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.2343312501907349,
      "learning_rate": 7.924921793534933e-06,
      "loss": 1.2628,
      "step": 2750
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.3123588562011719,
      "learning_rate": 7.913335650561929e-06,
      "loss": 1.2817,
      "step": 2760
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.3881012201309204,
      "learning_rate": 7.901749507588924e-06,
      "loss": 1.2461,
      "step": 2770
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.283443808555603,
      "learning_rate": 7.89016336461592e-06,
      "loss": 1.2736,
      "step": 2780
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.3242292404174805,
      "learning_rate": 7.878577221642916e-06,
      "loss": 1.2536,
      "step": 2790
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.3034474849700928,
      "learning_rate": 7.866991078669911e-06,
      "loss": 1.2494,
      "step": 2800
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.243727445602417,
      "learning_rate": 7.855404935696907e-06,
      "loss": 1.2427,
      "step": 2810
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.3450164794921875,
      "learning_rate": 7.843818792723902e-06,
      "loss": 1.2643,
      "step": 2820
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.4294393062591553,
      "learning_rate": 7.832232649750898e-06,
      "loss": 1.262,
      "step": 2830
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.413553237915039,
      "learning_rate": 7.820646506777895e-06,
      "loss": 1.2547,
      "step": 2840
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.3107550144195557,
      "learning_rate": 7.80906036380489e-06,
      "loss": 1.2387,
      "step": 2850
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.3627985715866089,
      "learning_rate": 7.797474220831886e-06,
      "loss": 1.2441,
      "step": 2860
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.2843525409698486,
      "learning_rate": 7.785888077858882e-06,
      "loss": 1.2678,
      "step": 2870
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.1956937313079834,
      "learning_rate": 7.774301934885877e-06,
      "loss": 1.2604,
      "step": 2880
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.4336110353469849,
      "learning_rate": 7.762715791912873e-06,
      "loss": 1.2554,
      "step": 2890
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.223803997039795,
      "learning_rate": 7.75112964893987e-06,
      "loss": 1.2484,
      "step": 2900
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.4403553009033203,
      "learning_rate": 7.739543505966864e-06,
      "loss": 1.255,
      "step": 2910
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.351047396659851,
      "learning_rate": 7.727957362993859e-06,
      "loss": 1.2353,
      "step": 2920
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5024089813232422,
      "learning_rate": 7.716371220020855e-06,
      "loss": 1.2607,
      "step": 2930
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.299880027770996,
      "learning_rate": 7.704785077047852e-06,
      "loss": 1.2835,
      "step": 2940
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.4588313102722168,
      "learning_rate": 7.693198934074846e-06,
      "loss": 1.2534,
      "step": 2950
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.3417565822601318,
      "learning_rate": 7.681612791101843e-06,
      "loss": 1.2288,
      "step": 2960
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.3168967962265015,
      "learning_rate": 7.67002664812884e-06,
      "loss": 1.2291,
      "step": 2970
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.223056674003601,
      "learning_rate": 7.658440505155834e-06,
      "loss": 1.2566,
      "step": 2980
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.3580880165100098,
      "learning_rate": 7.64685436218283e-06,
      "loss": 1.2644,
      "step": 2990
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.3440959453582764,
      "learning_rate": 7.635268219209827e-06,
      "loss": 1.2476,
      "step": 3000
    },
    {
      "epoch": 0.63,
      "eval_loss": 1.2491159439086914,
      "eval_runtime": 60.2595,
      "eval_samples_per_second": 102.888,
      "eval_steps_per_second": 12.861,
      "step": 3000
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.398789882659912,
      "learning_rate": 7.623682076236821e-06,
      "loss": 1.2305,
      "step": 3010
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.2427514791488647,
      "learning_rate": 7.612095933263817e-06,
      "loss": 1.2414,
      "step": 3020
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.2460520267486572,
      "learning_rate": 7.600509790290813e-06,
      "loss": 1.2453,
      "step": 3030
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.310631513595581,
      "learning_rate": 7.588923647317809e-06,
      "loss": 1.2483,
      "step": 3040
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.408805012702942,
      "learning_rate": 7.577337504344804e-06,
      "loss": 1.2074,
      "step": 3050
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.334288239479065,
      "learning_rate": 7.5657513613718006e-06,
      "loss": 1.251,
      "step": 3060
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.305930733680725,
      "learning_rate": 7.554165218398796e-06,
      "loss": 1.251,
      "step": 3070
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.4388633966445923,
      "learning_rate": 7.542579075425791e-06,
      "loss": 1.2508,
      "step": 3080
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.370788812637329,
      "learning_rate": 7.530992932452787e-06,
      "loss": 1.2571,
      "step": 3090
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.303532600402832,
      "learning_rate": 7.519406789479783e-06,
      "loss": 1.269,
      "step": 3100
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.3756309747695923,
      "learning_rate": 7.507820646506778e-06,
      "loss": 1.2607,
      "step": 3110
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.3260369300842285,
      "learning_rate": 7.4962345035337746e-06,
      "loss": 1.2225,
      "step": 3120
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.414718508720398,
      "learning_rate": 7.48464836056077e-06,
      "loss": 1.2436,
      "step": 3130
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.2363390922546387,
      "learning_rate": 7.473062217587766e-06,
      "loss": 1.2376,
      "step": 3140
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.259230613708496,
      "learning_rate": 7.461476074614762e-06,
      "loss": 1.2502,
      "step": 3150
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.397451400756836,
      "learning_rate": 7.4498899316417575e-06,
      "loss": 1.2514,
      "step": 3160
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.3317039012908936,
      "learning_rate": 7.438303788668752e-06,
      "loss": 1.2483,
      "step": 3170
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.3065232038497925,
      "learning_rate": 7.426717645695749e-06,
      "loss": 1.2455,
      "step": 3180
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.4810526371002197,
      "learning_rate": 7.415131502722744e-06,
      "loss": 1.2716,
      "step": 3190
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.314434289932251,
      "learning_rate": 7.40354535974974e-06,
      "loss": 1.212,
      "step": 3200
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.2311912775039673,
      "learning_rate": 7.391959216776735e-06,
      "loss": 1.2214,
      "step": 3210
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.2313491106033325,
      "learning_rate": 7.3803730738037315e-06,
      "loss": 1.223,
      "step": 3220
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.4492835998535156,
      "learning_rate": 7.368786930830727e-06,
      "loss": 1.2434,
      "step": 3230
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.3827974796295166,
      "learning_rate": 7.3572007878577225e-06,
      "loss": 1.2159,
      "step": 3240
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.3691293001174927,
      "learning_rate": 7.345614644884719e-06,
      "loss": 1.2542,
      "step": 3250
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5888055562973022,
      "learning_rate": 7.3340285019117144e-06,
      "loss": 1.2334,
      "step": 3260
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.413843035697937,
      "learning_rate": 7.322442358938709e-06,
      "loss": 1.2479,
      "step": 3270
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.2851001024246216,
      "learning_rate": 7.310856215965706e-06,
      "loss": 1.2389,
      "step": 3280
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.4129613637924194,
      "learning_rate": 7.299270072992701e-06,
      "loss": 1.2397,
      "step": 3290
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.327528953552246,
      "learning_rate": 7.2876839300196965e-06,
      "loss": 1.2404,
      "step": 3300
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.3374613523483276,
      "learning_rate": 7.276097787046693e-06,
      "loss": 1.2442,
      "step": 3310
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.4098230600357056,
      "learning_rate": 7.2645116440736884e-06,
      "loss": 1.2434,
      "step": 3320
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.359785795211792,
      "learning_rate": 7.252925501100684e-06,
      "loss": 1.24,
      "step": 3330
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5303146839141846,
      "learning_rate": 7.24133935812768e-06,
      "loss": 1.2418,
      "step": 3340
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.2719566822052002,
      "learning_rate": 7.229753215154676e-06,
      "loss": 1.2352,
      "step": 3350
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.3264471292495728,
      "learning_rate": 7.218167072181671e-06,
      "loss": 1.2487,
      "step": 3360
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.4324239492416382,
      "learning_rate": 7.206580929208668e-06,
      "loss": 1.2324,
      "step": 3370
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.4195739030838013,
      "learning_rate": 7.1949947862356624e-06,
      "loss": 1.243,
      "step": 3380
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.2976794242858887,
      "learning_rate": 7.183408643262658e-06,
      "loss": 1.245,
      "step": 3390
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.2134556770324707,
      "learning_rate": 7.171822500289654e-06,
      "loss": 1.2501,
      "step": 3400
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.4034926891326904,
      "learning_rate": 7.16023635731665e-06,
      "loss": 1.2501,
      "step": 3410
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.2896411418914795,
      "learning_rate": 7.148650214343645e-06,
      "loss": 1.2168,
      "step": 3420
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.2246516942977905,
      "learning_rate": 7.137064071370642e-06,
      "loss": 1.2563,
      "step": 3430
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.428499698638916,
      "learning_rate": 7.125477928397637e-06,
      "loss": 1.2749,
      "step": 3440
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.279649257659912,
      "learning_rate": 7.113891785424633e-06,
      "loss": 1.2473,
      "step": 3450
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.3014793395996094,
      "learning_rate": 7.102305642451628e-06,
      "loss": 1.238,
      "step": 3460
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.2327125072479248,
      "learning_rate": 7.090719499478625e-06,
      "loss": 1.2499,
      "step": 3470
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.2716156244277954,
      "learning_rate": 7.079133356505619e-06,
      "loss": 1.239,
      "step": 3480
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.3892238140106201,
      "learning_rate": 7.067547213532615e-06,
      "loss": 1.2453,
      "step": 3490
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.3814228773117065,
      "learning_rate": 7.055961070559611e-06,
      "loss": 1.2483,
      "step": 3500
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.327527403831482,
      "learning_rate": 7.044374927586607e-06,
      "loss": 1.2408,
      "step": 3510
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.359143614768982,
      "learning_rate": 7.032788784613602e-06,
      "loss": 1.2401,
      "step": 3520
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.2862249612808228,
