diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10360 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997456549385333,
+  "eval_steps": 500,
+  "global_step": 1474,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006782534972445952,
+      "grad_norm": 4.1065765912581975,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 1.6818,
+      "step": 1
+    },
+    {
+      "epoch": 0.0013565069944891904,
+      "grad_norm": 3.1768641966000803,
+      "learning_rate": 1.777777777777778e-06,
+      "loss": 1.6016,
+      "step": 2
+    },
+    {
+      "epoch": 0.0020347604917337857,
+      "grad_norm": 4.208842639816025,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.694,
+      "step": 3
+    },
+    {
+      "epoch": 0.0027130139889783808,
+      "grad_norm": 4.144597866071873,
+      "learning_rate": 3.555555555555556e-06,
+      "loss": 1.6743,
+      "step": 4
+    },
+    {
+      "epoch": 0.003391267486222976,
+      "grad_norm": 4.367675036005987,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.7215,
+      "step": 5
+    },
+    {
+      "epoch": 0.004069520983467571,
+      "grad_norm": 3.77451956658442,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.7412,
+      "step": 6
+    },
+    {
+      "epoch": 0.004747774480712166,
+      "grad_norm": 2.828341566261712,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 1.5605,
+      "step": 7
+    },
+    {
+      "epoch": 0.0054260279779567615,
+      "grad_norm": 2.1739524057415145,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 1.6683,
+      "step": 8
+    },
+    {
+      "epoch": 0.006104281475201356,
+      "grad_norm": 1.6980194638479158,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.6685,
+      "step": 9
+    },
+    {
+      "epoch": 0.006782534972445952,
+      "grad_norm": 1.5011794096474325,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.648,
+      "step": 10
+    },
+    {
+      "epoch": 0.007460788469690547,
+      "grad_norm": 2.322587769441417,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 1.686,
+      "step": 11
+    },
+    {
+      "epoch": 0.008139041966935143,
+      "grad_norm": 1.705227880799364,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 1.4842,
+      "step": 12
+    },
+    {
+      "epoch": 0.008817295464179737,
+      "grad_norm": 2.7772036835193217,
+      "learning_rate": 1.1555555555555556e-05,
+      "loss": 1.6403,
+      "step": 13
+    },
+    {
+      "epoch": 0.009495548961424332,
+      "grad_norm": 2.921483047712852,
+      "learning_rate": 1.2444444444444446e-05,
+      "loss": 1.6296,
+      "step": 14
+    },
+    {
+      "epoch": 0.010173802458668928,
+      "grad_norm": 2.8535382585231237,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.6325,
+      "step": 15
+    },
+    {
+      "epoch": 0.010852055955913523,
+      "grad_norm": 2.6523988159853955,
+      "learning_rate": 1.4222222222222224e-05,
+      "loss": 1.6585,
+      "step": 16
+    },
+    {
+      "epoch": 0.011530309453158118,
+      "grad_norm": 2.4275237868017268,
+      "learning_rate": 1.5111111111111112e-05,
+      "loss": 1.6138,
+      "step": 17
+    },
+    {
+      "epoch": 0.012208562950402712,
+      "grad_norm": 1.9022283612641862,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.6117,
+      "step": 18
+    },
+    {
+      "epoch": 0.012886816447647309,
+      "grad_norm": 1.4961663445429363,
+      "learning_rate": 1.688888888888889e-05,
+      "loss": 1.5869,
+      "step": 19
+    },
+    {
+      "epoch": 0.013565069944891903,
+      "grad_norm": 1.2530261801072107,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 1.6136,
+      "step": 20
+    },
+    {
+      "epoch": 0.014243323442136498,
+      "grad_norm": 1.3729319190613,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 1.5903,
+      "step": 21
+    },
+    {
+      "epoch": 0.014921576939381094,
+      "grad_norm": 1.6164121258819937,
+      "learning_rate": 1.9555555555555557e-05,
+      "loss": 1.5962,
+      "step": 22
+    },
+    {
+      "epoch": 0.015599830436625689,
+      "grad_norm": 1.868473824482503,
+      "learning_rate": 2.0444444444444446e-05,
+      "loss": 1.5636,
+      "step": 23
+    },
+    {
+      "epoch": 0.016278083933870285,
+      "grad_norm": 1.7300543140348683,
+      "learning_rate": 2.1333333333333335e-05,
+      "loss": 1.5771,
+      "step": 24
+    },
+    {
+      "epoch": 0.01695633743111488,
+      "grad_norm": 1.5738361080005434,
+      "learning_rate": 2.2222222222222227e-05,
+      "loss": 1.5466,
+      "step": 25
+    },
+    {
+      "epoch": 0.017634590928359475,
+      "grad_norm": 1.3215383127512057,
+      "learning_rate": 2.3111111111111112e-05,
+      "loss": 1.5627,
+      "step": 26
+    },
+    {
+      "epoch": 0.01831284442560407,
+      "grad_norm": 1.1786255386020765,
+      "learning_rate": 2.4e-05,
+      "loss": 1.5864,
+      "step": 27
+    },
+    {
+      "epoch": 0.018991097922848664,
+      "grad_norm": 1.0543471785756497,
+      "learning_rate": 2.4888888888888893e-05,
+      "loss": 1.515,
+      "step": 28
+    },
+    {
+      "epoch": 0.01966935142009326,
+      "grad_norm": 1.044341262007584,
+      "learning_rate": 2.577777777777778e-05,
+      "loss": 1.5272,
+      "step": 29
+    },
+    {
+      "epoch": 0.020347604917337857,
+      "grad_norm": 1.1429084132740057,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.5751,
+      "step": 30
+    },
+    {
+      "epoch": 0.02102585841458245,
+      "grad_norm": 1.1720155421313065,
+      "learning_rate": 2.755555555555556e-05,
+      "loss": 1.5623,
+      "step": 31
+    },
+    {
+      "epoch": 0.021704111911827046,
+      "grad_norm": 1.0700523128168293,
+      "learning_rate": 2.8444444444444447e-05,
+      "loss": 1.5182,
+      "step": 32
+    },
+    {
+      "epoch": 0.02238236540907164,
+      "grad_norm": 0.9851762418398425,
+      "learning_rate": 2.9333333333333333e-05,
+      "loss": 1.5367,
+      "step": 33
+    },
+    {
+      "epoch": 0.023060618906316235,
+      "grad_norm": 0.9364787529061317,
+      "learning_rate": 3.0222222222222225e-05,
+      "loss": 1.5335,
+      "step": 34
+    },
+    {
+      "epoch": 0.02373887240356083,
+      "grad_norm": 0.8677156668511112,
+      "learning_rate": 3.111111111111112e-05,
+      "loss": 1.53,
+      "step": 35
+    },
+    {
+      "epoch": 0.024417125900805425,
+      "grad_norm": 0.9026523779646922,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.497,
+      "step": 36
+    },
+    {
+      "epoch": 0.02509537939805002,
+      "grad_norm": 0.8913970222309001,
+      "learning_rate": 3.288888888888889e-05,
+      "loss": 1.4952,
+      "step": 37
+    },
+    {
+      "epoch": 0.025773632895294617,
+      "grad_norm": 0.8834806762413843,
+      "learning_rate": 3.377777777777778e-05,
+      "loss": 1.486,
+      "step": 38
+    },
+    {
+      "epoch": 0.02645188639253921,
+      "grad_norm": 0.8387084982910075,
+      "learning_rate": 3.466666666666667e-05,
+      "loss": 1.5111,
+      "step": 39
+    },
+    {
+      "epoch": 0.027130139889783807,
+      "grad_norm": 0.8002050518805109,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 1.4901,
+      "step": 40
+    },
+    {
+      "epoch": 0.027808393387028403,
+      "grad_norm": 0.8499910835361242,
+      "learning_rate": 3.644444444444445e-05,
+      "loss": 1.5206,
+      "step": 41
+    },
+    {
+      "epoch": 0.028486646884272996,
+      "grad_norm": 0.8341993828007787,
+      "learning_rate": 3.733333333333334e-05,
+      "loss": 1.5024,
+      "step": 42
+    },
+    {
+      "epoch": 0.029164900381517592,
+      "grad_norm": 0.880517862113079,
+      "learning_rate": 3.8222222222222226e-05,
+      "loss": 1.5103,
+      "step": 43
+    },
+    {
+      "epoch": 0.02984315387876219,
+      "grad_norm": 0.8096099002799838,
+      "learning_rate": 3.9111111111111115e-05,
+      "loss": 1.4883,
+      "step": 44
+    },
+    {
+      "epoch": 0.03052140737600678,
+      "grad_norm": 0.8050025671092632,
+      "learning_rate": 4e-05,
+      "loss": 1.4605,
+      "step": 45
+    },
+    {
+      "epoch": 0.031199660873251378,
+      "grad_norm": 0.8660853245091175,
+      "learning_rate": 3.999995166796149e-05,
+      "loss": 1.5147,
+      "step": 46
+    },
+    {
+      "epoch": 0.031877914370495974,
+      "grad_norm": 0.7562387786559109,
+      "learning_rate": 3.999980667207955e-05,
+      "loss": 1.4858,
+      "step": 47
+    },
+    {
+      "epoch": 0.03255616786774057,
+      "grad_norm": 0.7875069445374301,
+      "learning_rate": 3.9999565013054966e-05,
+      "loss": 1.4917,
+      "step": 48
+    },
+    {
+      "epoch": 0.03323442136498516,
+      "grad_norm": 0.80585752298433,
+      "learning_rate": 3.999922669205574e-05,
+      "loss": 1.4537,
+      "step": 49
+    },
+    {
+      "epoch": 0.03391267486222976,
+      "grad_norm": 0.8472273049831842,
+      "learning_rate": 3.9998791710717044e-05,
+      "loss": 1.4683,
+      "step": 50
+    },
+    {
+      "epoch": 0.03459092835947435,
+      "grad_norm": 0.8143071540357201,
+      "learning_rate": 3.999826007114122e-05,
+      "loss": 1.5072,
+      "step": 51
+    },
+    {
+      "epoch": 0.03526918185671895,
+      "grad_norm": 0.7861158344158029,
+      "learning_rate": 3.99976317758978e-05,
+      "loss": 1.468,
+      "step": 52
+    },
+    {
+      "epoch": 0.035947435353963546,
+      "grad_norm": 0.734537069409861,
+      "learning_rate": 3.999690682802347e-05,
+      "loss": 1.493,
+      "step": 53
+    },
+    {
+      "epoch": 0.03662568885120814,
+      "grad_norm": 0.7716856308294497,
+      "learning_rate": 3.9996085231022037e-05,
+      "loss": 1.4558,
+      "step": 54
+    },
+    {
+      "epoch": 0.03730394234845273,
+      "grad_norm": 0.8134849509429027,
+      "learning_rate": 3.9995166988864454e-05,
+      "loss": 1.4799,
+      "step": 55
+    },
+    {
+      "epoch": 0.03798219584569733,
+      "grad_norm": 0.7510660632610676,
+      "learning_rate": 3.999415210598877e-05,
+      "loss": 1.4794,
+      "step": 56
+    },
+    {
+      "epoch": 0.038660449342941924,
+      "grad_norm": 0.7615105320958904,
+      "learning_rate": 3.999304058730012e-05,
+      "loss": 1.4873,
+      "step": 57
+    },
+    {
+      "epoch": 0.03933870284018652,
+      "grad_norm": 0.7548985780285249,
+      "learning_rate": 3.9991832438170706e-05,
+      "loss": 1.4626,
+      "step": 58
+    },
+    {
+      "epoch": 0.04001695633743112,
+      "grad_norm": 0.7707075108441955,
+      "learning_rate": 3.999052766443975e-05,
+      "loss": 1.4955,
+      "step": 59
+    },
+    {
+      "epoch": 0.040695209834675714,
+      "grad_norm": 0.7796824365420911,
+      "learning_rate": 3.99891262724135e-05,
+      "loss": 1.4633,
+      "step": 60
+    },
+    {
+      "epoch": 0.0413734633319203,
+      "grad_norm": 1.6634166289908014,
+      "learning_rate": 3.998762826886516e-05,
+      "loss": 1.5228,
+      "step": 61
+    },
+    {
+      "epoch": 0.0420517168291649,
+      "grad_norm": 0.780351811457274,
+      "learning_rate": 3.998603366103489e-05,
+      "loss": 1.4538,
+      "step": 62
+    },
+    {
+      "epoch": 0.042729970326409496,
+      "grad_norm": 0.8635411707323375,
+      "learning_rate": 3.9984342456629754e-05,
+      "loss": 1.4477,
+      "step": 63
+    },
+    {
+      "epoch": 0.04340822382365409,
+      "grad_norm": 0.8326874431336305,
+      "learning_rate": 3.998255466382369e-05,
+      "loss": 1.4486,
+      "step": 64
+    },
+    {
+      "epoch": 0.04408647732089869,
+      "grad_norm": 0.7723770463913483,
+      "learning_rate": 3.998067029125746e-05,
+      "loss": 1.45,
+      "step": 65
+    },
+    {
+      "epoch": 0.04476473081814328,
+      "grad_norm": 0.7740863591041413,
+      "learning_rate": 3.9978689348038635e-05,
+      "loss": 1.4506,
+      "step": 66
+    },
+    {
+      "epoch": 0.045442984315387874,
+      "grad_norm": 0.7484072365132953,
+      "learning_rate": 3.99766118437415e-05,
+      "loss": 1.4502,
+      "step": 67
+    },
+    {
+      "epoch": 0.04612123781263247,
+      "grad_norm": 0.7800723070156931,
+      "learning_rate": 3.997443778840707e-05,
+      "loss": 1.4607,
+      "step": 68
+    },
+    {
+      "epoch": 0.04679949130987707,
+      "grad_norm": 0.7720573548329548,
+      "learning_rate": 3.997216719254298e-05,
+      "loss": 1.4343,
+      "step": 69
+    },
+    {
+      "epoch": 0.04747774480712166,
+      "grad_norm": 0.7419949116021992,
+      "learning_rate": 3.9969800067123503e-05,
+      "loss": 1.4575,
+      "step": 70
+    },
+    {
+      "epoch": 0.04815599830436626,
+      "grad_norm": 0.7757764657684928,
+      "learning_rate": 3.9967336423589425e-05,
+      "loss": 1.4346,
+      "step": 71
+    },
+    {
+      "epoch": 0.04883425180161085,
+      "grad_norm": 0.7372814045835184,
+      "learning_rate": 3.9964776273848044e-05,
+      "loss": 1.4572,
+      "step": 72
+    },
+    {
+      "epoch": 0.049512505298855446,
+      "grad_norm": 0.7861776737925739,
+      "learning_rate": 3.996211963027309e-05,
+      "loss": 1.4342,
+      "step": 73
+    },
+    {
+      "epoch": 0.05019075879610004,
+      "grad_norm": 0.7705824940277366,
+      "learning_rate": 3.9959366505704646e-05,
+      "loss": 1.4077,
+      "step": 74
+    },
+    {
+      "epoch": 0.05086901229334464,
+      "grad_norm": 0.7550842878169225,
+      "learning_rate": 3.995651691344914e-05,
+      "loss": 1.4187,
+      "step": 75
+    },
+    {
+      "epoch": 0.051547265790589235,
+      "grad_norm": 0.7935364668605709,
+      "learning_rate": 3.9953570867279225e-05,
+      "loss": 1.4359,
+      "step": 76
+    },
+    {
+      "epoch": 0.05222551928783383,
+      "grad_norm": 0.7682454471435735,
+      "learning_rate": 3.995052838143375e-05,
+      "loss": 1.4539,
+      "step": 77
+    },
+    {
+      "epoch": 0.05290377278507842,
+      "grad_norm": 0.7339626264298819,
+      "learning_rate": 3.994738947061766e-05,
+      "loss": 1.4596,
+      "step": 78
+    },
+    {
+      "epoch": 0.05358202628232302,
+      "grad_norm": 0.8142559965687818,
+      "learning_rate": 3.9944154150001956e-05,
+      "loss": 1.4757,
+      "step": 79
+    },
+    {
+      "epoch": 0.05426027977956761,
+      "grad_norm": 0.8023956266998655,
+      "learning_rate": 3.9940822435223596e-05,
+      "loss": 1.4541,
+      "step": 80
+    },
+    {
+      "epoch": 0.05493853327681221,
+      "grad_norm": 0.727937829181823,
+      "learning_rate": 3.993739434238545e-05,
+      "loss": 1.4323,
+      "step": 81
+    },
+    {
+      "epoch": 0.055616786774056806,
+      "grad_norm": 0.7209236130647971,
+      "learning_rate": 3.993386988805617e-05,
+      "loss": 1.435,
+      "step": 82
+    },
+    {
+      "epoch": 0.056295040271301396,
+      "grad_norm": 0.7707170493909989,
+      "learning_rate": 3.9930249089270185e-05,
+      "loss": 1.4307,
+      "step": 83
+    },
+    {
+      "epoch": 0.05697329376854599,
+      "grad_norm": 0.8208154253181018,
+      "learning_rate": 3.992653196352754e-05,
+      "loss": 1.4446,
+      "step": 84
+    },
+    {
+      "epoch": 0.05765154726579059,
+      "grad_norm": 0.7446657999243042,
+      "learning_rate": 3.9922718528793866e-05,
+      "loss": 1.4177,
+      "step": 85
+    },
+    {
+      "epoch": 0.058329800763035185,
+      "grad_norm": 0.7523922030736101,
+      "learning_rate": 3.991880880350026e-05,
+      "loss": 1.3962,
+      "step": 86
+    },
+    {
+      "epoch": 0.05900805426027978,
+      "grad_norm": 0.7154458001918853,
+      "learning_rate": 3.991480280654323e-05,
+      "loss": 1.4148,
+      "step": 87
+    },
+    {
+      "epoch": 0.05968630775752438,
+      "grad_norm": 0.7787531051472882,
+      "learning_rate": 3.991070055728458e-05,
+      "loss": 1.4224,
+      "step": 88
+    },
+    {
+      "epoch": 0.06036456125476897,
+      "grad_norm": 0.7284970405300326,
+      "learning_rate": 3.9906502075551314e-05,
+      "loss": 1.3971,
+      "step": 89
+    },
+    {
+      "epoch": 0.06104281475201356,
+      "grad_norm": 0.7290304582501952,
+      "learning_rate": 3.9902207381635544e-05,
+      "loss": 1.3923,
+      "step": 90
+    },
+    {
+      "epoch": 0.06172106824925816,
+      "grad_norm": 0.7250605798466435,
+      "learning_rate": 3.989781649629441e-05,
+      "loss": 1.4145,
+      "step": 91
+    },
+    {
+      "epoch": 0.062399321746502756,
+      "grad_norm": 0.8072790558594046,
+      "learning_rate": 3.989332944074995e-05,
+      "loss": 1.4557,
+      "step": 92
+    },
+    {
+      "epoch": 0.06307757524374735,
+      "grad_norm": 0.7814578631907936,
+      "learning_rate": 3.9888746236689014e-05,
+      "loss": 1.4243,
+      "step": 93
+    },
+    {
+      "epoch": 0.06375582874099195,
+      "grad_norm": 0.7348756701651429,
+      "learning_rate": 3.988406690626317e-05,
+      "loss": 1.4493,
+      "step": 94
+    },
+    {
+      "epoch": 0.06443408223823655,
+      "grad_norm": 0.7763605868250318,
+      "learning_rate": 3.987929147208857e-05,
+      "loss": 1.459,
+      "step": 95
+    },
+    {
+      "epoch": 0.06511233573548114,
+      "grad_norm": 0.7335967317515184,
+      "learning_rate": 3.987441995724587e-05,
+      "loss": 1.4323,
+      "step": 96
+    },
+    {
+      "epoch": 0.06579058923272574,
+      "grad_norm": 0.7970588880633319,
+      "learning_rate": 3.986945238528009e-05,
+      "loss": 1.431,
+      "step": 97
+    },
+    {
+      "epoch": 0.06646884272997032,
+      "grad_norm": 0.7882918020664678,
+      "learning_rate": 3.9864388780200514e-05,
+      "loss": 1.444,
+      "step": 98
+    },
+    {
+      "epoch": 0.06714709622721492,
+      "grad_norm": 0.7709405021384093,
+      "learning_rate": 3.985922916648058e-05,
+      "loss": 1.4113,
+      "step": 99
+    },
+    {
+      "epoch": 0.06782534972445951,
+      "grad_norm": 0.7372128644657954,
+      "learning_rate": 3.985397356905774e-05,
+      "loss": 1.4276,
+      "step": 100
+    },
+    {
+      "epoch": 0.06850360322170411,
+      "grad_norm": 0.8402512733533295,
+      "learning_rate": 3.984862201333339e-05,
+      "loss": 1.4214,
+      "step": 101
+    },
+    {
+      "epoch": 0.0691818567189487,
+      "grad_norm": 0.7775767209846817,
+      "learning_rate": 3.9843174525172686e-05,
+      "loss": 1.4241,
+      "step": 102
+    },
+    {
+      "epoch": 0.0698601102161933,
+      "grad_norm": 0.8803258285981507,
+      "learning_rate": 3.983763113090444e-05,
+      "loss": 1.5175,
+      "step": 103
+    },
+    {
+      "epoch": 0.0705383637134379,
+      "grad_norm": 0.9085210303547249,
+      "learning_rate": 3.9831991857321e-05,
+      "loss": 1.4321,
+      "step": 104
+    },
+    {
+      "epoch": 0.0712166172106825,
+      "grad_norm": 0.8413616396974392,
+      "learning_rate": 3.982625673167814e-05,
+      "loss": 1.4639,
+      "step": 105
+    },
+    {
+      "epoch": 0.07189487070792709,
+      "grad_norm": 0.7801398149333048,
+      "learning_rate": 3.982042578169489e-05,
+      "loss": 1.4014,
+      "step": 106
+    },
+    {
+      "epoch": 0.07257312420517169,
+      "grad_norm": 0.7866846001446234,
+      "learning_rate": 3.981449903555341e-05,
+      "loss": 1.4121,
+      "step": 107
+    },
+    {
+      "epoch": 0.07325137770241628,
+      "grad_norm": 0.7708520359423757,
+      "learning_rate": 3.980847652189888e-05,
+      "loss": 1.3839,
+      "step": 108
+    },
+    {
+      "epoch": 0.07392963119966087,
+      "grad_norm": 0.830761051025023,
+      "learning_rate": 3.980235826983933e-05,
+      "loss": 1.3914,
+      "step": 109
+    },
+    {
+      "epoch": 0.07460788469690546,
+      "grad_norm": 0.8253907698028418,
+      "learning_rate": 3.979614430894553e-05,
+      "loss": 1.3954,
+      "step": 110
+    },
+    {
+      "epoch": 0.07528613819415006,
+      "grad_norm": 0.7112323297589688,
+      "learning_rate": 3.9789834669250804e-05,
+      "loss": 1.4181,
+      "step": 111
+    },
+    {
+      "epoch": 0.07596439169139466,
+      "grad_norm": 0.8057491331242786,
+      "learning_rate": 3.978342938125094e-05,
+      "loss": 1.4377,
+      "step": 112
+    },
+    {
+      "epoch": 0.07664264518863925,
+      "grad_norm": 0.7298327963697059,
