diff --git "a/checkpoint-15000/trainer_state.json" "b/checkpoint-15000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-15000/trainer_state.json"
@@ -0,0 +1,10534 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.968804495252858,
+  "eval_steps": 500,
+  "global_step": 15000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006458696635019053,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9988374346056966e-05,
+      "loss": 1.2716,
+      "step": 10
+    },
+    {
+      "epoch": 0.0012917393270038106,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.997545695278693e-05,
+      "loss": 0.5754,
+      "step": 20
+    },
+    {
+      "epoch": 0.001937608990505716,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.996253955951689e-05,
+      "loss": 0.5303,
+      "step": 30
+    },
+    {
+      "epoch": 0.0025834786540076212,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9949622166246855e-05,
+      "loss": 0.4799,
+      "step": 40
+    },
+    {
+      "epoch": 0.0032293483175095264,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9936704772976815e-05,
+      "loss": 0.5005,
+      "step": 50
+    },
+    {
+      "epoch": 0.003875217981011432,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9923787379706776e-05,
+      "loss": 0.5047,
+      "step": 60
+    },
+    {
+      "epoch": 0.004521087644513337,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.991086998643674e-05,
+      "loss": 0.5359,
+      "step": 70
+    },
+    {
+      "epoch": 0.0051669573080152425,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.98979525931667e-05,
+      "loss": 0.5002,
+      "step": 80
+    },
+    {
+      "epoch": 0.005812826971517148,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.9885035199896665e-05,
+      "loss": 0.4836,
+      "step": 90
+    },
+    {
+      "epoch": 0.006458696635019053,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9872117806626625e-05,
+      "loss": 0.4985,
+      "step": 100
+    },
+    {
+      "epoch": 0.0071045662985209585,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.9859200413356586e-05,
+      "loss": 0.4608,
+      "step": 110
+    },
+    {
+      "epoch": 0.007750435962022864,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.9846283020086546e-05,
+      "loss": 0.4479,
+      "step": 120
+    },
+    {
+      "epoch": 0.008396305625524769,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.983336562681651e-05,
+      "loss": 0.4692,
+      "step": 130
+    },
+    {
+      "epoch": 0.009042175289026674,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.982044823354647e-05,
+      "loss": 0.4382,
+      "step": 140
+    },
+    {
+      "epoch": 0.00968804495252858,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9807530840276435e-05,
+      "loss": 0.4905,
+      "step": 150
+    },
+    {
+      "epoch": 0.010333914616030485,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9794613447006395e-05,
+      "loss": 0.4983,
+      "step": 160
+    },
+    {
+      "epoch": 0.01097978427953239,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9781696053736356e-05,
+      "loss": 0.4793,
+      "step": 170
+    },
+    {
+      "epoch": 0.011625653943034296,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.976877866046632e-05,
+      "loss": 0.4776,
+      "step": 180
+    },
+    {
+      "epoch": 0.012271523606536201,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.975586126719628e-05,
+      "loss": 0.4828,
+      "step": 190
+    },
+    {
+      "epoch": 0.012917393270038106,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9742943873926245e-05,
+      "loss": 0.4848,
+      "step": 200
+    },
+    {
+      "epoch": 0.013563262933540012,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9730026480656205e-05,
+      "loss": 0.456,
+      "step": 210
+    },
+    {
+      "epoch": 0.014209132597041917,
+      "grad_norm": 1.25,
+      "learning_rate": 1.971710908738617e-05,
+      "loss": 0.4839,
+      "step": 220
+    },
+    {
+      "epoch": 0.014855002260543822,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.970419169411613e-05,
+      "loss": 0.5081,
+      "step": 230
+    },
+    {
+      "epoch": 0.015500871924045728,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.969127430084609e-05,
+      "loss": 0.489,
+      "step": 240
+    },
+    {
+      "epoch": 0.01614674158754763,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9678356907576054e-05,
+      "loss": 0.462,
+      "step": 250
+    },
+    {
+      "epoch": 0.016792611251049538,
+      "grad_norm": 1.0,
+      "learning_rate": 1.9665439514306015e-05,
+      "loss": 0.4653,
+      "step": 260
+    },
+    {
+      "epoch": 0.017438480914551444,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.965252212103598e-05,
+      "loss": 0.435,
+      "step": 270
+    },
+    {
+      "epoch": 0.018084350578053347,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9639604727765936e-05,
+      "loss": 0.4497,
+      "step": 280
+    },
+    {
+      "epoch": 0.018730220241555254,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.96266873344959e-05,
+      "loss": 0.4648,
+      "step": 290
+    },
+    {
+      "epoch": 0.01937608990505716,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.961376994122586e-05,
+      "loss": 0.4975,
+      "step": 300
+    },
+    {
+      "epoch": 0.020021959568559063,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9600852547955825e-05,
+      "loss": 0.4695,
+      "step": 310
+    },
+    {
+      "epoch": 0.02066782923206097,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9587935154685785e-05,
+      "loss": 0.4662,
+      "step": 320
+    },
+    {
+      "epoch": 0.021313698895562876,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.957501776141575e-05,
+      "loss": 0.4465,
+      "step": 330
+    },
+    {
+      "epoch": 0.02195956855906478,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.956210036814571e-05,
+      "loss": 0.4287,
+      "step": 340
+    },
+    {
+      "epoch": 0.022605438222566686,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.954918297487567e-05,
+      "loss": 0.4362,
+      "step": 350
+    },
+    {
+      "epoch": 0.023251307886068592,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9536265581605635e-05,
+      "loss": 0.4555,
+      "step": 360
+    },
+    {
+      "epoch": 0.023897177549570495,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9523348188335595e-05,
+      "loss": 0.4631,
+      "step": 370
+    },
+    {
+      "epoch": 0.024543047213072402,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.951043079506556e-05,
+      "loss": 0.4535,
+      "step": 380
+    },
+    {
+      "epoch": 0.02518891687657431,
+      "grad_norm": 1.125,
+      "learning_rate": 1.949751340179552e-05,
+      "loss": 0.4533,
+      "step": 390
+    },
+    {
+      "epoch": 0.02583478654007621,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.948459600852548e-05,
+      "loss": 0.4782,
+      "step": 400
+    },
+    {
+      "epoch": 0.026480656203578118,
+      "grad_norm": 1.125,
+      "learning_rate": 1.9471678615255444e-05,
+      "loss": 0.47,
+      "step": 410
+    },
+    {
+      "epoch": 0.027126525867080024,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9458761221985405e-05,
+      "loss": 0.4593,
+      "step": 420
+    },
+    {
+      "epoch": 0.027772395530581927,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.944584382871537e-05,
+      "loss": 0.4724,
+      "step": 430
+    },
+    {
+      "epoch": 0.028418265194083834,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.943292643544533e-05,
+      "loss": 0.426,
+      "step": 440
+    },
+    {
+      "epoch": 0.02906413485758574,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.942000904217529e-05,
+      "loss": 0.3985,
+      "step": 450
+    },
+    {
+      "epoch": 0.029710004521087643,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.940709164890525e-05,
+      "loss": 0.4164,
+      "step": 460
+    },
+    {
+      "epoch": 0.03035587418458955,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.9394174255635215e-05,
+      "loss": 0.4387,
+      "step": 470
+    },
+    {
+      "epoch": 0.031001743848091456,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.9381256862365175e-05,
+      "loss": 0.4655,
+      "step": 480
+    },
+    {
+      "epoch": 0.03164761351159336,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.936833946909514e-05,
+      "loss": 0.4184,
+      "step": 490
+    },
+    {
+      "epoch": 0.03229348317509526,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.93554220758251e-05,
+      "loss": 0.4113,
+      "step": 500
+    },
+    {
+      "epoch": 0.03293935283859717,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.934250468255506e-05,
+      "loss": 0.4463,
+      "step": 510
+    },
+    {
+      "epoch": 0.033585222502099076,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9329587289285024e-05,
+      "loss": 0.455,
+      "step": 520
+    },
+    {
+      "epoch": 0.03423109216560098,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9316669896014985e-05,
+      "loss": 0.4951,
+      "step": 530
+    },
+    {
+      "epoch": 0.03487696182910289,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.930375250274495e-05,
+      "loss": 0.4477,
+      "step": 540
+    },
+    {
+      "epoch": 0.035522831492604795,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.929083510947491e-05,
+      "loss": 0.4095,
+      "step": 550
+    },
+    {
+      "epoch": 0.036168701156106695,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.9277917716204874e-05,
+      "loss": 0.4777,
+      "step": 560
+    },
+    {
+      "epoch": 0.0368145708196086,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9265000322934834e-05,
+      "loss": 0.4488,
+      "step": 570
+    },
+    {
+      "epoch": 0.03746044048311051,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9252082929664795e-05,
+      "loss": 0.4323,
+      "step": 580
+    },
+    {
+      "epoch": 0.038106310146612414,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.923916553639476e-05,
+      "loss": 0.4281,
+      "step": 590
+    },
+    {
+      "epoch": 0.03875217981011432,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.922624814312472e-05,
+      "loss": 0.4037,
+      "step": 600
+    },
+    {
+      "epoch": 0.03939804947361623,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9213330749854683e-05,
+      "loss": 0.4093,
+      "step": 610
+    },
+    {
+      "epoch": 0.04004391913711813,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.920041335658464e-05,
+      "loss": 0.3905,
+      "step": 620
+    },
+    {
+      "epoch": 0.04068978880062003,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9187495963314605e-05,
+      "loss": 0.3851,
+      "step": 630
+    },
+    {
+      "epoch": 0.04133565846412194,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9174578570044565e-05,
+      "loss": 0.436,
+      "step": 640
+    },
+    {
+      "epoch": 0.041981528127623846,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.916166117677453e-05,
+      "loss": 0.446,
+      "step": 650
+    },
+    {
+      "epoch": 0.04262739779112575,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.914874378350449e-05,
+      "loss": 0.4071,
+      "step": 660
+    },
+    {
+      "epoch": 0.04327326745462766,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9135826390234454e-05,
+      "loss": 0.4127,
+      "step": 670
+    },
+    {
+      "epoch": 0.04391913711812956,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9122908996964414e-05,
+      "loss": 0.4071,
+      "step": 680
+    },
+    {
+      "epoch": 0.044565006781631465,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9109991603694375e-05,
+      "loss": 0.4482,
+      "step": 690
+    },
+    {
+      "epoch": 0.04521087644513337,
+      "grad_norm": 1.125,
+      "learning_rate": 1.909707421042434e-05,
+      "loss": 0.4328,
+      "step": 700
+    },
+    {
+      "epoch": 0.04585674610863528,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.90841568171543e-05,
+      "loss": 0.3965,
+      "step": 710
+    },
+    {
+      "epoch": 0.046502615772137185,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.9071239423884263e-05,
+      "loss": 0.4269,
+      "step": 720
+    },
+    {
+      "epoch": 0.04714848543563909,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9058322030614224e-05,
+      "loss": 0.4305,
+      "step": 730
+    },
+    {
+      "epoch": 0.04779435509914099,
+      "grad_norm": 1.125,
+      "learning_rate": 1.9045404637344185e-05,
+      "loss": 0.4325,
+      "step": 740
+    },
+    {
+      "epoch": 0.0484402247626429,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.903248724407415e-05,
+      "loss": 0.4243,
+      "step": 750
+    },
+    {
+      "epoch": 0.049086094426144804,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.901956985080411e-05,
+      "loss": 0.4072,
+      "step": 760
+    },
+    {
+      "epoch": 0.04973196408964671,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9006652457534073e-05,
+      "loss": 0.3744,
+      "step": 770
+    },
+    {
+      "epoch": 0.05037783375314862,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.8993735064264034e-05,
+      "loss": 0.4234,
+      "step": 780
+    },
+    {
+      "epoch": 0.05102370341665052,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.8980817670993994e-05,
+      "loss": 0.4271,
+      "step": 790
+    },
+    {
+      "epoch": 0.05166957308015242,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8967900277723955e-05,
+      "loss": 0.4418,
+      "step": 800
+    },
+    {
+      "epoch": 0.05231544274365433,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.895498288445392e-05,
+      "loss": 0.3877,
+      "step": 810
+    },
+    {
+      "epoch": 0.052961312407156236,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.894206549118388e-05,
+      "loss": 0.3883,
+      "step": 820
+    },
+    {
+      "epoch": 0.05360718207065814,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8929148097913844e-05,
+      "loss": 0.3925,
+      "step": 830
+    },
+    {
+      "epoch": 0.05425305173416005,
+      "grad_norm": 1.0,
+      "learning_rate": 1.8916230704643804e-05,
+      "loss": 0.4525,
+      "step": 840
+    },
+    {
+      "epoch": 0.054898921397661955,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8903313311373765e-05,
+      "loss": 0.422,
+      "step": 850
+    },
+    {
+      "epoch": 0.055544791061163855,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.889039591810373e-05,
+      "loss": 0.4227,
+      "step": 860
+    },
+    {
+      "epoch": 0.05619066072466576,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.887747852483369e-05,
+      "loss": 0.4237,
+      "step": 870
+    },
+    {
+      "epoch": 0.05683653038816767,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8864561131563653e-05,
+      "loss": 0.4033,
+      "step": 880
+    },
+    {
+      "epoch": 0.057482400051669574,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8851643738293614e-05,
+      "loss": 0.4373,
+      "step": 890
+    },
+    {
+      "epoch": 0.05812826971517148,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.8838726345023578e-05,
+      "loss": 0.4513,
+      "step": 900
+    },
+    {
+      "epoch": 0.05877413937867338,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.882580895175354e-05,
+      "loss": 0.3729,
+      "step": 910
+    },
+    {
+      "epoch": 0.05942000904217529,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.88128915584835e-05,
+      "loss": 0.413,
+      "step": 920
+    },
+    {
+      "epoch": 0.06006587870567719,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.8799974165213463e-05,
+      "loss": 0.3959,
+      "step": 930
+    },
+    {
+      "epoch": 0.0607117483691791,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.8787056771943424e-05,
+      "loss": 0.407,
+      "step": 940
+    },
+    {
+      "epoch": 0.061357618032681006,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8774139378673388e-05,
+      "loss": 0.4124,
+      "step": 950
+    },
+    {
+      "epoch": 0.06200348769618291,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.8761221985403345e-05,
+      "loss": 0.4375,
+      "step": 960
+    },
+    {
+      "epoch": 0.06264935735968481,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.874830459213331e-05,
+      "loss": 0.403,
+      "step": 970
+    },
+    {
+      "epoch": 0.06329522702318673,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.873538719886327e-05,
+      "loss": 0.4728,
+      "step": 980
+    },
+    {
+      "epoch": 0.06394109668668863,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.8722469805593233e-05,
+      "loss": 0.4148,
+      "step": 990
+    },
+    {
+      "epoch": 0.06458696635019053,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.8709552412323194e-05,
+      "loss": 0.4078,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06523283601369244,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.8696635019053158e-05,
+      "loss": 0.3877,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06587870567719434,
+      "grad_norm": 1.125,
+      "learning_rate": 1.868371762578312e-05,
+      "loss": 0.4138,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06652457534069625,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.867080023251308e-05,
+      "loss": 0.4336,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06717044500419815,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8657882839243043e-05,
+      "loss": 0.4201,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06781631466770006,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8644965445973004e-05,
+      "loss": 0.3994,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06846218433120196,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8632048052702968e-05,
+      "loss": 0.4029,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06910805399470386,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.861913065943293e-05,
+      "loss": 0.3882,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06975392365820578,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.860621326616289e-05,
+      "loss": 0.3975,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07039979332170768,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.8593295872892853e-05,
+      "loss": 0.4082,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07104566298520959,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.8580378479622814e-05,
+      "loss": 0.4019,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07169153264871149,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8567461086352778e-05,
+      "loss": 0.4039,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07233740231221339,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8554543693082738e-05,
+      "loss": 0.3988,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0729832719757153,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.85416262998127e-05,
+      "loss": 0.3993,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0736291416392172,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.852870890654266e-05,
+      "loss": 0.3876,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07427501130271912,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8515791513272623e-05,
+      "loss": 0.3997,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07492088096622102,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8502874120002584e-05,
+      "loss": 0.4093,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07556675062972293,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.8489956726732548e-05,
+      "loss": 0.4015,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07621262029322483,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.847703933346251e-05,
+      "loss": 0.3889,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07685848995672673,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.846412194019247e-05,
+      "loss": 0.4231,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07750435962022864,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8451204546922433e-05,
+      "loss": 0.4312,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07815022928373054,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.8438287153652394e-05,
+      "loss": 0.4293,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07879609894723245,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.8425369760382358e-05,
+      "loss": 0.4149,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07944196861073435,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.8412452367112318e-05,
+      "loss": 0.3975,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08008783827423625,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.839953497384228e-05,
+      "loss": 0.3924,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08073370793773817,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.8386617580572243e-05,
+      "loss": 0.3891,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08137957760124007,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.8373700187302203e-05,
+      "loss": 0.4134,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08202544726474198,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.8360782794032167e-05,
+      "loss": 0.404,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08267131692824388,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.8347865400762128e-05,
+      "loss": 0.3847,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08331718659174578,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.8334948007492092e-05,
+      "loss": 0.4006,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08396305625524769,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.832203061422205e-05,
+      "loss": 0.3921,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08460892591874959,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.8309113220952013e-05,
+      "loss": 0.4274,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0852547955822515,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.8296195827681974e-05,
+      "loss": 0.3612,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0859006652457534,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.8283278434411938e-05,
+      "loss": 0.4426,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08654653490925532,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.82703610411419e-05,
+      "loss": 0.424,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08719240457275722,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8257443647871862e-05,
+      "loss": 0.3938,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08783827423625912,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8244526254601823e-05,
+      "loss": 0.3766,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08848414389976103,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.8231608861331784e-05,
+      "loss": 0.4252,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08913001356326293,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.8218691468061748e-05,
+      "loss": 0.3753,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08977588322676484,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.8205774074791708e-05,
+      "loss": 0.3706,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09042175289026674,
+      "grad_norm": 1.125,
+      "learning_rate": 1.8192856681521672e-05,
+      "loss": 0.436,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09106762255376864,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8179939288251633e-05,
+      "loss": 0.3956,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09171349221727056,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.8167021894981593e-05,
+      "loss": 0.3947,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09235936188077246,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.8154104501711557e-05,
+      "loss": 0.3785,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09300523154427437,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8141187108441518e-05,
+      "loss": 0.391,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09365110120777627,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.8128269715171482e-05,
+      "loss": 0.3503,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09429697087127818,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.8115352321901442e-05,
+      "loss": 0.3873,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09494284053478008,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.8102434928631403e-05,
+      "loss": 0.4106,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09558871019828198,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.8089517535361364e-05,
+      "loss": 0.382,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0962345798617839,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.8076600142091328e-05,
+      "loss": 0.3992,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0968804495252858,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.8063682748821288e-05,
+      "loss": 0.4325,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09752631918878771,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8050765355551252e-05,
+      "loss": 0.3857,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09817218885228961,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.8037847962281213e-05,
+      "loss": 0.3752,
+      "step": 1520
+    },
+    {
+      "epoch": 0.09881805851579151,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.8024930569011173e-05,
+      "loss": 0.382,
+      "step": 1530
+    },
+    {
+      "epoch": 0.09946392817929342,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.8012013175741137e-05,
+      "loss": 0.4044,
+      "step": 1540
+    },
+    {
+      "epoch": 0.10010979784279532,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.7999095782471098e-05,
+      "loss": 0.4115,
+      "step": 1550
+    },
+    {
+      "epoch": 0.10075566750629723,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.7986178389201062e-05,
+      "loss": 0.3554,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10140153716979913,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7973260995931023e-05,
+      "loss": 0.4131,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10204740683330105,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.7960343602660983e-05,
+      "loss": 0.3967,
+      "step": 1580
+    },
+    {
+      "epoch": 0.10269327649680295,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.7947426209390947e-05,
+      "loss": 0.4074,
+      "step": 1590
+    },
+    {
+      "epoch": 0.10333914616030485,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.7934508816120908e-05,
+      "loss": 0.3977,
+      "step": 1600
+    },
+    {
+      "epoch": 0.10398501582380676,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.7921591422850872e-05,
+      "loss": 0.3739,
+      "step": 1610
+    },
+    {
+      "epoch": 0.10463088548730866,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.7908674029580832e-05,
+      "loss": 0.3866,
+      "step": 1620
+    },
+    {
+      "epoch": 0.10527675515081057,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.7895756636310796e-05,
+      "loss": 0.4252,
+      "step": 1630
+    },
+    {
+      "epoch": 0.10592262481431247,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.7882839243040754e-05,
+      "loss": 0.4244,
+      "step": 1640
+    },
+    {
+      "epoch": 0.10656849447781437,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.7869921849770718e-05,
+      "loss": 0.3704,
+      "step": 1650
+    },
+    {
+      "epoch": 0.10721436414131628,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.7857004456500678e-05,
+      "loss": 0.3986,
+      "step": 1660
+    },
+    {
+      "epoch": 0.10786023380481818,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.7844087063230642e-05,
+      "loss": 0.3588,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1085061034683201,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.7831169669960603e-05,
+      "loss": 0.3863,
+      "step": 1680
+    },
+    {
+      "epoch": 0.109151973131822,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.7818252276690567e-05,
+      "loss": 0.3947,
+      "step": 1690
+    },
+    {
+      "epoch": 0.10979784279532391,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.7805334883420527e-05,
+      "loss": 0.3786,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11044371245882581,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.7792417490150488e-05,
+      "loss": 0.401,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11108958212232771,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.7779500096880452e-05,
+      "loss": 0.3877,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11173545178582962,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.7766582703610413e-05,
+      "loss": 0.3789,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11238132144933152,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.7753665310340376e-05,
+      "loss": 0.3633,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11302719111283344,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7740747917070337e-05,
+      "loss": 0.3516,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11367306077633534,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.7727830523800298e-05,
+      "loss": 0.3654,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11431893043983724,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.771491313053026e-05,
+      "loss": 0.4036,
+      "step": 1770
+    },
+    {
+      "epoch": 0.11496480010333915,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.7701995737260222e-05,
+      "loss": 0.3714,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11561066976684105,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.7689078343990186e-05,
+      "loss": 0.3859,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11625653943034296,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.7676160950720147e-05,
+      "loss": 0.3662,
+      "step": 1800
+    },
+    {
+      "epoch": 0.11690240909384486,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.7663243557450107e-05,
+      "loss": 0.3789,
+      "step": 1810
+    },
+    {
+      "epoch": 0.11754827875734676,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.7650326164180068e-05,
+      "loss": 0.381,
+      "step": 1820
+    },
+    {
+      "epoch": 0.11819414842084867,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.7637408770910032e-05,
+      "loss": 0.3797,
+      "step": 1830
+    },
+    {
+      "epoch": 0.11884001808435057,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.7624491377639993e-05,
+      "loss": 0.3987,
+      "step": 1840
+    },
+    {
+      "epoch": 0.11948588774785249,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.7611573984369957e-05,
+      "loss": 0.3614,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12013175741135439,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.7598656591099917e-05,