      "learning_rate": 7.021202641640599e-06,
      "loss": 1.2587,
      "step": 3530
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.3844903707504272,
      "learning_rate": 7.009616498667594e-06,
      "loss": 1.2313,
      "step": 3540
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.318165898323059,
      "learning_rate": 6.99803035569459e-06,
      "loss": 1.2517,
      "step": 3550
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.3430544137954712,
      "learning_rate": 6.986444212721586e-06,
      "loss": 1.2351,
      "step": 3560
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.3766462802886963,
      "learning_rate": 6.974858069748582e-06,
      "loss": 1.2374,
      "step": 3570
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.3596174716949463,
      "learning_rate": 6.963271926775576e-06,
      "loss": 1.2463,
      "step": 3580
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.2986698150634766,
      "learning_rate": 6.951685783802573e-06,
      "loss": 1.245,
      "step": 3590
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.3160896301269531,
      "learning_rate": 6.940099640829568e-06,
      "loss": 1.2205,
      "step": 3600
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.2758549451828003,
      "learning_rate": 6.928513497856564e-06,
      "loss": 1.2263,
      "step": 3610
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.3435688018798828,
      "learning_rate": 6.91692735488356e-06,
      "loss": 1.2054,
      "step": 3620
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.3898662328720093,
      "learning_rate": 6.905341211910556e-06,
      "loss": 1.2563,
      "step": 3630
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.3642399311065674,
      "learning_rate": 6.893755068937551e-06,
      "loss": 1.1986,
      "step": 3640
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.3172852993011475,
      "learning_rate": 6.8821689259645475e-06,
      "loss": 1.2369,
      "step": 3650
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.2481634616851807,
      "learning_rate": 6.870582782991543e-06,
      "loss": 1.2256,
      "step": 3660
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.3152412176132202,
      "learning_rate": 6.8589966400185386e-06,
      "loss": 1.2692,
      "step": 3670
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.3837155103683472,
      "learning_rate": 6.847410497045535e-06,
      "loss": 1.2395,
      "step": 3680
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.1218920946121216,
      "learning_rate": 6.83582435407253e-06,
      "loss": 1.2387,
      "step": 3690
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.3420273065567017,
      "learning_rate": 6.824238211099525e-06,
      "loss": 1.257,
      "step": 3700
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.4104361534118652,
      "learning_rate": 6.812652068126521e-06,
      "loss": 1.2445,
      "step": 3710
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.2503465414047241,
      "learning_rate": 6.801065925153517e-06,
      "loss": 1.2041,
      "step": 3720
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.4227067232131958,
      "learning_rate": 6.7894797821805126e-06,
      "loss": 1.244,
      "step": 3730
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.4678360223770142,
      "learning_rate": 6.777893639207508e-06,
      "loss": 1.2477,
      "step": 3740
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.2533615827560425,
      "learning_rate": 6.7663074962345045e-06,
      "loss": 1.2232,
      "step": 3750
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.3526480197906494,
      "learning_rate": 6.7547213532615e-06,
      "loss": 1.2248,
      "step": 3760
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.2819933891296387,
      "learning_rate": 6.743135210288495e-06,
      "loss": 1.2299,
      "step": 3770
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.2762075662612915,
      "learning_rate": 6.731549067315492e-06,
      "loss": 1.2518,
      "step": 3780
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.3507285118103027,
      "learning_rate": 6.7199629243424866e-06,
      "loss": 1.2286,
      "step": 3790
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.3762954473495483,
      "learning_rate": 6.708376781369482e-06,
      "loss": 1.2525,
      "step": 3800
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.4141082763671875,
      "learning_rate": 6.6967906383964785e-06,
      "loss": 1.2346,
      "step": 3810
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.3597639799118042,
      "learning_rate": 6.685204495423474e-06,
      "loss": 1.241,
      "step": 3820
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.4315990209579468,
      "learning_rate": 6.6736183524504695e-06,
      "loss": 1.2288,
      "step": 3830
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.4011619091033936,
      "learning_rate": 6.662032209477466e-06,
      "loss": 1.2436,
      "step": 3840
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.160510778427124,
      "learning_rate": 6.650446066504461e-06,
      "loss": 1.2224,
      "step": 3850
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.2349698543548584,
      "learning_rate": 6.638859923531457e-06,
      "loss": 1.2256,
      "step": 3860
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.3648158311843872,
      "learning_rate": 6.627273780558453e-06,
      "loss": 1.2215,
      "step": 3870
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.3032734394073486,
      "learning_rate": 6.615687637585449e-06,
      "loss": 1.2085,
      "step": 3880
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.414401650428772,
      "learning_rate": 6.6041014946124435e-06,
      "loss": 1.2434,
      "step": 3890
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.3182495832443237,
      "learning_rate": 6.59251535163944e-06,
      "loss": 1.2393,
      "step": 3900
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.3610124588012695,
      "learning_rate": 6.580929208666435e-06,
      "loss": 1.2338,
      "step": 3910
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.4505993127822876,
      "learning_rate": 6.569343065693431e-06,
      "loss": 1.247,
      "step": 3920
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.2760661840438843,
      "learning_rate": 6.557756922720427e-06,
      "loss": 1.2333,
      "step": 3930
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.301107406616211,
      "learning_rate": 6.546170779747423e-06,
      "loss": 1.2121,
      "step": 3940
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.3670973777770996,
      "learning_rate": 6.534584636774418e-06,
      "loss": 1.227,
      "step": 3950
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.4468539953231812,
      "learning_rate": 6.522998493801415e-06,
      "loss": 1.2407,
      "step": 3960
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.4032971858978271,
      "learning_rate": 6.51141235082841e-06,
      "loss": 1.2277,
      "step": 3970
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.357784390449524,
      "learning_rate": 6.499826207855405e-06,
      "loss": 1.2354,
      "step": 3980
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.3305027484893799,
      "learning_rate": 6.4882400648824004e-06,
      "loss": 1.2461,
      "step": 3990
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.3124698400497437,
      "learning_rate": 6.476653921909397e-06,
      "loss": 1.2154,
      "step": 4000
    },
    {
      "epoch": 0.83,
      "eval_loss": 1.2254222631454468,
      "eval_runtime": 60.262,
      "eval_samples_per_second": 102.884,
      "eval_steps_per_second": 12.861,
      "step": 4000
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.3544700145721436,
      "learning_rate": 6.465067778936392e-06,
      "loss": 1.2126,
      "step": 4010
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.5063039064407349,
      "learning_rate": 6.453481635963388e-06,
      "loss": 1.242,
      "step": 4020
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.3593188524246216,
      "learning_rate": 6.441895492990384e-06,
      "loss": 1.2336,
      "step": 4030
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.2792909145355225,
      "learning_rate": 6.43030935001738e-06,
      "loss": 1.2166,
      "step": 4040
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.3836535215377808,
      "learning_rate": 6.418723207044375e-06,
      "loss": 1.242,
      "step": 4050
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.3111571073532104,
      "learning_rate": 6.407137064071372e-06,
      "loss": 1.2166,
      "step": 4060
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.3556476831436157,
      "learning_rate": 6.395550921098367e-06,
      "loss": 1.2046,
      "step": 4070
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.4717987775802612,
      "learning_rate": 6.383964778125362e-06,
      "loss": 1.2319,
      "step": 4080
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.3269935846328735,
      "learning_rate": 6.372378635152359e-06,
      "loss": 1.2305,
      "step": 4090
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.2894436120986938,
      "learning_rate": 6.360792492179354e-06,
      "loss": 1.2194,
      "step": 4100
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.3850573301315308,
      "learning_rate": 6.349206349206349e-06,
      "loss": 1.2252,
      "step": 4110
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.3874263763427734,
      "learning_rate": 6.337620206233346e-06,
      "loss": 1.21,
      "step": 4120
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.3668591976165771,
      "learning_rate": 6.326034063260341e-06,
      "loss": 1.2083,
      "step": 4130
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.2573745250701904,
      "learning_rate": 6.314447920287337e-06,
      "loss": 1.2335,
      "step": 4140
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.2464340925216675,
      "learning_rate": 6.302861777314333e-06,
      "loss": 1.2337,
      "step": 4150
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.3820133209228516,
      "learning_rate": 6.291275634341329e-06,
      "loss": 1.2125,
      "step": 4160
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.3374152183532715,
      "learning_rate": 6.279689491368324e-06,
      "loss": 1.2191,
      "step": 4170
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.3884360790252686,
      "learning_rate": 6.2681033483953205e-06,
      "loss": 1.2114,
      "step": 4180
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.2888221740722656,
      "learning_rate": 6.256517205422315e-06,
      "loss": 1.2351,
      "step": 4190
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.272840142250061,
      "learning_rate": 6.244931062449311e-06,
      "loss": 1.2001,
      "step": 4200
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.3938418626785278,
      "learning_rate": 6.233344919476307e-06,
      "loss": 1.202,
      "step": 4210
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.445888876914978,
      "learning_rate": 6.221758776503303e-06,
      "loss": 1.2098,
      "step": 4220
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.320033311843872,
      "learning_rate": 6.210172633530298e-06,
      "loss": 1.1894,
      "step": 4230
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.476035475730896,
      "learning_rate": 6.198586490557294e-06,
      "loss": 1.2006,
      "step": 4240
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.3529783487319946,
      "learning_rate": 6.18700034758429e-06,
      "loss": 1.2424,
      "step": 4250
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.3271926641464233,
      "learning_rate": 6.1754142046112855e-06,
      "loss": 1.2142,
      "step": 4260
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.2901276350021362,
      "learning_rate": 6.163828061638281e-06,
      "loss": 1.2199,
      "step": 4270
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.3919955492019653,
      "learning_rate": 6.152241918665277e-06,
      "loss": 1.23,
      "step": 4280
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.247094750404358,
      "learning_rate": 6.140655775692272e-06,
      "loss": 1.2239,
      "step": 4290
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.2860187292099,
      "learning_rate": 6.129069632719268e-06,
      "loss": 1.2363,
      "step": 4300
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.3451370000839233,
      "learning_rate": 6.117483489746264e-06,