+      "learning_rate": 3.9776928475904e-05,
+      "loss": 1.4706,
+      "step": 113
+    },
+    {
+      "epoch": 0.07732089868588385,
+      "grad_norm": 0.9346906876193559,
+      "learning_rate": 3.9770331984630176e-05,
+      "loss": 1.4209,
+      "step": 114
+    },
+    {
+      "epoch": 0.07799915218312845,
+      "grad_norm": 0.800555245189771,
+      "learning_rate": 3.9763639939311664e-05,
+      "loss": 1.4042,
+      "step": 115
+    },
+    {
+      "epoch": 0.07867740568037304,
+      "grad_norm": 0.7820837234183521,
+      "learning_rate": 3.9756852372292475e-05,
+      "loss": 1.3983,
+      "step": 116
+    },
+    {
+      "epoch": 0.07935565917761764,
+      "grad_norm": 0.7759352883815379,
+      "learning_rate": 3.974996931637831e-05,
+      "loss": 1.4158,
+      "step": 117
+    },
+    {
+      "epoch": 0.08003391267486223,
+      "grad_norm": 0.8440621723868824,
+      "learning_rate": 3.974299080483638e-05,
+      "loss": 1.4007,
+      "step": 118
+    },
+    {
+      "epoch": 0.08071216617210683,
+      "grad_norm": 0.6400735452279184,
+      "learning_rate": 3.973591687139526e-05,
+      "loss": 1.4793,
+      "step": 119
+    },
+    {
+      "epoch": 0.08139041966935143,
+      "grad_norm": 0.8313238775827508,
+      "learning_rate": 3.97287475502447e-05,
+      "loss": 1.3954,
+      "step": 120
+    },
+    {
+      "epoch": 0.08206867316659601,
+      "grad_norm": 0.806251290610579,
+      "learning_rate": 3.97214828760355e-05,
+      "loss": 1.4117,
+      "step": 121
+    },
+    {
+      "epoch": 0.0827469266638406,
+      "grad_norm": 0.7712317181214216,
+      "learning_rate": 3.971412288387931e-05,
+      "loss": 1.3757,
+      "step": 122
+    },
+    {
+      "epoch": 0.0834251801610852,
+      "grad_norm": 0.7590372689247342,
+      "learning_rate": 3.970666760934846e-05,
+      "loss": 1.3929,
+      "step": 123
+    },
+    {
+      "epoch": 0.0841034336583298,
+      "grad_norm": 0.8632264568669091,
+      "learning_rate": 3.969911708847583e-05,
+      "loss": 1.4347,
+      "step": 124
+    },
+    {
+      "epoch": 0.0847816871555744,
+      "grad_norm": 0.8379449866062499,
+      "learning_rate": 3.9691471357754616e-05,
+      "loss": 1.3956,
+      "step": 125
+    },
+    {
+      "epoch": 0.08545994065281899,
+      "grad_norm": 0.8080362798715246,
+      "learning_rate": 3.9683730454138195e-05,
+      "loss": 1.3934,
+      "step": 126
+    },
+    {
+      "epoch": 0.08613819415006359,
+      "grad_norm": 0.7552807716389017,
+      "learning_rate": 3.967589441503994e-05,
+      "loss": 1.4253,
+      "step": 127
+    },
+    {
+      "epoch": 0.08681644764730818,
+      "grad_norm": 0.7685649759662768,
+      "learning_rate": 3.9667963278333006e-05,
+      "loss": 1.4015,
+      "step": 128
+    },
+    {
+      "epoch": 0.08749470114455278,
+      "grad_norm": 0.7984044689432437,
+      "learning_rate": 3.9659937082350214e-05,
+      "loss": 1.3644,
+      "step": 129
+    },
+    {
+      "epoch": 0.08817295464179738,
+      "grad_norm": 0.7882181163488372,
+      "learning_rate": 3.9651815865883794e-05,
+      "loss": 1.3886,
+      "step": 130
+    },
+    {
+      "epoch": 0.08885120813904197,
+      "grad_norm": 0.7755502826098504,
+      "learning_rate": 3.9643599668185246e-05,
+      "loss": 1.3836,
+      "step": 131
+    },
+    {
+      "epoch": 0.08952946163628656,
+      "grad_norm": 0.7595962684622689,
+      "learning_rate": 3.963528852896513e-05,
+      "loss": 1.3684,
+      "step": 132
+    },
+    {
+      "epoch": 0.09020771513353115,
+      "grad_norm": 0.771937384457462,
+      "learning_rate": 3.9626882488392864e-05,
+      "loss": 1.382,
+      "step": 133
+    },
+    {
+      "epoch": 0.09088596863077575,
+      "grad_norm": 0.7642977324026196,
+      "learning_rate": 3.961838158709657e-05,
+      "loss": 1.3969,
+      "step": 134
+    },
+    {
+      "epoch": 0.09156422212802034,
+      "grad_norm": 0.7782206736594208,
+      "learning_rate": 3.960978586616283e-05,
+      "loss": 1.3762,
+      "step": 135
+    },
+    {
+      "epoch": 0.09224247562526494,
+      "grad_norm": 0.7520850703006238,
+      "learning_rate": 3.9601095367136506e-05,
+      "loss": 1.4047,
+      "step": 136
+    },
+    {
+      "epoch": 0.09292072912250954,
+      "grad_norm": 0.7517437279702506,
+      "learning_rate": 3.959231013202057e-05,
+      "loss": 1.3862,
+      "step": 137
+    },
+    {
+      "epoch": 0.09359898261975413,
+      "grad_norm": 0.7310919198153829,
+      "learning_rate": 3.958343020327585e-05,
+      "loss": 1.4703,
+      "step": 138
+    },
+    {
+      "epoch": 0.09427723611699873,
+      "grad_norm": 0.7567376870493908,
+      "learning_rate": 3.957445562382085e-05,
+      "loss": 1.3923,
+      "step": 139
+    },
+    {
+      "epoch": 0.09495548961424333,
+      "grad_norm": 0.7996642480637515,
+      "learning_rate": 3.956538643703153e-05,
+      "loss": 1.3866,
+      "step": 140
+    },
+    {
+      "epoch": 0.09563374311148792,
+      "grad_norm": 0.5505747838701305,
+      "learning_rate": 3.9556222686741136e-05,
+      "loss": 1.4342,
+      "step": 141
+    },
+    {
+      "epoch": 0.09631199660873252,
+      "grad_norm": 0.790082512074581,
+      "learning_rate": 3.9546964417239926e-05,
+      "loss": 1.3961,
+      "step": 142
+    },
+    {
+      "epoch": 0.0969902501059771,
+      "grad_norm": 0.7523052325850244,
+      "learning_rate": 3.9537611673275017e-05,
+      "loss": 1.3622,
+      "step": 143
+    },
+    {
+      "epoch": 0.0976685036032217,
+      "grad_norm": 0.5987075742859339,
+      "learning_rate": 3.9528164500050116e-05,
+      "loss": 1.4587,
+      "step": 144
+    },
+    {
+      "epoch": 0.0983467571004663,
+      "grad_norm": 0.8487087393690602,
+      "learning_rate": 3.951862294322535e-05,
+      "loss": 1.393,
+      "step": 145
+    },
+    {
+      "epoch": 0.09902501059771089,
+      "grad_norm": 0.7950888183946537,
+      "learning_rate": 3.950898704891699e-05,
+      "loss": 1.3926,
+      "step": 146
+    },
+    {
+      "epoch": 0.09970326409495549,
+      "grad_norm": 0.708739946597553,
+      "learning_rate": 3.94992568636973e-05,
+      "loss": 1.3969,
+      "step": 147
+    },
+    {
+      "epoch": 0.10038151759220008,
+      "grad_norm": 0.7727410734149786,
+      "learning_rate": 3.9489432434594224e-05,
+      "loss": 1.3824,
+      "step": 148
+    },
+    {
+      "epoch": 0.10105977108944468,
+      "grad_norm": 0.7945940885527554,
+      "learning_rate": 3.9479513809091254e-05,
+      "loss": 1.3958,
+      "step": 149
+    },
+    {
+      "epoch": 0.10173802458668928,
+      "grad_norm": 0.7231623401610149,
+      "learning_rate": 3.9469501035127115e-05,
+      "loss": 1.3926,
+      "step": 150
+    },
+    {
+      "epoch": 0.10241627808393387,
+      "grad_norm": 0.736920583862729,
+      "learning_rate": 3.945939416109559e-05,
+      "loss": 1.4025,
+      "step": 151
+    },
+    {
+      "epoch": 0.10309453158117847,
+      "grad_norm": 0.7601601202084843,
+      "learning_rate": 3.9449193235845254e-05,
+      "loss": 1.367,
+      "step": 152
+    },
+    {
+      "epoch": 0.10377278507842307,
+      "grad_norm": 0.7889869577399472,
+      "learning_rate": 3.9438898308679264e-05,
+      "loss": 1.3798,
+      "step": 153
+    },
+    {
+      "epoch": 0.10445103857566766,
+      "grad_norm": 0.7143512752208158,
+      "learning_rate": 3.942850942935511e-05,
+      "loss": 1.3816,
+      "step": 154
+    },
+    {
+      "epoch": 0.10512929207291224,
+      "grad_norm": 0.7817356495329524,
+      "learning_rate": 3.941802664808434e-05,
+      "loss": 1.3656,
+      "step": 155
+    },
+    {
+      "epoch": 0.10580754557015684,
+      "grad_norm": 0.7956257676424011,
+      "learning_rate": 3.9407450015532404e-05,
+      "loss": 1.4003,
+      "step": 156
+    },
+    {
+      "epoch": 0.10648579906740144,
+      "grad_norm": 0.7032546203690988,
+      "learning_rate": 3.93967795828183e-05,
+      "loss": 1.3834,
+      "step": 157
+    },
+    {
+      "epoch": 0.10716405256464603,
+      "grad_norm": 0.8491721584532255,
+      "learning_rate": 3.9386015401514406e-05,
+      "loss": 1.3764,
+      "step": 158
+    },
+    {
+      "epoch": 0.10784230606189063,
+      "grad_norm": 0.7725869502257687,
+      "learning_rate": 3.9375157523646215e-05,
+      "loss": 1.3664,
+      "step": 159
+    },
+    {
+      "epoch": 0.10852055955913523,
+      "grad_norm": 0.7524439551144759,
+      "learning_rate": 3.9364206001692055e-05,
+      "loss": 1.3558,
+      "step": 160
+    },
+    {
+      "epoch": 0.10919881305637982,
+      "grad_norm": 0.7213983940170066,
+      "learning_rate": 3.935316088858287e-05,
+      "loss": 1.3582,
+      "step": 161
+    },
+    {
+      "epoch": 0.10987706655362442,
+      "grad_norm": 0.8731554191497123,
+      "learning_rate": 3.9342022237701945e-05,
+      "loss": 1.4107,
+      "step": 162
+    },
+    {
+      "epoch": 0.11055532005086902,
+      "grad_norm": 0.7752564142114415,
+      "learning_rate": 3.9330790102884646e-05,
+      "loss": 1.3571,
+      "step": 163
+    },
+    {
+      "epoch": 0.11123357354811361,
+      "grad_norm": 0.8516471203966379,
+      "learning_rate": 3.931946453841817e-05,
+      "loss": 1.3965,
+      "step": 164
+    },
+    {
+      "epoch": 0.11191182704535821,
+      "grad_norm": 0.7520205232084405,
+      "learning_rate": 3.930804559904128e-05,
+      "loss": 1.379,
+      "step": 165
+    },
+    {
+      "epoch": 0.11259008054260279,
+      "grad_norm": 0.8043059152508056,
+      "learning_rate": 3.929653333994404e-05,
+      "loss": 1.3993,
+      "step": 166
+    },
+    {
+      "epoch": 0.11326833403984739,
+      "grad_norm": 0.7874837588805026,
+      "learning_rate": 3.928492781676754e-05,
+      "loss": 1.3894,
+      "step": 167
+    },
+    {
+      "epoch": 0.11394658753709198,
+      "grad_norm": 0.7938048672188758,
+      "learning_rate": 3.927322908560364e-05,
+      "loss": 1.3838,
+      "step": 168
+    },
+    {
+      "epoch": 0.11462484103433658,
+      "grad_norm": 0.8203451910611482,
+      "learning_rate": 3.9261437202994696e-05,
+      "loss": 1.3837,
+      "step": 169
+    },
+    {
+      "epoch": 0.11530309453158118,
+      "grad_norm": 0.8329544923845756,
+      "learning_rate": 3.924955222593328e-05,
+      "loss": 1.3669,
+      "step": 170
+    },
+    {
+      "epoch": 0.11598134802882577,
+      "grad_norm": 0.8475717705924518,
+      "learning_rate": 3.92375742118619e-05,
+      "loss": 1.4013,
+      "step": 171
+    },
+    {
+      "epoch": 0.11665960152607037,
+      "grad_norm": 0.7646694203370581,
+      "learning_rate": 3.922550321867276e-05,
+      "loss": 1.3873,
+      "step": 172
+    },
+    {
+      "epoch": 0.11733785502331497,
+      "grad_norm": 0.7826878270694271,
+      "learning_rate": 3.921333930470741e-05,
+      "loss": 1.3869,
+      "step": 173
+    },
+    {
+      "epoch": 0.11801610852055956,
+      "grad_norm": 0.812404806202415,
+      "learning_rate": 3.920108252875653e-05,
+      "loss": 1.3644,
+      "step": 174
+    },
+    {
+      "epoch": 0.11869436201780416,
+      "grad_norm": 0.7434821957582678,
+      "learning_rate": 3.918873295005963e-05,
+      "loss": 1.3521,
+      "step": 175
+    },
+    {
+      "epoch": 0.11937261551504876,
+      "grad_norm": 0.819837272291158,
+      "learning_rate": 3.917629062830473e-05,
+      "loss": 1.3969,
+      "step": 176
+    },
+    {
+      "epoch": 0.12005086901229335,
+      "grad_norm": 0.883516555460024,
+      "learning_rate": 3.9163755623628105e-05,
+      "loss": 1.4598,
+      "step": 177
+    },
+    {
+      "epoch": 0.12072912250953793,
+      "grad_norm": 1.015649489713416,
+      "learning_rate": 3.9151127996614e-05,
+      "loss": 1.4091,
+      "step": 178
+    },
+    {
+      "epoch": 0.12140737600678253,
+      "grad_norm": 0.9239836623171834,
+      "learning_rate": 3.913840780829429e-05,
+      "loss": 1.4172,
+      "step": 179
+    },
+    {
+      "epoch": 0.12208562950402713,
+      "grad_norm": 0.8280501002137555,
+      "learning_rate": 3.9125595120148266e-05,
+      "loss": 1.3671,
+      "step": 180
+    },
+    {
+      "epoch": 0.12276388300127172,
+      "grad_norm": 0.8483926444765467,
+      "learning_rate": 3.911268999410224e-05,
+      "loss": 1.3692,
+      "step": 181
+    },
+    {
+      "epoch": 0.12344213649851632,
+      "grad_norm": 0.8355083213687582,
+      "learning_rate": 3.909969249252933e-05,
+      "loss": 1.3698,
+      "step": 182
+    },
+    {
+      "epoch": 0.12412038999576092,
+      "grad_norm": 0.8486653810483453,
+      "learning_rate": 3.9086602678249095e-05,
+      "loss": 1.3583,
+      "step": 183
+    },
+    {
+      "epoch": 0.12479864349300551,
+      "grad_norm": 0.938098681006725,
+      "learning_rate": 3.907342061452729e-05,
+      "loss": 1.4199,
+      "step": 184
+    },
+    {
+      "epoch": 0.1254768969902501,
+      "grad_norm": 0.8317954566954049,
+      "learning_rate": 3.906014636507551e-05,
+      "loss": 1.4094,
+      "step": 185
+    },
+    {
+      "epoch": 0.1261551504874947,
+      "grad_norm": 1.095456794855618,
+      "learning_rate": 3.904677999405091e-05,
+      "loss": 1.3697,
+      "step": 186
+    },
+    {
+      "epoch": 0.1268334039847393,
+      "grad_norm": 0.9483196201551992,
+      "learning_rate": 3.9033321566055885e-05,
+      "loss": 1.3756,
+      "step": 187
+    },
+    {
+      "epoch": 0.1275116574819839,
+      "grad_norm": 0.8024238008384962,
+      "learning_rate": 3.9019771146137764e-05,
+      "loss": 1.4041,
+      "step": 188
+    },
+    {
+      "epoch": 0.1281899109792285,
+      "grad_norm": 0.9413726100672268,
+      "learning_rate": 3.900612879978848e-05,
+      "loss": 1.3563,
+      "step": 189
+    },
+    {
+      "epoch": 0.1288681644764731,
+      "grad_norm": 0.9097638693583244,
+      "learning_rate": 3.8992394592944286e-05,
+      "loss": 1.4097,
+      "step": 190
+    },
+    {
+      "epoch": 0.1295464179737177,
+      "grad_norm": 0.8772276516195574,
+      "learning_rate": 3.8978568591985397e-05,
+      "loss": 1.3731,
+      "step": 191
+    },
+    {
+      "epoch": 0.13022467147096228,
+      "grad_norm": 0.9098273829678695,
+      "learning_rate": 3.89646508637357e-05,
+      "loss": 1.3672,
+      "step": 192
+    },
+    {
+      "epoch": 0.13090292496820688,
+      "grad_norm": 0.8183191888811304,
+      "learning_rate": 3.89506414754624e-05,
+      "loss": 1.3835,
+      "step": 193
+    },
+    {
+      "epoch": 0.13158117846545148,
+      "grad_norm": 0.8919803577719078,
+      "learning_rate": 3.893654049487574e-05,
+      "loss": 1.3874,
+      "step": 194
+    },
+    {
+      "epoch": 0.13225943196269604,
+      "grad_norm": 0.8416172745913636,
+      "learning_rate": 3.892234799012862e-05,
+      "loss": 1.3894,
+      "step": 195
+    },
+    {
+      "epoch": 0.13293768545994064,
+      "grad_norm": 0.8447411846074018,
+      "learning_rate": 3.890806402981632e-05,
+      "loss": 1.358,
+      "step": 196
+    },
+    {
+      "epoch": 0.13361593895718524,
+      "grad_norm": 1.13463194606766,
+      "learning_rate": 3.889368868297613e-05,
+      "loss": 1.4096,
+      "step": 197
+    },
+    {
+      "epoch": 0.13429419245442983,
+      "grad_norm": 1.0136148913804681,
+      "learning_rate": 3.887922201908703e-05,
+      "loss": 1.3233,
+      "step": 198
+    },
+    {
+      "epoch": 0.13497244595167443,
+      "grad_norm": 0.9353648682943572,
+      "learning_rate": 3.886466410806936e-05,
+      "loss": 1.3681,
+      "step": 199
+    },
+    {
+      "epoch": 0.13565069944891903,
+      "grad_norm": 0.9194457839730446,
+      "learning_rate": 3.885001502028446e-05,
+      "loss": 1.3777,
+      "step": 200
+    },
+    {
+      "epoch": 0.13632895294616362,
+      "grad_norm": 0.9414436636519163,
+      "learning_rate": 3.883527482653436e-05,
+      "loss": 1.3672,
+      "step": 201
+    },
+    {
+      "epoch": 0.13700720644340822,
+      "grad_norm": 0.9477037701457524,
+      "learning_rate": 3.882044359806144e-05,
+      "loss": 1.3587,
+      "step": 202
+    },
+    {
+      "epoch": 0.13768545994065282,
+      "grad_norm": 0.8296501298762375,
+      "learning_rate": 3.880552140654803e-05,
+      "loss": 1.3871,
+      "step": 203
+    },
+    {
+      "epoch": 0.1383637134378974,
+      "grad_norm": 1.0146984451817604,
+      "learning_rate": 3.879050832411613e-05,
+      "loss": 1.3864,
+      "step": 204
+    },
+    {
+      "epoch": 0.139041966935142,
+      "grad_norm": 1.0688397482821446,
+      "learning_rate": 3.877540442332703e-05,
+      "loss": 1.3566,
+      "step": 205
+    },
+    {
+      "epoch": 0.1397202204323866,
+      "grad_norm": 0.8967082496420603,
+      "learning_rate": 3.876020977718096e-05,
+      "loss": 1.3318,
+      "step": 206
+    },
+    {
+      "epoch": 0.1403984739296312,
+      "grad_norm": 0.8597915304069587,
+      "learning_rate": 3.8744924459116734e-05,
+      "loss": 1.359,
+      "step": 207
+    },
+    {
+      "epoch": 0.1410767274268758,
+      "grad_norm": 0.9861201591743611,
+      "learning_rate": 3.8729548543011423e-05,
+      "loss": 1.3637,
+      "step": 208
+    },
+    {
+      "epoch": 0.1417549809241204,
+      "grad_norm": 0.8462602408390785,
+      "learning_rate": 3.8714082103179956e-05,
+      "loss": 1.3951,
+      "step": 209
+    },
+    {
+      "epoch": 0.142433234421365,
+      "grad_norm": 1.0342160063623491,
+      "learning_rate": 3.86985252143748e-05,
+      "loss": 1.3534,
+      "step": 210
+    },
+    {
+      "epoch": 0.1431114879186096,
+      "grad_norm": 0.9700626844877009,
+      "learning_rate": 3.868287795178556e-05,
+      "loss": 1.4074,
+      "step": 211
+    },
+    {
+      "epoch": 0.14378974141585418,
+      "grad_norm": 0.8355599061006139,
+      "learning_rate": 3.8667140391038646e-05,
+      "loss": 1.3945,
+      "step": 212
+    },
+    {
+      "epoch": 0.14446799491309878,
+      "grad_norm": 0.8588794535196841,
+      "learning_rate": 3.8651312608196897e-05,
+      "loss": 1.3439,
+      "step": 213
+    },
+    {
+      "epoch": 0.14514624841034338,
+      "grad_norm": 0.9737294357813214,
+      "learning_rate": 3.863539467975922e-05,
+      "loss": 1.4077,
+      "step": 214
+    },
+    {
+      "epoch": 0.14582450190758797,
+      "grad_norm": 0.6834058192292372,
+      "learning_rate": 3.86193866826602e-05,
+      "loss": 1.4534,
+      "step": 215
+    },
+    {
+      "epoch": 0.14650275540483257,
+      "grad_norm": 0.945727982315802,
+      "learning_rate": 3.860328869426975e-05,
+      "loss": 1.3838,
+      "step": 216
+    },
+    {
+      "epoch": 0.14718100890207717,
+      "grad_norm": 0.8345543003512037,
+      "learning_rate": 3.8587100792392744e-05,
+      "loss": 1.3811,
+      "step": 217
+    },
+    {
+      "epoch": 0.14785926239932173,
+      "grad_norm": 0.8309922574724689,
+      "learning_rate": 3.8570823055268605e-05,
+      "loss": 1.3462,
+      "step": 218
+    },
+    {
+      "epoch": 0.14853751589656633,
+      "grad_norm": 0.8519439545936296,
+      "learning_rate": 3.855445556157093e-05,
+      "loss": 1.3804,
+      "step": 219
+    },
+    {
+      "epoch": 0.14921576939381093,
+      "grad_norm": 0.8373458633419311,
+      "learning_rate": 3.853799839040719e-05,
+      "loss": 1.3872,
+      "step": 220
+    },
+    {
+      "epoch": 0.14989402289105552,
+      "grad_norm": 0.8204422865846579,
+      "learning_rate": 3.852145162131824e-05,
+      "loss": 1.357,
+      "step": 221
+    },
+    {
+      "epoch": 0.15057227638830012,
+      "grad_norm": 0.7749768398156192,
+      "learning_rate": 3.850481533427797e-05,
+      "loss": 1.3422,
+      "step": 222
+    },
+    {
+      "epoch": 0.15125052988554472,
+      "grad_norm": 0.8751657869837003,
+      "learning_rate": 3.848808960969296e-05,
+      "loss": 1.3583,
+      "step": 223
+    },
+    {
+      "epoch": 0.1519287833827893,
+      "grad_norm": 0.8599619287852476,
+      "learning_rate": 3.847127452840204e-05,
+      "loss": 1.345,
+      "step": 224
+    },
+    {
+      "epoch": 0.1526070368800339,
+      "grad_norm": 0.7790197241344738,
+      "learning_rate": 3.8454370171675926e-05,
+      "loss": 1.3525,