+      "loss": 0.3693,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1207776270748563,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.7585739197829878e-05,
+      "loss": 0.4046,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1214234967383582,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.7572821804559842e-05,
+      "loss": 0.3848,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1220693664018601,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.7559904411289802e-05,
+      "loss": 0.3783,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12271523606536201,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.7546987018019766e-05,
+      "loss": 0.3475,
+      "step": 1900
+    },
+    {
+      "epoch": 0.12336110572886391,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.7534069624749727e-05,
+      "loss": 0.379,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12400697539236583,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.7521152231479688e-05,
+      "loss": 0.3901,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12465284505586773,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.750823483820965e-05,
+      "loss": 0.3238,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12529871471936962,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.7495317444939612e-05,
+      "loss": 0.3564,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12594458438287154,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7482400051669576e-05,
+      "loss": 0.374,
+      "step": 1950
+    },
+    {
+      "epoch": 0.12659045404637345,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7469482658399537e-05,
+      "loss": 0.3689,
+      "step": 1960
+    },
+    {
+      "epoch": 0.12723632370987534,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.74565652651295e-05,
+      "loss": 0.3852,
+      "step": 1970
+    },
+    {
+      "epoch": 0.12788219337337725,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7443647871859458e-05,
+      "loss": 0.3713,
+      "step": 1980
+    },
+    {
+      "epoch": 0.12852806303687916,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.7430730478589422e-05,
+      "loss": 0.3635,
+      "step": 1990
+    },
+    {
+      "epoch": 0.12917393270038105,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.7417813085319383e-05,
+      "loss": 0.3597,
+      "step": 2000
+    },
+    {
+      "epoch": 0.12981980236388296,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.7404895692049346e-05,
+      "loss": 0.363,
+      "step": 2010
+    },
+    {
+      "epoch": 0.13046567202738488,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.7391978298779307e-05,
+      "loss": 0.3515,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1311115416908868,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.737906090550927e-05,
+      "loss": 0.3909,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13175741135438868,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.736614351223923e-05,
+      "loss": 0.3673,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1324032810178906,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7353226118969192e-05,
+      "loss": 0.3656,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1330491506813925,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.7340308725699156e-05,
+      "loss": 0.3491,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1336950203448944,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.7327391332429117e-05,
+      "loss": 0.3741,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1343408900083963,
+      "grad_norm": 1.25,
+      "learning_rate": 1.731447393915908e-05,
+      "loss": 0.3811,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13498675967189822,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.730155654588904e-05,
+      "loss": 0.361,
+      "step": 2090
+    },
+    {
+      "epoch": 0.13563262933540013,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.7288639152619002e-05,
+      "loss": 0.3211,
+      "step": 2100
+    },
+    {
+      "epoch": 0.13627849899890201,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7275721759348966e-05,
+      "loss": 0.3374,
+      "step": 2110
+    },
+    {
+      "epoch": 0.13692436866240393,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.7262804366078927e-05,
+      "loss": 0.3585,
+      "step": 2120
+    },
+    {
+      "epoch": 0.13757023832590584,
+      "grad_norm": 1.125,
+      "learning_rate": 1.724988697280889e-05,
+      "loss": 0.3484,
+      "step": 2130
+    },
+    {
+      "epoch": 0.13821610798940773,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.723696957953885e-05,
+      "loss": 0.3619,
+      "step": 2140
+    },
+    {
+      "epoch": 0.13886197765290964,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7224052186268812e-05,
+      "loss": 0.3664,
+      "step": 2150
+    },
+    {
+      "epoch": 0.13950784731641155,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.7211134792998772e-05,
+      "loss": 0.3225,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14015371697991344,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.7198217399728736e-05,
+      "loss": 0.3962,
+      "step": 2170
+    },
+    {
+      "epoch": 0.14079958664341535,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.7185300006458697e-05,
+      "loss": 0.3453,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14144545630691727,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.717238261318866e-05,
+      "loss": 0.3801,
+      "step": 2190
+    },
+    {
+      "epoch": 0.14209132597041918,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.715946521991862e-05,
+      "loss": 0.3448,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14273719563392107,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7146547826648582e-05,
+      "loss": 0.337,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14338306529742298,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.7133630433378546e-05,
+      "loss": 0.3336,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1440289349609249,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.7120713040108507e-05,
+      "loss": 0.3712,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14467480462442678,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.710779564683847e-05,
+      "loss": 0.3873,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1453206742879287,
+      "grad_norm": 1.125,
+      "learning_rate": 1.709487825356843e-05,
+      "loss": 0.3807,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1459665439514306,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.7081960860298392e-05,
+      "loss": 0.3768,
+      "step": 2260
+    },
+    {
+      "epoch": 0.14661241361493252,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.7069043467028356e-05,
+      "loss": 0.3609,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1472582832784344,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.7056126073758316e-05,
+      "loss": 0.3956,
+      "step": 2280
+    },
+    {
+      "epoch": 0.14790415294193632,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.704320868048828e-05,
+      "loss": 0.348,
+      "step": 2290
+    },
+    {
+      "epoch": 0.14855002260543823,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.703029128721824e-05,
+      "loss": 0.3529,
+      "step": 2300
+    },
+    {
+      "epoch": 0.14919589226894012,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.7017373893948205e-05,
+      "loss": 0.374,
+      "step": 2310
+    },
+    {
+      "epoch": 0.14984176193244203,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.7004456500678162e-05,
+      "loss": 0.3567,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15048763159594394,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.6991539107408126e-05,
+      "loss": 0.3336,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15113350125944586,
+      "grad_norm": 1.25,
+      "learning_rate": 1.6978621714138087e-05,
+      "loss": 0.3622,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15177937092294774,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.696570432086805e-05,
+      "loss": 0.3463,
+      "step": 2350
+    },
+    {
+      "epoch": 0.15242524058644966,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.695278692759801e-05,
+      "loss": 0.3618,
+      "step": 2360
+    },
+    {
+      "epoch": 0.15307111024995157,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.6939869534327975e-05,
+      "loss": 0.361,
+      "step": 2370
+    },
+    {
+      "epoch": 0.15371697991345346,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.6926952141057936e-05,
+      "loss": 0.3525,
+      "step": 2380
+    },
+    {
+      "epoch": 0.15436284957695537,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.6914034747787897e-05,
+      "loss": 0.3987,
+      "step": 2390
+    },
+    {
+      "epoch": 0.15500871924045728,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.690111735451786e-05,
+      "loss": 0.3815,
+      "step": 2400
+    },
+    {
+      "epoch": 0.15565458890395917,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.688819996124782e-05,
+      "loss": 0.3661,
+      "step": 2410
+    },
+    {
+      "epoch": 0.15630045856746108,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.6875282567977785e-05,
+      "loss": 0.3854,
+      "step": 2420
+    },
+    {
+      "epoch": 0.156946328230963,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6862365174707746e-05,
+      "loss": 0.3461,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1575921978944649,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.6849447781437706e-05,
+      "loss": 0.3339,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1582380675579668,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.683653038816767e-05,
+      "loss": 0.3447,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1588839372214687,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.682361299489763e-05,
+      "loss": 0.3569,
+      "step": 2460
+    },
+    {
+      "epoch": 0.15952980688497062,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.6810695601627595e-05,
+      "loss": 0.3774,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1601756765484725,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.6797778208357556e-05,
+      "loss": 0.3591,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16082154621197442,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.6784860815087516e-05,
+      "loss": 0.353,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16146741587547633,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.6771943421817477e-05,
+      "loss": 0.3687,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16211328553897825,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.675902602854744e-05,
+      "loss": 0.3569,
+      "step": 2510
+    },
+    {
+      "epoch": 0.16275915520248013,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.67461086352774e-05,
+      "loss": 0.3871,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16340502486598205,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.6733191242007365e-05,
+      "loss": 0.3491,
+      "step": 2530
+    },
+    {
+      "epoch": 0.16405089452948396,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.6720273848737326e-05,
+      "loss": 0.3472,
+      "step": 2540
+    },
+    {
+      "epoch": 0.16469676419298585,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.6707356455467287e-05,
+      "loss": 0.3888,
+      "step": 2550
+    },
+    {
+      "epoch": 0.16534263385648776,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.669443906219725e-05,
+      "loss": 0.3699,
+      "step": 2560
+    },
+    {
+      "epoch": 0.16598850351998967,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.668152166892721e-05,
+      "loss": 0.3709,
+      "step": 2570
+    },
+    {
+      "epoch": 0.16663437318349156,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6668604275657175e-05,
+      "loss": 0.3407,
+      "step": 2580
+    },
+    {
+      "epoch": 0.16728024284699347,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.6655686882387136e-05,
+      "loss": 0.3095,
+      "step": 2590
+    },
+    {
+      "epoch": 0.16792611251049538,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.6642769489117096e-05,
+      "loss": 0.3539,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1685719821739973,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.662985209584706e-05,
+      "loss": 0.3644,
+      "step": 2610
+    },
+    {
+      "epoch": 0.16921785183749918,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.661693470257702e-05,
+      "loss": 0.338,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1698637215010011,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6604017309306985e-05,
+      "loss": 0.3815,
+      "step": 2630
+    },
+    {
+      "epoch": 0.170509591164503,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6591099916036945e-05,
+      "loss": 0.3428,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1711554608280049,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.657818252276691e-05,
+      "loss": 0.3627,
+      "step": 2650
+    },
+    {
+      "epoch": 0.1718013304915068,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6565265129496867e-05,
+      "loss": 0.339,
+      "step": 2660
+    },
+    {
+      "epoch": 0.17244720015500872,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.655234773622683e-05,
+      "loss": 0.3156,
+      "step": 2670
+    },
+    {
+      "epoch": 0.17309306981851064,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.653943034295679e-05,
+      "loss": 0.3778,
+      "step": 2680
+    },
+    {
+      "epoch": 0.17373893948201252,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.6526512949686755e-05,
+      "loss": 0.3228,
+      "step": 2690
+    },
+    {
+      "epoch": 0.17438480914551444,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.6513595556416716e-05,
+      "loss": 0.3097,
+      "step": 2700
+    },
+    {
+      "epoch": 0.17503067880901635,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.6500678163146676e-05,
+      "loss": 0.3373,
+      "step": 2710
+    },
+    {
+      "epoch": 0.17567654847251823,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.648776076987664e-05,
+      "loss": 0.2963,
+      "step": 2720
+    },
+    {
+      "epoch": 0.17632241813602015,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.64748433766066e-05,
+      "loss": 0.3356,
+      "step": 2730
+    },
+    {
+      "epoch": 0.17696828779952206,
+      "grad_norm": 1.375,
+      "learning_rate": 1.6461925983336565e-05,
+      "loss": 0.3537,
+      "step": 2740
+    },
+    {
+      "epoch": 0.17761415746302398,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.6449008590066526e-05,
+      "loss": 0.3363,
+      "step": 2750
+    },
+    {
+      "epoch": 0.17826002712652586,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.643609119679649e-05,
+      "loss": 0.3279,
+      "step": 2760
+    },
+    {
+      "epoch": 0.17890589679002777,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.642317380352645e-05,
+      "loss": 0.3453,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1795517664535297,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.3551,
+      "step": 2780
+    },
+    {
+      "epoch": 0.18019763611703157,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.6397339016986375e-05,
+      "loss": 0.3309,
+      "step": 2790
+    },
+    {
+      "epoch": 0.1808435057805335,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.6384421623716335e-05,
+      "loss": 0.3477,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1814893754440354,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.63715042304463e-05,
+      "loss": 0.3189,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18213524510753729,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.635858683717626e-05,
+      "loss": 0.3501,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1827811147710392,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.634566944390622e-05,
+      "loss": 0.376,
+      "step": 2830
+    },
+    {
+      "epoch": 0.1834269844345411,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.633275205063618e-05,
+      "loss": 0.3182,
+      "step": 2840
+    },
+    {
+      "epoch": 0.18407285409804303,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.6319834657366145e-05,
+      "loss": 0.348,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1847187237615449,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.6306917264096106e-05,
+      "loss": 0.3668,
+      "step": 2860
+    },
+    {
+      "epoch": 0.18536459342504683,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.629399987082607e-05,
+      "loss": 0.3126,
+      "step": 2870
+    },
+    {
+      "epoch": 0.18601046308854874,
+      "grad_norm": 1.375,
+      "learning_rate": 1.628108247755603e-05,
+      "loss": 0.3166,
+      "step": 2880
+    },
+    {
+      "epoch": 0.18665633275205062,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.626816508428599e-05,
+      "loss": 0.3214,
+      "step": 2890
+    },
+    {
+      "epoch": 0.18730220241555254,
+      "grad_norm": 1.0,
+      "learning_rate": 1.6255247691015955e-05,
+      "loss": 0.3319,
+      "step": 2900
+    },
+    {
+      "epoch": 0.18794807207905445,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6242330297745915e-05,
+      "loss": 0.3513,
+      "step": 2910
+    },
+    {
+      "epoch": 0.18859394174255636,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.622941290447588e-05,
+      "loss": 0.3384,
+      "step": 2920
+    },
+    {
+      "epoch": 0.18923981140605825,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.621649551120584e-05,
+      "loss": 0.3425,
+      "step": 2930
+    },
+    {
+      "epoch": 0.18988568106956016,
+      "grad_norm": 1.25,
+      "learning_rate": 1.62035781179358e-05,
+      "loss": 0.3595,
+      "step": 2940
+    },
+    {
+      "epoch": 0.19053155073306208,
+      "grad_norm": 1.25,
+      "learning_rate": 1.6190660724665765e-05,
+      "loss": 0.316,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19117742039656396,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6177743331395725e-05,
+      "loss": 0.3667,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19182329006006588,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.616482593812569e-05,
+      "loss": 0.3272,
+      "step": 2970
+    },
+    {
+      "epoch": 0.1924691597235678,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.615190854485565e-05,
+      "loss": 0.3412,
+      "step": 2980
+    },
+    {
+      "epoch": 0.19311502938706968,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.6138991151585614e-05,
+      "loss": 0.3202,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1937608990505716,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.612607375831557e-05,
+      "loss": 0.3119,
+      "step": 3000
+    },
+    {
+      "epoch": 0.1944067687140735,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.6113156365045535e-05,
+      "loss": 0.3172,
+      "step": 3010
+    },
+    {
+      "epoch": 0.19505263837757542,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.6100238971775496e-05,
+      "loss": 0.3259,
+      "step": 3020
+    },
+    {
+      "epoch": 0.1956985080410773,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.608732157850546e-05,
+      "loss": 0.3456,
+      "step": 3030
+    },
+    {
+      "epoch": 0.19634437770457921,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.607440418523542e-05,
+      "loss": 0.3434,
+      "step": 3040
+    },
+    {
+      "epoch": 0.19699024736808113,
+      "grad_norm": 1.375,
+      "learning_rate": 1.606148679196538e-05,
+      "loss": 0.3182,
+      "step": 3050
+    },
+    {
+      "epoch": 0.19763611703158301,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.6048569398695345e-05,
+      "loss": 0.3398,
+      "step": 3060
+    },
+    {
+      "epoch": 0.19828198669508493,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.6035652005425305e-05,
+      "loss": 0.3491,
+      "step": 3070
+    },
+    {
+      "epoch": 0.19892785635858684,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.602273461215527e-05,
+      "loss": 0.3299,
+      "step": 3080
+    },
+    {
+      "epoch": 0.19957372602208875,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.600981721888523e-05,
+      "loss": 0.3862,
+      "step": 3090
+    },
+    {
+      "epoch": 0.20021959568559064,
+      "grad_norm": 1.25,
+      "learning_rate": 1.5996899825615194e-05,
+      "loss": 0.3252,
+      "step": 3100
+    },
+    {
+      "epoch": 0.20086546534909255,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5983982432345154e-05,
+      "loss": 0.3303,
+      "step": 3110
+    },
+    {
+      "epoch": 0.20151133501259447,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5971065039075115e-05,
+      "loss": 0.293,
+      "step": 3120
+    },
+    {
+      "epoch": 0.20215720467609635,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.595814764580508e-05,
+      "loss": 0.3045,
+      "step": 3130
+    },
+    {
+      "epoch": 0.20280307433959827,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.594523025253504e-05,
+      "loss": 0.3352,
+      "step": 3140
+    },
+    {
+      "epoch": 0.20344894400310018,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5932312859265004e-05,
+      "loss": 0.3099,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2040948136666021,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.5919395465994964e-05,
+      "loss": 0.3435,
+      "step": 3160
+    },
+    {
+      "epoch": 0.20474068333010398,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5906478072724925e-05,
+      "loss": 0.3232,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2053865529936059,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5893560679454885e-05,
+      "loss": 0.3215,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2060324226571078,
+      "grad_norm": 1.5,
+      "learning_rate": 1.588064328618485e-05,
+      "loss": 0.2984,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2066782923206097,
+      "grad_norm": 1.25,
+      "learning_rate": 1.586772589291481e-05,
+      "loss": 0.3075,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2073241619841116,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.5854808499644774e-05,
+      "loss": 0.3388,
+      "step": 3210
+    },
+    {
+      "epoch": 0.20797003164761352,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.5841891106374735e-05,
+      "loss": 0.3372,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2086159013111154,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.5828973713104695e-05,
+      "loss": 0.3434,
+      "step": 3230
+    },
+    {
+      "epoch": 0.20926177097461732,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.581605631983466e-05,
+      "loss": 0.3172,
+      "step": 3240
+    },
+    {
+      "epoch": 0.20990764063811923,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.580313892656462e-05,
+      "loss": 0.2827,
+      "step": 3250
+    },
+    {
+      "epoch": 0.21055351030162114,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.5790221533294584e-05,
+      "loss": 0.3381,
+      "step": 3260
+    },
+    {
+      "epoch": 0.21119937996512303,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5777304140024544e-05,
+      "loss": 0.3342,
+      "step": 3270
+    },
+    {
+      "epoch": 0.21184524962862494,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.5764386746754505e-05,
+      "loss": 0.325,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21249111929212686,
+      "grad_norm": 1.25,
+      "learning_rate": 1.575146935348447e-05,
+      "loss": 0.3414,
+      "step": 3290
+    },
+    {
+      "epoch": 0.21313698895562874,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.573855196021443e-05,
+      "loss": 0.3553,
+      "step": 3300
+    },
+    {
+      "epoch": 0.21378285861913066,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5725634566944394e-05,
+      "loss": 0.3177,
+      "step": 3310
+    },
+    {
+      "epoch": 0.21442872828263257,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.5712717173674354e-05,
+      "loss": 0.3091,
+      "step": 3320
+    },
+    {
+      "epoch": 0.21507459794613448,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.5699799780404318e-05,
+      "loss": 0.3097,
+      "step": 3330
+    },
+    {
+      "epoch": 0.21572046760963637,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5686882387134275e-05,
+      "loss": 0.3581,
+      "step": 3340
+    },
+    {
+      "epoch": 0.21636633727313828,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.567396499386424e-05,
+      "loss": 0.3502,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2170122069366402,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.56610476005942e-05,
+      "loss": 0.3187,
+      "step": 3360
+    },
+    {
+      "epoch": 0.21765807660014208,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.5648130207324164e-05,
+      "loss": 0.3103,
+      "step": 3370
+    },
+    {
+      "epoch": 0.218303946263644,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5635212814054124e-05,
+      "loss": 0.3385,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2189498159271459,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5622295420784085e-05,
+      "loss": 0.3202,
+      "step": 3390
+    },
+    {
+      "epoch": 0.21959568559064782,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.560937802751405e-05,
+      "loss": 0.3226,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2202415552541497,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.559646063424401e-05,
+      "loss": 0.309,
+      "step": 3410
+    },
+    {
+      "epoch": 0.22088742491765162,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.5583543240973974e-05,
+      "loss": 0.359,
+      "step": 3420
+    },
+    {
+      "epoch": 0.22153329458115353,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.5570625847703934e-05,
+      "loss": 0.3195,
+      "step": 3430
+    },
+    {
+      "epoch": 0.22217916424465542,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.5557708454433898e-05,
+      "loss": 0.3031,
+      "step": 3440
+    },
+    {
+      "epoch": 0.22282503390815733,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.554479106116386e-05,
+      "loss": 0.347,
+      "step": 3450
+    },
+    {
+      "epoch": 0.22347090357165925,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.553187366789382e-05,
+      "loss": 0.3115,
+      "step": 3460
+    },
+    {
+      "epoch": 0.22411677323516113,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.5518956274623783e-05,
+      "loss": 0.3012,
+      "step": 3470
+    },
+    {
+      "epoch": 0.22476264289866305,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.5506038881353744e-05,
+      "loss": 0.343,
+      "step": 3480
+    },
+    {
+      "epoch": 0.22540851256216496,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.5493121488083708e-05,
+      "loss": 0.3255,
+      "step": 3490
+    },
+    {
+      "epoch": 0.22605438222566687,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.548020409481367e-05,
+      "loss": 0.3245,
+      "step": 3500
+    },
+    {
+      "epoch": 0.22670025188916876,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.546728670154363e-05,
+      "loss": 0.3372,
+      "step": 3510
+    },
+    {
+      "epoch": 0.22734612155267067,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.545436930827359e-05,
+      "loss": 0.3241,
+      "step": 3520
+    },
+    {
+      "epoch": 0.22799199121617258,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.5441451915003554e-05,
+      "loss": 0.3369,
+      "step": 3530
+    },
+    {
+      "epoch": 0.22863786087967447,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.5428534521733514e-05,
+      "loss": 0.3031,
+      "step": 3540
+    },
+    {
+      "epoch": 0.22928373054317638,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.541561712846348e-05,
+      "loss": 0.3163,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2299296002066783,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.540269973519344e-05,
+      "loss": 0.3225,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2305754698701802,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.53897823419234e-05,
+      "loss": 0.2971,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2312213395336821,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.5376864948653364e-05,
+      "loss": 0.2844,
+      "step": 3580
+    },
+    {
+      "epoch": 0.231867209197184,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5363947555383324e-05,
+      "loss": 0.3211,
+      "step": 3590
+    },
+    {
+      "epoch": 0.23251307886068592,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.5351030162113288e-05,
+      "loss": 0.338,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2331589485241878,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.533811276884325e-05,
+      "loss": 0.3553,
+      "step": 3610
+    },
+    {
+      "epoch": 0.23380481818768972,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.532519537557321e-05,
+      "loss": 0.31,
+      "step": 3620
+    },
+    {
+      "epoch": 0.23445068785119164,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.5312277982303173e-05,
+      "loss": 0.3114,
+      "step": 3630
+    },
+    {
+      "epoch": 0.23509655751469352,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.5299360589033134e-05,
+      "loss": 0.3469,
+      "step": 3640
+    },
+    {
+      "epoch": 0.23574242717819544,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.5286443195763098e-05,
+      "loss": 0.3169,
+      "step": 3650
+    },
+    {
+      "epoch": 0.23638829684169735,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.527352580249306e-05,
+      "loss": 0.3241,
+      "step": 3660
+    },
+    {
+      "epoch": 0.23703416650519926,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.5260608409223022e-05,
+      "loss": 0.3186,
+      "step": 3670
+    },
+    {
+      "epoch": 0.23768003616870115,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5247691015952981e-05,
+      "loss": 0.3087,
+      "step": 3680
+    },
+    {
+      "epoch": 0.23832590583220306,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.5234773622682944e-05,
+      "loss": 0.3515,
+      "step": 3690
+    },
+    {
+      "epoch": 0.23897177549570497,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5221856229412906e-05,
+      "loss": 0.3323,
+      "step": 3700
+    },
+    {
+      "epoch": 0.23961764515920686,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.5208938836142868e-05,
+      "loss": 0.3137,
+      "step": 3710
+    },
+    {
+      "epoch": 0.24026351482270877,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.519602144287283e-05,
+      "loss": 0.321,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2409093844862107,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5183104049602791e-05,
+      "loss": 0.3145,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2415552541497126,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.5170186656332753e-05,