      "loss": 1.2448,
      "step": 4310
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.3651721477508545,
      "learning_rate": 6.1058973467732595e-06,
      "loss": 1.2115,
      "step": 4320
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.351797103881836,
      "learning_rate": 6.094311203800255e-06,
      "loss": 1.2122,
      "step": 4330
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.3105823993682861,
      "learning_rate": 6.082725060827251e-06,
      "loss": 1.2185,
      "step": 4340
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.2687830924987793,
      "learning_rate": 6.071138917854247e-06,
      "loss": 1.2127,
      "step": 4350
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.3442779779434204,
      "learning_rate": 6.0595527748812425e-06,
      "loss": 1.2063,
      "step": 4360
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.4820213317871094,
      "learning_rate": 6.047966631908239e-06,
      "loss": 1.2495,
      "step": 4370
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.4442673921585083,
      "learning_rate": 6.036380488935234e-06,
      "loss": 1.2231,
      "step": 4380
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.328535795211792,
      "learning_rate": 6.024794345962229e-06,
      "loss": 1.2015,
      "step": 4390
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.2676218748092651,
      "learning_rate": 6.013208202989225e-06,
      "loss": 1.2337,
      "step": 4400
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.3665215969085693,
      "learning_rate": 6.001622060016221e-06,
      "loss": 1.2256,
      "step": 4410
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.399819016456604,
      "learning_rate": 5.9900359170432165e-06,
      "loss": 1.2025,
      "step": 4420
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.2961410284042358,
      "learning_rate": 5.978449774070213e-06,
      "loss": 1.2154,
      "step": 4430
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.279374122619629,
      "learning_rate": 5.966863631097208e-06,
      "loss": 1.2248,
      "step": 4440
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.3612409830093384,
      "learning_rate": 5.955277488124204e-06,
      "loss": 1.2473,
      "step": 4450
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.4078785181045532,
      "learning_rate": 5.9436913451512e-06,
      "loss": 1.1953,
      "step": 4460
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.2474583387374878,
      "learning_rate": 5.932105202178196e-06,
      "loss": 1.2183,
      "step": 4470
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.3541713953018188,
      "learning_rate": 5.920519059205191e-06,
      "loss": 1.2096,
      "step": 4480
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.420148253440857,
      "learning_rate": 5.908932916232186e-06,
      "loss": 1.2091,
      "step": 4490
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.2937144041061401,
      "learning_rate": 5.897346773259182e-06,
      "loss": 1.2188,
      "step": 4500
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.4640212059020996,
      "learning_rate": 5.885760630286178e-06,
      "loss": 1.2019,
      "step": 4510
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.3564051389694214,
      "learning_rate": 5.874174487313173e-06,
      "loss": 1.2241,
      "step": 4520
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.3135415315628052,
      "learning_rate": 5.86258834434017e-06,
      "loss": 1.227,
      "step": 4530
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.2767616510391235,
      "learning_rate": 5.851002201367165e-06,
      "loss": 1.228,
      "step": 4540
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.318343162536621,
      "learning_rate": 5.839416058394161e-06,
      "loss": 1.2212,
      "step": 4550
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.4655839204788208,
      "learning_rate": 5.827829915421157e-06,
      "loss": 1.2271,
      "step": 4560
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.3469165563583374,
      "learning_rate": 5.816243772448153e-06,
      "loss": 1.2359,
      "step": 4570
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.2466676235198975,
      "learning_rate": 5.804657629475147e-06,
      "loss": 1.2127,
      "step": 4580
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.3727000951766968,
      "learning_rate": 5.793071486502145e-06,
      "loss": 1.2326,
      "step": 4590
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.25077223777771,
      "learning_rate": 5.781485343529139e-06,
      "loss": 1.2383,
      "step": 4600
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.2925761938095093,
      "learning_rate": 5.769899200556135e-06,
      "loss": 1.229,
      "step": 4610
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.2797605991363525,
      "learning_rate": 5.758313057583131e-06,
      "loss": 1.1943,
      "step": 4620
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.3911211490631104,
      "learning_rate": 5.746726914610127e-06,
      "loss": 1.2158,
      "step": 4630
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.3184133768081665,
      "learning_rate": 5.735140771637122e-06,
      "loss": 1.2256,
      "step": 4640
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.242498755455017,
      "learning_rate": 5.723554628664119e-06,
      "loss": 1.2232,
      "step": 4650
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.2783172130584717,
      "learning_rate": 5.711968485691114e-06,
      "loss": 1.2151,
      "step": 4660
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.1488908529281616,
      "learning_rate": 5.70038234271811e-06,
      "loss": 1.2222,
      "step": 4670
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.2965818643569946,
      "learning_rate": 5.688796199745106e-06,
      "loss": 1.2135,
      "step": 4680
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.4008314609527588,
      "learning_rate": 5.6772100567721015e-06,
      "loss": 1.21,
      "step": 4690
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.3333206176757812,
      "learning_rate": 5.665623913799096e-06,
      "loss": 1.235,
      "step": 4700
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.3453115224838257,
      "learning_rate": 5.654037770826093e-06,
      "loss": 1.2175,
      "step": 4710
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.3863331079483032,
      "learning_rate": 5.642451627853088e-06,
      "loss": 1.2171,
      "step": 4720
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.3186782598495483,
      "learning_rate": 5.630865484880084e-06,
      "loss": 1.2179,
      "step": 4730
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.394921898841858,
      "learning_rate": 5.61927934190708e-06,
      "loss": 1.2374,
      "step": 4740
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.2399722337722778,
      "learning_rate": 5.6076931989340755e-06,
      "loss": 1.201,
      "step": 4750
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.2864711284637451,
      "learning_rate": 5.596107055961071e-06,
      "loss": 1.2222,
      "step": 4760
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.3668363094329834,
      "learning_rate": 5.584520912988067e-06,
      "loss": 1.2037,
      "step": 4770
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4248427152633667,
      "learning_rate": 5.572934770015063e-06,
      "loss": 1.2148,
      "step": 4780
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.3241183757781982,
      "learning_rate": 5.561348627042058e-06,
      "loss": 1.217,
      "step": 4790
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.2951161861419678,
      "learning_rate": 5.549762484069053e-06,
      "loss": 1.2082,
      "step": 4800
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4023364782333374,
      "learning_rate": 5.5381763410960495e-06,
      "loss": 1.1816,
      "step": 4810
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.3557953834533691,
      "learning_rate": 5.526590198123045e-06,
      "loss": 1.184,
      "step": 4820
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.282929539680481,
      "learning_rate": 5.515004055150041e-06,
      "loss": 1.1849,
      "step": 4830
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.3420430421829224,
      "learning_rate": 5.503417912177037e-06,
      "loss": 1.1783,
      "step": 4840
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.4730277061462402,
      "learning_rate": 5.4918317692040325e-06,
      "loss": 1.1656,
      "step": 4850
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.3748600482940674,
      "learning_rate": 5.480245626231028e-06,
      "loss": 1.2053,
      "step": 4860
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.395768165588379,
      "learning_rate": 5.468659483258024e-06,
      "loss": 1.1576,
      "step": 4870
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.3293030261993408,
      "learning_rate": 5.45707334028502e-06,
      "loss": 1.1857,
      "step": 4880
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.3349307775497437,
      "learning_rate": 5.4454871973120146e-06,
      "loss": 1.1906,
      "step": 4890
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.4291343688964844,
      "learning_rate": 5.433901054339012e-06,
      "loss": 1.1987,
      "step": 4900
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.542372703552246,
      "learning_rate": 5.4223149113660065e-06,
      "loss": 1.1728,
      "step": 4910
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.3494104146957397,
      "learning_rate": 5.410728768393002e-06,
      "loss": 1.1773,
      "step": 4920
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.3300869464874268,
      "learning_rate": 5.399142625419998e-06,
      "loss": 1.1648,
      "step": 4930
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.3492801189422607,
      "learning_rate": 5.387556482446994e-06,
      "loss": 1.1953,
      "step": 4940
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.3387207984924316,
      "learning_rate": 5.375970339473989e-06,
      "loss": 1.1847,
      "step": 4950
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.323564052581787,
      "learning_rate": 5.364384196500986e-06,
      "loss": 1.2085,
      "step": 4960
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.3828566074371338,
      "learning_rate": 5.352798053527981e-06,
      "loss": 1.1617,
      "step": 4970
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.4634027481079102,
      "learning_rate": 5.341211910554977e-06,
      "loss": 1.1663,
      "step": 4980
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.2878636121749878,
      "learning_rate": 5.329625767581973e-06,
      "loss": 1.1809,
      "step": 4990
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.2984651327133179,
      "learning_rate": 5.318039624608968e-06,
      "loss": 1.2281,
      "step": 5000
    },
    {
      "epoch": 1.04,
      "eval_loss": 1.2093191146850586,
      "eval_runtime": 60.3546,
      "eval_samples_per_second": 102.726,
      "eval_steps_per_second": 12.841,
      "step": 5000
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.4071316719055176,
      "learning_rate": 5.306453481635963e-06,
      "loss": 1.1754,
      "step": 5010
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.2421307563781738,
      "learning_rate": 5.294867338662959e-06,
      "loss": 1.1785,
      "step": 5020
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.3964232206344604,
      "learning_rate": 5.283281195689955e-06,
      "loss": 1.1833,
      "step": 5030
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.4068059921264648,
      "learning_rate": 5.271695052716951e-06,
      "loss": 1.1738,
      "step": 5040
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.273362398147583,
      "learning_rate": 5.260108909743946e-06,
      "loss": 1.1752,
      "step": 5050
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.3210726976394653,
      "learning_rate": 5.248522766770943e-06,
      "loss": 1.1819,
      "step": 5060
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.339066743850708,
      "learning_rate": 5.236936623797938e-06,
      "loss": 1.1803,
      "step": 5070
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.2689433097839355,
      "learning_rate": 5.225350480824933e-06,
      "loss": 1.1899,
      "step": 5080
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.339124083518982,
      "learning_rate": 5.21376433785193e-06,
      "loss": 1.1743,
      "step": 5090