+      "step": 225
+    },
+    {
+      "epoch": 0.1532852903772785,
+      "grad_norm": 0.7798362164571245,
+      "learning_rate": 3.843737662121682e-05,
+      "loss": 1.3326,
+      "step": 226
+    },
+    {
+      "epoch": 0.1539635438745231,
+      "grad_norm": 0.8117309588898491,
+      "learning_rate": 3.842029395915803e-05,
+      "loss": 1.3441,
+      "step": 227
+    },
+    {
+      "epoch": 0.1546417973717677,
+      "grad_norm": 0.7712267857854244,
+      "learning_rate": 3.8403122268063524e-05,
+      "loss": 1.3814,
+      "step": 228
+    },
+    {
+      "epoch": 0.1553200508690123,
+      "grad_norm": 0.8094984928131681,
+      "learning_rate": 3.83858616309276e-05,
+      "loss": 1.3731,
+      "step": 229
+    },
+    {
+      "epoch": 0.1559983043662569,
+      "grad_norm": 0.8020318070438405,
+      "learning_rate": 3.836851213117443e-05,
+      "loss": 1.3545,
+      "step": 230
+    },
+    {
+      "epoch": 0.1566765578635015,
+      "grad_norm": 0.7792455803212042,
+      "learning_rate": 3.835107385265768e-05,
+      "loss": 1.3835,
+      "step": 231
+    },
+    {
+      "epoch": 0.15735481136074608,
+      "grad_norm": 0.7843157924687033,
+      "learning_rate": 3.833354687966011e-05,
+      "loss": 1.3248,
+      "step": 232
+    },
+    {
+      "epoch": 0.15803306485799068,
+      "grad_norm": 0.8493476520147848,
+      "learning_rate": 3.831593129689315e-05,
+      "loss": 1.3557,
+      "step": 233
+    },
+    {
+      "epoch": 0.15871131835523528,
+      "grad_norm": 0.882277870795216,
+      "learning_rate": 3.82982271894965e-05,
+      "loss": 1.3543,
+      "step": 234
+    },
+    {
+      "epoch": 0.15938957185247987,
+      "grad_norm": 0.7572047887431174,
+      "learning_rate": 3.828043464303773e-05,
+      "loss": 1.3606,
+      "step": 235
+    },
+    {
+      "epoch": 0.16006782534972447,
+      "grad_norm": 0.8062972354097508,
+      "learning_rate": 3.826255374351183e-05,
+      "loss": 1.3674,
+      "step": 236
+    },
+    {
+      "epoch": 0.16074607884696906,
+      "grad_norm": 0.8508022675381113,
+      "learning_rate": 3.824458457734085e-05,
+      "loss": 1.3741,
+      "step": 237
+    },
+    {
+      "epoch": 0.16142433234421366,
+      "grad_norm": 0.778304420325275,
+      "learning_rate": 3.822652723137341e-05,
+      "loss": 1.3494,
+      "step": 238
+    },
+    {
+      "epoch": 0.16210258584145826,
+      "grad_norm": 0.8271572580761374,
+      "learning_rate": 3.8208381792884374e-05,
+      "loss": 1.3585,
+      "step": 239
+    },
+    {
+      "epoch": 0.16278083933870285,
+      "grad_norm": 0.7771004938802668,
+      "learning_rate": 3.8190148349574316e-05,
+      "loss": 1.3623,
+      "step": 240
+    },
+    {
+      "epoch": 0.16345909283594742,
+      "grad_norm": 0.8119254585705749,
+      "learning_rate": 3.81718269895692e-05,
+      "loss": 1.3507,
+      "step": 241
+    },
+    {
+      "epoch": 0.16413734633319202,
+      "grad_norm": 0.8054041105152854,
+      "learning_rate": 3.81534178014199e-05,
+      "loss": 1.3704,
+      "step": 242
+    },
+    {
+      "epoch": 0.16481559983043662,
+      "grad_norm": 0.8341651339274286,
+      "learning_rate": 3.8134920874101756e-05,
+      "loss": 1.3525,
+      "step": 243
+    },
+    {
+      "epoch": 0.1654938533276812,
+      "grad_norm": 0.8668088532427671,
+      "learning_rate": 3.8116336297014195e-05,
+      "loss": 1.3853,
+      "step": 244
+    },
+    {
+      "epoch": 0.1661721068249258,
+      "grad_norm": 0.8904485590262735,
+      "learning_rate": 3.809766415998028e-05,
+      "loss": 1.3615,
+      "step": 245
+    },
+    {
+      "epoch": 0.1668503603221704,
+      "grad_norm": 0.7958380694134233,
+      "learning_rate": 3.8078904553246234e-05,
+      "loss": 1.3637,
+      "step": 246
+    },
+    {
+      "epoch": 0.167528613819415,
+      "grad_norm": 0.7809146687054412,
+      "learning_rate": 3.806005756748108e-05,
+      "loss": 1.3862,
+      "step": 247
+    },
+    {
+      "epoch": 0.1682068673166596,
+      "grad_norm": 0.7448426981896193,
+      "learning_rate": 3.804112329377613e-05,
+      "loss": 1.3377,
+      "step": 248
+    },
+    {
+      "epoch": 0.1688851208139042,
+      "grad_norm": 0.7503956304358436,
+      "learning_rate": 3.8022101823644605e-05,
+      "loss": 1.3715,
+      "step": 249
+    },
+    {
+      "epoch": 0.1695633743111488,
+      "grad_norm": 0.7755324225568783,
+      "learning_rate": 3.800299324902112e-05,
+      "loss": 1.3544,
+      "step": 250
+    },
+    {
+      "epoch": 0.1702416278083934,
+      "grad_norm": 0.7640633646163552,
+      "learning_rate": 3.7983797662261335e-05,
+      "loss": 1.3456,
+      "step": 251
+    },
+    {
+      "epoch": 0.17091988130563798,
+      "grad_norm": 0.7504456766281555,
+      "learning_rate": 3.796451515614142e-05,
+      "loss": 1.3534,
+      "step": 252
+    },
+    {
+      "epoch": 0.17159813480288258,
+      "grad_norm": 0.8207771425862805,
+      "learning_rate": 3.794514582385767e-05,
+      "loss": 1.3276,
+      "step": 253
+    },
+    {
+      "epoch": 0.17227638830012718,
+      "grad_norm": 0.7625453955788795,
+      "learning_rate": 3.792568975902601e-05,
+      "loss": 1.3439,
+      "step": 254
+    },
+    {
+      "epoch": 0.17295464179737177,
+      "grad_norm": 0.7351975940767888,
+      "learning_rate": 3.790614705568156e-05,
+      "loss": 1.318,
+      "step": 255
+    },
+    {
+      "epoch": 0.17363289529461637,
+      "grad_norm": 0.809184694000721,
+      "learning_rate": 3.7886517808278205e-05,
+      "loss": 1.3334,
+      "step": 256
+    },
+    {
+      "epoch": 0.17431114879186096,
+      "grad_norm": 0.8061789671241181,
+      "learning_rate": 3.7866802111688084e-05,
+      "loss": 1.3489,
+      "step": 257
+    },
+    {
+      "epoch": 0.17498940228910556,
+      "grad_norm": 0.7936271064037789,
+      "learning_rate": 3.784700006120119e-05,
+      "loss": 1.3528,
+      "step": 258
+    },
+    {
+      "epoch": 0.17566765578635016,
+      "grad_norm": 0.7552641018922746,
+      "learning_rate": 3.7827111752524866e-05,
+      "loss": 1.3287,
+      "step": 259
+    },
+    {
+      "epoch": 0.17634590928359475,
+      "grad_norm": 0.8374689219052238,
+      "learning_rate": 3.780713728178335e-05,
+      "loss": 1.3752,
+      "step": 260
+    },
+    {
+      "epoch": 0.17702416278083935,
+      "grad_norm": 0.729554288427445,
+      "learning_rate": 3.7787076745517353e-05,
+      "loss": 1.3975,
+      "step": 261
+    },
+    {
+      "epoch": 0.17770241627808395,
+      "grad_norm": 0.8419610549000647,
+      "learning_rate": 3.776693024068351e-05,
+      "loss": 1.3363,
+      "step": 262
+    },
+    {
+      "epoch": 0.17838066977532852,
+      "grad_norm": 0.7889473275556681,
+      "learning_rate": 3.774669786465401e-05,
+      "loss": 1.3525,
+      "step": 263
+    },
+    {
+      "epoch": 0.1790589232725731,
+      "grad_norm": 0.7378097589904369,
+      "learning_rate": 3.772637971521604e-05,
+      "loss": 1.3348,
+      "step": 264
+    },
+    {
+      "epoch": 0.1797371767698177,
+      "grad_norm": 0.7913075811035389,
+      "learning_rate": 3.770597589057136e-05,
+      "loss": 1.3615,
+      "step": 265
+    },
+    {
+      "epoch": 0.1804154302670623,
+      "grad_norm": 0.7520950872769981,
+      "learning_rate": 3.768548648933581e-05,
+      "loss": 1.3906,
+      "step": 266
+    },
+    {
+      "epoch": 0.1810936837643069,
+      "grad_norm": 0.8509685860713678,
+      "learning_rate": 3.7664911610538844e-05,
+      "loss": 1.338,
+      "step": 267
+    },
+    {
+      "epoch": 0.1817719372615515,
+      "grad_norm": 0.855152434807828,
+      "learning_rate": 3.764425135362305e-05,
+      "loss": 1.3232,
+      "step": 268
+    },
+    {
+      "epoch": 0.1824501907587961,
+      "grad_norm": 0.8008198108696746,
+      "learning_rate": 3.762350581844366e-05,
+      "loss": 1.3109,
+      "step": 269
+    },
+    {
+      "epoch": 0.1831284442560407,
+      "grad_norm": 0.8705798073505797,
+      "learning_rate": 3.7602675105268065e-05,
+      "loss": 1.3843,
+      "step": 270
+    },
+    {
+      "epoch": 0.1838066977532853,
+      "grad_norm": 0.844301193211463,
+      "learning_rate": 3.758175931477537e-05,
+      "loss": 1.3401,
+      "step": 271
+    },
+    {
+      "epoch": 0.18448495125052988,
+      "grad_norm": 0.8314454058935925,
+      "learning_rate": 3.756075854805583e-05,
+      "loss": 1.3741,
+      "step": 272
+    },
+    {
+      "epoch": 0.18516320474777448,
+      "grad_norm": 0.8322405908547285,
+      "learning_rate": 3.7539672906610445e-05,
+      "loss": 1.3488,
+      "step": 273
+    },
+    {
+      "epoch": 0.18584145824501908,
+      "grad_norm": 0.7929593381557983,
+      "learning_rate": 3.751850249235041e-05,
+      "loss": 1.3128,
+      "step": 274
+    },
+    {
+      "epoch": 0.18651971174226367,
+      "grad_norm": 0.7936714435193957,
+      "learning_rate": 3.7497247407596665e-05,
+      "loss": 1.3248,
+      "step": 275
+    },
+    {
+      "epoch": 0.18719796523950827,
+      "grad_norm": 0.7968038025658714,
+      "learning_rate": 3.747590775507936e-05,
+      "loss": 1.3509,
+      "step": 276
+    },
+    {
+      "epoch": 0.18787621873675286,
+      "grad_norm": 0.8409932538922028,
+      "learning_rate": 3.745448363793738e-05,
+      "loss": 1.339,
+      "step": 277
+    },
+    {
+      "epoch": 0.18855447223399746,
+      "grad_norm": 0.7837545633394652,
+      "learning_rate": 3.7432975159717865e-05,
+      "loss": 1.3483,
+      "step": 278
+    },
+    {
+      "epoch": 0.18923272573124206,
+      "grad_norm": 0.7997013180068606,
+      "learning_rate": 3.741138242437566e-05,
+      "loss": 1.3549,
+      "step": 279
+    },
+    {
+      "epoch": 0.18991097922848665,
+      "grad_norm": 0.8235720011611928,
+      "learning_rate": 3.738970553627287e-05,
+      "loss": 1.3973,
+      "step": 280
+    },
+    {
+      "epoch": 0.19058923272573125,
+      "grad_norm": 0.7698450279657325,
+      "learning_rate": 3.73679446001783e-05,
+      "loss": 1.3507,
+      "step": 281
+    },
+    {
+      "epoch": 0.19126748622297585,
+      "grad_norm": 0.8025960582420557,
+      "learning_rate": 3.7346099721266995e-05,
+      "loss": 1.3481,
+      "step": 282
+    },
+    {
+      "epoch": 0.19194573972022044,
+      "grad_norm": 0.7590548934859671,
+      "learning_rate": 3.7324171005119716e-05,
+      "loss": 1.3411,
+      "step": 283
+    },
+    {
+      "epoch": 0.19262399321746504,
+      "grad_norm": 0.8147227638445041,
+      "learning_rate": 3.7302158557722415e-05,
+      "loss": 1.3631,
+      "step": 284
+    },
+    {
+      "epoch": 0.19330224671470964,
+      "grad_norm": 0.7386164064451395,
+      "learning_rate": 3.728006248546573e-05,
+      "loss": 1.3441,
+      "step": 285
+    },
+    {
+      "epoch": 0.1939805002119542,
+      "grad_norm": 0.7890002607630037,
+      "learning_rate": 3.7257882895144485e-05,
+      "loss": 1.3517,
+      "step": 286
+    },
+    {
+      "epoch": 0.1946587537091988,
+      "grad_norm": 0.7939882773368258,
+      "learning_rate": 3.7235619893957175e-05,
+      "loss": 1.3272,
+      "step": 287
+    },
+    {
+      "epoch": 0.1953370072064434,
+      "grad_norm": 0.7612534430811938,
+      "learning_rate": 3.7213273589505405e-05,
+      "loss": 1.3538,
+      "step": 288
+    },
+    {
+      "epoch": 0.196015260703688,
+      "grad_norm": 0.8166840142238,
+      "learning_rate": 3.719084408979343e-05,
+      "loss": 1.3856,
+      "step": 289
+    },
+    {
+      "epoch": 0.1966935142009326,
+      "grad_norm": 0.9655100351840458,
+      "learning_rate": 3.7168331503227586e-05,
+      "loss": 1.4738,
+      "step": 290
+    },
+    {
+      "epoch": 0.1973717676981772,
+      "grad_norm": 0.748970675337018,
+      "learning_rate": 3.71457359386158e-05,
+      "loss": 1.3158,
+      "step": 291
+    },
+    {
+      "epoch": 0.19805002119542178,
+      "grad_norm": 0.7913831795192827,
+      "learning_rate": 3.712305750516704e-05,
+      "loss": 1.2998,
+      "step": 292
+    },
+    {
+      "epoch": 0.19872827469266638,
+      "grad_norm": 0.8313988747400662,
+      "learning_rate": 3.71002963124908e-05,
+      "loss": 1.3098,
+      "step": 293
+    },
+    {
+      "epoch": 0.19940652818991098,
+      "grad_norm": 0.7698811838943792,
+      "learning_rate": 3.707745247059656e-05,
+      "loss": 1.3304,
+      "step": 294
+    },
+    {
+      "epoch": 0.20008478168715557,
+      "grad_norm": 0.8167400009802671,
+      "learning_rate": 3.705452608989327e-05,
+      "loss": 1.3425,
+      "step": 295
+    },
+    {
+      "epoch": 0.20076303518440017,
+      "grad_norm": 0.7853365125954755,
+      "learning_rate": 3.7031517281188795e-05,
+      "loss": 1.3229,
+      "step": 296
+    },
+    {
+      "epoch": 0.20144128868164476,
+      "grad_norm": 0.7951101430292243,
+      "learning_rate": 3.70084261556894e-05,
+      "loss": 1.3465,
+      "step": 297
+    },
+    {
+      "epoch": 0.20211954217888936,
+      "grad_norm": 0.7980331185038868,
+      "learning_rate": 3.698525282499921e-05,
+      "loss": 1.3304,
+      "step": 298
+    },
+    {
+      "epoch": 0.20279779567613396,
+      "grad_norm": 0.8765891197449154,
+      "learning_rate": 3.696199740111964e-05,
+      "loss": 1.3502,
+      "step": 299
+    },
+    {
+      "epoch": 0.20347604917337855,
+      "grad_norm": 0.7838726576701436,
+      "learning_rate": 3.6938659996448916e-05,
+      "loss": 1.343,
+      "step": 300
+    },
+    {
+      "epoch": 0.20415430267062315,
+      "grad_norm": 1.536887455459171,
+      "learning_rate": 3.691524072378145e-05,
+      "loss": 1.3951,
+      "step": 301
+    },
+    {
+      "epoch": 0.20483255616786775,
+      "grad_norm": 1.1072730222345826,
+      "learning_rate": 3.689173969630737e-05,
+      "loss": 1.3276,
+      "step": 302
+    },
+    {
+      "epoch": 0.20551080966511234,
+      "grad_norm": 1.0341965163359792,
+      "learning_rate": 3.6868157027611935e-05,
+      "loss": 1.3409,
+      "step": 303
+    },
+    {
+      "epoch": 0.20618906316235694,
+      "grad_norm": 0.855269607966128,
+      "learning_rate": 3.684449283167499e-05,
+      "loss": 1.3783,
+      "step": 304
+    },
+    {
+      "epoch": 0.20686731665960154,
+      "grad_norm": 0.9711680096158264,
+      "learning_rate": 3.6820747222870415e-05,
+      "loss": 1.3349,
+      "step": 305
+    },
+    {
+      "epoch": 0.20754557015684613,
+      "grad_norm": 0.9413940555791582,
+      "learning_rate": 3.679692031596557e-05,
+      "loss": 1.3565,
+      "step": 306
+    },
+    {
+      "epoch": 0.20822382365409073,
+      "grad_norm": 0.8505193076721568,
+      "learning_rate": 3.677301222612077e-05,
+      "loss": 1.3549,
+      "step": 307
+    },
+    {
+      "epoch": 0.20890207715133532,
+      "grad_norm": 0.8654824704034854,
+      "learning_rate": 3.674902306888868e-05,
+      "loss": 1.3463,
+      "step": 308
+    },
+    {
+      "epoch": 0.2095803306485799,
+      "grad_norm": 0.9017551080146448,
+      "learning_rate": 3.672495296021378e-05,
+      "loss": 1.3675,
+      "step": 309
+    },
+    {
+      "epoch": 0.2102585841458245,
+      "grad_norm": 0.8610924662593219,
+      "learning_rate": 3.670080201643183e-05,
+      "loss": 1.3418,
+      "step": 310
+    },
+    {
+      "epoch": 0.2109368376430691,
+      "grad_norm": 0.7567901953649514,
+      "learning_rate": 3.667657035426924e-05,
+      "loss": 1.3037,
+      "step": 311
+    },
+    {
+      "epoch": 0.21161509114031368,
+      "grad_norm": 0.9227793753462682,
+      "learning_rate": 3.6652258090842596e-05,
+      "loss": 1.3307,
+      "step": 312
+    },
+    {
+      "epoch": 0.21229334463755828,
+      "grad_norm": 0.8477420323577082,
+      "learning_rate": 3.6627865343658004e-05,
+      "loss": 1.3449,
+      "step": 313
+    },
+    {
+      "epoch": 0.21297159813480288,
+      "grad_norm": 0.7928792403999231,
+      "learning_rate": 3.6603392230610596e-05,
+      "loss": 1.3504,
+      "step": 314
+    },
+    {
+      "epoch": 0.21364985163204747,
+      "grad_norm": 0.8706330117973166,
+      "learning_rate": 3.657883886998391e-05,
+      "loss": 1.3384,
+      "step": 315
+    },
+    {
+      "epoch": 0.21432810512929207,
+      "grad_norm": 0.7912639457114111,
+      "learning_rate": 3.6554205380449344e-05,
+      "loss": 1.3058,
+      "step": 316
+    },
+    {
+      "epoch": 0.21500635862653666,
+      "grad_norm": 0.8074035387342698,
+      "learning_rate": 3.6529491881065584e-05,
+      "loss": 1.3212,
+      "step": 317
+    },
+    {
+      "epoch": 0.21568461212378126,
+      "grad_norm": 0.8193503607565463,
+      "learning_rate": 3.6504698491277996e-05,
+      "loss": 1.3591,
+      "step": 318
+    },
+    {
+      "epoch": 0.21636286562102586,
+      "grad_norm": 0.7995273140530428,
+      "learning_rate": 3.64798253309181e-05,
+      "loss": 1.3558,
+      "step": 319
+    },
+    {
+      "epoch": 0.21704111911827045,
+      "grad_norm": 0.8687660030654587,
+      "learning_rate": 3.645487252020294e-05,
+      "loss": 1.3528,
+      "step": 320
+    },
+    {
+      "epoch": 0.21771937261551505,
+      "grad_norm": 0.8264620665129695,
+      "learning_rate": 3.642984017973454e-05,
+      "loss": 1.317,
+      "step": 321
+    },
+    {
+      "epoch": 0.21839762611275965,
+      "grad_norm": 2.2301642268612523,
+      "learning_rate": 3.640472843049931e-05,
+      "loss": 1.4527,
+      "step": 322
+    },
+    {
+      "epoch": 0.21907587961000424,
+      "grad_norm": 1.0084140949380813,
+      "learning_rate": 3.637953739386744e-05,
+      "loss": 1.3177,
+      "step": 323
+    },
+    {
+      "epoch": 0.21975413310724884,
+      "grad_norm": 1.0372803017251364,
+      "learning_rate": 3.6354267191592356e-05,
+      "loss": 1.3312,
+      "step": 324
+    },
+    {
+      "epoch": 0.22043238660449344,
+      "grad_norm": 0.9457280722406972,
+      "learning_rate": 3.63289179458101e-05,
+      "loss": 1.331,
+      "step": 325
+    },
+    {
+      "epoch": 0.22111064010173803,
+      "grad_norm": 0.8999936637331983,
+      "learning_rate": 3.630348977903873e-05,
+      "loss": 1.3561,
+      "step": 326
+    },
+    {
+      "epoch": 0.22178889359898263,
+      "grad_norm": 0.8861915476360576,
+      "learning_rate": 3.627798281417778e-05,
+      "loss": 1.3401,
+      "step": 327
+    },
+    {
+      "epoch": 0.22246714709622722,
+      "grad_norm": 0.8583266540025729,
+      "learning_rate": 3.6252397174507595e-05,
+      "loss": 1.3225,
+      "step": 328
+    },
+    {
+      "epoch": 0.22314540059347182,
+      "grad_norm": 0.8424597827495318,
+      "learning_rate": 3.622673298368879e-05,
+      "loss": 1.3603,
+      "step": 329
+    },
+    {
+      "epoch": 0.22382365409071642,
+      "grad_norm": 0.8314710447674626,
+      "learning_rate": 3.620099036576163e-05,
+      "loss": 1.3519,
+      "step": 330
+    },
+    {
+      "epoch": 0.224501907587961,
+      "grad_norm": 0.9112220546676985,
+      "learning_rate": 3.617516944514544e-05,
+      "loss": 1.3362,
+      "step": 331
+    },
+    {
+      "epoch": 0.22518016108520558,
+      "grad_norm": 0.7831649871164149,
+      "learning_rate": 3.614927034663799e-05,
+      "loss": 1.3413,
+      "step": 332
+    },
+    {
+      "epoch": 0.22585841458245018,
+      "grad_norm": 0.8642522956740931,
+      "learning_rate": 3.6123293195414907e-05,
+      "loss": 1.3426,
+      "step": 333
+    },
+    {
+      "epoch": 0.22653666807969478,
+      "grad_norm": 0.8860961029717345,
+      "learning_rate": 3.609723811702905e-05,
+      "loss": 1.3591,
+      "step": 334
+    },
+    {
+      "epoch": 0.22721492157693937,
+      "grad_norm": 0.8060705554104319,
+      "learning_rate": 3.6071105237409926e-05,
+      "loss": 1.3497,
+      "step": 335
+    },
+    {
+      "epoch": 0.22789317507418397,
+      "grad_norm": 0.8286691463885103,
+      "learning_rate": 3.6044894682863076e-05,
+      "loss": 1.3406,
+      "step": 336
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 0.8518464543991984,
+      "learning_rate": 3.601860658006945e-05,
+      "loss": 1.3464,
+      "step": 337
+    },
+    {
+      "epoch": 0.22924968206867316,