+      "loss": 0.3122,
+      "step": 3740
+    },
+    {
+      "epoch": 0.2422011238132145,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5157269263062716e-05,
+      "loss": 0.306,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2428469934767164,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.5144351869792678e-05,
+      "loss": 0.2876,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2434928631402183,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.513143447652264e-05,
+      "loss": 0.3036,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2441387328037202,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.5118517083252603e-05,
+      "loss": 0.3037,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2447846024672221,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.5105599689982561e-05,
+      "loss": 0.347,
+      "step": 3790
+    },
+    {
+      "epoch": 0.24543047213072403,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.5092682296712524e-05,
+      "loss": 0.3208,
+      "step": 3800
+    },
+    {
+      "epoch": 0.24607634179422594,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.5079764903442486e-05,
+      "loss": 0.3294,
+      "step": 3810
+    },
+    {
+      "epoch": 0.24672221145772782,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5066847510172448e-05,
+      "loss": 0.3173,
+      "step": 3820
+    },
+    {
+      "epoch": 0.24736808112122974,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.505393011690241e-05,
+      "loss": 0.3107,
+      "step": 3830
+    },
+    {
+      "epoch": 0.24801395078473165,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.5041012723632373e-05,
+      "loss": 0.2944,
+      "step": 3840
+    },
+    {
+      "epoch": 0.24865982044823354,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.5028095330362334e-05,
+      "loss": 0.329,
+      "step": 3850
+    },
+    {
+      "epoch": 0.24930569011173545,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5015177937092296e-05,
+      "loss": 0.2837,
+      "step": 3860
+    },
+    {
+      "epoch": 0.24995155977523736,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.5002260543822258e-05,
+      "loss": 0.2982,
+      "step": 3870
+    },
+    {
+      "epoch": 0.25059742943873925,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.498934315055222e-05,
+      "loss": 0.346,
+      "step": 3880
+    },
+    {
+      "epoch": 0.2512432991022412,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.4976425757282183e-05,
+      "loss": 0.2859,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2518891687657431,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.4963508364012143e-05,
+      "loss": 0.3087,
+      "step": 3900
+    },
+    {
+      "epoch": 0.25253503842924496,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.4950590970742106e-05,
+      "loss": 0.3128,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2531809080927469,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.4937673577472068e-05,
+      "loss": 0.3046,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2538267777562488,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.492475618420203e-05,
+      "loss": 0.2898,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2544726474197507,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4911838790931992e-05,
+      "loss": 0.3,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2551185170832526,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.4898921397661955e-05,
+      "loss": 0.2813,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2557643867467545,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.4886004004391914e-05,
+      "loss": 0.3018,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2564102564102564,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.4873086611121876e-05,
+      "loss": 0.3272,
+      "step": 3970
+    },
+    {
+      "epoch": 0.25705612607375833,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.4860169217851838e-05,
+      "loss": 0.3298,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2577019957372602,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.48472518245818e-05,
+      "loss": 0.2867,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2583478654007621,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.4834334431311763e-05,
+      "loss": 0.2922,
+      "step": 4000
+    },
+    {
+      "epoch": 0.25899373506426404,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.4821417038041725e-05,
+      "loss": 0.2894,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2596396047277659,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.4808499644771686e-05,
+      "loss": 0.2941,
+      "step": 4020
+    },
+    {
+      "epoch": 0.26028547439126787,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.4795582251501648e-05,
+      "loss": 0.3188,
+      "step": 4030
+    },
+    {
+      "epoch": 0.26093134405476975,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.478266485823161e-05,
+      "loss": 0.3216,
+      "step": 4040
+    },
+    {
+      "epoch": 0.26157721371827164,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.4769747464961573e-05,
+      "loss": 0.3011,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2622230833817736,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.4756830071691535e-05,
+      "loss": 0.2879,
+      "step": 4060
+    },
+    {
+      "epoch": 0.26286895304527547,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.4743912678421495e-05,
+      "loss": 0.3318,
+      "step": 4070
+    },
+    {
+      "epoch": 0.26351482270877735,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.4730995285151458e-05,
+      "loss": 0.3285,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2641606923722793,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.471807789188142e-05,
+      "loss": 0.3014,
+      "step": 4090
+    },
+    {
+      "epoch": 0.2648065620357812,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4705160498611382e-05,
+      "loss": 0.3336,
+      "step": 4100
+    },
+    {
+      "epoch": 0.26545243169928306,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4692243105341345e-05,
+      "loss": 0.3544,
+      "step": 4110
+    },
+    {
+      "epoch": 0.266098301362785,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.4679325712071307e-05,
+      "loss": 0.3025,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2667441710262869,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4666408318801266e-05,
+      "loss": 0.2966,
+      "step": 4130
+    },
+    {
+      "epoch": 0.2673900406897888,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.4653490925531228e-05,
+      "loss": 0.2699,
+      "step": 4140
+    },
+    {
+      "epoch": 0.2680359103532907,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.464057353226119e-05,
+      "loss": 0.2826,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2686817800167926,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4627656138991153e-05,
+      "loss": 0.3193,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2693276496802945,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4614738745721115e-05,
+      "loss": 0.296,
+      "step": 4170
+    },
+    {
+      "epoch": 0.26997351934379643,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.4601821352451076e-05,
+      "loss": 0.2911,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2706193890072983,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.4588903959181038e-05,
+      "loss": 0.2901,
+      "step": 4190
+    },
+    {
+      "epoch": 0.27126525867080026,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.4575986565911e-05,
+      "loss": 0.3118,
+      "step": 4200
+    },
+    {
+      "epoch": 0.27191112833430214,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4563069172640962e-05,
+      "loss": 0.2657,
+      "step": 4210
+    },
+    {
+      "epoch": 0.27255699799780403,
+      "grad_norm": 1.25,
+      "learning_rate": 1.4550151779370925e-05,
+      "loss": 0.3434,
+      "step": 4220
+    },
+    {
+      "epoch": 0.27320286766130597,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.4537234386100887e-05,
+      "loss": 0.3357,
+      "step": 4230
+    },
+    {
+      "epoch": 0.27384873732480786,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.4524316992830848e-05,
+      "loss": 0.2931,
+      "step": 4240
+    },
+    {
+      "epoch": 0.27449460698830974,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.451139959956081e-05,
+      "loss": 0.2707,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2751404766518117,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.4498482206290772e-05,
+      "loss": 0.279,
+      "step": 4260
+    },
+    {
+      "epoch": 0.27578634631531357,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.4485564813020735e-05,
+      "loss": 0.2904,
+      "step": 4270
+    },
+    {
+      "epoch": 0.27643221597881545,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4472647419750697e-05,
+      "loss": 0.3024,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2770780856423174,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4459730026480659e-05,
+      "loss": 0.2895,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2777239553058193,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4446812633210618e-05,
+      "loss": 0.3145,
+      "step": 4300
+    },
+    {
+      "epoch": 0.27836982496932117,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.443389523994058e-05,
+      "loss": 0.2871,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2790156946328231,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4420977846670543e-05,
+      "loss": 0.3401,
+      "step": 4320
+    },
+    {
+      "epoch": 0.279661564296325,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4408060453400505e-05,
+      "loss": 0.2879,
+      "step": 4330
+    },
+    {
+      "epoch": 0.2803074339598269,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4395143060130467e-05,
+      "loss": 0.2919,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2809533036233288,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.4382225666860428e-05,
+      "loss": 0.2694,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2815991732868307,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.436930827359039e-05,
+      "loss": 0.2994,
+      "step": 4360
+    },
+    {
+      "epoch": 0.28224504295033265,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.4356390880320352e-05,
+      "loss": 0.2842,
+      "step": 4370
+    },
+    {
+      "epoch": 0.28289091261383453,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4343473487050315e-05,
+      "loss": 0.2762,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2835367822773364,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.4330556093780277e-05,
+      "loss": 0.3197,
+      "step": 4390
+    },
+    {
+      "epoch": 0.28418265194083836,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.431763870051024e-05,
+      "loss": 0.3159,
+      "step": 4400
+    },
+    {
+      "epoch": 0.28482852160434025,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.43047213072402e-05,
+      "loss": 0.3014,
+      "step": 4410
+    },
+    {
+      "epoch": 0.28547439126784213,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.4291803913970162e-05,
+      "loss": 0.2796,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2861202609313441,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4278886520700124e-05,
+      "loss": 0.2716,
+      "step": 4430
+    },
+    {
+      "epoch": 0.28676613059484596,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.4265969127430087e-05,
+      "loss": 0.2985,
+      "step": 4440
+    },
+    {
+      "epoch": 0.28741200025834784,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.4253051734160049e-05,
+      "loss": 0.2737,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2880578699218498,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4240134340890011e-05,
+      "loss": 0.2981,
+      "step": 4460
+    },
+    {
+      "epoch": 0.28870373958535167,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.422721694761997e-05,
+      "loss": 0.306,
+      "step": 4470
+    },
+    {
+      "epoch": 0.28934960924885356,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.4214299554349932e-05,
+      "loss": 0.2849,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2899954789123555,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.4201382161079895e-05,
+      "loss": 0.2926,
+      "step": 4490
+    },
+    {
+      "epoch": 0.2906413485758574,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4188464767809857e-05,
+      "loss": 0.2816,
+      "step": 4500
+    },
+    {
+      "epoch": 0.2912872182393593,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.417554737453982e-05,
+      "loss": 0.282,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2919330879028612,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.416262998126978e-05,
+      "loss": 0.3208,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2925789575663631,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4149712587999742e-05,
+      "loss": 0.2701,
+      "step": 4530
+    },
+    {
+      "epoch": 0.29322482722986504,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4136795194729705e-05,
+      "loss": 0.3363,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2938706968933669,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4123877801459667e-05,
+      "loss": 0.2998,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2945165665568688,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.4110960408189629e-05,
+      "loss": 0.2947,
+      "step": 4560
+    },
+    {
+      "epoch": 0.29516243622037075,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.4098043014919591e-05,
+      "loss": 0.3119,
+      "step": 4570
+    },
+    {
+      "epoch": 0.29580830588387264,
+      "grad_norm": 1.125,
+      "learning_rate": 1.4085125621649552e-05,
+      "loss": 0.2674,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2964541755473745,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.4072208228379514e-05,
+      "loss": 0.2893,
+      "step": 4590
+    },
+    {
+      "epoch": 0.29710004521087646,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4059290835109477e-05,
+      "loss": 0.3205,
+      "step": 4600
+    },
+    {
+      "epoch": 0.29774591487437835,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4046373441839439e-05,
+      "loss": 0.2671,
+      "step": 4610
+    },
+    {
+      "epoch": 0.29839178453788023,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.4033456048569401e-05,
+      "loss": 0.2869,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2990376542013822,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.4020538655299363e-05,
+      "loss": 0.2731,
+      "step": 4630
+    },
+    {
+      "epoch": 0.29968352386488406,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.4007621262029322e-05,
+      "loss": 0.2805,
+      "step": 4640
+    },
+    {
+      "epoch": 0.30032939352838595,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.3994703868759285e-05,
+      "loss": 0.3045,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3009752631918879,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3981786475489247e-05,
+      "loss": 0.3243,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3016211328553898,
+      "grad_norm": 1.375,
+      "learning_rate": 1.396886908221921e-05,
+      "loss": 0.2969,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3022670025188917,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.3955951688949172e-05,
+      "loss": 0.2728,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3029128721823936,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.3943034295679132e-05,
+      "loss": 0.2671,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3035587418458955,
+      "grad_norm": 1.125,
+      "learning_rate": 1.3930116902409094e-05,
+      "loss": 0.301,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3042046115093974,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.3917199509139057e-05,
+      "loss": 0.2767,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3048504811728993,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3904282115869019e-05,
+      "loss": 0.2887,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3054963508364012,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.3891364722598981e-05,
+      "loss": 0.2729,
+      "step": 4730
+    },
+    {
+      "epoch": 0.30614222049990314,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3878447329328944e-05,
+      "loss": 0.2962,
+      "step": 4740
+    },
+    {
+      "epoch": 0.306788090163405,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.3865529936058904e-05,
+      "loss": 0.2769,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3074339598269069,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3852612542788866e-05,
+      "loss": 0.296,
+      "step": 4760
+    },
+    {
+      "epoch": 0.30807982949040885,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.3839695149518829e-05,
+      "loss": 0.2779,
+      "step": 4770
+    },
+    {
+      "epoch": 0.30872569915391074,
+      "grad_norm": 1.375,
+      "learning_rate": 1.3826777756248791e-05,
+      "loss": 0.285,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3093715688174126,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.3813860362978753e-05,
+      "loss": 0.2754,
+      "step": 4790
+    },
+    {
+      "epoch": 0.31001743848091456,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.3800942969708716e-05,
+      "loss": 0.2928,
+      "step": 4800
+    },
+    {
+      "epoch": 0.31066330814441645,
+      "grad_norm": 1.5,
+      "learning_rate": 1.3788025576438675e-05,
+      "loss": 0.2771,
+      "step": 4810
+    },
+    {
+      "epoch": 0.31130917780791834,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.3775108183168637e-05,
+      "loss": 0.2937,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3119550474714203,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.3762190789898599e-05,
+      "loss": 0.3024,
+      "step": 4830
+    },
+    {
+      "epoch": 0.31260091713492216,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3749273396628561e-05,
+      "loss": 0.2915,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3132467867984241,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.3736356003358524e-05,
+      "loss": 0.2942,
+      "step": 4850
+    },
+    {
+      "epoch": 0.313892656461926,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3723438610088484e-05,
+      "loss": 0.2808,
+      "step": 4860
+    },
+    {
+      "epoch": 0.3145385261254279,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.3710521216818447e-05,
+      "loss": 0.2821,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3151843957889298,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.3697603823548409e-05,
+      "loss": 0.27,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3158302654524317,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.3684686430278371e-05,
+      "loss": 0.2942,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3164761351159336,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3671769037008333e-05,
+      "loss": 0.2887,
+      "step": 4900
+    },
+    {
+      "epoch": 0.31712200477943553,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.3658851643738296e-05,
+      "loss": 0.3196,
+      "step": 4910
+    },
+    {
+      "epoch": 0.3177678744429374,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.3645934250468256e-05,
+      "loss": 0.2701,
+      "step": 4920
+    },
+    {
+      "epoch": 0.3184137441064393,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3633016857198219e-05,
+      "loss": 0.2718,
+      "step": 4930
+    },
+    {
+      "epoch": 0.31905961376994124,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.3620099463928181e-05,
+      "loss": 0.2818,
+      "step": 4940
+    },
+    {
+      "epoch": 0.3197054834334431,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.3607182070658143e-05,
+      "loss": 0.2695,
+      "step": 4950
+    },
+    {
+      "epoch": 0.320351353096945,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3594264677388106e-05,
+      "loss": 0.2907,
+      "step": 4960
+    },
+    {
+      "epoch": 0.32099722276044695,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.3581347284118068e-05,
+      "loss": 0.264,
+      "step": 4970
+    },
+    {
+      "epoch": 0.32164309242394884,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.3568429890848027e-05,
+      "loss": 0.3022,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3222889620874507,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.3555512497577989e-05,
+      "loss": 0.294,
+      "step": 4990
+    },
+    {
+      "epoch": 0.32293483175095267,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.3542595104307951e-05,
+      "loss": 0.2754,
+      "step": 5000
+    },
+    {
+      "epoch": 0.32358070141445455,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3529677711037914e-05,
+      "loss": 0.3165,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3242265710779565,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3516760317767876e-05,
+      "loss": 0.3159,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3248724407414584,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.3503842924497836e-05,
+      "loss": 0.2921,
+      "step": 5030
+    },
+    {
+      "epoch": 0.32551831040496026,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.3490925531227799e-05,
+      "loss": 0.3127,
+      "step": 5040
+    },
+    {
+      "epoch": 0.3261641800684622,
+      "grad_norm": 1.5,
+      "learning_rate": 1.3478008137957761e-05,
+      "loss": 0.2831,
+      "step": 5050
+    },
+    {
+      "epoch": 0.3268100497319641,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.3465090744687723e-05,
+      "loss": 0.3182,
+      "step": 5060
+    },
+    {
+      "epoch": 0.327455919395466,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3452173351417686e-05,
+      "loss": 0.268,
+      "step": 5070
+    },
+    {
+      "epoch": 0.3281017890589679,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.3439255958147648e-05,
+      "loss": 0.2848,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3287476587224698,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.3426338564877609e-05,
+      "loss": 0.2831,
+      "step": 5090
+    },
+    {
+      "epoch": 0.3293935283859717,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.341342117160757e-05,
+      "loss": 0.2941,
+      "step": 5100
+    },
+    {
+      "epoch": 0.33003939804947363,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.3400503778337533e-05,
+      "loss": 0.2867,
+      "step": 5110
+    },
+    {
+      "epoch": 0.3306852677129755,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.3387586385067495e-05,
+      "loss": 0.2909,
+      "step": 5120
+    },
+    {
+      "epoch": 0.3313311373764774,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.3374668991797458e-05,
+      "loss": 0.2824,
+      "step": 5130
+    },
+    {
+      "epoch": 0.33197700703997934,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.336175159852742e-05,
+      "loss": 0.2925,
+      "step": 5140
+    },
+    {
+      "epoch": 0.33262287670348123,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.3348834205257379e-05,
+      "loss": 0.3143,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3332687463669831,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.3335916811987341e-05,
+      "loss": 0.3114,
+      "step": 5160
+    },
+    {
+      "epoch": 0.33391461603048506,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.3322999418717303e-05,
+      "loss": 0.2604,
+      "step": 5170
+    },
+    {
+      "epoch": 0.33456048569398694,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.3310082025447266e-05,
+      "loss": 0.3184,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3352063553574889,
+      "grad_norm": 1.25,
+      "learning_rate": 1.3297164632177228e-05,
+      "loss": 0.2632,
+      "step": 5190
+    },
+    {
+      "epoch": 0.33585222502099077,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.3284247238907189e-05,
+      "loss": 0.3283,
+      "step": 5200
+    },
+    {
+      "epoch": 0.33649809468449265,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.3271329845637151e-05,
+      "loss": 0.2789,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3371439643479946,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3258412452367113e-05,
+      "loss": 0.2834,
+      "step": 5220
+    },
+    {
+      "epoch": 0.3377898340114965,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.3245495059097076e-05,
+      "loss": 0.2747,
+      "step": 5230
+    },
+    {
+      "epoch": 0.33843570367499837,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.3232577665827038e-05,
+      "loss": 0.2696,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3390815733385003,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.3219660272557e-05,
+      "loss": 0.2821,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3397274430020022,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.320674287928696e-05,
+      "loss": 0.2733,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3403733126655041,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.3193825486016923e-05,
+      "loss": 0.2673,
+      "step": 5270
+    },
+    {
+      "epoch": 0.341019182329006,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.3180908092746885e-05,
+      "loss": 0.253,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3416650519925079,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.3167990699476848e-05,
+      "loss": 0.3147,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3423109216560098,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.315507330620681e-05,
+      "loss": 0.2985,
+      "step": 5300
+    },
+    {
+      "epoch": 0.34295679131951173,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.3142155912936772e-05,
+      "loss": 0.2319,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3436026609830136,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.3129238519666731e-05,
+      "loss": 0.3024,
+      "step": 5320
+    },
+    {
+      "epoch": 0.34424853064651556,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.3116321126396693e-05,
+      "loss": 0.2603,
+      "step": 5330
+    },
+    {
+      "epoch": 0.34489440031001745,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.3103403733126656e-05,
+      "loss": 0.2716,
+      "step": 5340
+    },
+    {
+      "epoch": 0.34554026997351933,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.3090486339856618e-05,
+      "loss": 0.2717,
+      "step": 5350
+    },
+    {
+      "epoch": 0.3461861396370213,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.307756894658658e-05,
+      "loss": 0.2988,
+      "step": 5360
+    },
+    {
+      "epoch": 0.34683200930052316,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.306465155331654e-05,
+      "loss": 0.3058,
+      "step": 5370
+    },
+    {
+      "epoch": 0.34747787896402504,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.3051734160046503e-05,
+      "loss": 0.2729,
+      "step": 5380
+    },
+    {
+      "epoch": 0.348123748627527,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.3038816766776465e-05,
+      "loss": 0.2547,
+      "step": 5390
+    },
+    {
+      "epoch": 0.34876961829102887,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.3025899373506428e-05,
+      "loss": 0.3013,
+      "step": 5400
+    },
+    {
+      "epoch": 0.34941548795453076,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.301298198023639e-05,
+      "loss": 0.2928,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3500613576180327,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3000064586966352e-05,
+      "loss": 0.2673,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3507072272815346,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.2987147193696313e-05,
+      "loss": 0.2968,
+      "step": 5430
+    },
+    {
+      "epoch": 0.35135309694503647,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.2974229800426275e-05,
+      "loss": 0.2581,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3519989666085384,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.2961312407156237e-05,
+      "loss": 0.2805,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3526448362720403,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.29483950138862e-05,
+      "loss": 0.2609,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3532907059355422,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2935477620616162e-05,
+      "loss": 0.2851,
+      "step": 5470
+    },
+    {
+      "epoch": 0.3539365755990441,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.2922560227346124e-05,
+      "loss": 0.2649,
+      "step": 5480
+    },
+    {
+      "epoch": 0.354582445262546,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2909642834076083e-05,
+      "loss": 0.3208,
+      "step": 5490
+    },
+    {
+      "epoch": 0.35522831492604795,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.2896725440806046e-05,
+      "loss": 0.2663,
+      "step": 5500
+    },
+    {
+      "epoch": 0.35587418458954984,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.2883808047536008e-05,
+      "loss": 0.266,
+      "step": 5510
+    },
+    {
+      "epoch": 0.3565200542530517,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.287089065426597e-05,
+      "loss": 0.2681,
+      "step": 5520
+    },
+    {
+      "epoch": 0.35716592391655366,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2857973260995932e-05,
+      "loss": 0.3058,
+      "step": 5530
+    },
+    {
+      "epoch": 0.35781179358005555,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.2845055867725893e-05,
+      "loss": 0.271,
+      "step": 5540
+    },
+    {
+      "epoch": 0.35845766324355743,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2832138474455855e-05,
+      "loss": 0.2786,
+      "step": 5550
+    },
+    {
+      "epoch": 0.3591035329070594,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.2819221081185818e-05,
+      "loss": 0.2897,
+      "step": 5560
+    },
+    {
+      "epoch": 0.35974940257056126,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.280630368791578e-05,
+      "loss": 0.262,
+      "step": 5570
+    },
+    {
+      "epoch": 0.36039527223406315,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.2793386294645742e-05,
+      "loss": 0.2695,
+      "step": 5580
+    },
+    {
+      "epoch": 0.3610411418975651,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.2780468901375704e-05,
+      "loss": 0.2609,
+      "step": 5590
+    },
+    {
+      "epoch": 0.361687011561067,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.2767551508105665e-05,
+      "loss": 0.2691,
+      "step": 5600
+    },
+    {
+      "epoch": 0.36233288122456886,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2754634114835627e-05,
+      "loss": 0.2583,
+      "step": 5610
+    },
+    {
+      "epoch": 0.3629787508880708,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.274171672156559e-05,