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.321183681488037,
      "learning_rate": 5.202178194878925e-06,
      "loss": 1.1832,
      "step": 5100
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.3904688358306885,
      "learning_rate": 5.19059205190592e-06,
      "loss": 1.1951,
      "step": 5110
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.421046495437622,
      "learning_rate": 5.179005908932917e-06,
      "loss": 1.1841,
      "step": 5120
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.26221764087677,
      "learning_rate": 5.167419765959912e-06,
      "loss": 1.1932,
      "step": 5130
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.5197116136550903,
      "learning_rate": 5.155833622986908e-06,
      "loss": 1.177,
      "step": 5140
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.3857975006103516,
      "learning_rate": 5.144247480013904e-06,
      "loss": 1.1803,
      "step": 5150
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.3418445587158203,
      "learning_rate": 5.1326613370409e-06,
      "loss": 1.1969,
      "step": 5160
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.4151560068130493,
      "learning_rate": 5.121075194067895e-06,
      "loss": 1.179,
      "step": 5170
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.4019502401351929,
      "learning_rate": 5.1094890510948916e-06,
      "loss": 1.1797,
      "step": 5180
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.3561010360717773,
      "learning_rate": 5.097902908121887e-06,
      "loss": 1.1906,
      "step": 5190
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.344114899635315,
      "learning_rate": 5.086316765148882e-06,
      "loss": 1.1851,
      "step": 5200
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.3627663850784302,
      "learning_rate": 5.074730622175878e-06,
      "loss": 1.154,
      "step": 5210
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.3944242000579834,
      "learning_rate": 5.063144479202874e-06,
      "loss": 1.1701,
      "step": 5220
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.3660093545913696,
      "learning_rate": 5.051558336229869e-06,
      "loss": 1.1861,
      "step": 5230
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.2477737665176392,
      "learning_rate": 5.0399721932568656e-06,
      "loss": 1.1917,
      "step": 5240
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.381079912185669,
      "learning_rate": 5.028386050283861e-06,
      "loss": 1.176,
      "step": 5250
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.277708649635315,
      "learning_rate": 5.016799907310857e-06,
      "loss": 1.2262,
      "step": 5260
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.4118417501449585,
      "learning_rate": 5.005213764337853e-06,
      "loss": 1.194,
      "step": 5270
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.283445119857788,
      "learning_rate": 4.9936276213648485e-06,
      "loss": 1.1641,
      "step": 5280
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.3364849090576172,
      "learning_rate": 4.982041478391844e-06,
      "loss": 1.1904,
      "step": 5290
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.3707315921783447,
      "learning_rate": 4.9704553354188395e-06,
      "loss": 1.2228,
      "step": 5300
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.3826156854629517,
      "learning_rate": 4.958869192445835e-06,
      "loss": 1.1808,
      "step": 5310
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.336873173713684,
      "learning_rate": 4.947283049472831e-06,
      "loss": 1.1765,
      "step": 5320
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.1747794151306152,
      "learning_rate": 4.935696906499827e-06,
      "loss": 1.1805,
      "step": 5330
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.4768750667572021,
      "learning_rate": 4.924110763526822e-06,
      "loss": 1.2086,
      "step": 5340
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.5451874732971191,
      "learning_rate": 4.912524620553818e-06,
      "loss": 1.1714,
      "step": 5350
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.420366644859314,
      "learning_rate": 4.9009384775808135e-06,
      "loss": 1.1848,
      "step": 5360
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.3127976655960083,
      "learning_rate": 4.889352334607809e-06,
      "loss": 1.164,
      "step": 5370
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.3515974283218384,
      "learning_rate": 4.8777661916348054e-06,
      "loss": 1.1703,
      "step": 5380
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.3823376893997192,
      "learning_rate": 4.866180048661801e-06,
      "loss": 1.192,
      "step": 5390
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.300921082496643,
      "learning_rate": 4.8545939056887965e-06,
      "loss": 1.1903,
      "step": 5400
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.3634570837020874,
      "learning_rate": 4.843007762715792e-06,
      "loss": 1.1683,
      "step": 5410
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.4684809446334839,
      "learning_rate": 4.831421619742788e-06,
      "loss": 1.1805,
      "step": 5420
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.3468976020812988,
      "learning_rate": 4.819835476769784e-06,
      "loss": 1.1739,
      "step": 5430
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.4344850778579712,
      "learning_rate": 4.8082493337967794e-06,
      "loss": 1.1568,
      "step": 5440
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.379841685295105,
      "learning_rate": 4.796663190823776e-06,
      "loss": 1.2019,
      "step": 5450
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.5211639404296875,
      "learning_rate": 4.7850770478507705e-06,
      "loss": 1.1752,
      "step": 5460
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.3554086685180664,
      "learning_rate": 4.773490904877767e-06,
      "loss": 1.1694,
      "step": 5470
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.3625938892364502,
      "learning_rate": 4.761904761904762e-06,
      "loss": 1.1554,
      "step": 5480
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.2949503660202026,
      "learning_rate": 4.750318618931758e-06,
      "loss": 1.1843,
      "step": 5490
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.2827516794204712,
      "learning_rate": 4.738732475958754e-06,
      "loss": 1.1819,
      "step": 5500
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.2222222089767456,
      "learning_rate": 4.727146332985749e-06,
      "loss": 1.1943,
      "step": 5510
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.419891357421875,
      "learning_rate": 4.715560190012745e-06,
      "loss": 1.1813,
      "step": 5520
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.2174855470657349,
      "learning_rate": 4.703974047039741e-06,
      "loss": 1.1664,
      "step": 5530
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.2947397232055664,
      "learning_rate": 4.692387904066736e-06,
      "loss": 1.1629,
      "step": 5540
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.3705869913101196,
      "learning_rate": 4.680801761093732e-06,
      "loss": 1.1929,
      "step": 5550
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.3978757858276367,
      "learning_rate": 4.669215618120728e-06,
      "loss": 1.1859,
      "step": 5560
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.4621931314468384,
      "learning_rate": 4.657629475147724e-06,
      "loss": 1.1839,
      "step": 5570
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.29670250415802,
      "learning_rate": 4.646043332174719e-06,
      "loss": 1.1709,
      "step": 5580
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.2986425161361694,
      "learning_rate": 4.634457189201716e-06,
      "loss": 1.1616,
      "step": 5590
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.3808799982070923,
      "learning_rate": 4.62287104622871e-06,
      "loss": 1.1606,
      "step": 5600
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.2885197401046753,
      "learning_rate": 4.611284903255707e-06,
      "loss": 1.173,
      "step": 5610
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.4067176580429077,
      "learning_rate": 4.599698760282702e-06,
      "loss": 1.1567,
      "step": 5620
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.2855522632598877,
      "learning_rate": 4.588112617309698e-06,
      "loss": 1.1577,
      "step": 5630
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.365801453590393,
      "learning_rate": 4.576526474336694e-06,
      "loss": 1.1897,
      "step": 5640
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.3854248523712158,
      "learning_rate": 4.564940331363689e-06,
      "loss": 1.1714,
      "step": 5650
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.288671851158142,
      "learning_rate": 4.553354188390685e-06,
      "loss": 1.1691,
      "step": 5660
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.3368744850158691,
      "learning_rate": 4.541768045417681e-06,
      "loss": 1.1676,
      "step": 5670
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.3967853784561157,
      "learning_rate": 4.530181902444676e-06,
      "loss": 1.1816,
      "step": 5680
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.2882468700408936,
      "learning_rate": 4.518595759471673e-06,
      "loss": 1.1732,
      "step": 5690
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.3411914110183716,
      "learning_rate": 4.507009616498668e-06,
      "loss": 1.1943,
      "step": 5700
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.3860743045806885,
      "learning_rate": 4.495423473525664e-06,
      "loss": 1.1771,
      "step": 5710
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.3196132183074951,
      "learning_rate": 4.483837330552659e-06,
      "loss": 1.1661,
      "step": 5720
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.3677616119384766,
      "learning_rate": 4.4722511875796556e-06,
      "loss": 1.1815,
      "step": 5730
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.4438225030899048,
      "learning_rate": 4.460665044606651e-06,
      "loss": 1.193,
      "step": 5740
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.2998027801513672,
      "learning_rate": 4.449078901633647e-06,
      "loss": 1.1665,
      "step": 5750
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.2200978994369507,
      "learning_rate": 4.437492758660642e-06,
      "loss": 1.1706,
      "step": 5760
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.2530813217163086,
      "learning_rate": 4.425906615687638e-06,
      "loss": 1.1836,
      "step": 5770
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.353306770324707,
      "learning_rate": 4.414320472714634e-06,
      "loss": 1.1792,
      "step": 5780
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.3246160745620728,
      "learning_rate": 4.4027343297416296e-06,
      "loss": 1.1785,
      "step": 5790
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.5756295919418335,
      "learning_rate": 4.391148186768625e-06,
      "loss": 1.207,
      "step": 5800
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.3847354650497437,
      "learning_rate": 4.379562043795621e-06,
      "loss": 1.1711,
      "step": 5810
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.2891080379486084,
      "learning_rate": 4.367975900822616e-06,
      "loss": 1.1649,
      "step": 5820
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.4148257970809937,
      "learning_rate": 4.3563897578496125e-06,
      "loss": 1.1836,
      "step": 5830
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.3125419616699219,
      "learning_rate": 4.344803614876608e-06,
      "loss": 1.1831,
      "step": 5840
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.3058887720108032,
      "learning_rate": 4.3332174719036036e-06,
      "loss": 1.1615,
      "step": 5850
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.2639198303222656,
      "learning_rate": 4.321631328930599e-06,
      "loss": 1.1642,
      "step": 5860
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.4334591627120972,
      "learning_rate": 4.310045185957595e-06,
      "loss": 1.1628,
      "step": 5870
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.3660703897476196,
      "learning_rate": 4.298459042984591e-06,
      "loss": 1.1861,
      "step": 5880
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.3341001272201538,
      "learning_rate": 4.2868729000115865e-06,