+      "grad_norm": 0.7881771078373155,
+      "learning_rate": 3.599224105608481e-05,
+      "loss": 1.3343,
+      "step": 338
+    },
+    {
+      "epoch": 0.22992793556591776,
+      "grad_norm": 0.8275811584205034,
+      "learning_rate": 3.59657982383391e-05,
+      "loss": 1.3507,
+      "step": 339
+    },
+    {
+      "epoch": 0.23060618906316235,
+      "grad_norm": 0.8533833802525195,
+      "learning_rate": 3.5939278254635854e-05,
+      "loss": 1.3532,
+      "step": 340
+    },
+    {
+      "epoch": 0.23128444256040695,
+      "grad_norm": 0.8377041485898244,
+      "learning_rate": 3.591268123315156e-05,
+      "loss": 1.3204,
+      "step": 341
+    },
+    {
+      "epoch": 0.23196269605765155,
+      "grad_norm": 0.7389800588833796,
+      "learning_rate": 3.5886007302435046e-05,
+      "loss": 1.3377,
+      "step": 342
+    },
+    {
+      "epoch": 0.23264094955489614,
+      "grad_norm": 0.8426270847369397,
+      "learning_rate": 3.585925659140685e-05,
+      "loss": 1.367,
+      "step": 343
+    },
+    {
+      "epoch": 0.23331920305214074,
+      "grad_norm": 0.7758726182250993,
+      "learning_rate": 3.583242922935862e-05,
+      "loss": 1.3328,
+      "step": 344
+    },
+    {
+      "epoch": 0.23399745654938534,
+      "grad_norm": 0.7722773102342985,
+      "learning_rate": 3.580552534595246e-05,
+      "loss": 1.3421,
+      "step": 345
+    },
+    {
+      "epoch": 0.23467571004662993,
+      "grad_norm": 0.7905307857352721,
+      "learning_rate": 3.577854507122032e-05,
+      "loss": 1.3386,
+      "step": 346
+    },
+    {
+      "epoch": 0.23535396354387453,
+      "grad_norm": 0.7243404498317927,
+      "learning_rate": 3.575148853556337e-05,
+      "loss": 1.3251,
+      "step": 347
+    },
+    {
+      "epoch": 0.23603221704111912,
+      "grad_norm": 0.8417562243598162,
+      "learning_rate": 3.572435586975138e-05,
+      "loss": 1.3504,
+      "step": 348
+    },
+    {
+      "epoch": 0.23671047053836372,
+      "grad_norm": 2.8634053953757204,
+      "learning_rate": 3.5697147204922026e-05,
+      "loss": 1.4622,
+      "step": 349
+    },
+    {
+      "epoch": 0.23738872403560832,
+      "grad_norm": 1.168899136110835,
+      "learning_rate": 3.5669862672580344e-05,
+      "loss": 1.3077,
+      "step": 350
+    },
+    {
+      "epoch": 0.2380669775328529,
+      "grad_norm": 1.2982026940957527,
+      "learning_rate": 3.564250240459805e-05,
+      "loss": 1.3518,
+      "step": 351
+    },
+    {
+      "epoch": 0.2387452310300975,
+      "grad_norm": 1.0615794582481726,
+      "learning_rate": 3.561506653321288e-05,
+      "loss": 1.3315,
+      "step": 352
+    },
+    {
+      "epoch": 0.2394234845273421,
+      "grad_norm": 0.9813507773031921,
+      "learning_rate": 3.5587555191028015e-05,
+      "loss": 1.3452,
+      "step": 353
+    },
+    {
+      "epoch": 0.2401017380245867,
+      "grad_norm": 0.9987807615736051,
+      "learning_rate": 3.5559968511011356e-05,
+      "loss": 1.3405,
+      "step": 354
+    },
+    {
+      "epoch": 0.24077999152183127,
+      "grad_norm": 0.9976124653212103,
+      "learning_rate": 3.5532306626494965e-05,
+      "loss": 1.3258,
+      "step": 355
+    },
+    {
+      "epoch": 0.24145824501907587,
+      "grad_norm": 1.0166956051519895,
+      "learning_rate": 3.5504569671174366e-05,
+      "loss": 1.3536,
+      "step": 356
+    },
+    {
+      "epoch": 0.24213649851632046,
+      "grad_norm": 0.8408913593507406,
+      "learning_rate": 3.547675777910791e-05,
+      "loss": 1.3342,
+      "step": 357
+    },
+    {
+      "epoch": 0.24281475201356506,
+      "grad_norm": 0.929344141469499,
+      "learning_rate": 3.544887108471616e-05,
+      "loss": 1.2994,
+      "step": 358
+    },
+    {
+      "epoch": 0.24349300551080966,
+      "grad_norm": 0.8734694907852633,
+      "learning_rate": 3.542090972278118e-05,
+      "loss": 1.3284,
+      "step": 359
+    },
+    {
+      "epoch": 0.24417125900805425,
+      "grad_norm": 0.9818367995858515,
+      "learning_rate": 3.539287382844594e-05,
+      "loss": 1.368,
+      "step": 360
+    },
+    {
+      "epoch": 0.24484951250529885,
+      "grad_norm": 0.8507298377945505,
+      "learning_rate": 3.5364763537213614e-05,
+      "loss": 1.3086,
+      "step": 361
+    },
+    {
+      "epoch": 0.24552776600254345,
+      "grad_norm": 0.8320634906336775,
+      "learning_rate": 3.533657898494699e-05,
+      "loss": 1.2986,
+      "step": 362
+    },
+    {
+      "epoch": 0.24620601949978804,
+      "grad_norm": 0.8998882822885591,
+      "learning_rate": 3.5308320307867755e-05,
+      "loss": 1.342,
+      "step": 363
+    },
+    {
+      "epoch": 0.24688427299703264,
+      "grad_norm": 0.88201944435703,
+      "learning_rate": 3.5279987642555845e-05,
+      "loss": 1.2974,
+      "step": 364
+    },
+    {
+      "epoch": 0.24756252649427724,
+      "grad_norm": 0.8058457005541839,
+      "learning_rate": 3.5251581125948806e-05,
+      "loss": 1.3408,
+      "step": 365
+    },
+    {
+      "epoch": 0.24824077999152183,
+      "grad_norm": 0.7771041142421848,
+      "learning_rate": 3.522310089534114e-05,
+      "loss": 1.3301,
+      "step": 366
+    },
+    {
+      "epoch": 0.24891903348876643,
+      "grad_norm": 1.5325889692952437,
+      "learning_rate": 3.5194547088383584e-05,
+      "loss": 1.3882,
+      "step": 367
+    },
+    {
+      "epoch": 0.24959728698601102,
+      "grad_norm": 0.9580037839333668,
+      "learning_rate": 3.516591984308253e-05,
+      "loss": 1.2997,
+      "step": 368
+    },
+    {
+      "epoch": 0.2502755404832556,
+      "grad_norm": 0.8748551180692621,
+      "learning_rate": 3.513721929779928e-05,
+      "loss": 1.3375,
+      "step": 369
+    },
+    {
+      "epoch": 0.2509537939805002,
+      "grad_norm": 0.7995787378451331,
+      "learning_rate": 3.510844559124942e-05,
+      "loss": 1.3261,
+      "step": 370
+    },
+    {
+      "epoch": 0.2516320474777448,
+      "grad_norm": 0.860615621983403,
+      "learning_rate": 3.507959886250213e-05,
+      "loss": 1.3141,
+      "step": 371
+    },
+    {
+      "epoch": 0.2523103009749894,
+      "grad_norm": 0.8475296631685789,
+      "learning_rate": 3.505067925097955e-05,
+      "loss": 1.3364,
+      "step": 372
+    },
+    {
+      "epoch": 0.252988554472234,
+      "grad_norm": 0.83958333049959,
+      "learning_rate": 3.5021686896456045e-05,
+      "loss": 1.3569,
+      "step": 373
+    },
+    {
+      "epoch": 0.2536668079694786,
+      "grad_norm": 0.820796767618206,
+      "learning_rate": 3.499262193905757e-05,
+      "loss": 1.2947,
+      "step": 374
+    },
+    {
+      "epoch": 0.2543450614667232,
+      "grad_norm": 0.8913128103227287,
+      "learning_rate": 3.4963484519261013e-05,
+      "loss": 1.3486,
+      "step": 375
+    },
+    {
+      "epoch": 0.2550233149639678,
+      "grad_norm": 0.8231441290517508,
+      "learning_rate": 3.493427477789343e-05,
+      "loss": 1.3342,
+      "step": 376
+    },
+    {
+      "epoch": 0.2557015684612124,
+      "grad_norm": 0.7871230510129541,
+      "learning_rate": 3.490499285613148e-05,
+      "loss": 1.3012,
+      "step": 377
+    },
+    {
+      "epoch": 0.256379821958457,
+      "grad_norm": 0.8024588319974968,
+      "learning_rate": 3.487563889550066e-05,
+      "loss": 1.3323,
+      "step": 378
+    },
+    {
+      "epoch": 0.2570580754557016,
+      "grad_norm": 0.8487882299189274,
+      "learning_rate": 3.4846213037874625e-05,
+      "loss": 1.3555,
+      "step": 379
+    },
+    {
+      "epoch": 0.2577363289529462,
+      "grad_norm": 0.8115184825040263,
+      "learning_rate": 3.4816715425474566e-05,
+      "loss": 1.3002,
+      "step": 380
+    },
+    {
+      "epoch": 0.2584145824501908,
+      "grad_norm": 0.7904471470376874,
+      "learning_rate": 3.478714620086844e-05,
+      "loss": 1.3433,
+      "step": 381
+    },
+    {
+      "epoch": 0.2590928359474354,
+      "grad_norm": 0.8085473654607371,
+      "learning_rate": 3.475750550697035e-05,
+      "loss": 1.2955,
+      "step": 382
+    },
+    {
+      "epoch": 0.25977108944467997,
+      "grad_norm": 0.7731171074438241,
+      "learning_rate": 3.47277934870398e-05,
+      "loss": 1.3215,
+      "step": 383
+    },
+    {
+      "epoch": 0.26044934294192457,
+      "grad_norm": 0.7876037215909811,
+      "learning_rate": 3.469801028468105e-05,
+      "loss": 1.3349,
+      "step": 384
+    },
+    {
+      "epoch": 0.26112759643916916,
+      "grad_norm": 0.7638575055784281,
+      "learning_rate": 3.4668156043842386e-05,
+      "loss": 1.3088,
+      "step": 385
+    },
+    {
+      "epoch": 0.26180584993641376,
+      "grad_norm": 0.7662086462010528,
+      "learning_rate": 3.4638230908815434e-05,
+      "loss": 1.3234,
+      "step": 386
+    },
+    {
+      "epoch": 0.26248410343365836,
+      "grad_norm": 0.7646960186961371,
+      "learning_rate": 3.460823502423448e-05,
+      "loss": 1.309,
+      "step": 387
+    },
+    {
+      "epoch": 0.26316235693090295,
+      "grad_norm": 0.7612754885045987,
+      "learning_rate": 3.457816853507575e-05,
+      "loss": 1.3099,
+      "step": 388
+    },
+    {
+      "epoch": 0.2638406104281475,
+      "grad_norm": 0.8200614112402728,
+      "learning_rate": 3.454803158665669e-05,
+      "loss": 1.3278,
+      "step": 389
+    },
+    {
+      "epoch": 0.2645188639253921,
+      "grad_norm": 0.7794386935749952,
+      "learning_rate": 3.4517824324635354e-05,
+      "loss": 1.3352,
+      "step": 390
+    },
+    {
+      "epoch": 0.2651971174226367,
+      "grad_norm": 0.7855863081380292,
+      "learning_rate": 3.448754689500957e-05,
+      "loss": 1.3191,
+      "step": 391
+    },
+    {
+      "epoch": 0.2658753709198813,
+      "grad_norm": 0.7680839532766452,
+      "learning_rate": 3.445719944411633e-05,
+      "loss": 1.2971,
+      "step": 392
+    },
+    {
+      "epoch": 0.2665536244171259,
+      "grad_norm": 0.8047274090370727,
+      "learning_rate": 3.442678211863107e-05,
+      "loss": 1.325,
+      "step": 393
+    },
+    {
+      "epoch": 0.2672318779143705,
+      "grad_norm": 0.7870281495612244,
+      "learning_rate": 3.4396295065566904e-05,
+      "loss": 1.3109,
+      "step": 394
+    },
+    {
+      "epoch": 0.26791013141161507,
+      "grad_norm": 1.3837649994463146,
+      "learning_rate": 3.4365738432273974e-05,
+      "loss": 1.42,
+      "step": 395
+    },
+    {
+      "epoch": 0.26858838490885967,
+      "grad_norm": 0.9524291373055933,
+      "learning_rate": 3.433511236643873e-05,
+      "loss": 1.3104,
+      "step": 396
+    },
+    {
+      "epoch": 0.26926663840610426,
+      "grad_norm": 0.8643652629290977,
+      "learning_rate": 3.430441701608319e-05,
+      "loss": 1.312,
+      "step": 397
+    },
+    {
+      "epoch": 0.26994489190334886,
+      "grad_norm": 0.8159960685032611,
+      "learning_rate": 3.427365252956423e-05,
+      "loss": 1.2978,
+      "step": 398
+    },
+    {
+      "epoch": 0.27062314540059346,
+      "grad_norm": 0.8439809279263578,
+      "learning_rate": 3.42428190555729e-05,
+      "loss": 1.3347,
+      "step": 399
+    },
+    {
+      "epoch": 0.27130139889783805,
+      "grad_norm": 0.8193404775640883,
+      "learning_rate": 3.421191674313365e-05,
+      "loss": 1.2928,
+      "step": 400
+    },
+    {
+      "epoch": 0.27197965239508265,
+      "grad_norm": 0.7966926170821803,
+      "learning_rate": 3.418094574160366e-05,
+      "loss": 1.3213,
+      "step": 401
+    },
+    {
+      "epoch": 0.27265790589232725,
+      "grad_norm": 0.8172754714007294,
+      "learning_rate": 3.4149906200672086e-05,
+      "loss": 1.3247,
+      "step": 402
+    },
+    {
+      "epoch": 0.27333615938957184,
+      "grad_norm": 0.7747521741583795,
+      "learning_rate": 3.4118798270359375e-05,
+      "loss": 1.3473,
+      "step": 403
+    },
+    {
+      "epoch": 0.27401441288681644,
+      "grad_norm": 0.7711076847067602,
+      "learning_rate": 3.4087622101016494e-05,
+      "loss": 1.3203,
+      "step": 404
+    },
+    {
+      "epoch": 0.27469266638406103,
+      "grad_norm": 0.821559468991808,
+      "learning_rate": 3.405637784332421e-05,
+      "loss": 1.3021,
+      "step": 405
+    },
+    {
+      "epoch": 0.27537091988130563,
+      "grad_norm": 0.7989113298586838,
+      "learning_rate": 3.402506564829239e-05,
+      "loss": 1.3076,
+      "step": 406
+    },
+    {
+      "epoch": 0.2760491733785502,
+      "grad_norm": 0.7851515411605702,
+      "learning_rate": 3.3993685667259276e-05,
+      "loss": 1.3526,
+      "step": 407
+    },
+    {
+      "epoch": 0.2767274268757948,
+      "grad_norm": 0.7849015476169263,
+      "learning_rate": 3.3962238051890684e-05,
+      "loss": 1.3253,
+      "step": 408
+    },
+    {
+      "epoch": 0.2774056803730394,
+      "grad_norm": 0.7673100708701389,
+      "learning_rate": 3.393072295417937e-05,
+      "loss": 1.2977,
+      "step": 409
+    },
+    {
+      "epoch": 0.278083933870284,
+      "grad_norm": 0.7576984686596503,
+      "learning_rate": 3.3899140526444236e-05,
+      "loss": 1.2995,
+      "step": 410
+    },
+    {
+      "epoch": 0.2787621873675286,
+      "grad_norm": 0.7688682595138551,
+      "learning_rate": 3.386749092132956e-05,
+      "loss": 1.3604,
+      "step": 411
+    },
+    {
+      "epoch": 0.2794404408647732,
+      "grad_norm": 0.7657672122532467,
+      "learning_rate": 3.383577429180436e-05,
+      "loss": 1.3254,
+      "step": 412
+    },
+    {
+      "epoch": 0.2801186943620178,
+      "grad_norm": 0.7550560758993546,
+      "learning_rate": 3.380399079116157e-05,
+      "loss": 1.3276,
+      "step": 413
+    },
+    {
+      "epoch": 0.2807969478592624,
+      "grad_norm": 0.7778218998319307,
+      "learning_rate": 3.377214057301732e-05,
+      "loss": 1.306,
+      "step": 414
+    },
+    {
+      "epoch": 0.281475201356507,
+      "grad_norm": 0.797574211706488,
+      "learning_rate": 3.374022379131021e-05,
+      "loss": 1.3245,
+      "step": 415
+    },
+    {
+      "epoch": 0.2821534548537516,
+      "grad_norm": 0.7301612230929835,
+      "learning_rate": 3.370824060030055e-05,
+      "loss": 1.3027,
+      "step": 416
+    },
+    {
+      "epoch": 0.2828317083509962,
+      "grad_norm": 0.7314764016540719,
+      "learning_rate": 3.367619115456963e-05,
+      "loss": 1.3009,
+      "step": 417
+    },
+    {
+      "epoch": 0.2835099618482408,
+      "grad_norm": 0.7689083615586159,
+      "learning_rate": 3.364407560901894e-05,
+      "loss": 1.3151,
+      "step": 418
+    },
+    {
+      "epoch": 0.2841882153454854,
+      "grad_norm": 0.7716865683673857,
+      "learning_rate": 3.361189411886947e-05,
+      "loss": 1.3477,
+      "step": 419
+    },
+    {
+      "epoch": 0.28486646884273,
+      "grad_norm": 0.729080211495635,
+      "learning_rate": 3.357964683966093e-05,
+      "loss": 1.333,
+      "step": 420
+    },
+    {
+      "epoch": 0.2855447223399746,
+      "grad_norm": 0.7899182195999055,
+      "learning_rate": 3.354733392725098e-05,
+      "loss": 1.3211,
+      "step": 421
+    },
+    {
+      "epoch": 0.2862229758372192,
+      "grad_norm": 0.7711518181659062,
+      "learning_rate": 3.3514955537814514e-05,
+      "loss": 1.2805,
+      "step": 422
+    },
+    {
+      "epoch": 0.28690122933446377,
+      "grad_norm": 0.7867286449711313,
+      "learning_rate": 3.348251182784289e-05,
+      "loss": 1.2831,
+      "step": 423
+    },
+    {
+      "epoch": 0.28757948283170837,
+      "grad_norm": 1.0065878825834589,
+      "learning_rate": 3.345000295414317e-05,
+      "loss": 1.4279,
+      "step": 424
+    },
+    {
+      "epoch": 0.28825773632895296,
+      "grad_norm": 0.8330374919809896,
+      "learning_rate": 3.3417429073837375e-05,
+      "loss": 1.3308,
+      "step": 425
+    },
+    {
+      "epoch": 0.28893598982619756,
+      "grad_norm": 0.8614665742850024,
+      "learning_rate": 3.3384790344361704e-05,
+      "loss": 1.3173,
+      "step": 426
+    },
+    {
+      "epoch": 0.28961424332344216,
+      "grad_norm": 0.7841205349981816,
+      "learning_rate": 3.335208692346579e-05,
+      "loss": 1.3253,
+      "step": 427
+    },
+    {
+      "epoch": 0.29029249682068675,
+      "grad_norm": 0.7900221442726821,
+      "learning_rate": 3.3319318969211935e-05,
+      "loss": 1.3228,
+      "step": 428
+    },
+    {
+      "epoch": 0.29097075031793135,
+      "grad_norm": 0.7430038452394311,
+      "learning_rate": 3.3286486639974336e-05,
+      "loss": 1.2977,
+      "step": 429
+    },
+    {
+      "epoch": 0.29164900381517594,
+      "grad_norm": 0.788096881114239,
+      "learning_rate": 3.325359009443834e-05,
+      "loss": 1.3163,
+      "step": 430
+    },
+    {
+      "epoch": 0.29232725731242054,
+      "grad_norm": 0.763056977179756,
+      "learning_rate": 3.322062949159965e-05,
+      "loss": 1.3317,
+      "step": 431
+    },
+    {
+      "epoch": 0.29300551080966514,
+      "grad_norm": 0.7528583965841259,
+      "learning_rate": 3.318760499076359e-05,
+      "loss": 1.3428,
+      "step": 432
+    },
+    {
+      "epoch": 0.29368376430690973,
+      "grad_norm": 0.8233547389855966,
+      "learning_rate": 3.315451675154429e-05,
+      "loss": 1.3128,
+      "step": 433
+    },
+    {
+      "epoch": 0.29436201780415433,
+      "grad_norm": 0.8006537555521465,
+      "learning_rate": 3.312136493386397e-05,
+      "loss": 1.3215,
+      "step": 434
+    },
+    {
+      "epoch": 0.29504027130139887,
+      "grad_norm": 0.7790959381873427,
+      "learning_rate": 3.308814969795211e-05,
+      "loss": 1.3182,
+      "step": 435
+    },
+    {
+      "epoch": 0.29571852479864347,
+      "grad_norm": 0.8156005104406299,
+      "learning_rate": 3.305487120434473e-05,
+      "loss": 1.3138,
+      "step": 436
+    },
+    {
+      "epoch": 0.29639677829588806,
+      "grad_norm": 0.7995387221213647,
+      "learning_rate": 3.302152961388356e-05,
+      "loss": 1.3033,
+      "step": 437
+    },
+    {
+      "epoch": 0.29707503179313266,
+      "grad_norm": 0.7773890592212905,
+      "learning_rate": 3.298812508771531e-05,
+      "loss": 1.3119,
+      "step": 438
+    },
+    {
+      "epoch": 0.29775328529037726,
+      "grad_norm": 0.7765905341629566,
+      "learning_rate": 3.295465778729086e-05,
+      "loss": 1.3189,
+      "step": 439
+    },
+    {
+      "epoch": 0.29843153878762185,
+      "grad_norm": 0.7395137371426747,
+      "learning_rate": 3.29211278743645e-05,
+      "loss": 1.297,
+      "step": 440
+    },
+    {
+      "epoch": 0.29910979228486645,
+      "grad_norm": 0.7686752854804622,
+      "learning_rate": 3.288753551099314e-05,
+      "loss": 1.3332,
+      "step": 441
+    },
+    {
+      "epoch": 0.29978804578211105,
+      "grad_norm": 0.7899946804799971,
+      "learning_rate": 3.2853880859535505e-05,
+      "loss": 1.3049,
+      "step": 442
+    },
+    {
+      "epoch": 0.30046629927935564,
+      "grad_norm": 0.7603115262725664,
+      "learning_rate": 3.28201640826514e-05,
+      "loss": 1.3001,
+      "step": 443
+    },
+    {
+      "epoch": 0.30114455277660024,
+      "grad_norm": 0.785989206001199,
+      "learning_rate": 3.278638534330087e-05,
+      "loss": 1.3182,
+      "step": 444
+    },
+    {
+      "epoch": 0.30182280627384483,
+      "grad_norm": 0.7428802548334874,
+      "learning_rate": 3.2752544804743454e-05,
+      "loss": 1.2838,
+      "step": 445
+    },
+    {
+      "epoch": 0.30250105977108943,
+      "grad_norm": 0.7667601472507088,
+      "learning_rate": 3.2718642630537374e-05,
+      "loss": 1.2867,
+      "step": 446
+    },
+    {
+      "epoch": 0.303179313268334,
+      "grad_norm": 0.8357980934067926,
+      "learning_rate": 3.268467898453875e-05,
+      "loss": 1.3245,
+      "step": 447
+    },
+    {
+      "epoch": 0.3038575667655786,
+      "grad_norm": 0.7511945493729915,
+      "learning_rate": 3.2650654030900795e-05,
+      "loss": 1.3121,
+      "step": 448
+    },
+    {
+      "epoch": 0.3045358202628232,
+      "grad_norm": 0.805566590221847,
+      "learning_rate": 3.2616567934073055e-05,
+      "loss": 1.2974,
+      "step": 449
+    },
+    {
+      "epoch": 0.3052140737600678,
+      "grad_norm": 0.7858413638773709,
+      "learning_rate": 3.2582420858800596e-05,