+      "loss": 0.2713,
+      "step": 5620
+    },
+    {
+      "epoch": 0.3636246205515727,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2728799328295552e-05,
+      "loss": 0.2948,
+      "step": 5630
+    },
+    {
+      "epoch": 0.36427049021507457,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.2715881935025514e-05,
+      "loss": 0.268,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3649163598785765,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.2702964541755477e-05,
+      "loss": 0.3048,
+      "step": 5650
+    },
+    {
+      "epoch": 0.3655622295420784,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.2690047148485435e-05,
+      "loss": 0.3148,
+      "step": 5660
+    },
+    {
+      "epoch": 0.36620809920558034,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2677129755215398e-05,
+      "loss": 0.2606,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3668539688690822,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.266421236194536e-05,
+      "loss": 0.2501,
+      "step": 5680
+    },
+    {
+      "epoch": 0.3674998385325841,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2651294968675322e-05,
+      "loss": 0.2725,
+      "step": 5690
+    },
+    {
+      "epoch": 0.36814570819608605,
+      "grad_norm": 1.375,
+      "learning_rate": 1.2638377575405285e-05,
+      "loss": 0.2557,
+      "step": 5700
+    },
+    {
+      "epoch": 0.36879157785958794,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.2625460182135245e-05,
+      "loss": 0.2807,
+      "step": 5710
+    },
+    {
+      "epoch": 0.3694374475230898,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.2612542788865207e-05,
+      "loss": 0.2331,
+      "step": 5720
+    },
+    {
+      "epoch": 0.37008331718659176,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.259962539559517e-05,
+      "loss": 0.2674,
+      "step": 5730
+    },
+    {
+      "epoch": 0.37072918685009365,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.2586708002325132e-05,
+      "loss": 0.2602,
+      "step": 5740
+    },
+    {
+      "epoch": 0.37137505651359554,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2573790609055094e-05,
+      "loss": 0.2682,
+      "step": 5750
+    },
+    {
+      "epoch": 0.3720209261770975,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.2560873215785057e-05,
+      "loss": 0.2859,
+      "step": 5760
+    },
+    {
+      "epoch": 0.37266679584059936,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.2547955822515017e-05,
+      "loss": 0.2611,
+      "step": 5770
+    },
+    {
+      "epoch": 0.37331266550410125,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.253503842924498e-05,
+      "loss": 0.3003,
+      "step": 5780
+    },
+    {
+      "epoch": 0.3739585351676032,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.2522121035974942e-05,
+      "loss": 0.2663,
+      "step": 5790
+    },
+    {
+      "epoch": 0.3746044048311051,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2509203642704904e-05,
+      "loss": 0.2889,
+      "step": 5800
+    },
+    {
+      "epoch": 0.37525027449460696,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.2496286249434866e-05,
+      "loss": 0.284,
+      "step": 5810
+    },
+    {
+      "epoch": 0.3758961441581089,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2483368856164825e-05,
+      "loss": 0.2516,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3765420138216108,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2470451462894788e-05,
+      "loss": 0.2857,
+      "step": 5830
+    },
+    {
+      "epoch": 0.37718788348511273,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.245753406962475e-05,
+      "loss": 0.2398,
+      "step": 5840
+    },
+    {
+      "epoch": 0.3778337531486146,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2444616676354712e-05,
+      "loss": 0.2716,
+      "step": 5850
+    },
+    {
+      "epoch": 0.3784796228121165,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.2431699283084674e-05,
+      "loss": 0.2544,
+      "step": 5860
+    },
+    {
+      "epoch": 0.37912549247561844,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.2418781889814637e-05,
+      "loss": 0.264,
+      "step": 5870
+    },
+    {
+      "epoch": 0.3797713621391203,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2405864496544597e-05,
+      "loss": 0.2627,
+      "step": 5880
+    },
+    {
+      "epoch": 0.3804172318026222,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.239294710327456e-05,
+      "loss": 0.2469,
+      "step": 5890
+    },
+    {
+      "epoch": 0.38106310146612415,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.2380029710004522e-05,
+      "loss": 0.237,
+      "step": 5900
+    },
+    {
+      "epoch": 0.38170897112962604,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2367112316734484e-05,
+      "loss": 0.2634,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3823548407931279,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.2354194923464447e-05,
+      "loss": 0.2816,
+      "step": 5920
+    },
+    {
+      "epoch": 0.38300071045662987,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2341277530194409e-05,
+      "loss": 0.2863,
+      "step": 5930
+    },
+    {
+      "epoch": 0.38364658012013175,
+      "grad_norm": 1.25,
+      "learning_rate": 1.232836013692437e-05,
+      "loss": 0.3067,
+      "step": 5940
+    },
+    {
+      "epoch": 0.38429244978363364,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2315442743654332e-05,
+      "loss": 0.2695,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3849383194471356,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2302525350384294e-05,
+      "loss": 0.2605,
+      "step": 5960
+    },
+    {
+      "epoch": 0.38558418911063747,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.2289607957114256e-05,
+      "loss": 0.2201,
+      "step": 5970
+    },
+    {
+      "epoch": 0.38623005877413935,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.2276690563844219e-05,
+      "loss": 0.2648,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3868759284376413,
+      "grad_norm": 1.5,
+      "learning_rate": 1.2263773170574177e-05,
+      "loss": 0.2734,
+      "step": 5990
+    },
+    {
+      "epoch": 0.3875217981011432,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.225085577730414e-05,
+      "loss": 0.2623,
+      "step": 6000
+    },
+    {
+      "epoch": 0.3881676677646451,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.2237938384034102e-05,
+      "loss": 0.2523,
+      "step": 6010
+    },
+    {
+      "epoch": 0.388813537428147,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.2225020990764064e-05,
+      "loss": 0.2922,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3894594070916489,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2212103597494027e-05,
+      "loss": 0.2432,
+      "step": 6030
+    },
+    {
+      "epoch": 0.39010527675515083,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.2199186204223989e-05,
+      "loss": 0.2515,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3907511464186527,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.218626881095395e-05,
+      "loss": 0.27,
+      "step": 6050
+    },
+    {
+      "epoch": 0.3913970160821546,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.2173351417683912e-05,
+      "loss": 0.2588,
+      "step": 6060
+    },
+    {
+      "epoch": 0.39204288574565654,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.2160434024413874e-05,
+      "loss": 0.2737,
+      "step": 6070
+    },
+    {
+      "epoch": 0.39268875540915843,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2147516631143836e-05,
+      "loss": 0.2602,
+      "step": 6080
+    },
+    {
+      "epoch": 0.3933346250726603,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.2134599237873799e-05,
+      "loss": 0.2524,
+      "step": 6090
+    },
+    {
+      "epoch": 0.39398049473616226,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.2121681844603761e-05,
+      "loss": 0.2579,
+      "step": 6100
+    },
+    {
+      "epoch": 0.39462636439966414,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.2108764451333722e-05,
+      "loss": 0.2479,
+      "step": 6110
+    },
+    {
+      "epoch": 0.39527223406316603,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.2095847058063684e-05,
+      "loss": 0.287,
+      "step": 6120
+    },
+    {
+      "epoch": 0.39591810372666797,
+      "grad_norm": 1.125,
+      "learning_rate": 1.2082929664793646e-05,
+      "loss": 0.2955,
+      "step": 6130
+    },
+    {
+      "epoch": 0.39656397339016985,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.2070012271523608e-05,
+      "loss": 0.2678,
+      "step": 6140
+    },
+    {
+      "epoch": 0.3972098430536718,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.205709487825357e-05,
+      "loss": 0.277,
+      "step": 6150
+    },
+    {
+      "epoch": 0.3978557127171737,
+      "grad_norm": 1.25,
+      "learning_rate": 1.204417748498353e-05,
+      "loss": 0.27,
+      "step": 6160
+    },
+    {
+      "epoch": 0.39850158238067557,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.2031260091713492e-05,
+      "loss": 0.236,
+      "step": 6170
+    },
+    {
+      "epoch": 0.3991474520441775,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.2018342698443454e-05,
+      "loss": 0.2735,
+      "step": 6180
+    },
+    {
+      "epoch": 0.3997933217076794,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2005425305173417e-05,
+      "loss": 0.2839,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4004391913711813,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.1992507911903379e-05,
+      "loss": 0.2568,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4010850610346832,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1979590518633341e-05,
+      "loss": 0.2549,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4017309306981851,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1966673125363302e-05,
+      "loss": 0.2614,
+      "step": 6220
+    },
+    {
+      "epoch": 0.402376800361687,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.1953755732093264e-05,
+      "loss": 0.2681,
+      "step": 6230
+    },
+    {
+      "epoch": 0.40302267002518893,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.1940838338823226e-05,
+      "loss": 0.2931,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4036685396886908,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1927920945553189e-05,
+      "loss": 0.3026,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4043144093521927,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.1915003552283151e-05,
+      "loss": 0.2875,
+      "step": 6260
+    },
+    {
+      "epoch": 0.40496027901569465,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.1902086159013113e-05,
+      "loss": 0.254,
+      "step": 6270
+    },
+    {
+      "epoch": 0.40560614867919653,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.1889168765743074e-05,
+      "loss": 0.2546,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4062520183426984,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.1876251372473036e-05,
+      "loss": 0.278,
+      "step": 6290
+    },
+    {
+      "epoch": 0.40689788800620036,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.1863333979202998e-05,
+      "loss": 0.263,
+      "step": 6300
+    },
+    {
+      "epoch": 0.40754375766970224,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.185041658593296e-05,
+      "loss": 0.2772,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4081896273332042,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1837499192662923e-05,
+      "loss": 0.2636,
+      "step": 6320
+    },
+    {
+      "epoch": 0.40883549699670607,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.1824581799392882e-05,
+      "loss": 0.2763,
+      "step": 6330
+    },
+    {
+      "epoch": 0.40948136666020796,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1811664406122844e-05,
+      "loss": 0.282,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4101272363237099,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.1798747012852806e-05,
+      "loss": 0.2645,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4107731059872118,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.1785829619582769e-05,
+      "loss": 0.252,
+      "step": 6360
+    },
+    {
+      "epoch": 0.41141897565071367,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.1772912226312731e-05,
+      "loss": 0.2339,
+      "step": 6370
+    },
+    {
+      "epoch": 0.4120648453142156,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.1759994833042693e-05,
+      "loss": 0.2572,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4127107149777175,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.1747077439772654e-05,
+      "loss": 0.2661,
+      "step": 6390
+    },
+    {
+      "epoch": 0.4133565846412194,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.1734160046502616e-05,
+      "loss": 0.2379,
+      "step": 6400
+    },
+    {
+      "epoch": 0.4140024543047213,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1721242653232578e-05,
+      "loss": 0.2577,
+      "step": 6410
+    },
+    {
+      "epoch": 0.4146483239682232,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.170832525996254e-05,
+      "loss": 0.2605,
+      "step": 6420
+    },
+    {
+      "epoch": 0.4152941936317251,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.1695407866692503e-05,
+      "loss": 0.2856,
+      "step": 6430
+    },
+    {
+      "epoch": 0.41594006329522704,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.1682490473422465e-05,
+      "loss": 0.2529,
+      "step": 6440
+    },
+    {
+      "epoch": 0.4165859329587289,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.1669573080152426e-05,
+      "loss": 0.25,
+      "step": 6450
+    },
+    {
+      "epoch": 0.4172318026222308,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.1656655686882388e-05,
+      "loss": 0.2936,
+      "step": 6460
+    },
+    {
+      "epoch": 0.41787767228573275,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.164373829361235e-05,
+      "loss": 0.2371,
+      "step": 6470
+    },
+    {
+      "epoch": 0.41852354194923463,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.1630820900342313e-05,
+      "loss": 0.281,
+      "step": 6480
+    },
+    {
+      "epoch": 0.4191694116127366,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1617903507072275e-05,
+      "loss": 0.2726,
+      "step": 6490
+    },
+    {
+      "epoch": 0.41981528127623846,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.1604986113802234e-05,
+      "loss": 0.2745,
+      "step": 6500
+    },
+    {
+      "epoch": 0.42046115093974035,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.1592068720532196e-05,
+      "loss": 0.2778,
+      "step": 6510
+    },
+    {
+      "epoch": 0.4211070206032423,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1579151327262159e-05,
+      "loss": 0.2831,
+      "step": 6520
+    },
+    {
+      "epoch": 0.4217528902667442,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.1566233933992121e-05,
+      "loss": 0.2526,
+      "step": 6530
+    },
+    {
+      "epoch": 0.42239875993024606,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.1553316540722083e-05,
+      "loss": 0.2756,
+      "step": 6540
+    },
+    {
+      "epoch": 0.423044629593748,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.1540399147452045e-05,
+      "loss": 0.2947,
+      "step": 6550
+    },
+    {
+      "epoch": 0.4236904992572499,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.1527481754182006e-05,
+      "loss": 0.2641,
+      "step": 6560
+    },
+    {
+      "epoch": 0.42433636892075177,
+      "grad_norm": 1.375,
+      "learning_rate": 1.1514564360911968e-05,
+      "loss": 0.2662,
+      "step": 6570
+    },
+    {
+      "epoch": 0.4249822385842537,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.150164696764193e-05,
+      "loss": 0.2915,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4256281082477556,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.1488729574371893e-05,
+      "loss": 0.2619,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4262739779112575,
+      "grad_norm": 1.25,
+      "learning_rate": 1.1475812181101855e-05,
+      "loss": 0.2574,
+      "step": 6600
+    },
+    {
+      "epoch": 0.4269198475747594,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.1462894787831817e-05,
+      "loss": 0.2781,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4275657172382613,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.1449977394561778e-05,
+      "loss": 0.2637,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4282115869017632,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.143706000129174e-05,
+      "loss": 0.2976,
+      "step": 6630
+    },
+    {
+      "epoch": 0.42885745656526514,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1424142608021703e-05,
+      "loss": 0.2766,
+      "step": 6640
+    },
+    {
+      "epoch": 0.429503326228767,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.1411225214751665e-05,
+      "loss": 0.2573,
+      "step": 6650
+    },
+    {
+      "epoch": 0.43014919589226897,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1398307821481627e-05,
+      "loss": 0.2483,
+      "step": 6660
+    },
+    {
+      "epoch": 0.43079506555577085,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.1385390428211586e-05,
+      "loss": 0.2724,
+      "step": 6670
+    },
+    {
+      "epoch": 0.43144093521927274,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.1372473034941548e-05,
+      "loss": 0.2643,
+      "step": 6680
+    },
+    {
+      "epoch": 0.4320868048827747,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.135955564167151e-05,
+      "loss": 0.2381,
+      "step": 6690
+    },
+    {
+      "epoch": 0.43273267454627656,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.1346638248401473e-05,
+      "loss": 0.2742,
+      "step": 6700
+    },
+    {
+      "epoch": 0.43337854420977845,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.1333720855131435e-05,
+      "loss": 0.2807,
+      "step": 6710
+    },
+    {
+      "epoch": 0.4340244138732804,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.1320803461861398e-05,
+      "loss": 0.2549,
+      "step": 6720
+    },
+    {
+      "epoch": 0.4346702835367823,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.1307886068591358e-05,
+      "loss": 0.2546,
+      "step": 6730
+    },
+    {
+      "epoch": 0.43531615320028416,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.129496867532132e-05,
+      "loss": 0.2787,
+      "step": 6740
+    },
+    {
+      "epoch": 0.4359620228637861,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.2977,
+      "step": 6750
+    },
+    {
+      "epoch": 0.436607892527288,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1269133888781245e-05,
+      "loss": 0.2579,
+      "step": 6760
+    },
+    {
+      "epoch": 0.4372537621907899,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.1256216495511207e-05,
+      "loss": 0.2587,
+      "step": 6770
+    },
+    {
+      "epoch": 0.4378996318542918,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.124329910224117e-05,
+      "loss": 0.2931,
+      "step": 6780
+    },
+    {
+      "epoch": 0.4385455015177937,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.123038170897113e-05,
+      "loss": 0.2854,
+      "step": 6790
+    },
+    {
+      "epoch": 0.43919137118129564,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.1217464315701093e-05,
+      "loss": 0.2496,
+      "step": 6800
+    },
+    {
+      "epoch": 0.43983724084479753,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.1204546922431055e-05,
+      "loss": 0.2726,
+      "step": 6810
+    },
+    {
+      "epoch": 0.4404831105082994,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.1191629529161017e-05,
+      "loss": 0.2489,
+      "step": 6820
+    },
+    {
+      "epoch": 0.44112898017180135,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.117871213589098e-05,
+      "loss": 0.2744,
+      "step": 6830
+    },
+    {
+      "epoch": 0.44177484983530324,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1165794742620938e-05,
+      "loss": 0.2838,
+      "step": 6840
+    },
+    {
+      "epoch": 0.4424207194988051,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.11528773493509e-05,
+      "loss": 0.2332,
+      "step": 6850
+    },
+    {
+      "epoch": 0.44306658916230707,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.1139959956080863e-05,
+      "loss": 0.2553,
+      "step": 6860
+    },
+    {
+      "epoch": 0.44371245882580895,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1127042562810825e-05,
+      "loss": 0.2643,
+      "step": 6870
+    },
+    {
+      "epoch": 0.44435832848931084,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.1114125169540788e-05,
+      "loss": 0.265,
+      "step": 6880
+    },
+    {
+      "epoch": 0.4450041981528128,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.110120777627075e-05,
+      "loss": 0.2502,
+      "step": 6890
+    },
+    {
+      "epoch": 0.44565006781631467,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.108829038300071e-05,
+      "loss": 0.2605,
+      "step": 6900
+    },
+    {
+      "epoch": 0.44629593747981655,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.1075372989730673e-05,
+      "loss": 0.2276,
+      "step": 6910
+    },
+    {
+      "epoch": 0.4469418071433185,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.1062455596460635e-05,
+      "loss": 0.2222,
+      "step": 6920
+    },
+    {
+      "epoch": 0.4475876768068204,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.1049538203190597e-05,
+      "loss": 0.2718,
+      "step": 6930
+    },
+    {
+      "epoch": 0.44823354647032226,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.103662080992056e-05,
+      "loss": 0.244,
+      "step": 6940
+    },
+    {
+      "epoch": 0.4488794161338242,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.1023703416650522e-05,
+      "loss": 0.258,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4495252857973261,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.1010786023380482e-05,
+      "loss": 0.243,
+      "step": 6960
+    },
+    {
+      "epoch": 0.45017115546082803,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.0997868630110445e-05,
+      "loss": 0.2232,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4508170251243299,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0984951236840407e-05,
+      "loss": 0.2422,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4514628947878318,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.097203384357037e-05,
+      "loss": 0.2531,
+      "step": 6990
+    },
+    {
+      "epoch": 0.45210876445133374,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0959116450300332e-05,
+      "loss": 0.2793,
+      "step": 7000
+    },
+    {
+      "epoch": 0.45275463411483563,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.094619905703029e-05,
+      "loss": 0.2557,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4534005037783375,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.0933281663760253e-05,
+      "loss": 0.2386,
+      "step": 7020
+    },
+    {
+      "epoch": 0.45404637344183946,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.0920364270490215e-05,
+      "loss": 0.2314,
+      "step": 7030
+    },
+    {
+      "epoch": 0.45469224310534134,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.0907446877220177e-05,
+      "loss": 0.2889,
+      "step": 7040
+    },
+    {
+      "epoch": 0.45533811276884323,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.089452948395014e-05,
+      "loss": 0.3016,
+      "step": 7050
+    },
+    {
+      "epoch": 0.45598398243234517,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.0881612090680102e-05,
+      "loss": 0.2784,
+      "step": 7060
+    },
+    {
+      "epoch": 0.45662985209584706,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.0868694697410063e-05,
+      "loss": 0.2575,
+      "step": 7070
+    },
+    {
+      "epoch": 0.45727572175934894,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.0855777304140025e-05,
+      "loss": 0.2651,
+      "step": 7080
+    },
+    {
+      "epoch": 0.4579215914228509,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.0842859910869987e-05,
+      "loss": 0.2405,
+      "step": 7090
+    },
+    {
+      "epoch": 0.45856746108635277,
+      "grad_norm": 1.125,
+      "learning_rate": 1.082994251759995e-05,
+      "loss": 0.2219,
+      "step": 7100
+    },
+    {
+      "epoch": 0.45921333074985465,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.0817025124329912e-05,
+      "loss": 0.2152,
+      "step": 7110
+    },
+    {
+      "epoch": 0.4598592004133566,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.0804107731059874e-05,
+      "loss": 0.2447,
+      "step": 7120
+    },
+    {
+      "epoch": 0.4605050700768585,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0791190337789835e-05,
+      "loss": 0.2324,
+      "step": 7130
+    },
+    {
+      "epoch": 0.4611509397403604,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.0778272944519797e-05,
+      "loss": 0.2615,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4617968094038623,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.076535555124976e-05,
+      "loss": 0.2184,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4624426790673642,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.0752438157979721e-05,
+      "loss": 0.2694,
+      "step": 7160
+    },
+    {
+      "epoch": 0.46308854873086613,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.0739520764709684e-05,
+      "loss": 0.2464,
+      "step": 7170
+    },
+    {
+      "epoch": 0.463734418394368,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0726603371439643e-05,
+      "loss": 0.2583,
+      "step": 7180
+    },
+    {
+      "epoch": 0.4643802880578699,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.0713685978169605e-05,
+      "loss": 0.2577,
+      "step": 7190
+    },
+    {
+      "epoch": 0.46502615772137185,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.0700768584899567e-05,
+      "loss": 0.2363,
+      "step": 7200
+    },
+    {
+      "epoch": 0.46567202738487373,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.068785119162953e-05,
+      "loss": 0.2979,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4663178970483756,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.0674933798359492e-05,
+      "loss": 0.2314,
+      "step": 7220
+    },
+    {
+      "epoch": 0.46696376671187756,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.0662016405089454e-05,
+      "loss": 0.2572,
+      "step": 7230
+    },
+    {
+      "epoch": 0.46760963637537944,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0649099011819415e-05,
+      "loss": 0.2777,
+      "step": 7240
+    },
+    {
+      "epoch": 0.46825550603888133,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.0636181618549377e-05,
+      "loss": 0.2364,
+      "step": 7250
+    },
+    {
+      "epoch": 0.46890137570238327,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.062326422527934e-05,
+      "loss": 0.2629,
+      "step": 7260
+    },
+    {
+      "epoch": 0.46954724536588516,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.0610346832009302e-05,
+      "loss": 0.2379,
+      "step": 7270
+    },
+    {
+      "epoch": 0.47019311502938704,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.0597429438739264e-05,
+      "loss": 0.2396,
+      "step": 7280
+    },
+    {
+      "epoch": 0.470838984692889,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.0584512045469226e-05,
+      "loss": 0.2371,
+      "step": 7290
+    },
+    {
+      "epoch": 0.47148485435639087,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.0571594652199187e-05,
+      "loss": 0.2627,
+      "step": 7300
+    },
+    {
+      "epoch": 0.4721307240198928,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.0558677258929149e-05,
+      "loss": 0.2372,
+      "step": 7310
+    },
+    {
+      "epoch": 0.4727765936833947,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.0545759865659111e-05,
+      "loss": 0.2485,
+      "step": 7320
+    },
+    {
+      "epoch": 0.4734224633468966,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.0532842472389074e-05,
+      "loss": 0.256,
+      "step": 7330
+    },
+    {
+      "epoch": 0.4740683330103985,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0519925079119036e-05,
+      "loss": 0.2518,
+      "step": 7340
+    },
+    {
+      "epoch": 0.4747142026739004,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.0507007685848995e-05,
+      "loss": 0.2475,
+      "step": 7350
+    },
+    {
+      "epoch": 0.4753600723374023,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0494090292578957e-05,
+      "loss": 0.2587,
+      "step": 7360
+    },
+    {
+      "epoch": 0.47600594200090424,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.048117289930892e-05,
+      "loss": 0.2741,
+      "step": 7370
+    },
+    {
+      "epoch": 0.4766518116644061,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0468255506038882e-05,
+      "loss": 0.2456,
+      "step": 7380
+    },
+    {
+      "epoch": 0.477297681327908,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.0455338112768844e-05,
+      "loss": 0.2799,
+      "step": 7390
+    },
+    {
+      "epoch": 0.47794355099140995,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.0442420719498806e-05,
+      "loss": 0.2491,
+      "step": 7400
+    },
+    {
+      "epoch": 0.47858942065491183,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.0429503326228767e-05,
+      "loss": 0.2457,
+      "step": 7410
+    },
+    {
+      "epoch": 0.4792352903184137,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.041658593295873e-05,
+      "loss": 0.2494,
+      "step": 7420
+    },
+    {
+      "epoch": 0.47988115998191566,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.0403668539688691e-05,
+      "loss": 0.2613,
+      "step": 7430
+    },
+    {
+      "epoch": 0.48052702964541755,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.0390751146418654e-05,
+      "loss": 0.254,
+      "step": 7440
+    },
+    {
+      "epoch": 0.48117289930891943,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.0377833753148616e-05,
+      "loss": 0.234,
+      "step": 7450
+    },
+    {
+      "epoch": 0.4818187689724214,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.0364916359878577e-05,
+      "loss": 0.256,
+      "step": 7460
+    },
+    {
+      "epoch": 0.48246463863592326,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.0351998966608539e-05,
+      "loss": 0.2439,
+      "step": 7470
+    },
+    {
+      "epoch": 0.4831105082994252,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.0339081573338501e-05,
+      "loss": 0.234,
+      "step": 7480
+    },
+    {
+      "epoch": 0.4837563779629271,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.0326164180068464e-05,
+      "loss": 0.2987,
+      "step": 7490
+    },
+    {