      "loss": 1.1316,
      "step": 5890
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.348137378692627,
      "learning_rate": 4.275286757038582e-06,
      "loss": 1.147,
      "step": 5900
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.3110828399658203,
      "learning_rate": 4.2637006140655775e-06,
      "loss": 1.1877,
      "step": 5910
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.3300223350524902,
      "learning_rate": 4.252114471092574e-06,
      "loss": 1.1824,
      "step": 5920
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.299660086631775,
      "learning_rate": 4.2405283281195694e-06,
      "loss": 1.1811,
      "step": 5930
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.3512821197509766,
      "learning_rate": 4.228942185146565e-06,
      "loss": 1.1575,
      "step": 5940
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.4316037893295288,
      "learning_rate": 4.217356042173561e-06,
      "loss": 1.1925,
      "step": 5950
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.266357183456421,
      "learning_rate": 4.205769899200556e-06,
      "loss": 1.1641,
      "step": 5960
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.329677939414978,
      "learning_rate": 4.194183756227552e-06,
      "loss": 1.1879,
      "step": 5970
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.4943017959594727,
      "learning_rate": 4.182597613254548e-06,
      "loss": 1.1731,
      "step": 5980
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.2583523988723755,
      "learning_rate": 4.1710114702815434e-06,
      "loss": 1.1659,
      "step": 5990
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.5405082702636719,
      "learning_rate": 4.15942532730854e-06,
      "loss": 1.1796,
      "step": 6000
    },
    {
      "epoch": 1.25,
      "eval_loss": 1.1978083848953247,
      "eval_runtime": 60.2518,
      "eval_samples_per_second": 102.901,
      "eval_steps_per_second": 12.863,
      "step": 6000
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.3956575393676758,
      "learning_rate": 4.1478391843355345e-06,
      "loss": 1.1614,
      "step": 6010
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.3546531200408936,
      "learning_rate": 4.136253041362531e-06,
      "loss": 1.1583,
      "step": 6020
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.4522563219070435,
      "learning_rate": 4.124666898389526e-06,
      "loss": 1.191,
      "step": 6030
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.442973017692566,
      "learning_rate": 4.113080755416522e-06,
      "loss": 1.1779,
      "step": 6040
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.2976762056350708,
      "learning_rate": 4.101494612443518e-06,
      "loss": 1.1713,
      "step": 6050
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.338387131690979,
      "learning_rate": 4.089908469470514e-06,
      "loss": 1.1736,
      "step": 6060
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.4153416156768799,
      "learning_rate": 4.078322326497509e-06,
      "loss": 1.1773,
      "step": 6070
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.4394385814666748,
      "learning_rate": 4.066736183524505e-06,
      "loss": 1.1923,
      "step": 6080
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.3546093702316284,
      "learning_rate": 4.055150040551501e-06,
      "loss": 1.172,
      "step": 6090
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.2290021181106567,
      "learning_rate": 4.043563897578496e-06,
      "loss": 1.1694,
      "step": 6100
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.297818899154663,
      "learning_rate": 4.031977754605492e-06,
      "loss": 1.1683,
      "step": 6110
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.3360967636108398,
      "learning_rate": 4.020391611632488e-06,
      "loss": 1.1645,
      "step": 6120
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.4013689756393433,
      "learning_rate": 4.008805468659483e-06,
      "loss": 1.1583,
      "step": 6130
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.3673937320709229,
      "learning_rate": 3.99721932568648e-06,
      "loss": 1.1418,
      "step": 6140
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.4013173580169678,
      "learning_rate": 3.985633182713474e-06,
      "loss": 1.1562,
      "step": 6150
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.23403799533844,
      "learning_rate": 3.974047039740471e-06,
      "loss": 1.1731,
      "step": 6160
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.3346298933029175,
      "learning_rate": 3.962460896767466e-06,
      "loss": 1.1548,
      "step": 6170
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.352409839630127,
      "learning_rate": 3.950874753794462e-06,
      "loss": 1.1872,
      "step": 6180
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.491523265838623,
      "learning_rate": 3.939288610821458e-06,
      "loss": 1.1774,
      "step": 6190
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.393288016319275,
      "learning_rate": 3.927702467848454e-06,
      "loss": 1.2024,
      "step": 6200
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.455191969871521,
      "learning_rate": 3.916116324875449e-06,
      "loss": 1.1672,
      "step": 6210
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.3734558820724487,
      "learning_rate": 3.904530181902445e-06,
      "loss": 1.1686,
      "step": 6220
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.35550057888031,
      "learning_rate": 3.892944038929441e-06,
      "loss": 1.1663,
      "step": 6230
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.2901463508605957,
      "learning_rate": 3.881357895956437e-06,
      "loss": 1.1938,
      "step": 6240
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.272072196006775,
      "learning_rate": 3.869771752983432e-06,
      "loss": 1.1487,
      "step": 6250
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.3195065259933472,
      "learning_rate": 3.858185610010428e-06,
      "loss": 1.14,
      "step": 6260
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.3816322088241577,
      "learning_rate": 3.846599467037423e-06,
      "loss": 1.1517,
      "step": 6270
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.2947728633880615,
      "learning_rate": 3.83501332406442e-06,
      "loss": 1.1686,
      "step": 6280
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.3573113679885864,
      "learning_rate": 3.823427181091415e-06,
      "loss": 1.1961,
      "step": 6290
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.3107391595840454,
      "learning_rate": 3.8118410381184106e-06,
      "loss": 1.1891,
      "step": 6300
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.4028772115707397,
      "learning_rate": 3.8002548951454066e-06,
      "loss": 1.1514,
      "step": 6310
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.3728693723678589,
      "learning_rate": 3.788668752172402e-06,
      "loss": 1.1378,
      "step": 6320
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.3074133396148682,
      "learning_rate": 3.777082609199398e-06,
      "loss": 1.1887,
      "step": 6330
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.4246951341629028,
      "learning_rate": 3.7654964662263936e-06,
      "loss": 1.1914,
      "step": 6340
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.4968029260635376,
      "learning_rate": 3.753910323253389e-06,
      "loss": 1.1747,
      "step": 6350
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.4184929132461548,
      "learning_rate": 3.742324180280385e-06,
      "loss": 1.1756,
      "step": 6360
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.4123560190200806,
      "learning_rate": 3.730738037307381e-06,
      "loss": 1.1737,
      "step": 6370
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.4032166004180908,
      "learning_rate": 3.719151894334376e-06,
      "loss": 1.1663,
      "step": 6380
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.4489792585372925,
      "learning_rate": 3.707565751361372e-06,
      "loss": 1.1667,
      "step": 6390
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.5236833095550537,
      "learning_rate": 3.6959796083883676e-06,
      "loss": 1.1548,
      "step": 6400
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.4273195266723633,
      "learning_rate": 3.6843934654153635e-06,
      "loss": 1.1724,
      "step": 6410
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.3556939363479614,
      "learning_rate": 3.6728073224423595e-06,
      "loss": 1.1646,
      "step": 6420
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.4586809873580933,
      "learning_rate": 3.6612211794693546e-06,
      "loss": 1.1648,
      "step": 6430
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.3400959968566895,
      "learning_rate": 3.6496350364963505e-06,
      "loss": 1.149,
      "step": 6440
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.4101979732513428,
      "learning_rate": 3.6380488935233465e-06,
      "loss": 1.1516,
      "step": 6450
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.3828833103179932,
      "learning_rate": 3.626462750550342e-06,
      "loss": 1.1609,
      "step": 6460
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.4406893253326416,
      "learning_rate": 3.614876607577338e-06,
      "loss": 1.1927,
      "step": 6470
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.3856877088546753,
      "learning_rate": 3.603290464604334e-06,
      "loss": 1.1618,
      "step": 6480
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.5042824745178223,
      "learning_rate": 3.591704321631329e-06,
      "loss": 1.169,
      "step": 6490
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.2347859144210815,
      "learning_rate": 3.580118178658325e-06,
      "loss": 1.1693,
      "step": 6500
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.4108585119247437,
      "learning_rate": 3.568532035685321e-06,
      "loss": 1.1807,
      "step": 6510
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.4458739757537842,
      "learning_rate": 3.5569458927123164e-06,
      "loss": 1.1801,
      "step": 6520
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.3607912063598633,
      "learning_rate": 3.5453597497393123e-06,
      "loss": 1.1735,
      "step": 6530
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.4772734642028809,
      "learning_rate": 3.5337736067663074e-06,
      "loss": 1.1798,
      "step": 6540
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.3552745580673218,
      "learning_rate": 3.5221874637933034e-06,
      "loss": 1.1558,
      "step": 6550
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.3368369340896606,
      "learning_rate": 3.5106013208202993e-06,
      "loss": 1.1902,
      "step": 6560
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.4489260911941528,
      "learning_rate": 3.499015177847295e-06,
      "loss": 1.1933,
      "step": 6570
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.2601078748703003,
      "learning_rate": 3.487429034874291e-06,
      "loss": 1.16,
      "step": 6580
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.3712470531463623,
      "learning_rate": 3.4758428919012863e-06,
      "loss": 1.1848,
      "step": 6590
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.5175118446350098,
      "learning_rate": 3.464256748928282e-06,
      "loss": 1.2016,
      "step": 6600
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.3620766401290894,
      "learning_rate": 3.452670605955278e-06,
      "loss": 1.1572,
      "step": 6610
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.3859511613845825,
      "learning_rate": 3.4410844629822738e-06,
      "loss": 1.179,
      "step": 6620
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.3397611379623413,
      "learning_rate": 3.4294983200092693e-06,
      "loss": 1.1589,
      "step": 6630
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.3851667642593384,
      "learning_rate": 3.417912177036265e-06,
      "loss": 1.1679,
      "step": 6640
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.3727117776870728,
      "learning_rate": 3.4063260340632603e-06,
      "loss": 1.1722,
      "step": 6650
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.3881512880325317,
      "learning_rate": 3.3947398910902563e-06,
      "loss": 1.1577,
      "step": 6660
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.3602831363677979,
      "learning_rate": 3.3831537481172522e-06,
      "loss": 1.1661,
      "step": 6670