+      "loss": 1.3424,
+      "step": 450
+    },
+    {
+      "epoch": 0.3058923272573124,
+      "grad_norm": 0.7950960516687662,
+      "learning_rate": 3.254821297012318e-05,
+      "loss": 1.3151,
+      "step": 451
+    },
+    {
+      "epoch": 0.306570580754557,
+      "grad_norm": 0.7816783938657665,
+      "learning_rate": 3.25139444333745e-05,
+      "loss": 1.2928,
+      "step": 452
+    },
+    {
+      "epoch": 0.3072488342518016,
+      "grad_norm": 0.8106894175155682,
+      "learning_rate": 3.24796154141814e-05,
+      "loss": 1.2952,
+      "step": 453
+    },
+    {
+      "epoch": 0.3079270877490462,
+      "grad_norm": 0.8169374050183266,
+      "learning_rate": 3.2445226078463e-05,
+      "loss": 1.2978,
+      "step": 454
+    },
+    {
+      "epoch": 0.3086053412462908,
+      "grad_norm": 0.8107605434555087,
+      "learning_rate": 3.241077659243e-05,
+      "loss": 1.3047,
+      "step": 455
+    },
+    {
+      "epoch": 0.3092835947435354,
+      "grad_norm": 0.7854751274722761,
+      "learning_rate": 3.2376267122583774e-05,
+      "loss": 1.3084,
+      "step": 456
+    },
+    {
+      "epoch": 0.30996184824078,
+      "grad_norm": 0.8358304822334056,
+      "learning_rate": 3.234169783571562e-05,
+      "loss": 1.3186,
+      "step": 457
+    },
+    {
+      "epoch": 0.3106401017380246,
+      "grad_norm": 0.8100113368228923,
+      "learning_rate": 3.230706889890595e-05,
+      "loss": 1.3132,
+      "step": 458
+    },
+    {
+      "epoch": 0.3113183552352692,
+      "grad_norm": 0.7714940319152012,
+      "learning_rate": 3.227238047952348e-05,
+      "loss": 1.2946,
+      "step": 459
+    },
+    {
+      "epoch": 0.3119966087325138,
+      "grad_norm": 0.8032984924634528,
+      "learning_rate": 3.223763274522442e-05,
+      "loss": 1.3338,
+      "step": 460
+    },
+    {
+      "epoch": 0.3126748622297584,
+      "grad_norm": 0.8491174491833109,
+      "learning_rate": 3.220282586395163e-05,
+      "loss": 1.3083,
+      "step": 461
+    },
+    {
+      "epoch": 0.313353115727003,
+      "grad_norm": 0.7589680883885287,
+      "learning_rate": 3.2167960003933883e-05,
+      "loss": 1.2991,
+      "step": 462
+    },
+    {
+      "epoch": 0.31403136922424757,
+      "grad_norm": 0.7724678326422264,
+      "learning_rate": 3.2133035333684985e-05,
+      "loss": 1.2792,
+      "step": 463
+    },
+    {
+      "epoch": 0.31470962272149217,
+      "grad_norm": 0.8192866442851098,
+      "learning_rate": 3.209805202200298e-05,
+      "loss": 1.3157,
+      "step": 464
+    },
+    {
+      "epoch": 0.31538787621873676,
+      "grad_norm": 0.7625468583050007,
+      "learning_rate": 3.206301023796934e-05,
+      "loss": 1.2978,
+      "step": 465
+    },
+    {
+      "epoch": 0.31606612971598136,
+      "grad_norm": 0.7773487076566746,
+      "learning_rate": 3.202791015094817e-05,
+      "loss": 1.3055,
+      "step": 466
+    },
+    {
+      "epoch": 0.31674438321322596,
+      "grad_norm": 0.7720199775299047,
+      "learning_rate": 3.199275193058533e-05,
+      "loss": 1.3373,
+      "step": 467
+    },
+    {
+      "epoch": 0.31742263671047055,
+      "grad_norm": 0.7702978706001685,
+      "learning_rate": 3.195753574680767e-05,
+      "loss": 1.2876,
+      "step": 468
+    },
+    {
+      "epoch": 0.31810089020771515,
+      "grad_norm": 0.7529851813319876,
+      "learning_rate": 3.1922261769822185e-05,
+      "loss": 1.2996,
+      "step": 469
+    },
+    {
+      "epoch": 0.31877914370495974,
+      "grad_norm": 1.2344261301658803,
+      "learning_rate": 3.1886930170115193e-05,
+      "loss": 1.4106,
+      "step": 470
+    },
+    {
+      "epoch": 0.31945739720220434,
+      "grad_norm": 0.8316996725515984,
+      "learning_rate": 3.1851541118451524e-05,
+      "loss": 1.3197,
+      "step": 471
+    },
+    {
+      "epoch": 0.32013565069944894,
+      "grad_norm": 0.8224106215404863,
+      "learning_rate": 3.181609478587367e-05,
+      "loss": 1.3022,
+      "step": 472
+    },
+    {
+      "epoch": 0.32081390419669353,
+      "grad_norm": 0.8093904826429651,
+      "learning_rate": 3.1780591343701e-05,
+      "loss": 1.2954,
+      "step": 473
+    },
+    {
+      "epoch": 0.32149215769393813,
+      "grad_norm": 0.7965078928648566,
+      "learning_rate": 3.1745030963528875e-05,
+      "loss": 1.3085,
+      "step": 474
+    },
+    {
+      "epoch": 0.3221704111911827,
+      "grad_norm": 0.7828063964744848,
+      "learning_rate": 3.170941381722785e-05,
+      "loss": 1.2965,
+      "step": 475
+    },
+    {
+      "epoch": 0.3228486646884273,
+      "grad_norm": 0.8876462180955584,
+      "learning_rate": 3.167374007694288e-05,
+      "loss": 1.2933,
+      "step": 476
+    },
+    {
+      "epoch": 0.3235269181856719,
+      "grad_norm": 0.7800447657814774,
+      "learning_rate": 3.163800991509239e-05,
+      "loss": 1.3203,
+      "step": 477
+    },
+    {
+      "epoch": 0.3242051716829165,
+      "grad_norm": 0.8282634521942043,
+      "learning_rate": 3.1602223504367574e-05,
+      "loss": 1.3233,
+      "step": 478
+    },
+    {
+      "epoch": 0.3248834251801611,
+      "grad_norm": 0.8535697522844348,
+      "learning_rate": 3.156638101773143e-05,
+      "loss": 1.323,
+      "step": 479
+    },
+    {
+      "epoch": 0.3255616786774057,
+      "grad_norm": 0.7954436934176183,
+      "learning_rate": 3.1530482628418e-05,
+      "loss": 1.2734,
+      "step": 480
+    },
+    {
+      "epoch": 0.32623993217465025,
+      "grad_norm": 0.9165026567224229,
+      "learning_rate": 3.1494528509931525e-05,
+      "loss": 1.3222,
+      "step": 481
+    },
+    {
+      "epoch": 0.32691818567189485,
+      "grad_norm": 0.7503270606156048,
+      "learning_rate": 3.145851883604558e-05,
+      "loss": 1.319,
+      "step": 482
+    },
+    {
+      "epoch": 0.32759643916913944,
+      "grad_norm": 0.9473825774688579,
+      "learning_rate": 3.1422453780802266e-05,
+      "loss": 1.3094,
+      "step": 483
+    },
+    {
+      "epoch": 0.32827469266638404,
+      "grad_norm": 0.7753116526540578,
+      "learning_rate": 3.1386333518511346e-05,
+      "loss": 1.3089,
+      "step": 484
+    },
+    {
+      "epoch": 0.32895294616362863,
+      "grad_norm": 1.2406352929881868,
+      "learning_rate": 3.135015822374942e-05,
+      "loss": 1.4282,
+      "step": 485
+    },
+    {
+      "epoch": 0.32963119966087323,
+      "grad_norm": 0.8914851616392252,
+      "learning_rate": 3.131392807135904e-05,
+      "loss": 1.3157,
+      "step": 486
+    },
+    {
+      "epoch": 0.3303094531581178,
+      "grad_norm": 0.8038376927352104,
+      "learning_rate": 3.127764323644794e-05,
+      "loss": 1.3288,
+      "step": 487
+    },
+    {
+      "epoch": 0.3309877066553624,
+      "grad_norm": 0.8892964512690509,
+      "learning_rate": 3.124130389438811e-05,
+      "loss": 1.3258,
+      "step": 488
+    },
+    {
+      "epoch": 0.331665960152607,
+      "grad_norm": 0.8196194137202071,
+      "learning_rate": 3.120491022081501e-05,
+      "loss": 1.3147,
+      "step": 489
+    },
+    {
+      "epoch": 0.3323442136498516,
+      "grad_norm": 0.7444040418507955,
+      "learning_rate": 3.1168462391626667e-05,
+      "loss": 1.2869,
+      "step": 490
+    },
+    {
+      "epoch": 0.3330224671470962,
+      "grad_norm": 0.8593102441640206,
+      "learning_rate": 3.1131960582982884e-05,
+      "loss": 1.2752,
+      "step": 491
+    },
+    {
+      "epoch": 0.3337007206443408,
+      "grad_norm": 0.8307583231844464,
+      "learning_rate": 3.1095404971304334e-05,
+      "loss": 1.2886,
+      "step": 492
+    },
+    {
+      "epoch": 0.3343789741415854,
+      "grad_norm": 0.7950709207927126,
+      "learning_rate": 3.105879573327174e-05,
+      "loss": 1.2985,
+      "step": 493
+    },
+    {
+      "epoch": 0.33505722763883,
+      "grad_norm": 0.8339570103968118,
+      "learning_rate": 3.1022133045825024e-05,
+      "loss": 1.2889,
+      "step": 494
+    },
+    {
+      "epoch": 0.3357354811360746,
+      "grad_norm": 0.7492370042831757,
+      "learning_rate": 3.098541708616242e-05,
+      "loss": 1.2747,
+      "step": 495
+    },
+    {
+      "epoch": 0.3364137346333192,
+      "grad_norm": 0.8170116352141963,
+      "learning_rate": 3.094864803173964e-05,
+      "loss": 1.3251,
+      "step": 496
+    },
+    {
+      "epoch": 0.3370919881305638,
+      "grad_norm": 0.8032401198395067,
+      "learning_rate": 3.091182606026903e-05,
+      "loss": 1.3076,
+      "step": 497
+    },
+    {
+      "epoch": 0.3377702416278084,
+      "grad_norm": 0.7546853470769238,
+      "learning_rate": 3.087495134971867e-05,
+      "loss": 1.3316,
+      "step": 498
+    },
+    {
+      "epoch": 0.338448495125053,
+      "grad_norm": 0.8134389048041515,
+      "learning_rate": 3.083802407831158e-05,
+      "loss": 1.3006,
+      "step": 499
+    },
+    {
+      "epoch": 0.3391267486222976,
+      "grad_norm": 1.0717469671088464,
+      "learning_rate": 3.080104442452476e-05,
+      "loss": 1.4107,
+      "step": 500
+    },
+    {
+      "epoch": 0.3398050021195422,
+      "grad_norm": 0.8548690740301038,
+      "learning_rate": 3.0764012567088435e-05,
+      "loss": 1.3009,
+      "step": 501
+    },
+    {
+      "epoch": 0.3404832556167868,
+      "grad_norm": 0.8953045227864236,
+      "learning_rate": 3.0726928684985104e-05,
+      "loss": 1.311,
+      "step": 502
+    },
+    {
+      "epoch": 0.34116150911403137,
+      "grad_norm": 0.8364067994792729,
+      "learning_rate": 3.068979295744876e-05,
+      "loss": 1.3196,
+      "step": 503
+    },
+    {
+      "epoch": 0.34183976261127597,
+      "grad_norm": 0.7752202509932031,
+      "learning_rate": 3.06526055639639e-05,
+      "loss": 1.3337,
+      "step": 504
+    },
+    {
+      "epoch": 0.34251801610852056,
+      "grad_norm": 0.8529787549558605,
+      "learning_rate": 3.061536668426481e-05,
+      "loss": 1.2983,
+      "step": 505
+    },
+    {
+      "epoch": 0.34319626960576516,
+      "grad_norm": 0.8587010705698397,
+      "learning_rate": 3.0578076498334574e-05,
+      "loss": 1.2886,
+      "step": 506
+    },
+    {
+      "epoch": 0.34387452310300975,
+      "grad_norm": 0.7328588429654739,
+      "learning_rate": 3.054073518640428e-05,
+      "loss": 1.3107,
+      "step": 507
+    },
+    {
+      "epoch": 0.34455277660025435,
+      "grad_norm": 0.8162742121854552,
+      "learning_rate": 3.0503342928952073e-05,
+      "loss": 1.2911,
+      "step": 508
+    },
+    {
+      "epoch": 0.34523103009749895,
+      "grad_norm": 0.8794715093274978,
+      "learning_rate": 3.0465899906702366e-05,
+      "loss": 1.3158,
+      "step": 509
+    },
+    {
+      "epoch": 0.34590928359474354,
+      "grad_norm": 0.7708822837782366,
+      "learning_rate": 3.042840630062493e-05,
+      "loss": 1.3143,
+      "step": 510
+    },
+    {
+      "epoch": 0.34658753709198814,
+      "grad_norm": 0.7911466862190698,
+      "learning_rate": 3.0390862291933995e-05,
+      "loss": 1.294,
+      "step": 511
+    },
+    {
+      "epoch": 0.34726579058923274,
+      "grad_norm": 0.8538894880559746,
+      "learning_rate": 3.0353268062087412e-05,
+      "loss": 1.3222,
+      "step": 512
+    },
+    {
+      "epoch": 0.34794404408647733,
+      "grad_norm": 0.7852631755793705,
+      "learning_rate": 3.031562379278575e-05,
+      "loss": 1.2989,
+      "step": 513
+    },
+    {
+      "epoch": 0.34862229758372193,
+      "grad_norm": 0.8734426165594713,
+      "learning_rate": 3.027792966597145e-05,
+      "loss": 1.3079,
+      "step": 514
+    },
+    {
+      "epoch": 0.3493005510809665,
+      "grad_norm": 0.744360116551954,
+      "learning_rate": 3.0240185863827904e-05,
+      "loss": 1.2956,
+      "step": 515
+    },
+    {
+      "epoch": 0.3499788045782111,
+      "grad_norm": 0.76062354917638,
+      "learning_rate": 3.0202392568778598e-05,
+      "loss": 1.2919,
+      "step": 516
+    },
+    {
+      "epoch": 0.3506570580754557,
+      "grad_norm": 0.7970415706248092,
+      "learning_rate": 3.0164549963486238e-05,
+      "loss": 1.3153,
+      "step": 517
+    },
+    {
+      "epoch": 0.3513353115727003,
+      "grad_norm": 0.763709088739663,
+      "learning_rate": 3.012665823085185e-05,
+      "loss": 1.3278,
+      "step": 518
+    },
+    {
+      "epoch": 0.3520135650699449,
+      "grad_norm": 0.7546502587704492,
+      "learning_rate": 3.008871755401389e-05,
+      "loss": 1.2888,
+      "step": 519
+    },
+    {
+      "epoch": 0.3526918185671895,
+      "grad_norm": 0.8045823402075674,
+      "learning_rate": 3.0050728116347402e-05,
+      "loss": 1.3287,
+      "step": 520
+    },
+    {
+      "epoch": 0.3533700720644341,
+      "grad_norm": 0.7889243224775581,
+      "learning_rate": 3.0012690101463066e-05,
+      "loss": 1.2737,
+      "step": 521
+    },
+    {
+      "epoch": 0.3540483255616787,
+      "grad_norm": 0.7503150202965315,
+      "learning_rate": 2.9974603693206368e-05,
+      "loss": 1.3118,
+      "step": 522
+    },
+    {
+      "epoch": 0.3547265790589233,
+      "grad_norm": 0.7577330239057181,
+      "learning_rate": 2.9936469075656687e-05,
+      "loss": 1.3058,
+      "step": 523
+    },
+    {
+      "epoch": 0.3554048325561679,
+      "grad_norm": 0.7986442780873649,
+      "learning_rate": 2.9898286433126394e-05,
+      "loss": 1.3214,
+      "step": 524
+    },
+    {
+      "epoch": 0.3560830860534125,
+      "grad_norm": 0.7996687192682629,
+      "learning_rate": 2.9860055950159997e-05,
+      "loss": 1.3095,
+      "step": 525
+    },
+    {
+      "epoch": 0.35676133955065703,
+      "grad_norm": 0.7385651563498843,
+      "learning_rate": 2.9821777811533207e-05,
+      "loss": 1.3111,
+      "step": 526
+    },
+    {
+      "epoch": 0.3574395930479016,
+      "grad_norm": 0.7795490034948527,
+      "learning_rate": 2.9783452202252066e-05,
+      "loss": 1.3312,
+      "step": 527
+    },
+    {
+      "epoch": 0.3581178465451462,
+      "grad_norm": 0.8121483784755045,
+      "learning_rate": 2.974507930755206e-05,
+      "loss": 1.2942,
+      "step": 528
+    },
+    {
+      "epoch": 0.3587961000423908,
+      "grad_norm": 0.782162395784764,
+      "learning_rate": 2.970665931289722e-05,
+      "loss": 1.3046,
+      "step": 529
+    },
+    {
+      "epoch": 0.3594743535396354,
+      "grad_norm": 0.7766997025659732,
+      "learning_rate": 2.9668192403979198e-05,
+      "loss": 1.334,
+      "step": 530
+    },
+    {
+      "epoch": 0.36015260703688,
+      "grad_norm": 0.7738747592143685,
+      "learning_rate": 2.9629678766716418e-05,
+      "loss": 1.3246,
+      "step": 531
+    },
+    {
+      "epoch": 0.3608308605341246,
+      "grad_norm": 0.7648781245429924,
+      "learning_rate": 2.959111858725313e-05,
+      "loss": 1.305,
+      "step": 532
+    },
+    {
+      "epoch": 0.3615091140313692,
+      "grad_norm": 0.7418982711728055,
+      "learning_rate": 2.9552512051958548e-05,
+      "loss": 1.2996,
+      "step": 533
+    },
+    {
+      "epoch": 0.3621873675286138,
+      "grad_norm": 0.7420503991032353,
+      "learning_rate": 2.9513859347425925e-05,
+      "loss": 1.2734,
+      "step": 534
+    },
+    {
+      "epoch": 0.3628656210258584,
+      "grad_norm": 0.7367250178072308,
+      "learning_rate": 2.9475160660471663e-05,
+      "loss": 1.3182,
+      "step": 535
+    },
+    {
+      "epoch": 0.363543874523103,
+      "grad_norm": 0.7864827093672973,
+      "learning_rate": 2.9436416178134405e-05,
+      "loss": 1.3212,
+      "step": 536
+    },
+    {
+      "epoch": 0.3642221280203476,
+      "grad_norm": 0.7395342801664027,
+      "learning_rate": 2.9397626087674133e-05,
+      "loss": 1.3069,
+      "step": 537
+    },
+    {
+      "epoch": 0.3649003815175922,
+      "grad_norm": 0.7599790206682413,
+      "learning_rate": 2.9358790576571258e-05,
+      "loss": 1.2985,
+      "step": 538
+    },
+    {
+      "epoch": 0.3655786350148368,
+      "grad_norm": 0.7743513258616831,
+      "learning_rate": 2.9319909832525724e-05,
+      "loss": 1.2913,
+      "step": 539
+    },
+    {
+      "epoch": 0.3662568885120814,
+      "grad_norm": 0.7645342223650176,
+      "learning_rate": 2.928098404345609e-05,
+      "loss": 1.3443,
+      "step": 540
+    },
+    {
+      "epoch": 0.366935142009326,
+      "grad_norm": 0.8268396657883249,
+      "learning_rate": 2.9242013397498638e-05,
+      "loss": 1.2847,
+      "step": 541
+    },
+    {
+      "epoch": 0.3676133955065706,
+      "grad_norm": 0.7682172066640904,
+      "learning_rate": 2.9202998083006436e-05,
+      "loss": 1.2781,
+      "step": 542
+    },
+    {
+      "epoch": 0.36829164900381517,
+      "grad_norm": 0.77396077420331,
+      "learning_rate": 2.916393828854845e-05,
+      "loss": 1.3145,
+      "step": 543
+    },
+    {
+      "epoch": 0.36896990250105977,
+      "grad_norm": 0.7959331968002789,
+      "learning_rate": 2.9124834202908636e-05,
+      "loss": 1.2987,
+      "step": 544
+    },
+    {
+      "epoch": 0.36964815599830436,
+      "grad_norm": 0.7583473002627757,
+      "learning_rate": 2.908568601508501e-05,
+      "loss": 1.2757,
+      "step": 545
+    },
+    {
+      "epoch": 0.37032640949554896,
+      "grad_norm": 0.8118008723150614,
+      "learning_rate": 2.9046493914288744e-05,
+      "loss": 1.2702,
+      "step": 546
+    },
+    {
+      "epoch": 0.37100466299279355,
+      "grad_norm": 0.8101625355380598,
+      "learning_rate": 2.900725808994325e-05,
+      "loss": 1.2959,
+      "step": 547
+    },
+    {
+      "epoch": 0.37168291649003815,
+      "grad_norm": 0.7191114911111972,
+      "learning_rate": 2.8967978731683266e-05,
+      "loss": 1.259,
+      "step": 548
+    },
+    {
+      "epoch": 0.37236116998728275,
+      "grad_norm": 0.7891369500757057,
+      "learning_rate": 2.8928656029353933e-05,
+      "loss": 1.3053,
+      "step": 549
+    },
+    {
+      "epoch": 0.37303942348452734,
+      "grad_norm": 0.7523339999578558,
+      "learning_rate": 2.88892901730099e-05,
+      "loss": 1.2761,
+      "step": 550
+    },
+    {
+      "epoch": 0.37371767698177194,
+      "grad_norm": 0.7462057986119706,
+      "learning_rate": 2.8849881352914354e-05,
+      "loss": 1.3017,
+      "step": 551
+    },
+    {
+      "epoch": 0.37439593047901654,
+      "grad_norm": 0.7323775120616418,
+      "learning_rate": 2.8810429759538175e-05,
+      "loss": 1.285,
+      "step": 552
+    },
+    {
+      "epoch": 0.37507418397626113,
+      "grad_norm": 0.7863772550083453,
+      "learning_rate": 2.877093558355895e-05,
+      "loss": 1.3328,
+      "step": 553
+    },
+    {
+      "epoch": 0.37575243747350573,
+      "grad_norm": 0.7591796309879588,
+      "learning_rate": 2.873139901586008e-05,
+      "loss": 1.2943,
+      "step": 554
+    },
+    {
+      "epoch": 0.3764306909707503,
+      "grad_norm": 0.7427278153708342,
+      "learning_rate": 2.869182024752986e-05,
+      "loss": 1.2892,
+      "step": 555
+    },
+    {
+      "epoch": 0.3771089444679949,
+      "grad_norm": 0.7561087499786749,
+      "learning_rate": 2.8652199469860544e-05,
+      "loss": 1.2908,
+      "step": 556
+    },
+    {
+      "epoch": 0.3777871979652395,
+      "grad_norm": 0.7096955290839252,
+      "learning_rate": 2.8612536874347428e-05,
+      "loss": 1.271,
+      "step": 557
+    },
+    {
+      "epoch": 0.3784654514624841,
+      "grad_norm": 0.7528073877703341,
+      "learning_rate": 2.857283265268792e-05,
+      "loss": 1.28,
+      "step": 558
+    },
+    {
+      "epoch": 0.3791437049597287,
+      "grad_norm": 0.7314180873816569,
+      "learning_rate": 2.853308699678061e-05,
+      "loss": 1.2789,
+      "step": 559
+    },
+    {
+      "epoch": 0.3798219584569733,
+      "grad_norm": 0.7891519735832144,
+      "learning_rate": 2.8493300098724374e-05,
+      "loss": 1.3019,
+      "step": 560
+    },
+    {
+      "epoch": 0.3805002119542179,
+      "grad_norm": 0.792681198228241,
+      "learning_rate": 2.8453472150817382e-05,
+      "loss": 1.3276,
+      "step": 561
+    },
+    {
+      "epoch": 0.3811784654514625,
+      "grad_norm": 0.8203111963480278,
+      "learning_rate": 2.841360334555624e-05,
+      "loss": 1.3002,
+      "step": 562
+    },
+    {
+      "epoch": 0.3818567189487071,