+      "epoch": 0.484402247626429,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.0313246786798426e-05,
+      "loss": 0.2478,
+      "step": 7500
+    },
+    {
+      "epoch": 0.4850481172899309,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.0300329393528388e-05,
+      "loss": 0.2426,
+      "step": 7510
+    },
+    {
+      "epoch": 0.4856939869534328,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.0287412000258347e-05,
+      "loss": 0.2623,
+      "step": 7520
+    },
+    {
+      "epoch": 0.4863398566169347,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.027449460698831e-05,
+      "loss": 0.2589,
+      "step": 7530
+    },
+    {
+      "epoch": 0.4869857262804366,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.0261577213718272e-05,
+      "loss": 0.2339,
+      "step": 7540
+    },
+    {
+      "epoch": 0.4876315959439385,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.0248659820448234e-05,
+      "loss": 0.2472,
+      "step": 7550
+    },
+    {
+      "epoch": 0.4882774656074404,
+      "grad_norm": 1.375,
+      "learning_rate": 1.0235742427178196e-05,
+      "loss": 0.2423,
+      "step": 7560
+    },
+    {
+      "epoch": 0.48892333527094234,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.0222825033908158e-05,
+      "loss": 0.2387,
+      "step": 7570
+    },
+    {
+      "epoch": 0.4895692049344442,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.0209907640638119e-05,
+      "loss": 0.2336,
+      "step": 7580
+    },
+    {
+      "epoch": 0.4902150745979461,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.0196990247368081e-05,
+      "loss": 0.2365,
+      "step": 7590
+    },
+    {
+      "epoch": 0.49086094426144805,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.0184072854098044e-05,
+      "loss": 0.2368,
+      "step": 7600
+    },
+    {
+      "epoch": 0.49150681392494994,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.0171155460828006e-05,
+      "loss": 0.2262,
+      "step": 7610
+    },
+    {
+      "epoch": 0.4921526835884519,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.0158238067557968e-05,
+      "loss": 0.2606,
+      "step": 7620
+    },
+    {
+      "epoch": 0.49279855325195376,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0145320674287929e-05,
+      "loss": 0.2414,
+      "step": 7630
+    },
+    {
+      "epoch": 0.49344442291545565,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.0132403281017891e-05,
+      "loss": 0.244,
+      "step": 7640
+    },
+    {
+      "epoch": 0.4940902925789576,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.0119485887747853e-05,
+      "loss": 0.247,
+      "step": 7650
+    },
+    {
+      "epoch": 0.4947361622424595,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.0106568494477816e-05,
+      "loss": 0.2467,
+      "step": 7660
+    },
+    {
+      "epoch": 0.49538203190596136,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.0093651101207778e-05,
+      "loss": 0.2655,
+      "step": 7670
+    },
+    {
+      "epoch": 0.4960279015694633,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.008073370793774e-05,
+      "loss": 0.2706,
+      "step": 7680
+    },
+    {
+      "epoch": 0.4966737712329652,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.00678163146677e-05,
+      "loss": 0.2847,
+      "step": 7690
+    },
+    {
+      "epoch": 0.4973196408964671,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.0054898921397662e-05,
+      "loss": 0.2563,
+      "step": 7700
+    },
+    {
+      "epoch": 0.497965510559969,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.0041981528127624e-05,
+      "loss": 0.2613,
+      "step": 7710
+    },
+    {
+      "epoch": 0.4986113802234709,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.0029064134857586e-05,
+      "loss": 0.2822,
+      "step": 7720
+    },
+    {
+      "epoch": 0.4992572498869728,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.0016146741587548e-05,
+      "loss": 0.2212,
+      "step": 7730
+    },
+    {
+      "epoch": 0.49990311955047473,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.000322934831751e-05,
+      "loss": 0.2724,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5005489892139766,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.990311955047473e-06,
+      "loss": 0.1998,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5011948588774785,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.977394561777434e-06,
+      "loss": 0.2457,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5018407285409804,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.964477168507396e-06,
+      "loss": 0.275,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5024865982044824,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.951559775237358e-06,
+      "loss": 0.2456,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5031324678679843,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.938642381967319e-06,
+      "loss": 0.2707,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5037783375314862,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.925724988697281e-06,
+      "loss": 0.2261,
+      "step": 7800
+    },
+    {
+      "epoch": 0.504424207194988,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.912807595427243e-06,
+      "loss": 0.2636,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5050700768584899,
+      "grad_norm": 1.25,
+      "learning_rate": 9.899890202157206e-06,
+      "loss": 0.274,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5057159465219918,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.886972808887168e-06,
+      "loss": 0.2692,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5063618161854938,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.87405541561713e-06,
+      "loss": 0.2192,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5070076858489957,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.86113802234709e-06,
+      "loss": 0.2687,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5076535555124976,
+      "grad_norm": 1.25,
+      "learning_rate": 9.848220629077053e-06,
+      "loss": 0.2762,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5082994251759995,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.835303235807014e-06,
+      "loss": 0.2624,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5089452948395013,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.822385842536976e-06,
+      "loss": 0.2207,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5095911645030033,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.809468449266938e-06,
+      "loss": 0.2796,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5102370341665052,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.7965510559969e-06,
+      "loss": 0.2686,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5108829038300071,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.783633662726863e-06,
+      "loss": 0.2711,
+      "step": 7910
+    },
+    {
+      "epoch": 0.511528773493509,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.770716269456825e-06,
+      "loss": 0.2388,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5121746431570109,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.757798876186786e-06,
+      "loss": 0.266,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5128205128205128,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.744881482916748e-06,
+      "loss": 0.2336,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5134663824840148,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.73196408964671e-06,
+      "loss": 0.2289,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5141122521475167,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.719046696376671e-06,
+      "loss": 0.2497,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5147581218110185,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.706129303106633e-06,
+      "loss": 0.2412,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5154039914745204,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.693211909836595e-06,
+      "loss": 0.2535,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5160498611380223,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.680294516566558e-06,
+      "loss": 0.2646,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5166957308015242,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.66737712329652e-06,
+      "loss": 0.2541,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5173416004650262,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.654459730026482e-06,
+      "loss": 0.2489,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5179874701285281,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.641542336756443e-06,
+      "loss": 0.2325,
+      "step": 8020
+    },
+    {
+      "epoch": 0.51863333979203,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.628624943486405e-06,
+      "loss": 0.2153,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5192792094555319,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.615707550216366e-06,
+      "loss": 0.2447,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5199250791190337,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.602790156946328e-06,
+      "loss": 0.2733,
+      "step": 8050
+    },
+    {
+      "epoch": 0.5205709487825357,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.58987276367629e-06,
+      "loss": 0.2482,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5212168184460376,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.576955370406253e-06,
+      "loss": 0.2449,
+      "step": 8070
+    },
+    {
+      "epoch": 0.5218626881095395,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.564037977136215e-06,
+      "loss": 0.2589,
+      "step": 8080
+    },
+    {
+      "epoch": 0.5225085577730414,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.551120583866177e-06,
+      "loss": 0.2875,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5231544274365433,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.538203190596138e-06,
+      "loss": 0.2275,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5238002971000452,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.5252857973261e-06,
+      "loss": 0.271,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5244461667635472,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.512368404056062e-06,
+      "loss": 0.3096,
+      "step": 8120
+    },
+    {
+      "epoch": 0.525092036427049,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.499451010786023e-06,
+      "loss": 0.2679,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5257379060905509,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.486533617515985e-06,
+      "loss": 0.266,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5263837757540528,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.473616224245948e-06,
+      "loss": 0.2349,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5270296454175547,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.46069883097591e-06,
+      "loss": 0.2319,
+      "step": 8160
+    },
+    {
+      "epoch": 0.5276755150810566,
+      "grad_norm": 1.25,
+      "learning_rate": 9.447781437705872e-06,
+      "loss": 0.2353,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5283213847445586,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.434864044435835e-06,
+      "loss": 0.246,
+      "step": 8180
+    },
+    {
+      "epoch": 0.5289672544080605,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.421946651165795e-06,
+      "loss": 0.2627,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5296131240715624,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.409029257895757e-06,
+      "loss": 0.2387,
+      "step": 8200
+    },
+    {
+      "epoch": 0.5302589937350642,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.396111864625718e-06,
+      "loss": 0.2612,
+      "step": 8210
+    },
+    {
+      "epoch": 0.5309048633985661,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.38319447135568e-06,
+      "loss": 0.2385,
+      "step": 8220
+    },
+    {
+      "epoch": 0.5315507330620681,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.370277078085643e-06,
+      "loss": 0.2664,
+      "step": 8230
+    },
+    {
+      "epoch": 0.53219660272557,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.357359684815605e-06,
+      "loss": 0.2554,
+      "step": 8240
+    },
+    {
+      "epoch": 0.5328424723890719,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.344442291545567e-06,
+      "loss": 0.2444,
+      "step": 8250
+    },
+    {
+      "epoch": 0.5334883420525738,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.33152489827553e-06,
+      "loss": 0.224,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5341342117160757,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.31860750500549e-06,
+      "loss": 0.2496,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5347800813795776,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.305690111735452e-06,
+      "loss": 0.2633,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5354259510430796,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.292772718465415e-06,
+      "loss": 0.2992,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5360718207065814,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.279855325195375e-06,
+      "loss": 0.2912,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5367176903700833,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.266937931925338e-06,
+      "loss": 0.2553,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5373635600335852,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.2540205386553e-06,
+      "loss": 0.2505,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5380094296970871,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.241103145385262e-06,
+      "loss": 0.268,
+      "step": 8330
+    },
+    {
+      "epoch": 0.538655299360589,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.228185752115224e-06,
+      "loss": 0.2382,
+      "step": 8340
+    },
+    {
+      "epoch": 0.539301169024091,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.215268358845187e-06,
+      "loss": 0.2399,
+      "step": 8350
+    },
+    {
+      "epoch": 0.5399470386875929,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.202350965575147e-06,
+      "loss": 0.2322,
+      "step": 8360
+    },
+    {
+      "epoch": 0.5405929083510947,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.18943357230511e-06,
+      "loss": 0.2495,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5412387780145966,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.17651617903507e-06,
+      "loss": 0.2474,
+      "step": 8380
+    },
+    {
+      "epoch": 0.5418846476780985,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.163598785765032e-06,
+      "loss": 0.2348,
+      "step": 8390
+    },
+    {
+      "epoch": 0.5425305173416005,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.150681392494995e-06,
+      "loss": 0.2457,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5431763870051024,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.137763999224957e-06,
+      "loss": 0.2355,
+      "step": 8410
+    },
+    {
+      "epoch": 0.5438222566686043,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.12484660595492e-06,
+      "loss": 0.2834,
+      "step": 8420
+    },
+    {
+      "epoch": 0.5444681263321062,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.111929212684882e-06,
+      "loss": 0.2525,
+      "step": 8430
+    },
+    {
+      "epoch": 0.5451139959956081,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.099011819414842e-06,
+      "loss": 0.2204,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5457598656591099,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.086094426144805e-06,
+      "loss": 0.2495,
+      "step": 8450
+    },
+    {
+      "epoch": 0.5464057353226119,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.073177032874767e-06,
+      "loss": 0.2312,
+      "step": 8460
+    },
+    {
+      "epoch": 0.5470516049861138,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.060259639604727e-06,
+      "loss": 0.2916,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5476974746496157,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.04734224633469e-06,
+      "loss": 0.2585,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5483433443131176,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.034424853064652e-06,
+      "loss": 0.242,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5489892139766195,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.021507459794614e-06,
+      "loss": 0.2534,
+      "step": 8500
+    },
+    {
+      "epoch": 0.5496350836401214,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.008590066524577e-06,
+      "loss": 0.2025,
+      "step": 8510
+    },
+    {
+      "epoch": 0.5502809533036234,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.995672673254539e-06,
+      "loss": 0.2508,
+      "step": 8520
+    },
+    {
+      "epoch": 0.5509268229671253,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.9827552799845e-06,
+      "loss": 0.2332,
+      "step": 8530
+    },
+    {
+      "epoch": 0.5515726926306271,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.969837886714462e-06,
+      "loss": 0.2065,
+      "step": 8540
+    },
+    {
+      "epoch": 0.552218562294129,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.956920493444422e-06,
+      "loss": 0.2772,
+      "step": 8550
+    },
+    {
+      "epoch": 0.5528644319576309,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.944003100174385e-06,
+      "loss": 0.234,
+      "step": 8560
+    },
+    {
+      "epoch": 0.5535103016211329,
+      "grad_norm": 1.0,
+      "learning_rate": 8.931085706904347e-06,
+      "loss": 0.2368,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5541561712846348,
+      "grad_norm": 1.3828125,
+      "learning_rate": 8.91816831363431e-06,
+      "loss": 0.2195,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5548020409481367,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.905250920364272e-06,
+      "loss": 0.2484,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5554479106116386,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.892333527094234e-06,
+      "loss": 0.2895,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5560937802751404,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.879416133824194e-06,
+      "loss": 0.2546,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5567396499386423,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.866498740554157e-06,
+      "loss": 0.2201,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5573855196021443,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.853581347284119e-06,
+      "loss": 0.2487,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5580313892656462,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.84066395401408e-06,
+      "loss": 0.2372,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5586772589291481,
+      "grad_norm": 1.125,
+      "learning_rate": 8.827746560744042e-06,
+      "loss": 0.2551,
+      "step": 8650
+    },
+    {
+      "epoch": 0.55932312859265,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.814829167474004e-06,
+      "loss": 0.244,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5599689982561519,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.801911774203966e-06,
+      "loss": 0.2379,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5606148679196538,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.788994380933929e-06,
+      "loss": 0.2401,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5612607375831558,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.77607698766389e-06,
+      "loss": 0.2448,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5619066072466576,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.763159594393852e-06,
+      "loss": 0.25,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5625524769101595,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.750242201123814e-06,
+      "loss": 0.2246,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5631983465736614,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.737324807853775e-06,
+      "loss": 0.2884,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5638442162371633,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.724407414583737e-06,
+      "loss": 0.2578,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5644900859006653,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.711490021313699e-06,
+      "loss": 0.2126,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5651359555641672,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.698572628043661e-06,
+      "loss": 0.2831,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5657818252276691,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.685655234773624e-06,
+      "loss": 0.2257,
+      "step": 8760
+    },
+    {
+      "epoch": 0.566427694891171,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.672737841503586e-06,
+      "loss": 0.2834,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5670735645546728,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.659820448233547e-06,
+      "loss": 0.2592,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5677194342181747,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.646903054963509e-06,
+      "loss": 0.2134,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5683653038816767,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.633985661693471e-06,
+      "loss": 0.2348,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5690111735451786,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.621068268423432e-06,
+      "loss": 0.2694,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5696570432086805,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.608150875153394e-06,
+      "loss": 0.2366,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5703029128721824,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.595233481883356e-06,
+      "loss": 0.2335,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5709487825356843,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.582316088613319e-06,
+      "loss": 0.2394,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5715946521991863,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.569398695343281e-06,
+      "loss": 0.2623,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5722405218626881,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.556481302073242e-06,
+      "loss": 0.2294,
+      "step": 8860
+    },
+    {
+      "epoch": 0.57288639152619,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.543563908803204e-06,
+      "loss": 0.2336,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5735322611896919,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.530646515533166e-06,
+      "loss": 0.2838,
+      "step": 8880
+    },
+    {
+      "epoch": 0.5741781308531938,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.517729122263127e-06,
+      "loss": 0.2696,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5748240005166957,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.504811728993089e-06,
+      "loss": 0.2556,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5754698701801977,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.491894335723051e-06,
+      "loss": 0.2377,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5761157398436996,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.478976942453014e-06,
+      "loss": 0.2368,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5767616095072015,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.466059549182976e-06,
+      "loss": 0.2255,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5774074791707033,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.453142155912938e-06,
+      "loss": 0.2601,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5780533488342052,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.440224762642899e-06,
+      "loss": 0.2555,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5786992184977071,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.427307369372861e-06,
+      "loss": 0.2149,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5793450881612091,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.414389976102823e-06,
+      "loss": 0.2565,
+      "step": 8970
+    },
+    {
+      "epoch": 0.579990957824711,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.401472582832784e-06,
+      "loss": 0.2314,
+      "step": 8980
+    },
+    {
+      "epoch": 0.5806368274882129,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.388555189562746e-06,
+      "loss": 0.2334,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5812826971517148,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.375637796292709e-06,
+      "loss": 0.2555,
+      "step": 9000
+    },
+    {
+      "epoch": 0.5819285668152167,
+      "grad_norm": 1.25,
+      "learning_rate": 8.36272040302267e-06,
+      "loss": 0.2389,
+      "step": 9010
+    },
+    {
+      "epoch": 0.5825744364787186,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.349803009752633e-06,
+      "loss": 0.2529,
+      "step": 9020
+    },
+    {
+      "epoch": 0.5832203061422205,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.336885616482594e-06,
+      "loss": 0.258,
+      "step": 9030
+    },
+    {
+      "epoch": 0.5838661758057224,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.323968223212556e-06,
+      "loss": 0.2563,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5845120454692243,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.311050829942518e-06,
+      "loss": 0.2306,
+      "step": 9050
+    },
+    {
+      "epoch": 0.5851579151327262,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.298133436672479e-06,
+      "loss": 0.2295,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5858037847962281,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.285216043402441e-06,
+      "loss": 0.2589,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5864496544597301,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.272298650132403e-06,
+      "loss": 0.2552,
+      "step": 9080
+    },
+    {
+      "epoch": 0.587095524123232,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.259381256862366e-06,
+      "loss": 0.2173,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5877413937867338,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.246463863592328e-06,
+      "loss": 0.235,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5883872634502357,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.23354647032229e-06,
+      "loss": 0.2277,
+      "step": 9110
+    },
+    {
+      "epoch": 0.5890331331137376,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.220629077052251e-06,
+      "loss": 0.1985,
+      "step": 9120
+    },
+    {
+      "epoch": 0.5896790027772395,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.207711683782213e-06,
+      "loss": 0.2056,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5903248724407415,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.194794290512176e-06,
+      "loss": 0.2722,
+      "step": 9140
+    },
+    {
+      "epoch": 0.5909707421042434,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.181876897242136e-06,
+      "loss": 0.2612,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5916166117677453,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.168959503972098e-06,
+      "loss": 0.2506,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5922624814312472,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.15604211070206e-06,
+      "loss": 0.2495,
+      "step": 9170
+    },
+    {
+      "epoch": 0.592908351094749,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.143124717432023e-06,
+      "loss": 0.2629,
+      "step": 9180
+    },
+    {
+      "epoch": 0.593554220758251,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.130207324161985e-06,
+      "loss": 0.257,
+      "step": 9190
+    },
+    {
+      "epoch": 0.5942000904217529,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.117289930891946e-06,
+      "loss": 0.2712,
+      "step": 9200
+    },
+    {
+      "epoch": 0.5948459600852548,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.104372537621908e-06,
+      "loss": 0.2282,
+      "step": 9210
+    },
+    {
+      "epoch": 0.5954918297487567,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.09145514435187e-06,
+      "loss": 0.2401,
+      "step": 9220
+    },
+    {
+      "epoch": 0.5961376994122586,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.078537751081831e-06,
+      "loss": 0.2296,
+      "step": 9230
+    },
+    {
+      "epoch": 0.5967835690757605,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.065620357811793e-06,
+      "loss": 0.2449,
+      "step": 9240
+    },
+    {
+      "epoch": 0.5974294387392625,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.052702964541756e-06,
+      "loss": 0.2533,
+      "step": 9250
+    },
+    {
+      "epoch": 0.5980753084027643,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.039785571271718e-06,
+      "loss": 0.2334,
+      "step": 9260
+    },
+    {
+      "epoch": 0.5987211780662662,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.02686817800168e-06,
+      "loss": 0.2295,
+      "step": 9270
+    },
+    {
+      "epoch": 0.5993670477297681,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.013950784731643e-06,
+      "loss": 0.252,
+      "step": 9280
+    },
+    {
+      "epoch": 0.60001291739327,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.001033391461603e-06,
+      "loss": 0.2393,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6006587870567719,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.988115998191565e-06,
+      "loss": 0.2367,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6013046567202739,
+      "grad_norm": 1.125,
+      "learning_rate": 7.975198604921528e-06,
+      "loss": 0.2473,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6019505263837758,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.962281211651488e-06,
+      "loss": 0.2641,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6025963960472777,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.94936381838145e-06,
+      "loss": 0.2767,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6032422657107795,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.936446425111413e-06,
+      "loss": 0.2438,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6038881353742814,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.923529031841375e-06,
+      "loss": 0.2143,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6045340050377834,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.910611638571337e-06,
+      "loss": 0.2204,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6051798747012853,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.897694245301298e-06,
+      "loss": 0.2618,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6058257443647872,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.88477685203126e-06,
+      "loss": 0.2456,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6064716140282891,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.871859458761223e-06,