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.3330827951431274,
      "learning_rate": 3.3715676051442473e-06,
      "loss": 1.1514,
      "step": 6680
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.4532897472381592,
      "learning_rate": 3.3599814621712433e-06,
      "loss": 1.1774,
      "step": 6690
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.3237956762313843,
      "learning_rate": 3.3483953191982392e-06,
      "loss": 1.1711,
      "step": 6700
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.2269086837768555,
      "learning_rate": 3.3368091762252348e-06,
      "loss": 1.1624,
      "step": 6710
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.2189347743988037,
      "learning_rate": 3.3252230332522307e-06,
      "loss": 1.1687,
      "step": 6720
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.3546026945114136,
      "learning_rate": 3.3136368902792266e-06,
      "loss": 1.163,
      "step": 6730
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.2999235391616821,
      "learning_rate": 3.3020507473062217e-06,
      "loss": 1.1456,
      "step": 6740
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.30852210521698,
      "learning_rate": 3.2904646043332177e-06,
      "loss": 1.1892,
      "step": 6750
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.546865463256836,
      "learning_rate": 3.2788784613602136e-06,
      "loss": 1.1683,
      "step": 6760
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.6763235330581665,
      "learning_rate": 3.267292318387209e-06,
      "loss": 1.1826,
      "step": 6770
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.3352094888687134,
      "learning_rate": 3.255706175414205e-06,
      "loss": 1.1588,
      "step": 6780
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.5526775121688843,
      "learning_rate": 3.2441200324412002e-06,
      "loss": 1.16,
      "step": 6790
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.2830092906951904,
      "learning_rate": 3.232533889468196e-06,
      "loss": 1.1536,
      "step": 6800
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.3866894245147705,
      "learning_rate": 3.220947746495192e-06,
      "loss": 1.1466,
      "step": 6810
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.4593145847320557,
      "learning_rate": 3.2093616035221876e-06,
      "loss": 1.2051,
      "step": 6820
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.231041669845581,
      "learning_rate": 3.1977754605491836e-06,
      "loss": 1.1576,
      "step": 6830
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.3121471405029297,
      "learning_rate": 3.1861893175761795e-06,
      "loss": 1.1806,
      "step": 6840
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.513574242591858,
      "learning_rate": 3.1746031746031746e-06,
      "loss": 1.1721,
      "step": 6850
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.3530925512313843,
      "learning_rate": 3.1630170316301706e-06,
      "loss": 1.1623,
      "step": 6860
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.3993455171585083,
      "learning_rate": 3.1514308886571665e-06,
      "loss": 1.1674,
      "step": 6870
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.3361057043075562,
      "learning_rate": 3.139844745684162e-06,
      "loss": 1.1677,
      "step": 6880
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.4186984300613403,
      "learning_rate": 3.1282586027111576e-06,
      "loss": 1.1863,
      "step": 6890
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.5800833702087402,
      "learning_rate": 3.1166724597381535e-06,
      "loss": 1.1733,
      "step": 6900
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.3589065074920654,
      "learning_rate": 3.105086316765149e-06,
      "loss": 1.1497,
      "step": 6910
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.3705077171325684,
      "learning_rate": 3.093500173792145e-06,
      "loss": 1.1578,
      "step": 6920
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.3242499828338623,
      "learning_rate": 3.0819140308191405e-06,
      "loss": 1.1678,
      "step": 6930
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.4003245830535889,
      "learning_rate": 3.070327887846136e-06,
      "loss": 1.1653,
      "step": 6940
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.3863157033920288,
      "learning_rate": 3.058741744873132e-06,
      "loss": 1.1787,
      "step": 6950
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.401785969734192,
      "learning_rate": 3.0471556019001275e-06,
      "loss": 1.1697,
      "step": 6960
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.291735053062439,
      "learning_rate": 3.0355694589271235e-06,
      "loss": 1.1729,
      "step": 6970
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.374833583831787,
      "learning_rate": 3.0239833159541194e-06,
      "loss": 1.1623,
      "step": 6980
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.388802409172058,
      "learning_rate": 3.0123971729811145e-06,
      "loss": 1.1655,
      "step": 6990
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.3213307857513428,
      "learning_rate": 3.0008110300081105e-06,
      "loss": 1.1698,
      "step": 7000
    },
    {
      "epoch": 1.46,
      "eval_loss": 1.188651204109192,
      "eval_runtime": 60.2084,
      "eval_samples_per_second": 102.976,
      "eval_steps_per_second": 12.872,
      "step": 7000
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.284946322441101,
      "learning_rate": 2.9892248870351064e-06,
      "loss": 1.1619,
      "step": 7010
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.4130306243896484,
      "learning_rate": 2.977638744062102e-06,
      "loss": 1.1714,
      "step": 7020
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.4033979177474976,
      "learning_rate": 2.966052601089098e-06,
      "loss": 1.1564,
      "step": 7030
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.303514003753662,
      "learning_rate": 2.954466458116093e-06,
      "loss": 1.1506,
      "step": 7040
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.3504719734191895,
      "learning_rate": 2.942880315143089e-06,
      "loss": 1.1647,
      "step": 7050
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.319595456123352,
      "learning_rate": 2.931294172170085e-06,
      "loss": 1.1663,
      "step": 7060
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.6136802434921265,
      "learning_rate": 2.9197080291970804e-06,
      "loss": 1.1508,
      "step": 7070
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.470433235168457,
      "learning_rate": 2.9081218862240764e-06,
      "loss": 1.1777,
      "step": 7080
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.3718762397766113,
      "learning_rate": 2.8965357432510723e-06,
      "loss": 1.1589,
      "step": 7090
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.4196834564208984,
      "learning_rate": 2.8849496002780674e-06,
      "loss": 1.1799,
      "step": 7100
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.301185131072998,
      "learning_rate": 2.8733634573050634e-06,
      "loss": 1.1518,
      "step": 7110
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.3427518606185913,
      "learning_rate": 2.8617773143320593e-06,
      "loss": 1.1566,
      "step": 7120
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.4186878204345703,
      "learning_rate": 2.850191171359055e-06,
      "loss": 1.1825,
      "step": 7130
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.3203967809677124,
      "learning_rate": 2.8386050283860508e-06,
      "loss": 1.1566,
      "step": 7140
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.490299940109253,
      "learning_rate": 2.8270188854130463e-06,
      "loss": 1.1774,
      "step": 7150
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.4719760417938232,
      "learning_rate": 2.815432742440042e-06,
      "loss": 1.1621,
      "step": 7160
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.284829020500183,
      "learning_rate": 2.8038465994670378e-06,
      "loss": 1.1575,
      "step": 7170
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.3767919540405273,
      "learning_rate": 2.7922604564940333e-06,
      "loss": 1.1623,
      "step": 7180
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.3010685443878174,
      "learning_rate": 2.780674313521029e-06,
      "loss": 1.1686,
      "step": 7190
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.3428490161895752,
      "learning_rate": 2.7690881705480248e-06,
      "loss": 1.1589,
      "step": 7200
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.3504585027694702,
      "learning_rate": 2.7575020275750203e-06,
      "loss": 1.1815,
      "step": 7210
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.4558221101760864,
      "learning_rate": 2.7459158846020162e-06,
      "loss": 1.1641,
      "step": 7220
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.3460663557052612,
      "learning_rate": 2.734329741629012e-06,
      "loss": 1.137,
      "step": 7230
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.5156904458999634,
      "learning_rate": 2.7227435986560073e-06,
      "loss": 1.1416,
      "step": 7240
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.4338836669921875,
      "learning_rate": 2.7111574556830032e-06,
      "loss": 1.1661,
      "step": 7250
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.2937146425247192,
      "learning_rate": 2.699571312709999e-06,
      "loss": 1.1619,
      "step": 7260
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.3950343132019043,
      "learning_rate": 2.6879851697369947e-06,
      "loss": 1.1502,
      "step": 7270
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.2576100826263428,
      "learning_rate": 2.6763990267639907e-06,
      "loss": 1.1817,
      "step": 7280
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.3002537488937378,
      "learning_rate": 2.6648128837909866e-06,
      "loss": 1.148,
      "step": 7290
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.474173903465271,
      "learning_rate": 2.6532267408179817e-06,
      "loss": 1.16,
      "step": 7300
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.361527681350708,
      "learning_rate": 2.6416405978449777e-06,
      "loss": 1.1629,
      "step": 7310
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.3156787157058716,
      "learning_rate": 2.630054454871973e-06,
      "loss": 1.1701,
      "step": 7320
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.49290132522583,
      "learning_rate": 2.618468311898969e-06,
      "loss": 1.164,
      "step": 7330
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.3317513465881348,
      "learning_rate": 2.606882168925965e-06,
      "loss": 1.1531,
      "step": 7340
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.3225781917572021,
      "learning_rate": 2.59529602595296e-06,
      "loss": 1.1513,
      "step": 7350
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.4606074094772339,
      "learning_rate": 2.583709882979956e-06,
      "loss": 1.1596,
      "step": 7360
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.5031137466430664,
      "learning_rate": 2.572123740006952e-06,
      "loss": 1.1519,
      "step": 7370
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.413018822669983,
      "learning_rate": 2.5605375970339476e-06,
      "loss": 1.1715,
      "step": 7380
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.440049409866333,
      "learning_rate": 2.5489514540609435e-06,
      "loss": 1.1665,
      "step": 7390
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.4148393869400024,
      "learning_rate": 2.537365311087939e-06,
      "loss": 1.1515,
      "step": 7400
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.4196386337280273,
      "learning_rate": 2.5257791681149346e-06,
      "loss": 1.1542,
      "step": 7410
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.4080944061279297,
      "learning_rate": 2.5141930251419305e-06,
      "loss": 1.1662,
      "step": 7420
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.366580843925476,
      "learning_rate": 2.5026068821689265e-06,
      "loss": 1.1656,
      "step": 7430
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.329052209854126,
      "learning_rate": 2.491020739195922e-06,
      "loss": 1.1562,
      "step": 7440
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.4457416534423828,
      "learning_rate": 2.4794345962229175e-06,
      "loss": 1.1757,
      "step": 7450