+      "grad_norm": 0.7812193618738654,
+      "learning_rate": 2.8373693875634997e-05,
+      "loss": 1.2996,
+      "step": 563
+    },
+    {
+      "epoch": 0.3825349724459517,
+      "grad_norm": 0.796401799095726,
+      "learning_rate": 2.8333743933944268e-05,
+      "loss": 1.3071,
+      "step": 564
+    },
+    {
+      "epoch": 0.3832132259431963,
+      "grad_norm": 0.8106133032954245,
+      "learning_rate": 2.829375371357025e-05,
+      "loss": 1.3186,
+      "step": 565
+    },
+    {
+      "epoch": 0.3838914794404409,
+      "grad_norm": 0.7548783654866896,
+      "learning_rate": 2.8253723407793855e-05,
+      "loss": 1.2779,
+      "step": 566
+    },
+    {
+      "epoch": 0.3845697329376855,
+      "grad_norm": 0.7794102240472927,
+      "learning_rate": 2.8213653210089692e-05,
+      "loss": 1.3019,
+      "step": 567
+    },
+    {
+      "epoch": 0.3852479864349301,
+      "grad_norm": 0.7765921389066097,
+      "learning_rate": 2.81735433141252e-05,
+      "loss": 1.2782,
+      "step": 568
+    },
+    {
+      "epoch": 0.3859262399321747,
+      "grad_norm": 0.7981815510661463,
+      "learning_rate": 2.8133393913759684e-05,
+      "loss": 1.2772,
+      "step": 569
+    },
+    {
+      "epoch": 0.38660449342941927,
+      "grad_norm": 0.7543209706721942,
+      "learning_rate": 2.8093205203043377e-05,
+      "loss": 1.3033,
+      "step": 570
+    },
+    {
+      "epoch": 0.38728274692666387,
+      "grad_norm": 0.7711577947039863,
+      "learning_rate": 2.805297737621651e-05,
+      "loss": 1.2896,
+      "step": 571
+    },
+    {
+      "epoch": 0.3879610004239084,
+      "grad_norm": 0.7093645514898603,
+      "learning_rate": 2.801271062770838e-05,
+      "loss": 1.2946,
+      "step": 572
+    },
+    {
+      "epoch": 0.388639253921153,
+      "grad_norm": 0.7602736069559685,
+      "learning_rate": 2.7972405152136377e-05,
+      "loss": 1.2972,
+      "step": 573
+    },
+    {
+      "epoch": 0.3893175074183976,
+      "grad_norm": 0.7372141195181956,
+      "learning_rate": 2.793206114430509e-05,
+      "loss": 1.2788,
+      "step": 574
+    },
+    {
+      "epoch": 0.3899957609156422,
+      "grad_norm": 0.7620898187142349,
+      "learning_rate": 2.789167879920533e-05,
+      "loss": 1.4549,
+      "step": 575
+    },
+    {
+      "epoch": 0.3906740144128868,
+      "grad_norm": 0.8146306015267761,
+      "learning_rate": 2.7851258312013203e-05,
+      "loss": 1.2849,
+      "step": 576
+    },
+    {
+      "epoch": 0.3913522679101314,
+      "grad_norm": 0.7574593357968628,
+      "learning_rate": 2.781079987808917e-05,
+      "loss": 1.2847,
+      "step": 577
+    },
+    {
+      "epoch": 0.392030521407376,
+      "grad_norm": 0.7852134352094805,
+      "learning_rate": 2.7770303692977077e-05,
+      "loss": 1.3101,
+      "step": 578
+    },
+    {
+      "epoch": 0.3927087749046206,
+      "grad_norm": 0.7301643782984193,
+      "learning_rate": 2.7729769952403255e-05,
+      "loss": 1.3081,
+      "step": 579
+    },
+    {
+      "epoch": 0.3933870284018652,
+      "grad_norm": 0.5855026097478325,
+      "learning_rate": 2.7689198852275512e-05,
+      "loss": 1.4311,
+      "step": 580
+    },
+    {
+      "epoch": 0.3940652818991098,
+      "grad_norm": 0.8453462322081536,
+      "learning_rate": 2.764859058868228e-05,
+      "loss": 1.2846,
+      "step": 581
+    },
+    {
+      "epoch": 0.3947435353963544,
+      "grad_norm": 0.7932115499525736,
+      "learning_rate": 2.7607945357891553e-05,
+      "loss": 1.2994,
+      "step": 582
+    },
+    {
+      "epoch": 0.39542178889359897,
+      "grad_norm": 0.7878322433522394,
+      "learning_rate": 2.756726335635002e-05,
+      "loss": 1.305,
+      "step": 583
+    },
+    {
+      "epoch": 0.39610004239084357,
+      "grad_norm": 0.7140626220319872,
+      "learning_rate": 2.7526544780682083e-05,
+      "loss": 1.3043,
+      "step": 584
+    },
+    {
+      "epoch": 0.39677829588808816,
+      "grad_norm": 0.8063047757487747,
+      "learning_rate": 2.7485789827688934e-05,
+      "loss": 1.2635,
+      "step": 585
+    },
+    {
+      "epoch": 0.39745654938533276,
+      "grad_norm": 0.8190532547225947,
+      "learning_rate": 2.7444998694347547e-05,
+      "loss": 1.2914,
+      "step": 586
+    },
+    {
+      "epoch": 0.39813480288257735,
+      "grad_norm": 0.7347525125016025,
+      "learning_rate": 2.7404171577809808e-05,
+      "loss": 1.2758,
+      "step": 587
+    },
+    {
+      "epoch": 0.39881305637982195,
+      "grad_norm": 0.8512983415572944,
+      "learning_rate": 2.7363308675401478e-05,
+      "loss": 1.2933,
+      "step": 588
+    },
+    {
+      "epoch": 0.39949130987706655,
+      "grad_norm": 0.7655041356648618,
+      "learning_rate": 2.7322410184621295e-05,
+      "loss": 1.2713,
+      "step": 589
+    },
+    {
+      "epoch": 0.40016956337431114,
+      "grad_norm": 0.7355254071516576,
+      "learning_rate": 2.7281476303140014e-05,
+      "loss": 1.2871,
+      "step": 590
+    },
+    {
+      "epoch": 0.40084781687155574,
+      "grad_norm": 0.7414572019400784,
+      "learning_rate": 2.7240507228799415e-05,
+      "loss": 1.2798,
+      "step": 591
+    },
+    {
+      "epoch": 0.40152607036880034,
+      "grad_norm": 0.7934461535907612,
+      "learning_rate": 2.7199503159611396e-05,
+      "loss": 1.2904,
+      "step": 592
+    },
+    {
+      "epoch": 0.40220432386604493,
+      "grad_norm": 0.7512655190197016,
+      "learning_rate": 2.7158464293756975e-05,
+      "loss": 1.289,
+      "step": 593
+    },
+    {
+      "epoch": 0.40288257736328953,
+      "grad_norm": 0.7490830945846155,
+      "learning_rate": 2.711739082958536e-05,
+      "loss": 1.3159,
+      "step": 594
+    },
+    {
+      "epoch": 0.4035608308605341,
+      "grad_norm": 0.8836221765235461,
+      "learning_rate": 2.7076282965612967e-05,
+      "loss": 1.2967,
+      "step": 595
+    },
+    {
+      "epoch": 0.4042390843577787,
+      "grad_norm": 0.7625471094203554,
+      "learning_rate": 2.7035140900522506e-05,
+      "loss": 1.2746,
+      "step": 596
+    },
+    {
+      "epoch": 0.4049173378550233,
+      "grad_norm": 0.7754942595291234,
+      "learning_rate": 2.6993964833161937e-05,
+      "loss": 1.259,
+      "step": 597
+    },
+    {
+      "epoch": 0.4055955913522679,
+      "grad_norm": 0.8302288117169206,
+      "learning_rate": 2.6952754962543604e-05,
+      "loss": 1.299,
+      "step": 598
+    },
+    {
+      "epoch": 0.4062738448495125,
+      "grad_norm": 0.7458232064348682,
+      "learning_rate": 2.6911511487843217e-05,
+      "loss": 1.2856,
+      "step": 599
+    },
+    {
+      "epoch": 0.4069520983467571,
+      "grad_norm": 0.7533008733753166,
+      "learning_rate": 2.6870234608398872e-05,
+      "loss": 1.2563,
+      "step": 600
+    },
+    {
+      "epoch": 0.4076303518440017,
+      "grad_norm": 0.713259841427573,
+      "learning_rate": 2.682892452371017e-05,
+      "loss": 1.2735,
+      "step": 601
+    },
+    {
+      "epoch": 0.4083086053412463,
+      "grad_norm": 0.7407768024928815,
+      "learning_rate": 2.6787581433437156e-05,
+      "loss": 1.3237,
+      "step": 602
+    },
+    {
+      "epoch": 0.4089868588384909,
+      "grad_norm": 0.7811252651845391,
+      "learning_rate": 2.674620553739941e-05,
+      "loss": 1.2713,
+      "step": 603
+    },
+    {
+      "epoch": 0.4096651123357355,
+      "grad_norm": 0.7829564243679579,
+      "learning_rate": 2.6704797035575083e-05,
+      "loss": 1.2687,
+      "step": 604
+    },
+    {
+      "epoch": 0.4103433658329801,
+      "grad_norm": 0.7540347265094556,
+      "learning_rate": 2.6663356128099902e-05,
+      "loss": 1.2961,
+      "step": 605
+    },
+    {
+      "epoch": 0.4110216193302247,
+      "grad_norm": 0.7119414126159732,
+      "learning_rate": 2.6621883015266217e-05,
+      "loss": 1.2603,
+      "step": 606
+    },
+    {
+      "epoch": 0.4116998728274693,
+      "grad_norm": 0.7666224044912271,
+      "learning_rate": 2.658037789752204e-05,
+      "loss": 1.2717,
+      "step": 607
+    },
+    {
+      "epoch": 0.4123781263247139,
+      "grad_norm": 0.7950232772588871,
+      "learning_rate": 2.6538840975470064e-05,
+      "loss": 1.3044,
+      "step": 608
+    },
+    {
+      "epoch": 0.4130563798219585,
+      "grad_norm": 0.7396309531748422,
+      "learning_rate": 2.6497272449866708e-05,
+      "loss": 1.2887,
+      "step": 609
+    },
+    {
+      "epoch": 0.41373463331920307,
+      "grad_norm": 0.7834865102164703,
+      "learning_rate": 2.6455672521621112e-05,
+      "loss": 1.3073,
+      "step": 610
+    },
+    {
+      "epoch": 0.41441288681644767,
+      "grad_norm": 0.7303842980777212,
+      "learning_rate": 2.6414041391794226e-05,
+      "loss": 1.2773,
+      "step": 611
+    },
+    {
+      "epoch": 0.41509114031369226,
+      "grad_norm": 0.7418178497569553,
+      "learning_rate": 2.637237926159779e-05,
+      "loss": 1.2734,
+      "step": 612
+    },
+    {
+      "epoch": 0.41576939381093686,
+      "grad_norm": 0.7576949634136853,
+      "learning_rate": 2.633068633239335e-05,
+      "loss": 1.2813,
+      "step": 613
+    },
+    {
+      "epoch": 0.41644764730818146,
+      "grad_norm": 0.7651480859750399,
+      "learning_rate": 2.6288962805691357e-05,
+      "loss": 1.281,
+      "step": 614
+    },
+    {
+      "epoch": 0.41712590080542605,
+      "grad_norm": 0.7986560095498149,
+      "learning_rate": 2.6247208883150107e-05,
+      "loss": 1.2633,
+      "step": 615
+    },
+    {
+      "epoch": 0.41780415430267065,
+      "grad_norm": 0.7696603647611979,
+      "learning_rate": 2.6205424766574826e-05,
+      "loss": 1.2987,
+      "step": 616
+    },
+    {
+      "epoch": 0.41848240779991525,
+      "grad_norm": 0.7517581233977016,
+      "learning_rate": 2.6163610657916655e-05,
+      "loss": 1.2808,
+      "step": 617
+    },
+    {
+      "epoch": 0.4191606612971598,
+      "grad_norm": 0.7753358993235157,
+      "learning_rate": 2.6121766759271716e-05,
+      "loss": 1.2774,
+      "step": 618
+    },
+    {
+      "epoch": 0.4198389147944044,
+      "grad_norm": 0.7566013758054129,
+      "learning_rate": 2.6079893272880102e-05,
+      "loss": 1.28,
+      "step": 619
+    },
+    {
+      "epoch": 0.420517168291649,
+      "grad_norm": 0.7965400573019791,
+      "learning_rate": 2.6037990401124903e-05,
+      "loss": 1.3203,
+      "step": 620
+    },
+    {
+      "epoch": 0.4211954217888936,
+      "grad_norm": 0.8457299678893127,
+      "learning_rate": 2.5996058346531247e-05,
+      "loss": 1.2675,
+      "step": 621
+    },
+    {
+      "epoch": 0.4218736752861382,
+      "grad_norm": 0.7236788297057595,
+      "learning_rate": 2.5954097311765294e-05,
+      "loss": 1.2608,
+      "step": 622
+    },
+    {
+      "epoch": 0.42255192878338277,
+      "grad_norm": 0.802431150827969,
+      "learning_rate": 2.5912107499633278e-05,
+      "loss": 1.2753,
+      "step": 623
+    },
+    {
+      "epoch": 0.42323018228062737,
+      "grad_norm": 0.824996773908104,
+      "learning_rate": 2.5870089113080533e-05,
+      "loss": 1.3015,
+      "step": 624
+    },
+    {
+      "epoch": 0.42390843577787196,
+      "grad_norm": 0.7944101948619141,
+      "learning_rate": 2.5828042355190475e-05,
+      "loss": 1.3058,
+      "step": 625
+    },
+    {
+      "epoch": 0.42458668927511656,
+      "grad_norm": 0.7380566963430693,
+      "learning_rate": 2.5785967429183652e-05,
+      "loss": 1.2724,
+      "step": 626
+    },
+    {
+      "epoch": 0.42526494277236115,
+      "grad_norm": 0.7848691720302051,
+      "learning_rate": 2.574386453841678e-05,
+      "loss": 1.2688,
+      "step": 627
+    },
+    {
+      "epoch": 0.42594319626960575,
+      "grad_norm": 0.7668918737753835,
+      "learning_rate": 2.570173388638169e-05,
+      "loss": 1.2937,
+      "step": 628
+    },
+    {
+      "epoch": 0.42662144976685035,
+      "grad_norm": 0.7983440220905413,
+      "learning_rate": 2.5659575676704426e-05,
+      "loss": 1.2901,
+      "step": 629
+    },
+    {
+      "epoch": 0.42729970326409494,
+      "grad_norm": 0.761042939856984,
+      "learning_rate": 2.56173901131442e-05,
+      "loss": 1.2911,
+      "step": 630
+    },
+    {
+      "epoch": 0.42797795676133954,
+      "grad_norm": 0.7645593903471752,
+      "learning_rate": 2.5575177399592447e-05,
+      "loss": 1.2787,
+      "step": 631
+    },
+    {
+      "epoch": 0.42865621025858414,
+      "grad_norm": 0.8295918351947635,
+      "learning_rate": 2.5532937740071814e-05,
+      "loss": 1.2907,
+      "step": 632
+    },
+    {
+      "epoch": 0.42933446375582873,
+      "grad_norm": 0.7930301114347823,
+      "learning_rate": 2.5490671338735178e-05,
+      "loss": 1.2738,
+      "step": 633
+    },
+    {
+      "epoch": 0.43001271725307333,
+      "grad_norm": 0.7896063345267272,
+      "learning_rate": 2.5448378399864683e-05,
+      "loss": 1.2968,
+      "step": 634
+    },
+    {
+      "epoch": 0.4306909707503179,
+      "grad_norm": 0.7951954101418609,
+      "learning_rate": 2.540605912787073e-05,
+      "loss": 1.2791,
+      "step": 635
+    },
+    {
+      "epoch": 0.4313692242475625,
+      "grad_norm": 0.8325761502518559,
+      "learning_rate": 2.5363713727290974e-05,
+      "loss": 1.3083,
+      "step": 636
+    },
+    {
+      "epoch": 0.4320474777448071,
+      "grad_norm": 0.8414502480322035,
+      "learning_rate": 2.5321342402789377e-05,
+      "loss": 1.2975,
+      "step": 637
+    },
+    {
+      "epoch": 0.4327257312420517,
+      "grad_norm": 0.8131875545774248,
+      "learning_rate": 2.5278945359155183e-05,
+      "loss": 1.29,
+      "step": 638
+    },
+    {
+      "epoch": 0.4334039847392963,
+      "grad_norm": 0.7968566988162644,
+      "learning_rate": 2.5236522801301945e-05,
+      "loss": 1.2808,
+      "step": 639
+    },
+    {
+      "epoch": 0.4340822382365409,
+      "grad_norm": 0.797923868742449,
+      "learning_rate": 2.5194074934266538e-05,
+      "loss": 1.276,
+      "step": 640
+    },
+    {
+      "epoch": 0.4347604917337855,
+      "grad_norm": 0.8148462451491864,
+      "learning_rate": 2.5151601963208152e-05,
+      "loss": 1.2816,
+      "step": 641
+    },
+    {
+      "epoch": 0.4354387452310301,
+      "grad_norm": 0.7835535568430211,
+      "learning_rate": 2.510910409340732e-05,
+      "loss": 1.3027,
+      "step": 642
+    },
+    {
+      "epoch": 0.4361169987282747,
+      "grad_norm": 0.7496355105193699,
+      "learning_rate": 2.50665815302649e-05,
+      "loss": 1.2834,
+      "step": 643
+    },
+    {
+      "epoch": 0.4367952522255193,
+      "grad_norm": 0.7875835891135449,
+      "learning_rate": 2.502403447930112e-05,
+      "loss": 1.2517,
+      "step": 644
+    },
+    {
+      "epoch": 0.4374735057227639,
+      "grad_norm": 0.7389635498409806,
+      "learning_rate": 2.4981463146154546e-05,
+      "loss": 1.2756,
+      "step": 645
+    },
+    {
+      "epoch": 0.4381517592200085,
+      "grad_norm": 0.7604996907814987,
+      "learning_rate": 2.493886773658111e-05,
+      "loss": 1.2869,
+      "step": 646
+    },
+    {
+      "epoch": 0.4388300127172531,
+      "grad_norm": 0.7293596088195266,
+      "learning_rate": 2.48962484564531e-05,
+      "loss": 1.2868,
+      "step": 647
+    },
+    {
+      "epoch": 0.4395082662144977,
+      "grad_norm": 0.780363247786965,
+      "learning_rate": 2.4853605511758193e-05,
+      "loss": 1.3103,
+      "step": 648
+    },
+    {
+      "epoch": 0.4401865197117423,
+      "grad_norm": 0.8259123145424743,
+      "learning_rate": 2.4810939108598443e-05,
+      "loss": 1.301,
+      "step": 649
+    },
+    {
+      "epoch": 0.44086477320898687,
+      "grad_norm": 0.8247133906271952,
+      "learning_rate": 2.4768249453189256e-05,
+      "loss": 1.3135,
+      "step": 650
+    },
+    {
+      "epoch": 0.44154302670623147,
+      "grad_norm": 0.7141969140986962,
+      "learning_rate": 2.4725536751858453e-05,
+      "loss": 1.2542,
+      "step": 651
+    },
+    {
+      "epoch": 0.44222128020347606,
+      "grad_norm": 0.7521164292440358,
+      "learning_rate": 2.4682801211045214e-05,
+      "loss": 1.2935,
+      "step": 652
+    },
+    {
+      "epoch": 0.44289953370072066,
+      "grad_norm": 0.7843461839457703,
+      "learning_rate": 2.4640043037299135e-05,
+      "loss": 1.2799,
+      "step": 653
+    },
+    {
+      "epoch": 0.44357778719796526,
+      "grad_norm": 0.8020780634370414,
+      "learning_rate": 2.4597262437279172e-05,
+      "loss": 1.3121,
+      "step": 654
+    },
+    {
+      "epoch": 0.44425604069520985,
+      "grad_norm": 0.7418747882607136,
+      "learning_rate": 2.455445961775269e-05,
+      "loss": 1.2531,
+      "step": 655
+    },
+    {
+      "epoch": 0.44493429419245445,
+      "grad_norm": 0.8239106028350953,
+      "learning_rate": 2.451163478559444e-05,
+      "loss": 1.3214,
+      "step": 656
+    },
+    {
+      "epoch": 0.44561254768969905,
+      "grad_norm": 0.7157980505078755,
+      "learning_rate": 2.4468788147785574e-05,
+      "loss": 1.3993,
+      "step": 657
+    },
+    {
+      "epoch": 0.44629080118694364,
+      "grad_norm": 0.8222169942619503,
+      "learning_rate": 2.442591991141262e-05,
+      "loss": 1.2691,
+      "step": 658
+    },
+    {
+      "epoch": 0.44696905468418824,
+      "grad_norm": 0.7867959562129859,
+      "learning_rate": 2.4383030283666505e-05,
+      "loss": 1.2866,
+      "step": 659
+    },
+    {
+      "epoch": 0.44764730818143283,
+      "grad_norm": 0.7342001393007882,
+      "learning_rate": 2.434011947184154e-05,
+      "loss": 1.294,
+      "step": 660
+    },
+    {
+      "epoch": 0.44832556167867743,
+      "grad_norm": 0.5006326334519876,
+      "learning_rate": 2.429718768333443e-05,
+      "loss": 1.4205,
+      "step": 661
+    },
+    {
+      "epoch": 0.449003815175922,
+      "grad_norm": 0.8123981244739599,
+      "learning_rate": 2.4254235125643258e-05,
+      "loss": 1.2908,
+      "step": 662
+    },
+    {
+      "epoch": 0.44968206867316657,
+      "grad_norm": 0.8233710990841918,
+      "learning_rate": 2.421126200636649e-05,
+      "loss": 1.3067,
+      "step": 663
+    },
+    {
+      "epoch": 0.45036032217041116,
+      "grad_norm": 0.7474277882376499,
+      "learning_rate": 2.4168268533201978e-05,
+      "loss": 1.2803,
+      "step": 664
+    },
+    {
+      "epoch": 0.45103857566765576,
+      "grad_norm": 0.8240228952238716,
+      "learning_rate": 2.4125254913945933e-05,
+      "loss": 1.3009,
+      "step": 665
+    },
+    {
+      "epoch": 0.45171682916490036,
+      "grad_norm": 0.7825754554639681,
+      "learning_rate": 2.408222135649195e-05,
+      "loss": 1.2953,
+      "step": 666
+    },
+    {
+      "epoch": 0.45239508266214495,
+      "grad_norm": 0.7335894316828039,
+      "learning_rate": 2.4039168068829986e-05,
+      "loss": 1.2611,
+      "step": 667
+    },
+    {
+      "epoch": 0.45307333615938955,
+      "grad_norm": 0.7626283165326443,
+      "learning_rate": 2.399609525904536e-05,
+      "loss": 1.2842,
+      "step": 668
+    },
+    {
+      "epoch": 0.45375158965663415,
+      "grad_norm": 0.7958975092419973,
+      "learning_rate": 2.395300313531773e-05,
+      "loss": 1.303,
+      "step": 669
+    },
+    {
+      "epoch": 0.45442984315387874,
+      "grad_norm": 0.8028562108690003,
+      "learning_rate": 2.390989190592012e-05,
+      "loss": 1.2993,
+      "step": 670
+    },
+    {
+      "epoch": 0.45510809665112334,
+      "grad_norm": 0.7705990126826399,
+      "learning_rate": 2.3866761779217894e-05,
+      "loss": 1.2882,
+      "step": 671
+    },
+    {
+      "epoch": 0.45578635014836794,
+      "grad_norm": 0.7406487372590941,
+      "learning_rate": 2.3823612963667754e-05,
+      "loss": 1.2679,
+      "step": 672
+    },
+    {
+      "epoch": 0.45646460364561253,
+      "grad_norm": 0.8387719012384222,
+      "learning_rate": 2.37804456678167e-05,
+      "loss": 1.2859,
+      "step": 673
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.7747517984064684,
+      "learning_rate": 2.373726010030109e-05,
+      "loss": 1.2804,
+      "step": 674
+    },
+    {
+      "epoch": 0.4578211106401017,
+      "grad_norm": 0.7304383087457895,