+      "loss": 0.2472,
+      "step": 9390
+    },
+    {
+      "epoch": 0.607117483691791,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.858942065491185e-06,
+      "loss": 0.2297,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6077633533552929,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.846024672221146e-06,
+      "loss": 0.2648,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6084092230187949,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.833107278951108e-06,
+      "loss": 0.1953,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6090550926822967,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.82018988568107e-06,
+      "loss": 0.2508,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6097009623457986,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.807272492411032e-06,
+      "loss": 0.2198,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6103468320093005,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.794355099140995e-06,
+      "loss": 0.2324,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6109927016728024,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.781437705870955e-06,
+      "loss": 0.2298,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6116385713363043,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.768520312600918e-06,
+      "loss": 0.2689,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6122844409998063,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.75560291933088e-06,
+      "loss": 0.2756,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6129303106633082,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.742685526060842e-06,
+      "loss": 0.233,
+      "step": 9490
+    },
+    {
+      "epoch": 0.61357618032681,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.729768132790804e-06,
+      "loss": 0.2231,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6142220499903119,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.716850739520765e-06,
+      "loss": 0.2582,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6148679196538138,
+      "grad_norm": 1.0,
+      "learning_rate": 7.703933346250727e-06,
+      "loss": 0.2592,
+      "step": 9520
+    },
+    {
+      "epoch": 0.6155137893173158,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.69101595298069e-06,
+      "loss": 0.2449,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6161596589808177,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.67809855971065e-06,
+      "loss": 0.2509,
+      "step": 9540
+    },
+    {
+      "epoch": 0.6168055286443196,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.665181166440613e-06,
+      "loss": 0.225,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6174513983078215,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.652263773170575e-06,
+      "loss": 0.2244,
+      "step": 9560
+    },
+    {
+      "epoch": 0.6180972679713234,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.639346379900537e-06,
+      "loss": 0.2768,
+      "step": 9570
+    },
+    {
+      "epoch": 0.6187431376348252,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.6264289866304986e-06,
+      "loss": 0.2491,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6193890072983272,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.613511593360461e-06,
+      "loss": 0.2623,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6200348769618291,
+      "grad_norm": 1.125,
+      "learning_rate": 7.600594200090422e-06,
+      "loss": 0.2585,
+      "step": 9600
+    },
+    {
+      "epoch": 0.620680746625331,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.587676806820385e-06,
+      "loss": 0.2312,
+      "step": 9610
+    },
+    {
+      "epoch": 0.6213266162888329,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.574759413550347e-06,
+      "loss": 0.2359,
+      "step": 9620
+    },
+    {
+      "epoch": 0.6219724859523348,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.5618420202803075e-06,
+      "loss": 0.2409,
+      "step": 9630
+    },
+    {
+      "epoch": 0.6226183556158367,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.54892462701027e-06,
+      "loss": 0.2221,
+      "step": 9640
+    },
+    {
+      "epoch": 0.6232642252793387,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.536007233740232e-06,
+      "loss": 0.2712,
+      "step": 9650
+    },
+    {
+      "epoch": 0.6239100949428406,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.5230898404701935e-06,
+      "loss": 0.2658,
+      "step": 9660
+    },
+    {
+      "epoch": 0.6245559646063424,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.510172447200156e-06,
+      "loss": 0.2249,
+      "step": 9670
+    },
+    {
+      "epoch": 0.6252018342698443,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.497255053930117e-06,
+      "loss": 0.2309,
+      "step": 9680
+    },
+    {
+      "epoch": 0.6258477039333462,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.4843376606600795e-06,
+      "loss": 0.2182,
+      "step": 9690
+    },
+    {
+      "epoch": 0.6264935735968482,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.471420267390042e-06,
+      "loss": 0.2291,
+      "step": 9700
+    },
+    {
+      "epoch": 0.6271394432603501,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.458502874120002e-06,
+      "loss": 0.2266,
+      "step": 9710
+    },
+    {
+      "epoch": 0.627785312923852,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.445585480849965e-06,
+      "loss": 0.2655,
+      "step": 9720
+    },
+    {
+      "epoch": 0.6284311825873539,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.432668087579927e-06,
+      "loss": 0.2227,
+      "step": 9730
+    },
+    {
+      "epoch": 0.6290770522508558,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.4197506943098884e-06,
+      "loss": 0.2341,
+      "step": 9740
+    },
+    {
+      "epoch": 0.6297229219143576,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.406833301039851e-06,
+      "loss": 0.2642,
+      "step": 9750
+    },
+    {
+      "epoch": 0.6303687915778596,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.393915907769813e-06,
+      "loss": 0.2253,
+      "step": 9760
+    },
+    {
+      "epoch": 0.6310146612413615,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.3809985144997745e-06,
+      "loss": 0.2435,
+      "step": 9770
+    },
+    {
+      "epoch": 0.6316605309048634,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.368081121229737e-06,
+      "loss": 0.2608,
+      "step": 9780
+    },
+    {
+      "epoch": 0.6323064005683653,
+      "grad_norm": 1.421875,
+      "learning_rate": 7.355163727959699e-06,
+      "loss": 0.2654,
+      "step": 9790
+    },
+    {
+      "epoch": 0.6329522702318672,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.34224633468966e-06,
+      "loss": 0.2321,
+      "step": 9800
+    },
+    {
+      "epoch": 0.6335981398953691,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.329328941419622e-06,
+      "loss": 0.2545,
+      "step": 9810
+    },
+    {
+      "epoch": 0.6342440095588711,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.316411548149584e-06,
+      "loss": 0.2016,
+      "step": 9820
+    },
+    {
+      "epoch": 0.6348898792223729,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.303494154879546e-06,
+      "loss": 0.2529,
+      "step": 9830
+    },
+    {
+      "epoch": 0.6355357488858748,
+      "grad_norm": 1.0,
+      "learning_rate": 7.290576761609508e-06,
+      "loss": 0.2683,
+      "step": 9840
+    },
+    {
+      "epoch": 0.6361816185493767,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.277659368339469e-06,
+      "loss": 0.2553,
+      "step": 9850
+    },
+    {
+      "epoch": 0.6368274882128786,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.264741975069432e-06,
+      "loss": 0.2363,
+      "step": 9860
+    },
+    {
+      "epoch": 0.6374733578763806,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.251824581799394e-06,
+      "loss": 0.2267,
+      "step": 9870
+    },
+    {
+      "epoch": 0.6381192275398825,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.238907188529355e-06,
+      "loss": 0.241,
+      "step": 9880
+    },
+    {
+      "epoch": 0.6387650972033844,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.225989795259317e-06,
+      "loss": 0.2146,
+      "step": 9890
+    },
+    {
+      "epoch": 0.6394109668668863,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.213072401989279e-06,
+      "loss": 0.261,
+      "step": 9900
+    },
+    {
+      "epoch": 0.6400568365303881,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.200155008719241e-06,
+      "loss": 0.2299,
+      "step": 9910
+    },
+    {
+      "epoch": 0.64070270619389,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.187237615449203e-06,
+      "loss": 0.2633,
+      "step": 9920
+    },
+    {
+      "epoch": 0.641348575857392,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.174320222179165e-06,
+      "loss": 0.2379,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6419944455208939,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.161402828909127e-06,
+      "loss": 0.2405,
+      "step": 9940
+    },
+    {
+      "epoch": 0.6426403151843958,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.148485435639089e-06,
+      "loss": 0.2508,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6432861848478977,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.135568042369051e-06,
+      "loss": 0.24,
+      "step": 9960
+    },
+    {
+      "epoch": 0.6439320545113996,
+      "grad_norm": 1.125,
+      "learning_rate": 7.122650649099012e-06,
+      "loss": 0.2258,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6445779241749015,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.109733255828974e-06,
+      "loss": 0.2058,
+      "step": 9980
+    },
+    {
+      "epoch": 0.6452237938384034,
+      "grad_norm": 1.25,
+      "learning_rate": 7.096815862558936e-06,
+      "loss": 0.2515,
+      "step": 9990
+    },
+    {
+      "epoch": 0.6458696635019053,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.083898469288898e-06,
+      "loss": 0.269,
+      "step": 10000
+    },
+    {
+      "epoch": 0.6465155331654072,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.07098107601886e-06,
+      "loss": 0.2313,
+      "step": 10010
+    },
+    {
+      "epoch": 0.6471614028289091,
+      "grad_norm": 1.125,
+      "learning_rate": 7.058063682748822e-06,
+      "loss": 0.2284,
+      "step": 10020
+    },
+    {
+      "epoch": 0.647807272492411,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.045146289478784e-06,
+      "loss": 0.2329,
+      "step": 10030
+    },
+    {
+      "epoch": 0.648453142155913,
+      "grad_norm": 1.3828125,
+      "learning_rate": 7.032228896208746e-06,
+      "loss": 0.2498,
+      "step": 10040
+    },
+    {
+      "epoch": 0.6490990118194149,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.019311502938707e-06,
+      "loss": 0.2136,
+      "step": 10050
+    },
+    {
+      "epoch": 0.6497448814829168,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.006394109668669e-06,
+      "loss": 0.2427,
+      "step": 10060
+    },
+    {
+      "epoch": 0.6503907511464186,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.993476716398631e-06,
+      "loss": 0.2034,
+      "step": 10070
+    },
+    {
+      "epoch": 0.6510366208099205,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.980559323128593e-06,
+      "loss": 0.273,
+      "step": 10080
+    },
+    {
+      "epoch": 0.6516824904734224,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.967641929858555e-06,
+      "loss": 0.254,
+      "step": 10090
+    },
+    {
+      "epoch": 0.6523283601369244,
+      "grad_norm": 1.3203125,
+      "learning_rate": 6.954724536588517e-06,
+      "loss": 0.2433,
+      "step": 10100
+    },
+    {
+      "epoch": 0.6529742298004263,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.941807143318479e-06,
+      "loss": 0.2463,
+      "step": 10110
+    },
+    {
+      "epoch": 0.6536200994639282,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.928889750048441e-06,
+      "loss": 0.2632,
+      "step": 10120
+    },
+    {
+      "epoch": 0.6542659691274301,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.915972356778403e-06,
+      "loss": 0.2612,
+      "step": 10130
+    },
+    {
+      "epoch": 0.654911838790932,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.903054963508364e-06,
+      "loss": 0.2633,
+      "step": 10140
+    },
+    {
+      "epoch": 0.6555577084544338,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.890137570238326e-06,
+      "loss": 0.2134,
+      "step": 10150
+    },
+    {
+      "epoch": 0.6562035781179358,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.877220176968288e-06,
+      "loss": 0.2409,
+      "step": 10160
+    },
+    {
+      "epoch": 0.6568494477814377,
+      "grad_norm": 1.328125,
+      "learning_rate": 6.86430278369825e-06,
+      "loss": 0.2656,
+      "step": 10170
+    },
+    {
+      "epoch": 0.6574953174449396,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.851385390428212e-06,
+      "loss": 0.2124,
+      "step": 10180
+    },
+    {
+      "epoch": 0.6581411871084415,
+      "grad_norm": 1.359375,
+      "learning_rate": 6.838467997158174e-06,
+      "loss": 0.2451,
+      "step": 10190
+    },
+    {
+      "epoch": 0.6587870567719434,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.825550603888136e-06,
+      "loss": 0.2633,
+      "step": 10200
+    },
+    {
+      "epoch": 0.6594329264354454,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.812633210618098e-06,
+      "loss": 0.2689,
+      "step": 10210
+    },
+    {
+      "epoch": 0.6600787960989473,
+      "grad_norm": 1.125,
+      "learning_rate": 6.799715817348059e-06,
+      "loss": 0.2153,
+      "step": 10220
+    },
+    {
+      "epoch": 0.6607246657624491,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.786798424078021e-06,
+      "loss": 0.1888,
+      "step": 10230
+    },
+    {
+      "epoch": 0.661370535425951,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.7738810308079835e-06,
+      "loss": 0.2652,
+      "step": 10240
+    },
+    {
+      "epoch": 0.6620164050894529,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.760963637537945e-06,
+      "loss": 0.2182,
+      "step": 10250
+    },
+    {
+      "epoch": 0.6626622747529548,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.748046244267907e-06,
+      "loss": 0.2141,
+      "step": 10260
+    },
+    {
+      "epoch": 0.6633081444164568,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.7351288509978696e-06,
+      "loss": 0.2405,
+      "step": 10270
+    },
+    {
+      "epoch": 0.6639540140799587,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.722211457727831e-06,
+      "loss": 0.2127,
+      "step": 10280
+    },
+    {
+      "epoch": 0.6645998837434606,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.709294064457793e-06,
+      "loss": 0.2512,
+      "step": 10290
+    },
+    {
+      "epoch": 0.6652457534069625,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.696376671187756e-06,
+      "loss": 0.2451,
+      "step": 10300
+    },
+    {
+      "epoch": 0.6658916230704643,
+      "grad_norm": 1.375,
+      "learning_rate": 6.683459277917716e-06,
+      "loss": 0.2546,
+      "step": 10310
+    },
+    {
+      "epoch": 0.6665374927339662,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.6705418846476785e-06,
+      "loss": 0.208,
+      "step": 10320
+    },
+    {
+      "epoch": 0.6671833623974682,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.65762449137764e-06,
+      "loss": 0.2314,
+      "step": 10330
+    },
+    {
+      "epoch": 0.6678292320609701,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.644707098107602e-06,
+      "loss": 0.2281,
+      "step": 10340
+    },
+    {
+      "epoch": 0.668475101724472,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.6317897048375645e-06,
+      "loss": 0.2652,
+      "step": 10350
+    },
+    {
+      "epoch": 0.6691209713879739,
+      "grad_norm": 1.125,
+      "learning_rate": 6.618872311567526e-06,
+      "loss": 0.215,
+      "step": 10360
+    },
+    {
+      "epoch": 0.6697668410514758,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.605954918297488e-06,
+      "loss": 0.2365,
+      "step": 10370
+    },
+    {
+      "epoch": 0.6704127107149778,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.5930375250274505e-06,
+      "loss": 0.2547,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6710585803784797,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.580120131757411e-06,
+      "loss": 0.2333,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6717044500419815,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.567202738487373e-06,
+      "loss": 0.259,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6723503197054834,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.554285345217336e-06,
+      "loss": 0.2513,
+      "step": 10410
+    },
+    {
+      "epoch": 0.6729961893689853,
+      "grad_norm": 1.3203125,
+      "learning_rate": 6.541367951947297e-06,
+      "loss": 0.2835,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6736420590324872,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.5284505586772594e-06,
+      "loss": 0.2441,
+      "step": 10430
+    },
+    {
+      "epoch": 0.6742879286959892,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.515533165407222e-06,
+      "loss": 0.229,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6749337983594911,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.502615772137183e-06,
+      "loss": 0.214,
+      "step": 10450
+    },
+    {
+      "epoch": 0.675579668022993,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.4896983788671455e-06,
+      "loss": 0.2228,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6762255376864948,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.476780985597108e-06,
+      "loss": 0.2207,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6768714073499967,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.463863592327068e-06,
+      "loss": 0.2385,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6775172770134987,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.450946199057031e-06,
+      "loss": 0.2337,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6781631466770006,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.438028805786992e-06,
+      "loss": 0.2325,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6788090163405025,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.425111412516954e-06,
+      "loss": 0.2066,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6794548860040044,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.412194019246917e-06,
+      "loss": 0.2501,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6801007556675063,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.399276625976878e-06,
+      "loss": 0.2366,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6807466253310082,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.38635923270684e-06,
+      "loss": 0.2267,
+      "step": 10540
+    },
+    {
+      "epoch": 0.6813924949945102,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.373441839436803e-06,
+      "loss": 0.2168,
+      "step": 10550
+    },
+    {
+      "epoch": 0.682038364658012,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.360524446166763e-06,
+      "loss": 0.2523,
+      "step": 10560
+    },
+    {
+      "epoch": 0.6826842343215139,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.347607052896726e-06,
+      "loss": 0.2334,
+      "step": 10570
+    },
+    {
+      "epoch": 0.6833301039850158,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.334689659626688e-06,
+      "loss": 0.2319,
+      "step": 10580
+    },
+    {
+      "epoch": 0.6839759736485177,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.321772266356649e-06,
+      "loss": 0.2315,
+      "step": 10590
+    },
+    {
+      "epoch": 0.6846218433120196,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.308854873086612e-06,
+      "loss": 0.2355,
+      "step": 10600
+    },
+    {
+      "epoch": 0.6852677129755216,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.295937479816574e-06,
+      "loss": 0.2184,
+      "step": 10610
+    },
+    {
+      "epoch": 0.6859135826390235,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.283020086546535e-06,
+      "loss": 0.2648,
+      "step": 10620
+    },
+    {
+      "epoch": 0.6865594523025254,
+      "grad_norm": 1.3359375,
+      "learning_rate": 6.270102693276498e-06,
+      "loss": 0.2613,
+      "step": 10630
+    },
+    {
+      "epoch": 0.6872053219660272,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.25718530000646e-06,
+      "loss": 0.2433,
+      "step": 10640
+    },
+    {
+      "epoch": 0.6878511916295291,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.2442679067364205e-06,
+      "loss": 0.214,
+      "step": 10650
+    },
+    {
+      "epoch": 0.6884970612930311,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.231350513466383e-06,
+      "loss": 0.2556,
+      "step": 10660
+    },
+    {
+      "epoch": 0.689142930956533,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.218433120196344e-06,
+      "loss": 0.252,
+      "step": 10670
+    },
+    {
+      "epoch": 0.6897888006200349,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.2055157269263066e-06,
+      "loss": 0.2207,
+      "step": 10680
+    },
+    {
+      "epoch": 0.6904346702835368,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.192598333656269e-06,
+      "loss": 0.2363,
+      "step": 10690
+    },
+    {
+      "epoch": 0.6910805399470387,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.17968094038623e-06,
+      "loss": 0.2087,
+      "step": 10700
+    },
+    {
+      "epoch": 0.6917264096105405,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.166763547116193e-06,
+      "loss": 0.2266,
+      "step": 10710
+    },
+    {
+      "epoch": 0.6923722792740425,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.2275,
+      "step": 10720
+    },
+    {
+      "epoch": 0.6930181489375444,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.1409287605761155e-06,
+      "loss": 0.2398,
+      "step": 10730
+    },
+    {
+      "epoch": 0.6936640186010463,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.128011367306078e-06,
+      "loss": 0.2038,
+      "step": 10740
+    },
+    {
+      "epoch": 0.6943098882645482,
+      "grad_norm": 1.25,
+      "learning_rate": 6.11509397403604e-06,
+      "loss": 0.2484,
+      "step": 10750
+    },
+    {
+      "epoch": 0.6949557579280501,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.1021765807660015e-06,
+      "loss": 0.2225,
+      "step": 10760
+    },
+    {
+      "epoch": 0.695601627591552,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.089259187495964e-06,
+      "loss": 0.2146,
+      "step": 10770
+    },
+    {
+      "epoch": 0.696247497255054,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.076341794225926e-06,
+      "loss": 0.2215,
+      "step": 10780
+    },
+    {
+      "epoch": 0.6968933669185559,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.0634244009558875e-06,
+      "loss": 0.2484,
+      "step": 10790
+    },
+    {
+      "epoch": 0.6975392365820577,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.05050700768585e-06,
+      "loss": 0.2411,
+      "step": 10800
+    },
+    {
+      "epoch": 0.6981851062455596,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.037589614415812e-06,
+      "loss": 0.2753,
+      "step": 10810
+    },
+    {
+      "epoch": 0.6988309759090615,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.024672221145773e-06,
+      "loss": 0.2271,
+      "step": 10820
+    },
+    {
+      "epoch": 0.6994768455725635,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.011754827875735e-06,
+      "loss": 0.2329,
+      "step": 10830
+    },
+    {
+      "epoch": 0.7001227152360654,
+      "grad_norm": 0.87890625,
+      "learning_rate": 5.9988374346056964e-06,
+      "loss": 0.2663,
+      "step": 10840
+    },
+    {
+      "epoch": 0.7007685848995673,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.985920041335659e-06,
+      "loss": 0.2558,
+      "step": 10850
+    },
+    {
+      "epoch": 0.7014144545630692,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.973002648065621e-06,
+      "loss": 0.2417,
+      "step": 10860
+    },
+    {
+      "epoch": 0.702060324226571,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.9600852547955825e-06,
+      "loss": 0.2314,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7027061938900729,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.947167861525545e-06,
+      "loss": 0.2169,
+      "step": 10880
+    },
+    {
+      "epoch": 0.7033520635535749,
+      "grad_norm": 1.125,
+      "learning_rate": 5.934250468255507e-06,
+      "loss": 0.2502,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7039979332170768,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.921333074985468e-06,
+      "loss": 0.2163,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7046438028805787,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.90841568171543e-06,
+      "loss": 0.231,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7052896725440806,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.895498288445392e-06,
+      "loss": 0.2155,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7059355422075825,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.882580895175354e-06,
+      "loss": 0.2304,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7065814118710844,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.869663501905316e-06,
+      "loss": 0.2275,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7072272815345864,
+      "grad_norm": 1.328125,
+      "learning_rate": 5.856746108635278e-06,
+      "loss": 0.2555,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7078731511980882,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.84382871536524e-06,
+      "loss": 0.2211,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7085190208615901,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.830911322095202e-06,
+      "loss": 0.2312,
+      "step": 10970
+    },
+    {
+      "epoch": 0.709164890525092,
+      "grad_norm": 1.2734375,
+      "learning_rate": 5.817993928825163e-06,
+      "loss": 0.2354,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7098107601885939,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.805076535555125e-06,
+      "loss": 0.244,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7104566298520959,
+      "grad_norm": 1.2890625,
+      "learning_rate": 5.792159142285087e-06,
+      "loss": 0.2207,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7111024995155978,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.779241749015049e-06,
+      "loss": 0.2197,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7117483691790997,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.766324355745011e-06,
+      "loss": 0.2475,
+      "step": 11020
+    },
+    {
+      "epoch": 0.7123942388426016,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.753406962474973e-06,
+      "loss": 0.2199,
+      "step": 11030
+    },
+    {
+      "epoch": 0.7130401085061034,
+      "grad_norm": 1.25,
+      "learning_rate": 5.740489569204935e-06,
+      "loss": 0.2299,
+      "step": 11040
+    },
+    {
+      "epoch": 0.7136859781696053,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.727572175934897e-06,
+      "loss": 0.2194,
+      "step": 11050
+    },
+    {
+      "epoch": 0.7143318478331073,
+      "grad_norm": 0.86328125,
+      "learning_rate": 5.714654782664859e-06,
+      "loss": 0.2057,
+      "step": 11060
+    },
+    {
+      "epoch": 0.7149777174966092,
+      "grad_norm": 1.2578125,
+      "learning_rate": 5.701737389394821e-06,
+      "loss": 0.2703,
+      "step": 11070
+    },
+    {
+      "epoch": 0.7156235871601111,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.688819996124782e-06,
+      "loss": 0.223,
+      "step": 11080
+    },
+    {
+      "epoch": 0.716269456823613,
+      "grad_norm": 1.25,
+      "learning_rate": 5.675902602854744e-06,
+      "loss": 0.2439,
+      "step": 11090
+    },
+    {
+      "epoch": 0.7169153264871149,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.662985209584706e-06,
+      "loss": 0.2441,
+      "step": 11100
+    },
+    {
+      "epoch": 0.7175611961506168,
+      "grad_norm": 1.28125,
+      "learning_rate": 5.650067816314668e-06,
+      "loss": 0.2307,
+      "step": 11110
+    },
+    {
+      "epoch": 0.7182070658141188,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.6371504230446304e-06,
+      "loss": 0.228,
+      "step": 11120
+    },
+    {
+      "epoch": 0.7188529354776206,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.624233029774592e-06,
+      "loss": 0.2389,
+      "step": 11130
+    },
+    {
+      "epoch": 0.7194988051411225,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.611315636504554e-06,
+      "loss": 0.2133,
+      "step": 11140
+    },
+    {
+      "epoch": 0.7201446748046244,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.598398243234516e-06,
+      "loss": 0.2637,
+      "step": 11150
+    },
+    {
+      "epoch": 0.7207905444681263,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.585480849964478e-06,
+      "loss": 0.2252,
+      "step": 11160
+    },
+    {
+      "epoch": 0.7214364141316283,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.57256345669444e-06,
+      "loss": 0.2309,
+      "step": 11170
+    },
+    {
+      "epoch": 0.7220822837951302,
+      "grad_norm": 1.25,
+      "learning_rate": 5.559646063424401e-06,
+      "loss": 0.233,
+      "step": 11180
+    },
+    {
+      "epoch": 0.7227281534586321,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.546728670154363e-06,
+      "loss": 0.1879,
+      "step": 11190
+    },
+    {
+      "epoch": 0.723374023122134,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.533811276884325e-06,
+      "loss": 0.2332,
+      "step": 11200
+    },
+    {
+      "epoch": 0.7240198927856358,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.520893883614287e-06,
+      "loss": 0.2329,
+      "step": 11210
+    },
+    {
+      "epoch": 0.7246657624491377,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.507976490344249e-06,
+      "loss": 0.2142,
+      "step": 11220
+    },
+    {
+      "epoch": 0.7253116321126397,
+      "grad_norm": 1.2734375,
+      "learning_rate": 5.495059097074211e-06,
+      "loss": 0.2515,
+      "step": 11230
+    },
+    {
+      "epoch": 0.7259575017761416,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.482141703804173e-06,
+      "loss": 0.1971,
+      "step": 11240
+    },
+    {
+      "epoch": 0.7266033714396435,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.469224310534135e-06,
+      "loss": 0.227,
+      "step": 11250
+    },
+    {
+      "epoch": 0.7272492411031454,
+      "grad_norm": 1.3125,
+      "learning_rate": 5.456306917264097e-06,
+      "loss": 0.2482,
+      "step": 11260
+    },
+    {
+      "epoch": 0.7278951107666473,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.443389523994058e-06,