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.3773384094238281,
      "learning_rate": 2.4678484532499135e-06,
      "loss": 1.1724,
      "step": 7460
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.4378938674926758,
      "learning_rate": 2.456262310276909e-06,
      "loss": 1.1599,
      "step": 7470
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.4069546461105347,
      "learning_rate": 2.4446761673039045e-06,
      "loss": 1.1784,
      "step": 7480
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.4680144786834717,
      "learning_rate": 2.4330900243309005e-06,
      "loss": 1.1548,
      "step": 7490
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.4188005924224854,
      "learning_rate": 2.421503881357896e-06,
      "loss": 1.1599,
      "step": 7500
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.3987394571304321,
      "learning_rate": 2.409917738384892e-06,
      "loss": 1.1674,
      "step": 7510
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.3398780822753906,
      "learning_rate": 2.398331595411888e-06,
      "loss": 1.1457,
      "step": 7520
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.3323601484298706,
      "learning_rate": 2.3867454524388834e-06,
      "loss": 1.1583,
      "step": 7530
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.4326480627059937,
      "learning_rate": 2.375159309465879e-06,
      "loss": 1.1746,
      "step": 7540
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.4204566478729248,
      "learning_rate": 2.3635731664928745e-06,
      "loss": 1.1401,
      "step": 7550
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.554408073425293,
      "learning_rate": 2.3519870235198704e-06,
      "loss": 1.1531,
      "step": 7560
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.4092999696731567,
      "learning_rate": 2.340400880546866e-06,
      "loss": 1.1601,
      "step": 7570
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.502114176750183,
      "learning_rate": 2.328814737573862e-06,
      "loss": 1.1625,
      "step": 7580
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.319533348083496,
      "learning_rate": 2.317228594600858e-06,
      "loss": 1.1411,
      "step": 7590
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.3012977838516235,
      "learning_rate": 2.3056424516278534e-06,
      "loss": 1.1716,
      "step": 7600
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.3816500902175903,
      "learning_rate": 2.294056308654849e-06,
      "loss": 1.1468,
      "step": 7610
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.4806541204452515,
      "learning_rate": 2.2824701656818444e-06,
      "loss": 1.139,
      "step": 7620
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.4676132202148438,
      "learning_rate": 2.2708840227088404e-06,
      "loss": 1.1717,
      "step": 7630
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.4198739528656006,
      "learning_rate": 2.2592978797358363e-06,
      "loss": 1.1671,
      "step": 7640
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5252964496612549,
      "learning_rate": 2.247711736762832e-06,
      "loss": 1.149,
      "step": 7650
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.4139654636383057,
      "learning_rate": 2.2361255937898278e-06,
      "loss": 1.1606,
      "step": 7660
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.3876101970672607,
      "learning_rate": 2.2245394508168233e-06,
      "loss": 1.1421,
      "step": 7670
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.2850245237350464,
      "learning_rate": 2.212953307843819e-06,
      "loss": 1.1725,
      "step": 7680
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.2772204875946045,
      "learning_rate": 2.2013671648708148e-06,
      "loss": 1.1612,
      "step": 7690
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.4063725471496582,
      "learning_rate": 2.1897810218978103e-06,
      "loss": 1.18,
      "step": 7700
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.239322543144226,
      "learning_rate": 2.1781948789248063e-06,
      "loss": 1.1477,
      "step": 7710
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.381710171699524,
      "learning_rate": 2.1666087359518018e-06,
      "loss": 1.1786,
      "step": 7720
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.2689425945281982,
      "learning_rate": 2.1550225929787973e-06,
      "loss": 1.1655,
      "step": 7730
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.4614635705947876,
      "learning_rate": 2.1434364500057933e-06,
      "loss": 1.1742,
      "step": 7740
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.383084774017334,
      "learning_rate": 2.1318503070327888e-06,
      "loss": 1.1628,
      "step": 7750
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.4612348079681396,
      "learning_rate": 2.1202641640597847e-06,
      "loss": 1.1829,
      "step": 7760
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.4572889804840088,
      "learning_rate": 2.1086780210867807e-06,
      "loss": 1.1262,
      "step": 7770
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.3573559522628784,
      "learning_rate": 2.097091878113776e-06,
      "loss": 1.1657,
      "step": 7780
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.3430317640304565,
      "learning_rate": 2.0855057351407717e-06,
      "loss": 1.1563,
      "step": 7790
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.3933885097503662,
      "learning_rate": 2.0739195921677672e-06,
      "loss": 1.1655,
      "step": 7800
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.3937543630599976,
      "learning_rate": 2.062333449194763e-06,
      "loss": 1.1554,
      "step": 7810
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.422379493713379,
      "learning_rate": 2.050747306221759e-06,
      "loss": 1.1785,
      "step": 7820
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.4572478532791138,
      "learning_rate": 2.0391611632487547e-06,
      "loss": 1.1618,
      "step": 7830
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.5157662630081177,
      "learning_rate": 2.0275750202757506e-06,
      "loss": 1.1577,
      "step": 7840
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.4413377046585083,
      "learning_rate": 2.015988877302746e-06,
      "loss": 1.171,
      "step": 7850
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.472377061843872,
      "learning_rate": 2.0044027343297417e-06,
      "loss": 1.1562,
      "step": 7860
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.3218101263046265,
      "learning_rate": 1.992816591356737e-06,
      "loss": 1.1528,
      "step": 7870
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.460989236831665,
      "learning_rate": 1.981230448383733e-06,
      "loss": 1.1568,
      "step": 7880
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.3286033868789673,
      "learning_rate": 1.969644305410729e-06,
      "loss": 1.1467,
      "step": 7890
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.386688470840454,
      "learning_rate": 1.9580581624377246e-06,
      "loss": 1.162,
      "step": 7900
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.3312408924102783,
      "learning_rate": 1.9464720194647206e-06,
      "loss": 1.1483,
      "step": 7910
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.3570996522903442,
      "learning_rate": 1.934885876491716e-06,
      "loss": 1.1552,
      "step": 7920
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.502630352973938,
      "learning_rate": 1.9232997335187116e-06,
      "loss": 1.1627,
      "step": 7930
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.2717859745025635,
      "learning_rate": 1.9117135905457076e-06,
      "loss": 1.1518,
      "step": 7940
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.4142167568206787,
      "learning_rate": 1.9001274475727033e-06,
      "loss": 1.1686,
      "step": 7950
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.3600444793701172,
      "learning_rate": 1.888541304599699e-06,
      "loss": 1.1585,
      "step": 7960
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.3272130489349365,
      "learning_rate": 1.8769551616266945e-06,
      "loss": 1.1369,
      "step": 7970
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.4064016342163086,
      "learning_rate": 1.8653690186536905e-06,
      "loss": 1.1551,
      "step": 7980
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.4075134992599487,
      "learning_rate": 1.853782875680686e-06,
      "loss": 1.1771,
      "step": 7990
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.4398192167282104,
      "learning_rate": 1.8421967327076818e-06,
      "loss": 1.1643,
      "step": 8000
    },
    {
      "epoch": 1.67,
      "eval_loss": 1.1820555925369263,
      "eval_runtime": 60.2709,
      "eval_samples_per_second": 102.869,
      "eval_steps_per_second": 12.859,
      "step": 8000
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.490997076034546,
      "learning_rate": 1.8306105897346773e-06,
      "loss": 1.1809,
      "step": 8010
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.3520011901855469,
      "learning_rate": 1.8190244467616732e-06,
      "loss": 1.1466,
      "step": 8020
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.3630290031433105,
      "learning_rate": 1.807438303788669e-06,
      "loss": 1.1692,
      "step": 8030
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.489708423614502,
      "learning_rate": 1.7958521608156645e-06,
      "loss": 1.1581,
      "step": 8040
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.386827826499939,
      "learning_rate": 1.7842660178426604e-06,
      "loss": 1.1536,
      "step": 8050
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.3416543006896973,
      "learning_rate": 1.7726798748696562e-06,
      "loss": 1.1627,
      "step": 8060
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.4776194095611572,
      "learning_rate": 1.7610937318966517e-06,
      "loss": 1.1496,
      "step": 8070
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.3277490139007568,
      "learning_rate": 1.7495075889236474e-06,
      "loss": 1.1672,
      "step": 8080
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.3496487140655518,
      "learning_rate": 1.7379214459506432e-06,
      "loss": 1.1578,
      "step": 8090
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.4114041328430176,
      "learning_rate": 1.726335302977639e-06,
      "loss": 1.1716,
      "step": 8100
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.4332119226455688,
      "learning_rate": 1.7147491600046346e-06,
      "loss": 1.1527,
      "step": 8110
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.508005976676941,
      "learning_rate": 1.7031630170316302e-06,
      "loss": 1.1629,
      "step": 8120
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.3349090814590454,
      "learning_rate": 1.6915768740586261e-06,
      "loss": 1.1536,
      "step": 8130
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.4674837589263916,
      "learning_rate": 1.6799907310856216e-06,
      "loss": 1.1704,
      "step": 8140
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.3746360540390015,
      "learning_rate": 1.6684045881126174e-06,
      "loss": 1.1495,
      "step": 8150
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.370354175567627,
      "learning_rate": 1.6568184451396133e-06,
      "loss": 1.1597,
      "step": 8160
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.3676754236221313,
      "learning_rate": 1.6452323021666088e-06,
      "loss": 1.1516,
      "step": 8170
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.399137020111084,
      "learning_rate": 1.6336461591936046e-06,
      "loss": 1.175,
      "step": 8180
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.4263873100280762,
      "learning_rate": 1.6220600162206001e-06,
      "loss": 1.168,
      "step": 8190
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.2758104801177979,
      "learning_rate": 1.610473873247596e-06,
      "loss": 1.1683,
      "step": 8200
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.4730124473571777,
      "learning_rate": 1.5988877302745918e-06,
      "loss": 1.179,
      "step": 8210
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.3805128335952759,
      "learning_rate": 1.5873015873015873e-06,
      "loss": 1.1105,
      "step": 8220
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.3347926139831543,
      "learning_rate": 1.5757154443285833e-06,
      "loss": 1.1447,
      "step": 8230