+      "learning_rate": 2.3694056469845564e-05,
+      "loss": 1.2883,
+      "step": 675
+    },
+    {
+      "epoch": 0.4584993641373463,
+      "grad_norm": 0.8324980955959124,
+      "learning_rate": 2.3650834985262087e-05,
+      "loss": 1.2525,
+      "step": 676
+    },
+    {
+      "epoch": 0.4591776176345909,
+      "grad_norm": 0.7458679753068804,
+      "learning_rate": 2.3607595855448894e-05,
+      "loss": 1.2732,
+      "step": 677
+    },
+    {
+      "epoch": 0.4598558711318355,
+      "grad_norm": 0.734114365691043,
+      "learning_rate": 2.356433928938952e-05,
+      "loss": 1.277,
+      "step": 678
+    },
+    {
+      "epoch": 0.4605341246290801,
+      "grad_norm": 0.805471283388345,
+      "learning_rate": 2.3521065496151766e-05,
+      "loss": 1.3127,
+      "step": 679
+    },
+    {
+      "epoch": 0.4612123781263247,
+      "grad_norm": 0.7409705986781921,
+      "learning_rate": 2.347777468488669e-05,
+      "loss": 1.3094,
+      "step": 680
+    },
+    {
+      "epoch": 0.4618906316235693,
+      "grad_norm": 0.7838969300568545,
+      "learning_rate": 2.343446706482762e-05,
+      "loss": 1.2851,
+      "step": 681
+    },
+    {
+      "epoch": 0.4625688851208139,
+      "grad_norm": 0.7433118851032263,
+      "learning_rate": 2.3391142845289098e-05,
+      "loss": 1.2757,
+      "step": 682
+    },
+    {
+      "epoch": 0.4632471386180585,
+      "grad_norm": 0.7725357938863449,
+      "learning_rate": 2.3347802235665924e-05,
+      "loss": 1.3011,
+      "step": 683
+    },
+    {
+      "epoch": 0.4639253921153031,
+      "grad_norm": 0.7630007346948985,
+      "learning_rate": 2.330444544543208e-05,
+      "loss": 1.2868,
+      "step": 684
+    },
+    {
+      "epoch": 0.4646036456125477,
+      "grad_norm": 0.7714636610331252,
+      "learning_rate": 2.3261072684139787e-05,
+      "loss": 1.2682,
+      "step": 685
+    },
+    {
+      "epoch": 0.4652818991097923,
+      "grad_norm": 0.8215968802215311,
+      "learning_rate": 2.3217684161418438e-05,
+      "loss": 1.2654,
+      "step": 686
+    },
+    {
+      "epoch": 0.4659601526070369,
+      "grad_norm": 0.7314483643688189,
+      "learning_rate": 2.317428008697361e-05,
+      "loss": 1.242,
+      "step": 687
+    },
+    {
+      "epoch": 0.4666384061042815,
+      "grad_norm": 0.7817360509036845,
+      "learning_rate": 2.3130860670586036e-05,
+      "loss": 1.2564,
+      "step": 688
+    },
+    {
+      "epoch": 0.4673166596015261,
+      "grad_norm": 0.7364807816212218,
+      "learning_rate": 2.308742612211061e-05,
+      "loss": 1.2618,
+      "step": 689
+    },
+    {
+      "epoch": 0.46799491309877067,
+      "grad_norm": 0.781087543645866,
+      "learning_rate": 2.304397665147537e-05,
+      "loss": 1.2725,
+      "step": 690
+    },
+    {
+      "epoch": 0.46867316659601527,
+      "grad_norm": 0.7258903133686357,
+      "learning_rate": 2.3000512468680444e-05,
+      "loss": 1.2811,
+      "step": 691
+    },
+    {
+      "epoch": 0.46935142009325986,
+      "grad_norm": 0.7357562054565555,
+      "learning_rate": 2.29570337837971e-05,
+      "loss": 1.2519,
+      "step": 692
+    },
+    {
+      "epoch": 0.47002967359050446,
+      "grad_norm": 0.7073247488405348,
+      "learning_rate": 2.291354080696668e-05,
+      "loss": 1.2674,
+      "step": 693
+    },
+    {
+      "epoch": 0.47070792708774906,
+      "grad_norm": 0.7416390462669545,
+      "learning_rate": 2.2870033748399613e-05,
+      "loss": 1.2871,
+      "step": 694
+    },
+    {
+      "epoch": 0.47138618058499365,
+      "grad_norm": 0.7412100881028095,
+      "learning_rate": 2.2826512818374386e-05,
+      "loss": 1.2598,
+      "step": 695
+    },
+    {
+      "epoch": 0.47206443408223825,
+      "grad_norm": 0.7263342004136881,
+      "learning_rate": 2.2782978227236515e-05,
+      "loss": 1.2836,
+      "step": 696
+    },
+    {
+      "epoch": 0.47274268757948285,
+      "grad_norm": 0.7377485093769998,
+      "learning_rate": 2.273943018539755e-05,
+      "loss": 1.2705,
+      "step": 697
+    },
+    {
+      "epoch": 0.47342094107672744,
+      "grad_norm": 0.7505195569359884,
+      "learning_rate": 2.2695868903334075e-05,
+      "loss": 1.3118,
+      "step": 698
+    },
+    {
+      "epoch": 0.47409919457397204,
+      "grad_norm": 0.7919153501492263,
+      "learning_rate": 2.2652294591586625e-05,
+      "loss": 1.2968,
+      "step": 699
+    },
+    {
+      "epoch": 0.47477744807121663,
+      "grad_norm": 0.765268302876366,
+      "learning_rate": 2.2608707460758742e-05,
+      "loss": 1.2666,
+      "step": 700
+    },
+    {
+      "epoch": 0.47545570156846123,
+      "grad_norm": 0.747503993072507,
+      "learning_rate": 2.2565107721515913e-05,
+      "loss": 1.2705,
+      "step": 701
+    },
+    {
+      "epoch": 0.4761339550657058,
+      "grad_norm": 0.7303954806914537,
+      "learning_rate": 2.2521495584584567e-05,
+      "loss": 1.415,
+      "step": 702
+    },
+    {
+      "epoch": 0.4768122085629504,
+      "grad_norm": 0.7966191203716216,
+      "learning_rate": 2.247787126075105e-05,
+      "loss": 1.2616,
+      "step": 703
+    },
+    {
+      "epoch": 0.477490462060195,
+      "grad_norm": 0.5339493405526096,
+      "learning_rate": 2.243423496086061e-05,
+      "loss": 1.3757,
+      "step": 704
+    },
+    {
+      "epoch": 0.4781687155574396,
+      "grad_norm": 0.7954246313983925,
+      "learning_rate": 2.239058689581638e-05,
+      "loss": 1.2622,
+      "step": 705
+    },
+    {
+      "epoch": 0.4788469690546842,
+      "grad_norm": 0.76273918772253,
+      "learning_rate": 2.2346927276578363e-05,
+      "loss": 1.2958,
+      "step": 706
+    },
+    {
+      "epoch": 0.4795252225519288,
+      "grad_norm": 0.741611495665349,
+      "learning_rate": 2.2303256314162392e-05,
+      "loss": 1.2703,
+      "step": 707
+    },
+    {
+      "epoch": 0.4802034760491734,
+      "grad_norm": 0.7802891637540444,
+      "learning_rate": 2.2259574219639128e-05,
+      "loss": 1.267,
+      "step": 708
+    },
+    {
+      "epoch": 0.48088172954641795,
+      "grad_norm": 0.7359657086413339,
+      "learning_rate": 2.221588120413305e-05,
+      "loss": 1.2536,
+      "step": 709
+    },
+    {
+      "epoch": 0.48155998304366254,
+      "grad_norm": 0.7282463841135269,
+      "learning_rate": 2.2172177478821397e-05,
+      "loss": 1.2905,
+      "step": 710
+    },
+    {
+      "epoch": 0.48223823654090714,
+      "grad_norm": 0.7559767659269689,
+      "learning_rate": 2.212846325493319e-05,
+      "loss": 1.295,
+      "step": 711
+    },
+    {
+      "epoch": 0.48291649003815174,
+      "grad_norm": 0.7345869412181345,
+      "learning_rate": 2.208473874374818e-05,
+      "loss": 1.28,
+      "step": 712
+    },
+    {
+      "epoch": 0.48359474353539633,
+      "grad_norm": 0.8395369552411838,
+      "learning_rate": 2.2041004156595845e-05,
+      "loss": 1.2943,
+      "step": 713
+    },
+    {
+      "epoch": 0.48427299703264093,
+      "grad_norm": 0.7656500828016878,
+      "learning_rate": 2.1997259704854362e-05,
+      "loss": 1.3387,
+      "step": 714
+    },
+    {
+      "epoch": 0.4849512505298855,
+      "grad_norm": 0.7708038244468336,
+      "learning_rate": 2.1953505599949577e-05,
+      "loss": 1.3023,
+      "step": 715
+    },
+    {
+      "epoch": 0.4856295040271301,
+      "grad_norm": 0.7218970860061366,
+      "learning_rate": 2.1909742053354005e-05,
+      "loss": 1.266,
+      "step": 716
+    },
+    {
+      "epoch": 0.4863077575243747,
+      "grad_norm": 0.7682310473494609,
+      "learning_rate": 2.1865969276585787e-05,
+      "loss": 1.2693,
+      "step": 717
+    },
+    {
+      "epoch": 0.4869860110216193,
+      "grad_norm": 0.7471960328816069,
+      "learning_rate": 2.1822187481207675e-05,
+      "loss": 1.2534,
+      "step": 718
+    },
+    {
+      "epoch": 0.4876642645188639,
+      "grad_norm": 0.7727623743996936,
+      "learning_rate": 2.1778396878826008e-05,
+      "loss": 1.2781,
+      "step": 719
+    },
+    {
+      "epoch": 0.4883425180161085,
+      "grad_norm": 0.7065341701899533,
+      "learning_rate": 2.17345976810897e-05,
+      "loss": 1.2897,
+      "step": 720
+    },
+    {
+      "epoch": 0.4890207715133531,
+      "grad_norm": 0.8056543929766073,
+      "learning_rate": 2.1690790099689197e-05,
+      "loss": 1.2828,
+      "step": 721
+    },
+    {
+      "epoch": 0.4896990250105977,
+      "grad_norm": 0.7330818708592541,
+      "learning_rate": 2.1646974346355472e-05,
+      "loss": 1.271,
+      "step": 722
+    },
+    {
+      "epoch": 0.4903772785078423,
+      "grad_norm": 0.7846831270623186,
+      "learning_rate": 2.1603150632858986e-05,
+      "loss": 1.2793,
+      "step": 723
+    },
+    {
+      "epoch": 0.4910555320050869,
+      "grad_norm": 0.7300693718708892,
+      "learning_rate": 2.15593191710087e-05,
+      "loss": 1.293,
+      "step": 724
+    },
+    {
+      "epoch": 0.4917337855023315,
+      "grad_norm": 0.7133388247822674,
+      "learning_rate": 2.1515480172650983e-05,
+      "loss": 1.2889,
+      "step": 725
+    },
+    {
+      "epoch": 0.4924120389995761,
+      "grad_norm": 0.733479372471087,
+      "learning_rate": 2.1471633849668666e-05,
+      "loss": 1.2909,
+      "step": 726
+    },
+    {
+      "epoch": 0.4930902924968207,
+      "grad_norm": 0.7711594632629859,
+      "learning_rate": 2.1427780413979954e-05,
+      "loss": 1.2554,
+      "step": 727
+    },
+    {
+      "epoch": 0.4937685459940653,
+      "grad_norm": 0.7206642982288171,
+      "learning_rate": 2.1383920077537445e-05,
+      "loss": 1.2825,
+      "step": 728
+    },
+    {
+      "epoch": 0.4944467994913099,
+      "grad_norm": 0.7256230446585492,
+      "learning_rate": 2.1340053052327087e-05,
+      "loss": 1.3023,
+      "step": 729
+    },
+    {
+      "epoch": 0.49512505298855447,
+      "grad_norm": 0.7698155651507559,
+      "learning_rate": 2.1296179550367152e-05,
+      "loss": 1.2886,
+      "step": 730
+    },
+    {
+      "epoch": 0.49580330648579907,
+      "grad_norm": 0.7605859306746997,
+      "learning_rate": 2.1252299783707233e-05,
+      "loss": 1.2605,
+      "step": 731
+    },
+    {
+      "epoch": 0.49648155998304366,
+      "grad_norm": 0.698780430218635,
+      "learning_rate": 2.120841396442717e-05,
+      "loss": 1.2415,
+      "step": 732
+    },
+    {
+      "epoch": 0.49715981348028826,
+      "grad_norm": 0.7881374449205716,
+      "learning_rate": 2.116452230463608e-05,
+      "loss": 1.2763,
+      "step": 733
+    },
+    {
+      "epoch": 0.49783806697753286,
+      "grad_norm": 0.7347578225034074,
+      "learning_rate": 2.1120625016471302e-05,
+      "loss": 1.2901,
+      "step": 734
+    },
+    {
+      "epoch": 0.49851632047477745,
+      "grad_norm": 0.7267350175824305,
+      "learning_rate": 2.107672231209738e-05,
+      "loss": 1.254,
+      "step": 735
+    },
+    {
+      "epoch": 0.49919457397202205,
+      "grad_norm": 0.9191272469301005,
+      "learning_rate": 2.1032814403705028e-05,
+      "loss": 1.4062,
+      "step": 736
+    },
+    {
+      "epoch": 0.49987282746926665,
+      "grad_norm": 0.8087334133463818,
+      "learning_rate": 2.098890150351013e-05,
+      "loss": 1.2988,
+      "step": 737
+    },
+    {
+      "epoch": 0.5005510809665112,
+      "grad_norm": 0.7873244004928791,
+      "learning_rate": 2.0944983823752664e-05,
+      "loss": 1.2786,
+      "step": 738
+    },
+    {
+      "epoch": 0.5012293344637558,
+      "grad_norm": 0.7551947535757042,
+      "learning_rate": 2.0901061576695754e-05,
+      "loss": 1.2648,
+      "step": 739
+    },
+    {
+      "epoch": 0.5019075879610004,
+      "grad_norm": 0.7238097481098938,
+      "learning_rate": 2.085713497462456e-05,
+      "loss": 1.2745,
+      "step": 740
+    },
+    {
+      "epoch": 0.502585841458245,
+      "grad_norm": 0.7863072916144123,
+      "learning_rate": 2.08132042298453e-05,
+      "loss": 1.2938,
+      "step": 741
+    },
+    {
+      "epoch": 0.5032640949554896,
+      "grad_norm": 0.7321035315725165,
+      "learning_rate": 2.076926955468423e-05,
+      "loss": 1.2626,
+      "step": 742
+    },
+    {
+      "epoch": 0.5039423484527342,
+      "grad_norm": 0.7758679023626488,
+      "learning_rate": 2.072533116148658e-05,
+      "loss": 1.2786,
+      "step": 743
+    },
+    {
+      "epoch": 0.5046206019499788,
+      "grad_norm": 0.7620697769692747,
+      "learning_rate": 2.0681389262615574e-05,
+      "loss": 1.2546,
+      "step": 744
+    },
+    {
+      "epoch": 0.5052988554472234,
+      "grad_norm": 0.6988722404611495,
+      "learning_rate": 2.0637444070451346e-05,
+      "loss": 1.2647,
+      "step": 745
+    },
+    {
+      "epoch": 0.505977108944468,
+      "grad_norm": 0.736111947921635,
+      "learning_rate": 2.0593495797389994e-05,
+      "loss": 1.3025,
+      "step": 746
+    },
+    {
+      "epoch": 0.5066553624417126,
+      "grad_norm": 0.7598553414031834,
+      "learning_rate": 2.054954465584246e-05,
+      "loss": 1.2898,
+      "step": 747
+    },
+    {
+      "epoch": 0.5073336159389572,
+      "grad_norm": 0.77145507844893,
+      "learning_rate": 2.0505590858233587e-05,
+      "loss": 1.2901,
+      "step": 748
+    },
+    {
+      "epoch": 0.5080118694362018,
+      "grad_norm": 0.7450436756941183,
+      "learning_rate": 2.0461634617001022e-05,
+      "loss": 1.2741,
+      "step": 749
+    },
+    {
+      "epoch": 0.5086901229334464,
+      "grad_norm": 0.7506494842537246,
+      "learning_rate": 2.0417676144594257e-05,
+      "loss": 1.2665,
+      "step": 750
+    },
+    {
+      "epoch": 0.509368376430691,
+      "grad_norm": 0.8068507358614899,
+      "learning_rate": 2.0373715653473534e-05,
+      "loss": 1.2781,
+      "step": 751
+    },
+    {
+      "epoch": 0.5100466299279356,
+      "grad_norm": 0.7559844586490145,
+      "learning_rate": 2.032975335610888e-05,
+      "loss": 1.2462,
+      "step": 752
+    },
+    {
+      "epoch": 0.5107248834251802,
+      "grad_norm": 0.7382892331154312,
+      "learning_rate": 2.0285789464979033e-05,
+      "loss": 1.2425,
+      "step": 753
+    },
+    {
+      "epoch": 0.5114031369224248,
+      "grad_norm": 0.7536737034033136,
+      "learning_rate": 2.0241824192570446e-05,
+      "loss": 1.2958,
+      "step": 754
+    },
+    {
+      "epoch": 0.5120813904196694,
+      "grad_norm": 0.7821684212875851,
+      "learning_rate": 2.0197857751376237e-05,
+      "loss": 1.2565,
+      "step": 755
+    },
+    {
+      "epoch": 0.512759643916914,
+      "grad_norm": 0.8176897966736769,
+      "learning_rate": 2.0153890353895186e-05,
+      "loss": 1.2899,
+      "step": 756
+    },
+    {
+      "epoch": 0.5134378974141586,
+      "grad_norm": 0.7559675257892411,
+      "learning_rate": 2.010992221263068e-05,
+      "loss": 1.2981,
+      "step": 757
+    },
+    {
+      "epoch": 0.5141161509114032,
+      "grad_norm": 0.7468879468648103,
+      "learning_rate": 2.006595354008971e-05,
+      "loss": 1.2732,
+      "step": 758
+    },
+    {
+      "epoch": 0.5147944044086478,
+      "grad_norm": 0.7181990484784445,
+      "learning_rate": 2.0021984548781844e-05,
+      "loss": 1.2893,
+      "step": 759
+    },
+    {
+      "epoch": 0.5154726579058924,
+      "grad_norm": 0.7481524052553766,
+      "learning_rate": 1.9978015451218166e-05,
+      "loss": 1.2711,
+      "step": 760
+    },
+    {
+      "epoch": 0.516150911403137,
+      "grad_norm": 0.7676149070117987,
+      "learning_rate": 1.9934046459910293e-05,
+      "loss": 1.2782,
+      "step": 761
+    },
+    {
+      "epoch": 0.5168291649003816,
+      "grad_norm": 0.7467675834428721,
+      "learning_rate": 1.989007778736933e-05,
+      "loss": 1.2741,
+      "step": 762
+    },
+    {
+      "epoch": 0.5175074183976262,
+      "grad_norm": 0.7079588693786695,
+      "learning_rate": 1.9846109646104824e-05,
+      "loss": 1.2785,
+      "step": 763
+    },
+    {
+      "epoch": 0.5181856718948707,
+      "grad_norm": 0.7567069792021545,
+      "learning_rate": 1.9802142248623767e-05,
+      "loss": 1.2952,
+      "step": 764
+    },
+    {
+      "epoch": 0.5188639253921153,
+      "grad_norm": 0.7160899538626819,
+      "learning_rate": 1.9758175807429564e-05,
+      "loss": 1.2492,
+      "step": 765
+    },
+    {
+      "epoch": 0.5195421788893599,
+      "grad_norm": 0.7375646034553129,
+      "learning_rate": 1.971421053502097e-05,
+      "loss": 1.2725,
+      "step": 766
+    },
+    {
+      "epoch": 0.5202204323866045,
+      "grad_norm": 0.7459565281487732,
+      "learning_rate": 1.9670246643891125e-05,
+      "loss": 1.2887,
+      "step": 767
+    },
+    {
+      "epoch": 0.5208986858838491,
+      "grad_norm": 0.7241813728113365,
+      "learning_rate": 1.962628434652647e-05,
+      "loss": 1.2787,
+      "step": 768
+    },
+    {
+      "epoch": 0.5215769393810937,
+      "grad_norm": 0.7377185301313828,
+      "learning_rate": 1.9582323855405753e-05,
+      "loss": 1.2604,
+      "step": 769
+    },
+    {
+      "epoch": 0.5222551928783383,
+      "grad_norm": 0.7619328845955535,
+      "learning_rate": 1.953836538299898e-05,
+      "loss": 1.2484,
+      "step": 770
+    },
+    {
+      "epoch": 0.5229334463755829,
+      "grad_norm": 0.771598320473821,
+      "learning_rate": 1.9494409141766416e-05,
+      "loss": 1.3023,
+      "step": 771
+    },
+    {
+      "epoch": 0.5236116998728275,
+      "grad_norm": 0.7575952299434909,
+      "learning_rate": 1.9450455344157547e-05,
+      "loss": 1.2932,
+      "step": 772
+    },
+    {
+      "epoch": 0.5242899533700721,
+      "grad_norm": 0.7497620136321677,
+      "learning_rate": 1.940650420261001e-05,
+      "loss": 1.2816,
+      "step": 773
+    },
+    {
+      "epoch": 0.5249682068673167,
+      "grad_norm": 0.8152292701670396,
+      "learning_rate": 1.936255592954866e-05,
+      "loss": 1.271,
+      "step": 774
+    },
+    {
+      "epoch": 0.5256464603645613,
+      "grad_norm": 0.731035301388366,
+      "learning_rate": 1.9318610737384436e-05,
+      "loss": 1.2606,
+      "step": 775
+    },
+    {
+      "epoch": 0.5263247138618059,
+      "grad_norm": 0.7408084710649866,
+      "learning_rate": 1.927466883851343e-05,
+      "loss": 1.2823,
+      "step": 776
+    },
+    {
+      "epoch": 0.5270029673590505,
+      "grad_norm": 0.769360474224272,
+      "learning_rate": 1.923073044531578e-05,
+      "loss": 1.2809,
+      "step": 777
+    },
+    {
+      "epoch": 0.527681220856295,
+      "grad_norm": 0.790407790205139,
+      "learning_rate": 1.918679577015471e-05,
+      "loss": 1.3015,
+      "step": 778
+    },
+    {
+      "epoch": 0.5283594743535396,
+      "grad_norm": 0.7512918107436672,
+      "learning_rate": 1.9142865025375447e-05,
+      "loss": 1.2853,
+      "step": 779
+    },
+    {
+      "epoch": 0.5290377278507842,
+      "grad_norm": 0.7367728132150647,
+      "learning_rate": 1.9098938423304246e-05,
+      "loss": 1.2742,
+      "step": 780
+    },
+    {
+      "epoch": 0.5297159813480288,
+      "grad_norm": 0.7137355054556672,
+      "learning_rate": 1.905501617624734e-05,
+      "loss": 1.2466,
+      "step": 781
+    },
+    {
+      "epoch": 0.5303942348452734,
+      "grad_norm": 0.749371561195395,
+      "learning_rate": 1.9011098496489876e-05,
+      "loss": 1.2653,
+      "step": 782
+    },
+    {
+      "epoch": 0.531072488342518,
+      "grad_norm": 0.7495797022110012,
+      "learning_rate": 1.896718559629498e-05,
+      "loss": 1.2784,
+      "step": 783
+    },
+    {
+      "epoch": 0.5317507418397626,
+      "grad_norm": 0.7192581639726054,
+      "learning_rate": 1.8923277687902625e-05,
+      "loss": 1.2711,
+      "step": 784
+    },
+    {
+      "epoch": 0.5324289953370072,
+      "grad_norm": 0.7261147473593981,
+      "learning_rate": 1.8879374983528708e-05,
+      "loss": 1.2525,
+      "step": 785
+    },
+    {
+      "epoch": 0.5331072488342518,
+      "grad_norm": 0.6962500047334699,
+      "learning_rate": 1.8835477695363926e-05,
+      "loss": 1.2657,
+      "step": 786
+    },
+    {
+      "epoch": 0.5337855023314964,
+      "grad_norm": 0.7553121004344828,
+      "learning_rate": 1.8791586035572832e-05,
+      "loss": 1.2738,
+      "step": 787
+    },
+    {