+      "loss": 0.2288,
+      "step": 11270
+    },
+    {
+      "epoch": 0.7285409804301491,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.43047213072402e-06,
+      "loss": 0.2573,
+      "step": 11280
+    },
+    {
+      "epoch": 0.7291868500936511,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.417554737453983e-06,
+      "loss": 0.2271,
+      "step": 11290
+    },
+    {
+      "epoch": 0.729832719757153,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.404637344183944e-06,
+      "loss": 0.2434,
+      "step": 11300
+    },
+    {
+      "epoch": 0.7304785894206549,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.391719950913906e-06,
+      "loss": 0.2314,
+      "step": 11310
+    },
+    {
+      "epoch": 0.7311244590841568,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.378802557643868e-06,
+      "loss": 0.2399,
+      "step": 11320
+    },
+    {
+      "epoch": 0.7317703287476587,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.36588516437383e-06,
+      "loss": 0.2411,
+      "step": 11330
+    },
+    {
+      "epoch": 0.7324161984111607,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.352967771103792e-06,
+      "loss": 0.257,
+      "step": 11340
+    },
+    {
+      "epoch": 0.7330620680746626,
+      "grad_norm": 1.3046875,
+      "learning_rate": 5.340050377833753e-06,
+      "loss": 0.2474,
+      "step": 11350
+    },
+    {
+      "epoch": 0.7337079377381645,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.327132984563715e-06,
+      "loss": 0.237,
+      "step": 11360
+    },
+    {
+      "epoch": 0.7343538074016663,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.3142155912936775e-06,
+      "loss": 0.2134,
+      "step": 11370
+    },
+    {
+      "epoch": 0.7349996770651682,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.301298198023639e-06,
+      "loss": 0.2061,
+      "step": 11380
+    },
+    {
+      "epoch": 0.7356455467286701,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.288380804753601e-06,
+      "loss": 0.2141,
+      "step": 11390
+    },
+    {
+      "epoch": 0.7362914163921721,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.2754634114835636e-06,
+      "loss": 0.2399,
+      "step": 11400
+    },
+    {
+      "epoch": 0.736937286055674,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.262546018213525e-06,
+      "loss": 0.2636,
+      "step": 11410
+    },
+    {
+      "epoch": 0.7375831557191759,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.249628624943487e-06,
+      "loss": 0.2606,
+      "step": 11420
+    },
+    {
+      "epoch": 0.7382290253826778,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.23671123167345e-06,
+      "loss": 0.2397,
+      "step": 11430
+    },
+    {
+      "epoch": 0.7388748950461796,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.22379383840341e-06,
+      "loss": 0.2462,
+      "step": 11440
+    },
+    {
+      "epoch": 0.7395207647096815,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.2108764451333725e-06,
+      "loss": 0.2722,
+      "step": 11450
+    },
+    {
+      "epoch": 0.7401666343731835,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.197959051863335e-06,
+      "loss": 0.2328,
+      "step": 11460
+    },
+    {
+      "epoch": 0.7408125040366854,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.185041658593296e-06,
+      "loss": 0.2192,
+      "step": 11470
+    },
+    {
+      "epoch": 0.7414583737001873,
+      "grad_norm": 0.89453125,
+      "learning_rate": 5.1721242653232585e-06,
+      "loss": 0.2083,
+      "step": 11480
+    },
+    {
+      "epoch": 0.7421042433636892,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.15920687205322e-06,
+      "loss": 0.2051,
+      "step": 11490
+    },
+    {
+      "epoch": 0.7427501130271911,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.146289478783182e-06,
+      "loss": 0.2385,
+      "step": 11500
+    },
+    {
+      "epoch": 0.7433959826906931,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.1333720855131445e-06,
+      "loss": 0.2129,
+      "step": 11510
+    },
+    {
+      "epoch": 0.744041852354195,
+      "grad_norm": 1.3203125,
+      "learning_rate": 5.120454692243105e-06,
+      "loss": 0.231,
+      "step": 11520
+    },
+    {
+      "epoch": 0.7446877220176968,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.1075372989730674e-06,
+      "loss": 0.2169,
+      "step": 11530
+    },
+    {
+      "epoch": 0.7453335916811987,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.09461990570303e-06,
+      "loss": 0.2281,
+      "step": 11540
+    },
+    {
+      "epoch": 0.7459794613447006,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.081702512432991e-06,
+      "loss": 0.1929,
+      "step": 11550
+    },
+    {
+      "epoch": 0.7466253310082025,
+      "grad_norm": 1.2890625,
+      "learning_rate": 5.0687851191629535e-06,
+      "loss": 0.2483,
+      "step": 11560
+    },
+    {
+      "epoch": 0.7472712006717045,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.055867725892916e-06,
+      "loss": 0.2081,
+      "step": 11570
+    },
+    {
+      "epoch": 0.7479170703352064,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.042950332622877e-06,
+      "loss": 0.2112,
+      "step": 11580
+    },
+    {
+      "epoch": 0.7485629399987083,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.0300329393528395e-06,
+      "loss": 0.2387,
+      "step": 11590
+    },
+    {
+      "epoch": 0.7492088096622102,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.017115546082802e-06,
+      "loss": 0.2498,
+      "step": 11600
+    },
+    {
+      "epoch": 0.749854679325712,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.004198152812762e-06,
+      "loss": 0.2093,
+      "step": 11610
+    },
+    {
+      "epoch": 0.7505005489892139,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.991280759542725e-06,
+      "loss": 0.2232,
+      "step": 11620
+    },
+    {
+      "epoch": 0.7511464186527159,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.978363366272687e-06,
+      "loss": 0.2384,
+      "step": 11630
+    },
+    {
+      "epoch": 0.7517922883162178,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.965445973002648e-06,
+      "loss": 0.2139,
+      "step": 11640
+    },
+    {
+      "epoch": 0.7524381579797197,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.952528579732611e-06,
+      "loss": 0.2457,
+      "step": 11650
+    },
+    {
+      "epoch": 0.7530840276432216,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.939611186462572e-06,
+      "loss": 0.2236,
+      "step": 11660
+    },
+    {
+      "epoch": 0.7537298973067235,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.926693793192534e-06,
+      "loss": 0.241,
+      "step": 11670
+    },
+    {
+      "epoch": 0.7543757669702255,
+      "grad_norm": 1.390625,
+      "learning_rate": 4.913776399922496e-06,
+      "loss": 0.2113,
+      "step": 11680
+    },
+    {
+      "epoch": 0.7550216366337273,
+      "grad_norm": 1.0,
+      "learning_rate": 4.900859006652458e-06,
+      "loss": 0.2375,
+      "step": 11690
+    },
+    {
+      "epoch": 0.7556675062972292,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.88794161338242e-06,
+      "loss": 0.2422,
+      "step": 11700
+    },
+    {
+      "epoch": 0.7563133759607311,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.875024220112382e-06,
+      "loss": 0.2266,
+      "step": 11710
+    },
+    {
+      "epoch": 0.756959245624233,
+      "grad_norm": 1.3671875,
+      "learning_rate": 4.862106826842344e-06,
+      "loss": 0.2342,
+      "step": 11720
+    },
+    {
+      "epoch": 0.7576051152877349,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.849189433572306e-06,
+      "loss": 0.205,
+      "step": 11730
+    },
+    {
+      "epoch": 0.7582509849512369,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.836272040302267e-06,
+      "loss": 0.229,
+      "step": 11740
+    },
+    {
+      "epoch": 0.7588968546147388,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.823354647032229e-06,
+      "loss": 0.2068,
+      "step": 11750
+    },
+    {
+      "epoch": 0.7595427242782407,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.810437253762192e-06,
+      "loss": 0.2329,
+      "step": 11760
+    },
+    {
+      "epoch": 0.7601885939417425,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.797519860492153e-06,
+      "loss": 0.2479,
+      "step": 11770
+    },
+    {
+      "epoch": 0.7608344636052444,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.7846024672221145e-06,
+      "loss": 0.219,
+      "step": 11780
+    },
+    {
+      "epoch": 0.7614803332687463,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.771685073952077e-06,
+      "loss": 0.214,
+      "step": 11790
+    },
+    {
+      "epoch": 0.7621262029322483,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.758767680682039e-06,
+      "loss": 0.2501,
+      "step": 11800
+    },
+    {
+      "epoch": 0.7627720725957502,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.7458502874120006e-06,
+      "loss": 0.2031,
+      "step": 11810
+    },
+    {
+      "epoch": 0.7634179422592521,
+      "grad_norm": 0.875,
+      "learning_rate": 4.732932894141963e-06,
+      "loss": 0.2355,
+      "step": 11820
+    },
+    {
+      "epoch": 0.764063811922754,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.720015500871924e-06,
+      "loss": 0.2322,
+      "step": 11830
+    },
+    {
+      "epoch": 0.7647096815862559,
+      "grad_norm": 1.6953125,
+      "learning_rate": 4.707098107601887e-06,
+      "loss": 0.2518,
+      "step": 11840
+    },
+    {
+      "epoch": 0.7653555512497578,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.694180714331848e-06,
+      "loss": 0.227,
+      "step": 11850
+    },
+    {
+      "epoch": 0.7660014209132597,
+      "grad_norm": 1.3671875,
+      "learning_rate": 4.68126332106181e-06,
+      "loss": 0.2343,
+      "step": 11860
+    },
+    {
+      "epoch": 0.7666472905767616,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.668345927791772e-06,
+      "loss": 0.2372,
+      "step": 11870
+    },
+    {
+      "epoch": 0.7672931602402635,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.655428534521734e-06,
+      "loss": 0.2584,
+      "step": 11880
+    },
+    {
+      "epoch": 0.7679390299037654,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.642511141251696e-06,
+      "loss": 0.2545,
+      "step": 11890
+    },
+    {
+      "epoch": 0.7685848995672673,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.629593747981658e-06,
+      "loss": 0.2579,
+      "step": 11900
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.616676354711619e-06,
+      "loss": 0.2326,
+      "step": 11910
+    },
+    {
+      "epoch": 0.7698766388942712,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.6037589614415815e-06,
+      "loss": 0.2593,
+      "step": 11920
+    },
+    {
+      "epoch": 0.770522508557773,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.590841568171544e-06,
+      "loss": 0.2103,
+      "step": 11930
+    },
+    {
+      "epoch": 0.7711683782212749,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.577924174901505e-06,
+      "loss": 0.2412,
+      "step": 11940
+    },
+    {
+      "epoch": 0.7718142478847768,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.565006781631467e-06,
+      "loss": 0.2634,
+      "step": 11950
+    },
+    {
+      "epoch": 0.7724601175482787,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.552089388361429e-06,
+      "loss": 0.2483,
+      "step": 11960
+    },
+    {
+      "epoch": 0.7731059872117807,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.539171995091391e-06,
+      "loss": 0.2265,
+      "step": 11970
+    },
+    {
+      "epoch": 0.7737518568752826,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.526254601821353e-06,
+      "loss": 0.2277,
+      "step": 11980
+    },
+    {
+      "epoch": 0.7743977265387845,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.513337208551315e-06,
+      "loss": 0.2311,
+      "step": 11990
+    },
+    {
+      "epoch": 0.7750435962022864,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.5004198152812765e-06,
+      "loss": 0.2154,
+      "step": 12000
+    },
+    {
+      "epoch": 0.7756894658657882,
+      "grad_norm": 1.0,
+      "learning_rate": 4.487502422011239e-06,
+      "loss": 0.2349,
+      "step": 12010
+    },
+    {
+      "epoch": 0.7763353355292902,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.4745850287412e-06,
+      "loss": 0.2194,
+      "step": 12020
+    },
+    {
+      "epoch": 0.7769812051927921,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.4616676354711625e-06,
+      "loss": 0.2572,
+      "step": 12030
+    },
+    {
+      "epoch": 0.777627074856294,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.448750242201124e-06,
+      "loss": 0.2068,
+      "step": 12040
+    },
+    {
+      "epoch": 0.7782729445197959,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.435832848931086e-06,
+      "loss": 0.2165,
+      "step": 12050
+    },
+    {
+      "epoch": 0.7789188141832978,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.4229154556610485e-06,
+      "loss": 0.2507,
+      "step": 12060
+    },
+    {
+      "epoch": 0.7795646838467997,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.40999806239101e-06,
+      "loss": 0.2445,
+      "step": 12070
+    },
+    {
+      "epoch": 0.7802105535103017,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.397080669120971e-06,
+      "loss": 0.2526,
+      "step": 12080
+    },
+    {
+      "epoch": 0.7808564231738035,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.384163275850934e-06,
+      "loss": 0.2274,
+      "step": 12090
+    },
+    {
+      "epoch": 0.7815022928373054,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.371245882580896e-06,
+      "loss": 0.2206,
+      "step": 12100
+    },
+    {
+      "epoch": 0.7821481625008073,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.3583284893108574e-06,
+      "loss": 0.2237,
+      "step": 12110
+    },
+    {
+      "epoch": 0.7827940321643092,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.345411096040819e-06,
+      "loss": 0.2292,
+      "step": 12120
+    },
+    {
+      "epoch": 0.7834399018278112,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.332493702770781e-06,
+      "loss": 0.2187,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7840857714913131,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.3195763095007435e-06,
+      "loss": 0.2322,
+      "step": 12140
+    },
+    {
+      "epoch": 0.784731641154815,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.306658916230705e-06,
+      "loss": 0.2421,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7853775108183169,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.293741522960667e-06,
+      "loss": 0.2402,
+      "step": 12160
+    },
+    {
+      "epoch": 0.7860233804818187,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.280824129690629e-06,
+      "loss": 0.2079,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7866692501453206,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.267906736420591e-06,
+      "loss": 0.2632,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7873151198088226,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.254989343150552e-06,
+      "loss": 0.2151,
+      "step": 12190
+    },
+    {
+      "epoch": 0.7879609894723245,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.242071949880515e-06,
+      "loss": 0.2426,
+      "step": 12200
+    },
+    {
+      "epoch": 0.7886068591358264,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.229154556610476e-06,
+      "loss": 0.2211,
+      "step": 12210
+    },
+    {
+      "epoch": 0.7892527287993283,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.216237163340438e-06,
+      "loss": 0.2445,
+      "step": 12220
+    },
+    {
+      "epoch": 0.7898985984628302,
+      "grad_norm": 1.25,
+      "learning_rate": 4.203319770070401e-06,
+      "loss": 0.2213,
+      "step": 12230
+    },
+    {
+      "epoch": 0.7905444681263321,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.190402376800362e-06,
+      "loss": 0.2377,
+      "step": 12240
+    },
+    {
+      "epoch": 0.791190337789834,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.177484983530324e-06,
+      "loss": 0.2229,
+      "step": 12250
+    },
+    {
+      "epoch": 0.7918362074533359,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.164567590260286e-06,
+      "loss": 0.2211,
+      "step": 12260
+    },
+    {
+      "epoch": 0.7924820771168378,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.151650196990248e-06,
+      "loss": 0.2128,
+      "step": 12270
+    },
+    {
+      "epoch": 0.7931279467803397,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.13873280372021e-06,
+      "loss": 0.2372,
+      "step": 12280
+    },
+    {
+      "epoch": 0.7937738164438416,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.125815410450171e-06,
+      "loss": 0.2202,
+      "step": 12290
+    },
+    {
+      "epoch": 0.7944196861073436,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.112898017180133e-06,
+      "loss": 0.2204,
+      "step": 12300
+    },
+    {
+      "epoch": 0.7950655557708455,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.099980623910096e-06,
+      "loss": 0.2391,
+      "step": 12310
+    },
+    {
+      "epoch": 0.7957114254343474,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.087063230640057e-06,
+      "loss": 0.2184,
+      "step": 12320
+    },
+    {
+      "epoch": 0.7963572950978492,
+      "grad_norm": 1.3203125,
+      "learning_rate": 4.074145837370019e-06,
+      "loss": 0.2371,
+      "step": 12330
+    },
+    {
+      "epoch": 0.7970031647613511,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.061228444099981e-06,
+      "loss": 0.2155,
+      "step": 12340
+    },
+    {
+      "epoch": 0.797649034424853,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.048311050829943e-06,
+      "loss": 0.2208,
+      "step": 12350
+    },
+    {
+      "epoch": 0.798294904088355,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.0353936575599046e-06,
+      "loss": 0.2519,
+      "step": 12360
+    },
+    {
+      "epoch": 0.7989407737518569,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.022476264289867e-06,
+      "loss": 0.2373,
+      "step": 12370
+    },
+    {
+      "epoch": 0.7995866434153588,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.009558871019828e-06,
+      "loss": 0.2572,
+      "step": 12380
+    },
+    {
+      "epoch": 0.8002325130788607,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.996641477749791e-06,
+      "loss": 0.2135,
+      "step": 12390
+    },
+    {
+      "epoch": 0.8008783827423626,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.983724084479753e-06,
+      "loss": 0.2309,
+      "step": 12400
+    },
+    {
+      "epoch": 0.8015242524058644,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.970806691209714e-06,
+      "loss": 0.2218,
+      "step": 12410
+    },
+    {
+      "epoch": 0.8021701220693664,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.957889297939676e-06,
+      "loss": 0.2109,
+      "step": 12420
+    },
+    {
+      "epoch": 0.8028159917328683,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.944971904669638e-06,
+      "loss": 0.2329,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8034618613963702,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.9320545113996e-06,
+      "loss": 0.2506,
+      "step": 12440
+    },
+    {
+      "epoch": 0.8041077310598721,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.919137118129562e-06,
+      "loss": 0.2371,
+      "step": 12450
+    },
+    {
+      "epoch": 0.804753600723374,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.906219724859523e-06,
+      "loss": 0.2433,
+      "step": 12460
+    },
+    {
+      "epoch": 0.805399470386876,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.8933023315894855e-06,
+      "loss": 0.2441,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8060453400503779,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.880384938319448e-06,
+      "loss": 0.2258,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8066912097138798,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.867467545049409e-06,
+      "loss": 0.245,
+      "step": 12490
+    },
+    {
+      "epoch": 0.8073370793773816,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.854550151779371e-06,
+      "loss": 0.2194,
+      "step": 12500
+    },
+    {
+      "epoch": 0.8079829490408835,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.841632758509333e-06,
+      "loss": 0.227,
+      "step": 12510
+    },
+    {
+      "epoch": 0.8086288187043854,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.828715365239295e-06,
+      "loss": 0.2206,
+      "step": 12520
+    },
+    {
+      "epoch": 0.8092746883678874,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.815797971969257e-06,
+      "loss": 0.2514,
+      "step": 12530
+    },
+    {
+      "epoch": 0.8099205580313893,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.802880578699219e-06,
+      "loss": 0.2319,
+      "step": 12540
+    },
+    {
+      "epoch": 0.8105664276948912,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.789963185429181e-06,
+      "loss": 0.1914,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8112122973583931,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.7770457921591423e-06,
+      "loss": 0.2361,
+      "step": 12560
+    },
+    {
+      "epoch": 0.811858167021895,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.7641283988891046e-06,
+      "loss": 0.2408,
+      "step": 12570
+    },
+    {
+      "epoch": 0.8125040366853968,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.7512110056190665e-06,
+      "loss": 0.2192,
+      "step": 12580
+    },
+    {
+      "epoch": 0.8131499063488988,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.7382936123490284e-06,
+      "loss": 0.2201,
+      "step": 12590
+    },
+    {
+      "epoch": 0.8137957760124007,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.72537621907899e-06,
+      "loss": 0.2429,
+      "step": 12600
+    },
+    {
+      "epoch": 0.8144416456759026,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.712458825808952e-06,
+      "loss": 0.2223,
+      "step": 12610
+    },
+    {
+      "epoch": 0.8150875153394045,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.699541432538914e-06,
+      "loss": 0.206,
+      "step": 12620
+    },
+    {
+      "epoch": 0.8157333850029064,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.686624039268876e-06,
+      "loss": 0.2308,
+      "step": 12630
+    },
+    {
+      "epoch": 0.8163792546664084,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.673706645998838e-06,
+      "loss": 0.2061,
+      "step": 12640
+    },
+    {
+      "epoch": 0.8170251243299103,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.6607892527287996e-06,
+      "loss": 0.2433,
+      "step": 12650
+    },
+    {
+      "epoch": 0.8176709939934121,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.6478718594587614e-06,
+      "loss": 0.221,
+      "step": 12660
+    },
+    {
+      "epoch": 0.818316863656914,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.6349544661887233e-06,
+      "loss": 0.1949,
+      "step": 12670
+    },
+    {
+      "epoch": 0.8189627333204159,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.6220370729186856e-06,
+      "loss": 0.274,
+      "step": 12680
+    },
+    {
+      "epoch": 0.8196086029839178,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.609119679648647e-06,
+      "loss": 0.1969,
+      "step": 12690
+    },
+    {
+      "epoch": 0.8202544726474198,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.596202286378609e-06,
+      "loss": 0.2273,
+      "step": 12700
+    },
+    {
+      "epoch": 0.8209003423109217,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.583284893108571e-06,
+      "loss": 0.2622,
+      "step": 12710
+    },
+    {
+      "epoch": 0.8215462119744236,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.570367499838533e-06,
+      "loss": 0.2085,
+      "step": 12720
+    },
+    {
+      "epoch": 0.8221920816379255,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.5574501065684945e-06,
+      "loss": 0.242,
+      "step": 12730
+    },
+    {
+      "epoch": 0.8228379513014273,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.5445327132984564e-06,
+      "loss": 0.2249,
+      "step": 12740
+    },
+    {
+      "epoch": 0.8234838209649292,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.5316153200284187e-06,
+      "loss": 0.2375,
+      "step": 12750
+    },
+    {
+      "epoch": 0.8241296906284312,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.5186979267583805e-06,
+      "loss": 0.231,
+      "step": 12760
+    },
+    {
+      "epoch": 0.8247755602919331,
+      "grad_norm": 1.375,
+      "learning_rate": 3.505780533488342e-06,
+      "loss": 0.2467,
+      "step": 12770
+    },
+    {
+      "epoch": 0.825421429955435,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.4928631402183043e-06,
+      "loss": 0.2118,
+      "step": 12780
+    },
+    {
+      "epoch": 0.8260672996189369,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.479945746948266e-06,
+      "loss": 0.2308,
+      "step": 12790
+    },
+    {
+      "epoch": 0.8267131692824388,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.467028353678228e-06,
+      "loss": 0.2351,
+      "step": 12800
+    },
+    {
+      "epoch": 0.8273590389459408,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.4541109604081903e-06,
+      "loss": 0.2173,
+      "step": 12810
+    },
+    {
+      "epoch": 0.8280049086094426,
+      "grad_norm": 1.125,
+      "learning_rate": 3.4411935671381517e-06,
+      "loss": 0.2532,
+      "step": 12820
+    },
+    {
+      "epoch": 0.8286507782729445,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.4282761738681136e-06,
+      "loss": 0.2222,
+      "step": 12830
+    },
+    {
+      "epoch": 0.8292966479364464,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.4153587805980755e-06,
+      "loss": 0.2353,
+      "step": 12840
+    },
+    {
+      "epoch": 0.8299425175999483,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.4024413873280378e-06,
+      "loss": 0.2213,
+      "step": 12850
+    },
+    {
+      "epoch": 0.8305883872634502,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.3895239940579992e-06,
+      "loss": 0.2482,
+      "step": 12860
+    },
+    {
+      "epoch": 0.8312342569269522,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.376606600787961e-06,
+      "loss": 0.2217,
+      "step": 12870
+    },
+    {
+      "epoch": 0.8318801265904541,
+      "grad_norm": 1.125,
+      "learning_rate": 3.3636892075179234e-06,
+      "loss": 0.2221,
+      "step": 12880
+    },
+    {
+      "epoch": 0.832525996253956,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.3507718142478852e-06,
+      "loss": 0.2003,
+      "step": 12890
+    },
+    {
+      "epoch": 0.8331718659174578,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.3378544209778467e-06,
+      "loss": 0.2185,
+      "step": 12900
+    },
+    {
+      "epoch": 0.8338177355809597,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.3249370277078086e-06,
+      "loss": 0.2438,
+      "step": 12910
+    },
+    {
+      "epoch": 0.8344636052444616,
+      "grad_norm": 1.3125,
+      "learning_rate": 3.312019634437771e-06,
+      "loss": 0.2259,
+      "step": 12920
+    },
+    {
+      "epoch": 0.8351094749079636,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.2991022411677327e-06,
+      "loss": 0.2109,
+      "step": 12930
+    },
+    {
+      "epoch": 0.8357553445714655,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.286184847897694e-06,
+      "loss": 0.2293,
+      "step": 12940
+    },
+    {
+      "epoch": 0.8364012142349674,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.2732674546276565e-06,
+      "loss": 0.2258,
+      "step": 12950
+    },
+    {
+      "epoch": 0.8370470838984693,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.2603500613576183e-06,
+      "loss": 0.2252,
+      "step": 12960
+    },
+    {
+      "epoch": 0.8376929535619712,
+      "grad_norm": 1.125,
+      "learning_rate": 3.24743266808758e-06,
+      "loss": 0.2463,
+      "step": 12970
+    },
+    {
+      "epoch": 0.8383388232254732,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.2345152748175425e-06,
+      "loss": 0.2751,
+      "step": 12980
+    },
+    {
+      "epoch": 0.838984692888975,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.221597881547504e-06,
+      "loss": 0.2129,
+      "step": 12990
+    },
+    {
+      "epoch": 0.8396305625524769,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.208680488277466e-06,
+      "loss": 0.2241,
+      "step": 13000
+    },
+    {
+      "epoch": 0.8402764322159788,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.1957630950074277e-06,
+      "loss": 0.1997,
+      "step": 13010
+    },
+    {
+      "epoch": 0.8409223018794807,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.18284570173739e-06,
+      "loss": 0.1954,
+      "step": 13020
+    },
+    {
+      "epoch": 0.8415681715429826,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.1699283084673514e-06,
+      "loss": 0.2343,
+      "step": 13030
+    },
+    {
+      "epoch": 0.8422140412064846,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.1570109151973133e-06,
+      "loss": 0.2288,
+      "step": 13040
+    },
+    {
+      "epoch": 0.8428599108699865,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.1440935219272756e-06,
+      "loss": 0.2359,
+      "step": 13050
+    },
+    {
+      "epoch": 0.8435057805334883,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.1311761286572374e-06,
+      "loss": 0.2185,
+      "step": 13060
+    },
+    {
+      "epoch": 0.8441516501969902,
+      "grad_norm": 1.0,
+      "learning_rate": 3.118258735387199e-06,
+      "loss": 0.2491,
+      "step": 13070
+    },
+    {
+      "epoch": 0.8447975198604921,
+      "grad_norm": 1.375,
+      "learning_rate": 3.1053413421171607e-06,
+      "loss": 0.2181,
+      "step": 13080
+    },
+    {
+      "epoch": 0.845443389523994,
+      "grad_norm": 1.0,
+      "learning_rate": 3.092423948847123e-06,
+      "loss": 0.2381,
+      "step": 13090
+    },
+    {
+      "epoch": 0.846089259187496,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.079506555577085e-06,
+      "loss": 0.2612,
+      "step": 13100
+    },
+    {
+      "epoch": 0.8467351288509979,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.0665891623070463e-06,
+      "loss": 0.229,
+      "step": 13110
+    },
+    {
+      "epoch": 0.8473809985144998,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.0536717690370086e-06,
+      "loss": 0.2439,
+      "step": 13120
+    },
+    {
+      "epoch": 0.8480268681780017,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.0407543757669705e-06,
+      "loss": 0.2199,
+      "step": 13130
+    },
+    {
+      "epoch": 0.8486727378415035,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.0278369824969324e-06,
+      "loss": 0.2042,