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.3892558813095093,
      "learning_rate": 1.5641293013555788e-06,
      "loss": 1.1397,
      "step": 8240
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.410576581954956,
      "learning_rate": 1.5525431583825745e-06,
      "loss": 1.159,
      "step": 8250
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.4224263429641724,
      "learning_rate": 1.5409570154095703e-06,
      "loss": 1.1605,
      "step": 8260
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.3661879301071167,
      "learning_rate": 1.529370872436566e-06,
      "loss": 1.1622,
      "step": 8270
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.5230523347854614,
      "learning_rate": 1.5177847294635617e-06,
      "loss": 1.1594,
      "step": 8280
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.5382969379425049,
      "learning_rate": 1.5061985864905573e-06,
      "loss": 1.1793,
      "step": 8290
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.2719836235046387,
      "learning_rate": 1.4946124435175532e-06,
      "loss": 1.1346,
      "step": 8300
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.3252252340316772,
      "learning_rate": 1.483026300544549e-06,
      "loss": 1.1519,
      "step": 8310
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.2065455913543701,
      "learning_rate": 1.4714401575715445e-06,
      "loss": 1.1581,
      "step": 8320
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.382408857345581,
      "learning_rate": 1.4598540145985402e-06,
      "loss": 1.1335,
      "step": 8330
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.4275171756744385,
      "learning_rate": 1.4482678716255362e-06,
      "loss": 1.1686,
      "step": 8340
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.332566261291504,
      "learning_rate": 1.4366817286525317e-06,
      "loss": 1.1786,
      "step": 8350
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.3520534038543701,
      "learning_rate": 1.4250955856795274e-06,
      "loss": 1.1677,
      "step": 8360
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.4580789804458618,
      "learning_rate": 1.4135094427065231e-06,
      "loss": 1.1404,
      "step": 8370
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.307956576347351,
      "learning_rate": 1.4019232997335189e-06,
      "loss": 1.1605,
      "step": 8380
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.3615368604660034,
      "learning_rate": 1.3903371567605144e-06,
      "loss": 1.1574,
      "step": 8390
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.4463071823120117,
      "learning_rate": 1.3787510137875101e-06,
      "loss": 1.1534,
      "step": 8400
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.432977557182312,
      "learning_rate": 1.367164870814506e-06,
      "loss": 1.1603,
      "step": 8410
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.3181747198104858,
      "learning_rate": 1.3555787278415016e-06,
      "loss": 1.1707,
      "step": 8420
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.481392502784729,
      "learning_rate": 1.3439925848684974e-06,
      "loss": 1.1743,
      "step": 8430
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.4732555150985718,
      "learning_rate": 1.3324064418954933e-06,
      "loss": 1.1741,
      "step": 8440
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.4790183305740356,
      "learning_rate": 1.3208202989224888e-06,
      "loss": 1.1821,
      "step": 8450
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.2341439723968506,
      "learning_rate": 1.3092341559494846e-06,
      "loss": 1.1661,
      "step": 8460
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.4015235900878906,
      "learning_rate": 1.29764801297648e-06,
      "loss": 1.1704,
      "step": 8470
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.2874525785446167,
      "learning_rate": 1.286061870003476e-06,
      "loss": 1.1584,
      "step": 8480
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.497933030128479,
      "learning_rate": 1.2744757270304718e-06,
      "loss": 1.1648,
      "step": 8490
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.3120431900024414,
      "learning_rate": 1.2628895840574673e-06,
      "loss": 1.187,
      "step": 8500
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.327337622642517,
      "learning_rate": 1.2513034410844632e-06,
      "loss": 1.1738,
      "step": 8510
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.224055528640747,
      "learning_rate": 1.2397172981114588e-06,
      "loss": 1.1518,
      "step": 8520
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.5562435388565063,
      "learning_rate": 1.2281311551384545e-06,
      "loss": 1.1634,
      "step": 8530
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.4716066122055054,
      "learning_rate": 1.2165450121654502e-06,
      "loss": 1.162,
      "step": 8540
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.395019769668579,
      "learning_rate": 1.204958869192446e-06,
      "loss": 1.1696,
      "step": 8550
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.557733416557312,
      "learning_rate": 1.1933727262194417e-06,
      "loss": 1.1699,
      "step": 8560
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.4722071886062622,
      "learning_rate": 1.1817865832464372e-06,
      "loss": 1.1678,
      "step": 8570
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.4380254745483398,
      "learning_rate": 1.170200440273433e-06,
      "loss": 1.1694,
      "step": 8580
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.3241981267929077,
      "learning_rate": 1.158614297300429e-06,
      "loss": 1.1539,
      "step": 8590
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.4241721630096436,
      "learning_rate": 1.1470281543274244e-06,
      "loss": 1.1493,
      "step": 8600
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.2441765069961548,
      "learning_rate": 1.1354420113544202e-06,
      "loss": 1.143,
      "step": 8610
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.3431496620178223,
      "learning_rate": 1.123855868381416e-06,
      "loss": 1.1343,
      "step": 8620
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.4855791330337524,
      "learning_rate": 1.1122697254084117e-06,
      "loss": 1.1424,
      "step": 8630
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.3640236854553223,
      "learning_rate": 1.1006835824354074e-06,
      "loss": 1.1751,
      "step": 8640
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.414133906364441,
      "learning_rate": 1.0890974394624031e-06,
      "loss": 1.1595,
      "step": 8650
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.3939659595489502,
      "learning_rate": 1.0775112964893987e-06,
      "loss": 1.1735,
      "step": 8660
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.5434401035308838,
      "learning_rate": 1.0659251535163944e-06,
      "loss": 1.1629,
      "step": 8670
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.3433867692947388,
      "learning_rate": 1.0543390105433903e-06,
      "loss": 1.1606,
      "step": 8680
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.4562678337097168,
      "learning_rate": 1.0427528675703859e-06,
      "loss": 1.168,
      "step": 8690
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.3237522840499878,
      "learning_rate": 1.0311667245973816e-06,
      "loss": 1.1599,
      "step": 8700
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.326677680015564,
      "learning_rate": 1.0195805816243773e-06,
      "loss": 1.1587,
      "step": 8710
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.3268414735794067,
      "learning_rate": 1.007994438651373e-06,
      "loss": 1.154,
      "step": 8720
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.4188928604125977,
      "learning_rate": 9.964082956783686e-07,
      "loss": 1.1608,
      "step": 8730
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.4449864625930786,
      "learning_rate": 9.848221527053645e-07,
      "loss": 1.164,
      "step": 8740
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.3237680196762085,
      "learning_rate": 9.732360097323603e-07,
      "loss": 1.1547,
      "step": 8750
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.4183787107467651,
      "learning_rate": 9.616498667593558e-07,
      "loss": 1.1632,
      "step": 8760
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.3433119058609009,
      "learning_rate": 9.500637237863516e-07,
      "loss": 1.1542,
      "step": 8770
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.5118272304534912,
      "learning_rate": 9.384775808133473e-07,
      "loss": 1.1631,
      "step": 8780
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.470158338546753,
      "learning_rate": 9.26891437840343e-07,
      "loss": 1.1638,
      "step": 8790
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.408806324005127,
      "learning_rate": 9.153052948673386e-07,
      "loss": 1.1676,
      "step": 8800
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.4310715198516846,
      "learning_rate": 9.037191518943345e-07,
      "loss": 1.1606,
      "step": 8810
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.4561901092529297,
      "learning_rate": 8.921330089213302e-07,
      "loss": 1.1639,
      "step": 8820
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.3165732622146606,
      "learning_rate": 8.805468659483258e-07,
      "loss": 1.1619,
      "step": 8830
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.4040567874908447,
      "learning_rate": 8.689607229753216e-07,
      "loss": 1.1413,
      "step": 8840
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.3458449840545654,
      "learning_rate": 8.573745800023173e-07,
      "loss": 1.1518,
      "step": 8850
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.441347599029541,
      "learning_rate": 8.457884370293131e-07,
      "loss": 1.136,
      "step": 8860
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.4389218091964722,
      "learning_rate": 8.342022940563087e-07,
      "loss": 1.1782,
      "step": 8870
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.6026980876922607,
      "learning_rate": 8.226161510833044e-07,
      "loss": 1.1606,
      "step": 8880
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.5719023942947388,
      "learning_rate": 8.110300081103001e-07,
      "loss": 1.1482,
      "step": 8890
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.302774429321289,
      "learning_rate": 7.994438651372959e-07,
      "loss": 1.1545,
      "step": 8900
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.3073673248291016,
      "learning_rate": 7.878577221642916e-07,
      "loss": 1.1514,
      "step": 8910
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.4172277450561523,
      "learning_rate": 7.762715791912873e-07,
      "loss": 1.156,
      "step": 8920
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.4633533954620361,
      "learning_rate": 7.64685436218283e-07,
      "loss": 1.1578,
      "step": 8930
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.3664774894714355,
      "learning_rate": 7.530992932452786e-07,
      "loss": 1.1529,
      "step": 8940
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4149627685546875,
      "learning_rate": 7.415131502722745e-07,
      "loss": 1.1512,
      "step": 8950
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4857783317565918,
      "learning_rate": 7.299270072992701e-07,
      "loss": 1.1465,
      "step": 8960
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4186878204345703,
      "learning_rate": 7.183408643262658e-07,
      "loss": 1.1556,
      "step": 8970
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4118033647537231,
      "learning_rate": 7.067547213532616e-07,
      "loss": 1.1484,
      "step": 8980
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.396365761756897,
      "learning_rate": 6.951685783802572e-07,
      "loss": 1.1646,
      "step": 8990
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.5443834066390991,
      "learning_rate": 6.83582435407253e-07,
      "loss": 1.1621,
      "step": 9000
    },
    {
      "epoch": 1.88,
      "eval_loss": 1.178328037261963,
      "eval_runtime": 60.0885,
      "eval_samples_per_second": 103.181,
      "eval_steps_per_second": 12.898,
      "step": 9000
    }
  ],
  "logging_steps": 10,
  "max_steps": 9590,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 1,
  "total_flos": 7.483761454915518e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}