+      "step": 13140
+    },
+    {
+      "epoch": 0.8493186075050055,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.014919589226894e-06,
+      "loss": 0.2303,
+      "step": 13150
+    },
+    {
+      "epoch": 0.8499644771685074,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.002002195956856e-06,
+      "loss": 0.212,
+      "step": 13160
+    },
+    {
+      "epoch": 0.8506103468320093,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.989084802686818e-06,
+      "loss": 0.2391,
+      "step": 13170
+    },
+    {
+      "epoch": 0.8512562164955112,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.97616740941678e-06,
+      "loss": 0.2093,
+      "step": 13180
+    },
+    {
+      "epoch": 0.8519020861590131,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.963250016146742e-06,
+      "loss": 0.2014,
+      "step": 13190
+    },
+    {
+      "epoch": 0.852547955822515,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.9503326228767036e-06,
+      "loss": 0.2446,
+      "step": 13200
+    },
+    {
+      "epoch": 0.853193825486017,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.9374152296066654e-06,
+      "loss": 0.2072,
+      "step": 13210
+    },
+    {
+      "epoch": 0.8538396951495189,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.9244978363366277e-06,
+      "loss": 0.2366,
+      "step": 13220
+    },
+    {
+      "epoch": 0.8544855648130207,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.9115804430665896e-06,
+      "loss": 0.2186,
+      "step": 13230
+    },
+    {
+      "epoch": 0.8551314344765226,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.898663049796551e-06,
+      "loss": 0.229,
+      "step": 13240
+    },
+    {
+      "epoch": 0.8557773041400245,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.885745656526513e-06,
+      "loss": 0.2373,
+      "step": 13250
+    },
+    {
+      "epoch": 0.8564231738035264,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.872828263256475e-06,
+      "loss": 0.2457,
+      "step": 13260
+    },
+    {
+      "epoch": 0.8570690434670284,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.859910869986437e-06,
+      "loss": 0.2358,
+      "step": 13270
+    },
+    {
+      "epoch": 0.8577149131305303,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.8469934767163985e-06,
+      "loss": 0.223,
+      "step": 13280
+    },
+    {
+      "epoch": 0.8583607827940322,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.834076083446361e-06,
+      "loss": 0.226,
+      "step": 13290
+    },
+    {
+      "epoch": 0.859006652457534,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.8211586901763227e-06,
+      "loss": 0.2203,
+      "step": 13300
+    },
+    {
+      "epoch": 0.8596525221210359,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.8082412969062845e-06,
+      "loss": 0.2594,
+      "step": 13310
+    },
+    {
+      "epoch": 0.8602983917845379,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.795323903636246e-06,
+      "loss": 0.226,
+      "step": 13320
+    },
+    {
+      "epoch": 0.8609442614480398,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.7824065103662083e-06,
+      "loss": 0.2076,
+      "step": 13330
+    },
+    {
+      "epoch": 0.8615901311115417,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.76948911709617e-06,
+      "loss": 0.231,
+      "step": 13340
+    },
+    {
+      "epoch": 0.8622360007750436,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.756571723826132e-06,
+      "loss": 0.226,
+      "step": 13350
+    },
+    {
+      "epoch": 0.8628818704385455,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.7436543305560943e-06,
+      "loss": 0.2411,
+      "step": 13360
+    },
+    {
+      "epoch": 0.8635277401020474,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.7307369372860557e-06,
+      "loss": 0.2044,
+      "step": 13370
+    },
+    {
+      "epoch": 0.8641736097655494,
+      "grad_norm": 0.875,
+      "learning_rate": 2.7178195440160176e-06,
+      "loss": 0.2203,
+      "step": 13380
+    },
+    {
+      "epoch": 0.8648194794290512,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.70490215074598e-06,
+      "loss": 0.1989,
+      "step": 13390
+    },
+    {
+      "epoch": 0.8654653490925531,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.6919847574759418e-06,
+      "loss": 0.2382,
+      "step": 13400
+    },
+    {
+      "epoch": 0.866111218756055,
+      "grad_norm": 1.25,
+      "learning_rate": 2.6790673642059032e-06,
+      "loss": 0.2477,
+      "step": 13410
+    },
+    {
+      "epoch": 0.8667570884195569,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.666149970935865e-06,
+      "loss": 0.2271,
+      "step": 13420
+    },
+    {
+      "epoch": 0.8674029580830588,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.6532325776658274e-06,
+      "loss": 0.2345,
+      "step": 13430
+    },
+    {
+      "epoch": 0.8680488277465608,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.6403151843957892e-06,
+      "loss": 0.1954,
+      "step": 13440
+    },
+    {
+      "epoch": 0.8686946974100627,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.6273977911257507e-06,
+      "loss": 0.2317,
+      "step": 13450
+    },
+    {
+      "epoch": 0.8693405670735646,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.614480397855713e-06,
+      "loss": 0.2256,
+      "step": 13460
+    },
+    {
+      "epoch": 0.8699864367370664,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.601563004585675e-06,
+      "loss": 0.2492,
+      "step": 13470
+    },
+    {
+      "epoch": 0.8706323064005683,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.5886456113156367e-06,
+      "loss": 0.197,
+      "step": 13480
+    },
+    {
+      "epoch": 0.8712781760640703,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.575728218045598e-06,
+      "loss": 0.2342,
+      "step": 13490
+    },
+    {
+      "epoch": 0.8719240457275722,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.5628108247755604e-06,
+      "loss": 0.2428,
+      "step": 13500
+    },
+    {
+      "epoch": 0.8725699153910741,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.5498934315055223e-06,
+      "loss": 0.2235,
+      "step": 13510
+    },
+    {
+      "epoch": 0.873215785054576,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.536976038235484e-06,
+      "loss": 0.2374,
+      "step": 13520
+    },
+    {
+      "epoch": 0.8738616547180779,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.5240586449654465e-06,
+      "loss": 0.2008,
+      "step": 13530
+    },
+    {
+      "epoch": 0.8745075243815797,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.511141251695408e-06,
+      "loss": 0.2279,
+      "step": 13540
+    },
+    {
+      "epoch": 0.8751533940450817,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.4982238584253698e-06,
+      "loss": 0.2138,
+      "step": 13550
+    },
+    {
+      "epoch": 0.8757992637085836,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.4853064651553317e-06,
+      "loss": 0.2408,
+      "step": 13560
+    },
+    {
+      "epoch": 0.8764451333720855,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.472389071885294e-06,
+      "loss": 0.2262,
+      "step": 13570
+    },
+    {
+      "epoch": 0.8770910030355874,
+      "grad_norm": 1.6171875,
+      "learning_rate": 2.4594716786152554e-06,
+      "loss": 0.2629,
+      "step": 13580
+    },
+    {
+      "epoch": 0.8777368726990893,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.4465542853452177e-06,
+      "loss": 0.2294,
+      "step": 13590
+    },
+    {
+      "epoch": 0.8783827423625913,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.433636892075179e-06,
+      "loss": 0.2014,
+      "step": 13600
+    },
+    {
+      "epoch": 0.8790286120260932,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.4207194988051414e-06,
+      "loss": 0.2162,
+      "step": 13610
+    },
+    {
+      "epoch": 0.8796744816895951,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.4078021055351033e-06,
+      "loss": 0.2502,
+      "step": 13620
+    },
+    {
+      "epoch": 0.8803203513530969,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.394884712265065e-06,
+      "loss": 0.2121,
+      "step": 13630
+    },
+    {
+      "epoch": 0.8809662210165988,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.381967318995027e-06,
+      "loss": 0.2182,
+      "step": 13640
+    },
+    {
+      "epoch": 0.8816120906801007,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.369049925724989e-06,
+      "loss": 0.2202,
+      "step": 13650
+    },
+    {
+      "epoch": 0.8822579603436027,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.3561325324549508e-06,
+      "loss": 0.2262,
+      "step": 13660
+    },
+    {
+      "epoch": 0.8829038300071046,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.3432151391849126e-06,
+      "loss": 0.2558,
+      "step": 13670
+    },
+    {
+      "epoch": 0.8835496996706065,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.3302977459148745e-06,
+      "loss": 0.2542,
+      "step": 13680
+    },
+    {
+      "epoch": 0.8841955693341084,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.3173803526448368e-06,
+      "loss": 0.221,
+      "step": 13690
+    },
+    {
+      "epoch": 0.8848414389976103,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.3044629593747982e-06,
+      "loss": 0.2304,
+      "step": 13700
+    },
+    {
+      "epoch": 0.8854873086611121,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.2915455661047605e-06,
+      "loss": 0.2542,
+      "step": 13710
+    },
+    {
+      "epoch": 0.8861331783246141,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.278628172834722e-06,
+      "loss": 0.2394,
+      "step": 13720
+    },
+    {
+      "epoch": 0.886779047988116,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.2657107795646843e-06,
+      "loss": 0.2272,
+      "step": 13730
+    },
+    {
+      "epoch": 0.8874249176516179,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.252793386294646e-06,
+      "loss": 0.2146,
+      "step": 13740
+    },
+    {
+      "epoch": 0.8880707873151198,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.239875993024608e-06,
+      "loss": 0.2356,
+      "step": 13750
+    },
+    {
+      "epoch": 0.8887166569786217,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.22695859975457e-06,
+      "loss": 0.2072,
+      "step": 13760
+    },
+    {
+      "epoch": 0.8893625266421237,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.2140412064845317e-06,
+      "loss": 0.2024,
+      "step": 13770
+    },
+    {
+      "epoch": 0.8900083963056256,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.2011238132144936e-06,
+      "loss": 0.2202,
+      "step": 13780
+    },
+    {
+      "epoch": 0.8906542659691274,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.1882064199444555e-06,
+      "loss": 0.2418,
+      "step": 13790
+    },
+    {
+      "epoch": 0.8913001356326293,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.1752890266744173e-06,
+      "loss": 0.2116,
+      "step": 13800
+    },
+    {
+      "epoch": 0.8919460052961312,
+      "grad_norm": 1.2890625,
+      "learning_rate": 2.162371633404379e-06,
+      "loss": 0.2606,
+      "step": 13810
+    },
+    {
+      "epoch": 0.8925918749596331,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.149454240134341e-06,
+      "loss": 0.2238,
+      "step": 13820
+    },
+    {
+      "epoch": 0.8932377446231351,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.136536846864303e-06,
+      "loss": 0.234,
+      "step": 13830
+    },
+    {
+      "epoch": 0.893883614286637,
+      "grad_norm": 1.25,
+      "learning_rate": 2.123619453594265e-06,
+      "loss": 0.2401,
+      "step": 13840
+    },
+    {
+      "epoch": 0.8945294839501389,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.1107020603242267e-06,
+      "loss": 0.2205,
+      "step": 13850
+    },
+    {
+      "epoch": 0.8951753536136408,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.097784667054189e-06,
+      "loss": 0.2125,
+      "step": 13860
+    },
+    {
+      "epoch": 0.8958212232771426,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.0848672737841504e-06,
+      "loss": 0.2512,
+      "step": 13870
+    },
+    {
+      "epoch": 0.8964670929406445,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.0719498805141127e-06,
+      "loss": 0.2168,
+      "step": 13880
+    },
+    {
+      "epoch": 0.8971129626041465,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.059032487244074e-06,
+      "loss": 0.2138,
+      "step": 13890
+    },
+    {
+      "epoch": 0.8977588322676484,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.0461150939740364e-06,
+      "loss": 0.2093,
+      "step": 13900
+    },
+    {
+      "epoch": 0.8984047019311503,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.033197700703998e-06,
+      "loss": 0.2138,
+      "step": 13910
+    },
+    {
+      "epoch": 0.8990505715946522,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.02028030743396e-06,
+      "loss": 0.2486,
+      "step": 13920
+    },
+    {
+      "epoch": 0.8996964412581541,
+      "grad_norm": 1.0,
+      "learning_rate": 2.007362914163922e-06,
+      "loss": 0.2278,
+      "step": 13930
+    },
+    {
+      "epoch": 0.9003423109216561,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.994445520893884e-06,
+      "loss": 0.2334,
+      "step": 13940
+    },
+    {
+      "epoch": 0.900988180585158,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9815281276238458e-06,
+      "loss": 0.2393,
+      "step": 13950
+    },
+    {
+      "epoch": 0.9016340502486598,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9686107343538076e-06,
+      "loss": 0.2319,
+      "step": 13960
+    },
+    {
+      "epoch": 0.9022799199121617,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9556933410837695e-06,
+      "loss": 0.2443,
+      "step": 13970
+    },
+    {
+      "epoch": 0.9029257895756636,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9427759478137314e-06,
+      "loss": 0.2578,
+      "step": 13980
+    },
+    {
+      "epoch": 0.9035716592391655,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9298585545436932e-06,
+      "loss": 0.2175,
+      "step": 13990
+    },
+    {
+      "epoch": 0.9042175289026675,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.916941161273655e-06,
+      "loss": 0.2073,
+      "step": 14000
+    },
+    {
+      "epoch": 0.9048633985661694,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.904023768003617e-06,
+      "loss": 0.2171,
+      "step": 14010
+    },
+    {
+      "epoch": 0.9055092682296713,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.8911063747335788e-06,
+      "loss": 0.222,
+      "step": 14020
+    },
+    {
+      "epoch": 0.9061551378931731,
+      "grad_norm": 1.125,
+      "learning_rate": 1.8781889814635407e-06,
+      "loss": 0.2034,
+      "step": 14030
+    },
+    {
+      "epoch": 0.906801007556675,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.8652715881935026e-06,
+      "loss": 0.2077,
+      "step": 14040
+    },
+    {
+      "epoch": 0.9074468772201769,
+      "grad_norm": 1.125,
+      "learning_rate": 1.8523541949234647e-06,
+      "loss": 0.2269,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9080927468836789,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8394368016534265e-06,
+      "loss": 0.2174,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9087386165471808,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.8265194083833884e-06,
+      "loss": 0.2188,
+      "step": 14070
+    },
+    {
+      "epoch": 0.9093844862106827,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.8136020151133503e-06,
+      "loss": 0.2332,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9100303558741846,
+      "grad_norm": 1.125,
+      "learning_rate": 1.8006846218433121e-06,
+      "loss": 0.1985,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9106762255376865,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.7877672285732742e-06,
+      "loss": 0.2411,
+      "step": 14100
+    },
+    {
+      "epoch": 0.9113220952011885,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.7748498353032359e-06,
+      "loss": 0.257,
+      "step": 14110
+    },
+    {
+      "epoch": 0.9119679648646903,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.761932442033198e-06,
+      "loss": 0.2211,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9126138345281922,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.7490150487631596e-06,
+      "loss": 0.2456,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9132597041916941,
+      "grad_norm": 1.25,
+      "learning_rate": 1.7360976554931217e-06,
+      "loss": 0.2163,
+      "step": 14140
+    },
+    {
+      "epoch": 0.913905573855196,
+      "grad_norm": 1.125,
+      "learning_rate": 1.7231802622230838e-06,
+      "loss": 0.2068,
+      "step": 14150
+    },
+    {
+      "epoch": 0.9145514435186979,
+      "grad_norm": 1.0,
+      "learning_rate": 1.7102628689530454e-06,
+      "loss": 0.216,
+      "step": 14160
+    },
+    {
+      "epoch": 0.9151973131821999,
+      "grad_norm": 1.375,
+      "learning_rate": 1.6973454756830075e-06,
+      "loss": 0.2269,
+      "step": 14170
+    },
+    {
+      "epoch": 0.9158431828457018,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.6844280824129691e-06,
+      "loss": 0.2435,
+      "step": 14180
+    },
+    {
+      "epoch": 0.9164890525092036,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.6715106891429312e-06,
+      "loss": 0.2096,
+      "step": 14190
+    },
+    {
+      "epoch": 0.9171349221727055,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.6585932958728929e-06,
+      "loss": 0.2148,
+      "step": 14200
+    },
+    {
+      "epoch": 0.9177807918362074,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.645675902602855e-06,
+      "loss": 0.212,
+      "step": 14210
+    },
+    {
+      "epoch": 0.9184266614997093,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.6327585093328168e-06,
+      "loss": 0.2286,
+      "step": 14220
+    },
+    {
+      "epoch": 0.9190725311632113,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.6198411160627787e-06,
+      "loss": 0.2263,
+      "step": 14230
+    },
+    {
+      "epoch": 0.9197184008267132,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.6069237227927406e-06,
+      "loss": 0.2046,
+      "step": 14240
+    },
+    {
+      "epoch": 0.9203642704902151,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.5940063295227024e-06,
+      "loss": 0.2153,
+      "step": 14250
+    },
+    {
+      "epoch": 0.921010140153717,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.5810889362526643e-06,
+      "loss": 0.2215,
+      "step": 14260
+    },
+    {
+      "epoch": 0.9216560098172188,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.5681715429826264e-06,
+      "loss": 0.2487,
+      "step": 14270
+    },
+    {
+      "epoch": 0.9223018794807208,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.555254149712588e-06,
+      "loss": 0.2326,
+      "step": 14280
+    },
+    {
+      "epoch": 0.9229477491442227,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.5423367564425501e-06,
+      "loss": 0.2476,
+      "step": 14290
+    },
+    {
+      "epoch": 0.9235936188077246,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.5294193631725118e-06,
+      "loss": 0.2299,
+      "step": 14300
+    },
+    {
+      "epoch": 0.9242394884712265,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.5165019699024739e-06,
+      "loss": 0.2373,
+      "step": 14310
+    },
+    {
+      "epoch": 0.9248853581347284,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5035845766324355e-06,
+      "loss": 0.2055,
+      "step": 14320
+    },
+    {
+      "epoch": 0.9255312277982303,
+      "grad_norm": 1.125,
+      "learning_rate": 1.4906671833623976e-06,
+      "loss": 0.2434,
+      "step": 14330
+    },
+    {
+      "epoch": 0.9261770974617323,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.4777497900923597e-06,
+      "loss": 0.2256,
+      "step": 14340
+    },
+    {
+      "epoch": 0.9268229671252342,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.4648323968223213e-06,
+      "loss": 0.2611,
+      "step": 14350
+    },
+    {
+      "epoch": 0.927468836788736,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.4519150035522834e-06,
+      "loss": 0.273,
+      "step": 14360
+    },
+    {
+      "epoch": 0.9281147064522379,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.438997610282245e-06,
+      "loss": 0.22,
+      "step": 14370
+    },
+    {
+      "epoch": 0.9287605761157398,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.4260802170122071e-06,
+      "loss": 0.2239,
+      "step": 14380
+    },
+    {
+      "epoch": 0.9294064457792417,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.413162823742169e-06,
+      "loss": 0.2509,
+      "step": 14390
+    },
+    {
+      "epoch": 0.9300523154427437,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4002454304721309e-06,
+      "loss": 0.2163,
+      "step": 14400
+    },
+    {
+      "epoch": 0.9306981851062456,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.3873280372020927e-06,
+      "loss": 0.2509,
+      "step": 14410
+    },
+    {
+      "epoch": 0.9313440547697475,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.3744106439320546e-06,
+      "loss": 0.2203,
+      "step": 14420
+    },
+    {
+      "epoch": 0.9319899244332494,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.3614932506620165e-06,
+      "loss": 0.2271,
+      "step": 14430
+    },
+    {
+      "epoch": 0.9326357940967512,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.3485758573919783e-06,
+      "loss": 0.2165,
+      "step": 14440
+    },
+    {
+      "epoch": 0.9332816637602532,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.3356584641219402e-06,
+      "loss": 0.2613,
+      "step": 14450
+    },
+    {
+      "epoch": 0.9339275334237551,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.3227410708519023e-06,
+      "loss": 0.2163,
+      "step": 14460
+    },
+    {
+      "epoch": 0.934573403087257,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.309823677581864e-06,
+      "loss": 0.2471,
+      "step": 14470
+    },
+    {
+      "epoch": 0.9352192727507589,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.296906284311826e-06,
+      "loss": 0.2347,
+      "step": 14480
+    },
+    {
+      "epoch": 0.9358651424142608,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.2839888910417877e-06,
+      "loss": 0.2137,
+      "step": 14490
+    },
+    {
+      "epoch": 0.9365110120777627,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.2710714977717498e-06,
+      "loss": 0.1864,
+      "step": 14500
+    },
+    {
+      "epoch": 0.9371568817412647,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.2581541045017118e-06,
+      "loss": 0.2092,
+      "step": 14510
+    },
+    {
+      "epoch": 0.9378027514047665,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.2452367112316735e-06,
+      "loss": 0.2157,
+      "step": 14520
+    },
+    {
+      "epoch": 0.9384486210682684,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2323193179616356e-06,
+      "loss": 0.2132,
+      "step": 14530
+    },
+    {
+      "epoch": 0.9390944907317703,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.2194019246915974e-06,
+      "loss": 0.2278,
+      "step": 14540
+    },
+    {
+      "epoch": 0.9397403603952722,
+      "grad_norm": 1.5,
+      "learning_rate": 1.2064845314215593e-06,
+      "loss": 0.2301,
+      "step": 14550
+    },
+    {
+      "epoch": 0.9403862300587741,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1935671381515212e-06,
+      "loss": 0.2384,
+      "step": 14560
+    },
+    {
+      "epoch": 0.9410320997222761,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.180649744881483e-06,
+      "loss": 0.2372,
+      "step": 14570
+    },
+    {
+      "epoch": 0.941677969385778,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.167732351611445e-06,
+      "loss": 0.2263,
+      "step": 14580
+    },
+    {
+      "epoch": 0.9423238390492799,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.1548149583414068e-06,
+      "loss": 0.2134,
+      "step": 14590
+    },
+    {
+      "epoch": 0.9429697087127817,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.1418975650713686e-06,
+      "loss": 0.218,
+      "step": 14600
+    },
+    {
+      "epoch": 0.9436155783762836,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.1289801718013305e-06,
+      "loss": 0.2383,
+      "step": 14610
+    },
+    {
+      "epoch": 0.9442614480397856,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.1160627785312924e-06,
+      "loss": 0.2381,
+      "step": 14620
+    },
+    {
+      "epoch": 0.9449073177032875,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.1031453852612543e-06,
+      "loss": 0.2117,
+      "step": 14630
+    },
+    {
+      "epoch": 0.9455531873667894,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.0902279919912161e-06,
+      "loss": 0.2245,
+      "step": 14640
+    },
+    {
+      "epoch": 0.9461990570302913,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0773105987211782e-06,
+      "loss": 0.2186,
+      "step": 14650
+    },
+    {
+      "epoch": 0.9468449266937932,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.06439320545114e-06,
+      "loss": 0.2165,
+      "step": 14660
+    },
+    {
+      "epoch": 0.947490796357295,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.051475812181102e-06,
+      "loss": 0.2171,
+      "step": 14670
+    },
+    {
+      "epoch": 0.948136666020797,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.0385584189110638e-06,
+      "loss": 0.2335,
+      "step": 14680
+    },
+    {
+      "epoch": 0.9487825356842989,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.2102,
+      "step": 14690
+    },
+    {
+      "epoch": 0.9494284053478008,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.0127236323709875e-06,
+      "loss": 0.2354,
+      "step": 14700
+    },
+    {
+      "epoch": 0.9500742750113027,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.998062391009496e-07,
+      "loss": 0.2335,
+      "step": 14710
+    },
+    {
+      "epoch": 0.9507201446748046,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.868888458309115e-07,
+      "loss": 0.2285,
+      "step": 14720
+    },
+    {
+      "epoch": 0.9513660143383065,
+      "grad_norm": 0.875,
+      "learning_rate": 9.739714525608734e-07,
+      "loss": 0.2098,
+      "step": 14730
+    },
+    {
+      "epoch": 0.9520118840018085,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.610540592908352e-07,
+      "loss": 0.2188,
+      "step": 14740
+    },
+    {
+      "epoch": 0.9526577536653104,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.481366660207971e-07,
+      "loss": 0.2318,
+      "step": 14750
+    },
+    {
+      "epoch": 0.9533036233288122,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.35219272750759e-07,
+      "loss": 0.2758,
+      "step": 14760
+    },
+    {
+      "epoch": 0.9539494929923141,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.223018794807209e-07,
+      "loss": 0.2229,
+      "step": 14770
+    },
+    {
+      "epoch": 0.954595362655816,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.093844862106828e-07,
+      "loss": 0.1928,
+      "step": 14780
+    },
+    {
+      "epoch": 0.955241232319318,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.964670929406447e-07,
+      "loss": 0.2239,
+      "step": 14790
+    },
+    {
+      "epoch": 0.9558871019828199,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.835496996706065e-07,
+      "loss": 0.2304,
+      "step": 14800
+    },
+    {
+      "epoch": 0.9565329716463218,
+      "grad_norm": 1.125,
+      "learning_rate": 8.706323064005684e-07,
+      "loss": 0.2196,
+      "step": 14810
+    },
+    {
+      "epoch": 0.9571788413098237,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.577149131305303e-07,
+      "loss": 0.2376,
+      "step": 14820
+    },
+    {
+      "epoch": 0.9578247109733256,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.447975198604922e-07,
+      "loss": 0.2115,
+      "step": 14830
+    },
+    {
+      "epoch": 0.9584705806368274,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.318801265904541e-07,
+      "loss": 0.2222,
+      "step": 14840
+    },
+    {
+      "epoch": 0.9591164503003294,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.18962733320416e-07,
+      "loss": 0.2235,
+      "step": 14850
+    },
+    {
+      "epoch": 0.9597623199638313,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.060453400503778e-07,
+      "loss": 0.223,
+      "step": 14860
+    },
+    {
+      "epoch": 0.9604081896273332,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.931279467803397e-07,
+      "loss": 0.242,
+      "step": 14870
+    },
+    {
+      "epoch": 0.9610540592908351,
+      "grad_norm": 1.0,
+      "learning_rate": 7.802105535103016e-07,
+      "loss": 0.24,
+      "step": 14880
+    },
+    {
+      "epoch": 0.961699928954337,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.672931602402637e-07,
+      "loss": 0.2555,
+      "step": 14890
+    },
+    {
+      "epoch": 0.9623457986178389,
+      "grad_norm": 1.375,
+      "learning_rate": 7.543757669702255e-07,
+      "loss": 0.2124,
+      "step": 14900
+    },
+    {
+      "epoch": 0.9629916682813409,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.414583737001874e-07,
+      "loss": 0.2361,
+      "step": 14910
+    },
+    {
+      "epoch": 0.9636375379448427,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.285409804301493e-07,
+      "loss": 0.2524,
+      "step": 14920
+    },
+    {
+      "epoch": 0.9642834076083446,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.156235871601111e-07,
+      "loss": 0.2455,
+      "step": 14930
+    },
+    {
+      "epoch": 0.9649292772718465,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.027061938900731e-07,
+      "loss": 0.2373,
+      "step": 14940
+    },
+    {
+      "epoch": 0.9655751469353484,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.89788800620035e-07,
+      "loss": 0.2175,
+      "step": 14950
+    },
+    {
+      "epoch": 0.9662210165988504,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.768714073499968e-07,
+      "loss": 0.2294,
+      "step": 14960
+    },
+    {
+      "epoch": 0.9668668862623523,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.639540140799587e-07,
+      "loss": 0.2087,
+      "step": 14970
+    },
+    {
+      "epoch": 0.9675127559258542,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.510366208099206e-07,
+      "loss": 0.2346,
+      "step": 14980
+    },
+    {
+      "epoch": 0.9681586255893561,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.381192275398824e-07,
+      "loss": 0.2342,
+      "step": 14990
+    },
+    {
+      "epoch": 0.968804495252858,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.252018342698444e-07,
+      "loss": 0.2246,
+      "step": 15000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 15483,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5678418788963066e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}