+      "epoch": 0.534463755828741,
+      "grad_norm": 0.7746187801268444,
+      "learning_rate": 1.8747700216292774e-05,
+      "loss": 1.2579,
+      "step": 788
+    },
+    {
+      "epoch": 0.5351420093259855,
+      "grad_norm": 0.7464557600044285,
+      "learning_rate": 1.8703820449632844e-05,
+      "loss": 1.2936,
+      "step": 789
+    },
+    {
+      "epoch": 0.5358202628232301,
+      "grad_norm": 0.7499003825986563,
+      "learning_rate": 1.865994694767292e-05,
+      "loss": 1.28,
+      "step": 790
+    },
+    {
+      "epoch": 0.5364985163204747,
+      "grad_norm": 0.7312593356679844,
+      "learning_rate": 1.861607992246256e-05,
+      "loss": 1.2885,
+      "step": 791
+    },
+    {
+      "epoch": 0.5371767698177193,
+      "grad_norm": 0.7614005096187081,
+      "learning_rate": 1.8572219586020056e-05,
+      "loss": 1.2682,
+      "step": 792
+    },
+    {
+      "epoch": 0.5378550233149639,
+      "grad_norm": 0.7017023223046736,
+      "learning_rate": 1.8528366150331344e-05,
+      "loss": 1.254,
+      "step": 793
+    },
+    {
+      "epoch": 0.5385332768122085,
+      "grad_norm": 0.7466796945484455,
+      "learning_rate": 1.848451982734902e-05,
+      "loss": 1.2767,
+      "step": 794
+    },
+    {
+      "epoch": 0.5392115303094531,
+      "grad_norm": 0.730197813705563,
+      "learning_rate": 1.8440680828991308e-05,
+      "loss": 1.2655,
+      "step": 795
+    },
+    {
+      "epoch": 0.5398897838066977,
+      "grad_norm": 0.7661939092830227,
+      "learning_rate": 1.839684936714101e-05,
+      "loss": 1.2594,
+      "step": 796
+    },
+    {
+      "epoch": 0.5405680373039423,
+      "grad_norm": 0.6534248100019572,
+      "learning_rate": 1.8353025653644535e-05,
+      "loss": 1.4072,
+      "step": 797
+    },
+    {
+      "epoch": 0.5412462908011869,
+      "grad_norm": 0.7630832597467281,
+      "learning_rate": 1.830920990031081e-05,
+      "loss": 1.2739,
+      "step": 798
+    },
+    {
+      "epoch": 0.5419245442984315,
+      "grad_norm": 0.7981120415918996,
+      "learning_rate": 1.826540231891031e-05,
+      "loss": 1.2751,
+      "step": 799
+    },
+    {
+      "epoch": 0.5426027977956761,
+      "grad_norm": 0.7436141589265621,
+      "learning_rate": 1.8221603121174e-05,
+      "loss": 1.28,
+      "step": 800
+    },
+    {
+      "epoch": 0.5432810512929207,
+      "grad_norm": 0.7369961328790705,
+      "learning_rate": 1.8177812518792332e-05,
+      "loss": 1.2706,
+      "step": 801
+    },
+    {
+      "epoch": 0.5439593047901653,
+      "grad_norm": 0.738256573102201,
+      "learning_rate": 1.813403072341422e-05,
+      "loss": 1.2885,
+      "step": 802
+    },
+    {
+      "epoch": 0.5446375582874099,
+      "grad_norm": 0.7472209862893595,
+      "learning_rate": 1.8090257946645998e-05,
+      "loss": 1.257,
+      "step": 803
+    },
+    {
+      "epoch": 0.5453158117846545,
+      "grad_norm": 0.710354620157084,
+      "learning_rate": 1.804649440005043e-05,
+      "loss": 1.2455,
+      "step": 804
+    },
+    {
+      "epoch": 0.5459940652818991,
+      "grad_norm": 0.7340068410967466,
+      "learning_rate": 1.8002740295145645e-05,
+      "loss": 1.3025,
+      "step": 805
+    },
+    {
+      "epoch": 0.5466723187791437,
+      "grad_norm": 0.48470218124317466,
+      "learning_rate": 1.7958995843404165e-05,
+      "loss": 1.4318,
+      "step": 806
+    },
+    {
+      "epoch": 0.5473505722763883,
+      "grad_norm": 0.7600875770027712,
+      "learning_rate": 1.7915261256251825e-05,
+      "loss": 1.2491,
+      "step": 807
+    },
+    {
+      "epoch": 0.5480288257736329,
+      "grad_norm": 0.8090050109964079,
+      "learning_rate": 1.787153674506682e-05,
+      "loss": 1.2578,
+      "step": 808
+    },
+    {
+      "epoch": 0.5487070792708775,
+      "grad_norm": 0.7219551531427646,
+      "learning_rate": 1.782782252117861e-05,
+      "loss": 1.2651,
+      "step": 809
+    },
+    {
+      "epoch": 0.5493853327681221,
+      "grad_norm": 0.7882648722353363,
+      "learning_rate": 1.7784118795866954e-05,
+      "loss": 1.3079,
+      "step": 810
+    },
+    {
+      "epoch": 0.5500635862653667,
+      "grad_norm": 0.7362810674678596,
+      "learning_rate": 1.774042578036088e-05,
+      "loss": 1.2399,
+      "step": 811
+    },
+    {
+      "epoch": 0.5507418397626113,
+      "grad_norm": 0.7673342227031724,
+      "learning_rate": 1.769674368583761e-05,
+      "loss": 1.2736,
+      "step": 812
+    },
+    {
+      "epoch": 0.5514200932598559,
+      "grad_norm": 0.7485637816315667,
+      "learning_rate": 1.765307272342165e-05,
+      "loss": 1.2742,
+      "step": 813
+    },
+    {
+      "epoch": 0.5520983467571005,
+      "grad_norm": 0.7991809227370478,
+      "learning_rate": 1.7609413104183623e-05,
+      "loss": 1.3009,
+      "step": 814
+    },
+    {
+      "epoch": 0.552776600254345,
+      "grad_norm": 0.72841400665626,
+      "learning_rate": 1.75657650391394e-05,
+      "loss": 1.2415,
+      "step": 815
+    },
+    {
+      "epoch": 0.5534548537515896,
+      "grad_norm": 0.7328231976918355,
+      "learning_rate": 1.7522128739248956e-05,
+      "loss": 1.2642,
+      "step": 816
+    },
+    {
+      "epoch": 0.5541331072488342,
+      "grad_norm": 0.7648044415971349,
+      "learning_rate": 1.7478504415415437e-05,
+      "loss": 1.2465,
+      "step": 817
+    },
+    {
+      "epoch": 0.5548113607460788,
+      "grad_norm": 0.7683617384717926,
+      "learning_rate": 1.743489227848409e-05,
+      "loss": 1.2578,
+      "step": 818
+    },
+    {
+      "epoch": 0.5554896142433234,
+      "grad_norm": 0.7388908337530937,
+      "learning_rate": 1.739129253924126e-05,
+      "loss": 1.2859,
+      "step": 819
+    },
+    {
+      "epoch": 0.556167867740568,
+      "grad_norm": 0.8056916671907673,
+      "learning_rate": 1.734770540841338e-05,
+      "loss": 1.2399,
+      "step": 820
+    },
+    {
+      "epoch": 0.5568461212378126,
+      "grad_norm": 0.7547650035174724,
+      "learning_rate": 1.730413109666593e-05,
+      "loss": 1.2685,
+      "step": 821
+    },
+    {
+      "epoch": 0.5575243747350572,
+      "grad_norm": 0.7299262736595057,
+      "learning_rate": 1.7260569814602452e-05,
+      "loss": 1.2188,
+      "step": 822
+    },
+    {
+      "epoch": 0.5582026282323018,
+      "grad_norm": 0.7525499939514374,
+      "learning_rate": 1.7217021772763495e-05,
+      "loss": 1.2618,
+      "step": 823
+    },
+    {
+      "epoch": 0.5588808817295464,
+      "grad_norm": 0.7098846627583694,
+      "learning_rate": 1.7173487181625618e-05,
+      "loss": 1.2711,
+      "step": 824
+    },
+    {
+      "epoch": 0.559559135226791,
+      "grad_norm": 0.7459859867180162,
+      "learning_rate": 1.712996625160039e-05,
+      "loss": 1.2846,
+      "step": 825
+    },
+    {
+      "epoch": 0.5602373887240356,
+      "grad_norm": 0.757384152710219,
+      "learning_rate": 1.708645919303332e-05,
+      "loss": 1.2668,
+      "step": 826
+    },
+    {
+      "epoch": 0.5609156422212802,
+      "grad_norm": 0.7823318074574718,
+      "learning_rate": 1.7042966216202906e-05,
+      "loss": 1.2735,
+      "step": 827
+    },
+    {
+      "epoch": 0.5615938957185248,
+      "grad_norm": 0.706030098280657,
+      "learning_rate": 1.6999487531319563e-05,
+      "loss": 1.2406,
+      "step": 828
+    },
+    {
+      "epoch": 0.5622721492157694,
+      "grad_norm": 0.7782884788037362,
+      "learning_rate": 1.6956023348524643e-05,
+      "loss": 1.2823,
+      "step": 829
+    },
+    {
+      "epoch": 0.562950402713014,
+      "grad_norm": 0.7701685920541365,
+      "learning_rate": 1.6912573877889393e-05,
+      "loss": 1.2658,
+      "step": 830
+    },
+    {
+      "epoch": 0.5636286562102586,
+      "grad_norm": 0.7828155055979443,
+      "learning_rate": 1.6869139329413967e-05,
+      "loss": 1.2987,
+      "step": 831
+    },
+    {
+      "epoch": 0.5643069097075032,
+      "grad_norm": 0.7585431050580741,
+      "learning_rate": 1.6825719913026398e-05,
+      "loss": 1.2746,
+      "step": 832
+    },
+    {
+      "epoch": 0.5649851632047478,
+      "grad_norm": 0.7975578679187447,
+      "learning_rate": 1.6782315838581566e-05,
+      "loss": 1.3008,
+      "step": 833
+    },
+    {
+      "epoch": 0.5656634167019924,
+      "grad_norm": 0.7407472825658853,
+      "learning_rate": 1.6738927315860216e-05,
+      "loss": 1.2584,
+      "step": 834
+    },
+    {
+      "epoch": 0.566341670199237,
+      "grad_norm": 0.7400394043223196,
+      "learning_rate": 1.6695554554567925e-05,
+      "loss": 1.2766,
+      "step": 835
+    },
+    {
+      "epoch": 0.5670199236964816,
+      "grad_norm": 0.7982859253619563,
+      "learning_rate": 1.665219776433409e-05,
+      "loss": 1.2735,
+      "step": 836
+    },
+    {
+      "epoch": 0.5676981771937262,
+      "grad_norm": 0.7558919073776923,
+      "learning_rate": 1.6608857154710905e-05,
+      "loss": 1.2239,
+      "step": 837
+    },
+    {
+      "epoch": 0.5683764306909708,
+      "grad_norm": 0.7633849662408336,
+      "learning_rate": 1.6565532935172387e-05,
+      "loss": 1.2812,
+      "step": 838
+    },
+    {
+      "epoch": 0.5690546841882154,
+      "grad_norm": 0.7259209453123879,
+      "learning_rate": 1.6522225315113313e-05,
+      "loss": 1.2505,
+      "step": 839
+    },
+    {
+      "epoch": 0.56973293768546,
+      "grad_norm": 0.7219112999111875,
+      "learning_rate": 1.6478934503848237e-05,
+      "loss": 1.2573,
+      "step": 840
+    },
+    {
+      "epoch": 0.5704111911827046,
+      "grad_norm": 0.761409047706854,
+      "learning_rate": 1.6435660710610485e-05,
+      "loss": 1.2477,
+      "step": 841
+    },
+    {
+      "epoch": 0.5710894446799492,
+      "grad_norm": 0.7445867519761595,
+      "learning_rate": 1.639240414455111e-05,
+      "loss": 1.2502,
+      "step": 842
+    },
+    {
+      "epoch": 0.5717676981771938,
+      "grad_norm": 0.7156396178869799,
+      "learning_rate": 1.6349165014737923e-05,
+      "loss": 1.2299,
+      "step": 843
+    },
+    {
+      "epoch": 0.5724459516744383,
+      "grad_norm": 0.7901111725085388,
+      "learning_rate": 1.630594353015444e-05,
+      "loss": 1.2958,
+      "step": 844
+    },
+    {
+      "epoch": 0.5731242051716829,
+      "grad_norm": 0.788495518441872,
+      "learning_rate": 1.626273989969892e-05,
+      "loss": 1.2464,
+      "step": 845
+    },
+    {
+      "epoch": 0.5738024586689275,
+      "grad_norm": 0.7125175624415945,
+      "learning_rate": 1.6219554332183307e-05,
+      "loss": 1.272,
+      "step": 846
+    },
+    {
+      "epoch": 0.5744807121661721,
+      "grad_norm": 0.5312981682163216,
+      "learning_rate": 1.617638703633225e-05,
+      "loss": 1.3833,
+      "step": 847
+    },
+    {
+      "epoch": 0.5751589656634167,
+      "grad_norm": 0.7154786352168268,
+      "learning_rate": 1.613323822078211e-05,
+      "loss": 1.2373,
+      "step": 848
+    },
+    {
+      "epoch": 0.5758372191606613,
+      "grad_norm": 0.7322890290043707,
+      "learning_rate": 1.609010809407988e-05,
+      "loss": 1.2353,
+      "step": 849
+    },
+    {
+      "epoch": 0.5765154726579059,
+      "grad_norm": 0.7402441049307272,
+      "learning_rate": 1.604699686468228e-05,
+      "loss": 1.2683,
+      "step": 850
+    },
+    {
+      "epoch": 0.5771937261551505,
+      "grad_norm": 0.6934244915813785,
+      "learning_rate": 1.600390474095465e-05,
+      "loss": 1.2503,
+      "step": 851
+    },
+    {
+      "epoch": 0.5778719796523951,
+      "grad_norm": 0.7306958031183086,
+      "learning_rate": 1.5960831931170024e-05,
+      "loss": 1.2754,
+      "step": 852
+    },
+    {
+      "epoch": 0.5785502331496397,
+      "grad_norm": 0.7414223143033765,
+      "learning_rate": 1.5917778643508052e-05,
+      "loss": 1.281,
+      "step": 853
+    },
+    {
+      "epoch": 0.5792284866468843,
+      "grad_norm": 0.7367648536751301,
+      "learning_rate": 1.587474508605407e-05,
+      "loss": 1.2624,
+      "step": 854
+    },
+    {
+      "epoch": 0.5799067401441289,
+      "grad_norm": 0.7930690411423662,
+      "learning_rate": 1.5831731466798032e-05,
+      "loss": 1.2484,
+      "step": 855
+    },
+    {
+      "epoch": 0.5805849936413735,
+      "grad_norm": 0.8057908289247886,
+      "learning_rate": 1.578873799363351e-05,
+      "loss": 1.276,
+      "step": 856
+    },
+    {
+      "epoch": 0.5812632471386181,
+      "grad_norm": 0.7374427881988856,
+      "learning_rate": 1.574576487435675e-05,
+      "loss": 1.2507,
+      "step": 857
+    },
+    {
+      "epoch": 0.5819415006358627,
+      "grad_norm": 0.7534989203331487,
+      "learning_rate": 1.5702812316665576e-05,
+      "loss": 1.2524,
+      "step": 858
+    },
+    {
+      "epoch": 0.5826197541331073,
+      "grad_norm": 0.7372448874629303,
+      "learning_rate": 1.5659880528158464e-05,
+      "loss": 1.2622,
+      "step": 859
+    },
+    {
+      "epoch": 0.5832980076303519,
+      "grad_norm": 0.7697321537663581,
+      "learning_rate": 1.56169697163335e-05,
+      "loss": 1.2572,
+      "step": 860
+    },
+    {
+      "epoch": 0.5839762611275965,
+      "grad_norm": 0.8052964355736111,
+      "learning_rate": 1.557408008858738e-05,
+      "loss": 1.2446,
+      "step": 861
+    },
+    {
+      "epoch": 0.5846545146248411,
+      "grad_norm": 0.7354976869190977,
+      "learning_rate": 1.553121185221443e-05,
+      "loss": 1.2636,
+      "step": 862
+    },
+    {
+      "epoch": 0.5853327681220857,
+      "grad_norm": 0.7144607440214421,
+      "learning_rate": 1.548836521440556e-05,
+      "loss": 1.2576,
+      "step": 863
+    },
+    {
+      "epoch": 0.5860110216193303,
+      "grad_norm": 0.7752669102183751,
+      "learning_rate": 1.5445540382247317e-05,
+      "loss": 1.2706,
+      "step": 864
+    },
+    {
+      "epoch": 0.5866892751165749,
+      "grad_norm": 0.7671360295238987,
+      "learning_rate": 1.5402737562720835e-05,
+      "loss": 1.2694,
+      "step": 865
+    },
+    {
+      "epoch": 0.5873675286138195,
+      "grad_norm": 0.7339904919637064,
+      "learning_rate": 1.5359956962700875e-05,
+      "loss": 1.2265,
+      "step": 866
+    },
+    {
+      "epoch": 0.5880457821110641,
+      "grad_norm": 0.7453945953489697,
+      "learning_rate": 1.531719878895479e-05,
+      "loss": 1.2794,
+      "step": 867
+    },
+    {
+      "epoch": 0.5887240356083087,
+      "grad_norm": 0.7049362289674199,
+      "learning_rate": 1.5274463248141554e-05,
+      "loss": 1.2199,
+      "step": 868
+    },
+    {
+      "epoch": 0.5894022891055531,
+      "grad_norm": 0.7199336413075009,
+      "learning_rate": 1.523175054681075e-05,
+      "loss": 1.2329,
+      "step": 869
+    },
+    {
+      "epoch": 0.5900805426027977,
+      "grad_norm": 0.7051800273850447,
+      "learning_rate": 1.5189060891401564e-05,
+      "loss": 1.2517,
+      "step": 870
+    },
+    {
+      "epoch": 0.5907587961000423,
+      "grad_norm": 0.7253522354142042,
+      "learning_rate": 1.5146394488241812e-05,
+      "loss": 1.2464,
+      "step": 871
+    },
+    {
+      "epoch": 0.5914370495972869,
+      "grad_norm": 0.7180955256697961,
+      "learning_rate": 1.5103751543546906e-05,
+      "loss": 1.2384,
+      "step": 872
+    },
+    {
+      "epoch": 0.5921153030945315,
+      "grad_norm": 0.7370359916598486,
+      "learning_rate": 1.5061132263418903e-05,
+      "loss": 1.2977,
+      "step": 873
+    },
+    {
+      "epoch": 0.5927935565917761,
+      "grad_norm": 0.7328367261254487,
+      "learning_rate": 1.5018536853845461e-05,
+      "loss": 1.3052,
+      "step": 874
+    },
+    {
+      "epoch": 0.5934718100890207,
+      "grad_norm": 0.7140757454830974,
+      "learning_rate": 1.4975965520698887e-05,
+      "loss": 1.2721,
+      "step": 875
+    },
+    {
+      "epoch": 0.5941500635862653,
+      "grad_norm": 0.75263882554648,
+      "learning_rate": 1.4933418469735103e-05,
+      "loss": 1.23,
+      "step": 876
+    },
+    {
+      "epoch": 0.5948283170835099,
+      "grad_norm": 0.706673856380447,
+      "learning_rate": 1.4890895906592682e-05,
+      "loss": 1.2545,
+      "step": 877
+    },
+    {
+      "epoch": 0.5955065705807545,
+      "grad_norm": 0.7394045603711208,
+      "learning_rate": 1.4848398036791856e-05,
+      "loss": 1.2518,
+      "step": 878
+    },
+    {
+      "epoch": 0.5961848240779991,
+      "grad_norm": 0.7455235410630398,
+      "learning_rate": 1.4805925065733468e-05,
+      "loss": 1.2528,
+      "step": 879
+    },
+    {
+      "epoch": 0.5968630775752437,
+      "grad_norm": 0.7224215309064259,
+      "learning_rate": 1.4763477198698062e-05,
+      "loss": 1.2824,
+      "step": 880
+    },
+    {
+      "epoch": 0.5975413310724883,
+      "grad_norm": 0.7093906902341754,
+      "learning_rate": 1.4721054640844826e-05,
+      "loss": 1.2311,
+      "step": 881
+    },
+    {
+      "epoch": 0.5982195845697329,
+      "grad_norm": 0.7113504579397316,
+      "learning_rate": 1.4678657597210633e-05,
+      "loss": 1.2458,
+      "step": 882
+    },
+    {
+      "epoch": 0.5988978380669775,
+      "grad_norm": 0.8035763524682514,
+      "learning_rate": 1.4636286272709033e-05,
+      "loss": 1.2345,
+      "step": 883
+    },
+    {
+      "epoch": 0.5995760915642221,
+      "grad_norm": 0.6926055300430728,
+      "learning_rate": 1.4593940872129268e-05,
+      "loss": 1.2695,
+      "step": 884
+    },
+    {
+      "epoch": 0.6002543450614667,
+      "grad_norm": 0.7055103538930207,
+      "learning_rate": 1.4551621600135319e-05,
+      "loss": 1.2355,
+      "step": 885
+    },
+    {
+      "epoch": 0.6009325985587113,
+      "grad_norm": 0.7267167745593457,
+      "learning_rate": 1.4509328661264824e-05,
+      "loss": 1.2909,
+      "step": 886
+    },
+    {
+      "epoch": 0.6016108520559559,
+      "grad_norm": 0.7405252682023891,
+      "learning_rate": 1.44670622599282e-05,
+      "loss": 1.2457,
+      "step": 887
+    },
+    {
+      "epoch": 0.6022891055532005,
+      "grad_norm": 0.7209599678458054,
+      "learning_rate": 1.4424822600407558e-05,
+      "loss": 1.2596,
+      "step": 888
+    },
+    {
+      "epoch": 0.6029673590504451,
+      "grad_norm": 0.6906841638341872,
+      "learning_rate": 1.4382609886855811e-05,
+      "loss": 1.2629,
+      "step": 889
+    },
+    {
+      "epoch": 0.6036456125476897,
+      "grad_norm": 0.7240507304691446,
+      "learning_rate": 1.4340424323295579e-05,
+      "loss": 1.2775,
+      "step": 890
+    },
+    {
+      "epoch": 0.6043238660449343,
+      "grad_norm": 0.708461486979943,
+      "learning_rate": 1.4298266113618311e-05,
+      "loss": 1.2555,
+      "step": 891
+    },
+    {
+      "epoch": 0.6050021195421789,
+      "grad_norm": 0.7043004010120361,
+      "learning_rate": 1.4256135461583225e-05,
+      "loss": 1.2327,
+      "step": 892
+    },
+    {
+      "epoch": 0.6056803730394235,
+      "grad_norm": 0.6954193461319649,
+      "learning_rate": 1.4214032570816346e-05,
+      "loss": 1.2673,
+      "step": 893
+    },
+    {
+      "epoch": 0.606358626536668,
+      "grad_norm": 0.7259164250225965,
+      "learning_rate": 1.4171957644809533e-05,
+      "loss": 1.2707,
+      "step": 894
+    },
+    {
+      "epoch": 0.6070368800339127,
+      "grad_norm": 0.7299474390160032,
+      "learning_rate": 1.4129910886919472e-05,
+      "loss": 1.2471,
+      "step": 895
+    },
+    {
+      "epoch": 0.6077151335311572,
+      "grad_norm": 0.6744713222322375,
+      "learning_rate": 1.4087892500366725e-05,
+      "loss": 1.2568,
+      "step": 896
+    },
+    {
+      "epoch": 0.6083933870284018,
+      "grad_norm": 0.7216386104384064,
+      "learning_rate": 1.4045902688234711e-05,
+      "loss": 1.2794,
+      "step": 897
+    },
+    {
+      "epoch": 0.6090716405256464,
+      "grad_norm": 0.7658293993436277,
+      "learning_rate": 1.4003941653468758e-05,
+      "loss": 1.2536,
+      "step": 898
+    },
+    {
+      "epoch": 0.609749894022891,
+      "grad_norm": 0.7234591048031078,
+      "learning_rate": 1.39620095988751e-05,
+      "loss": 1.2694,
+      "step": 899
+    },
+    {
+      "epoch": 0.6104281475201356,
+      "grad_norm": 0.6981886633012235,
+      "learning_rate": 1.3920106727119901e-05,
+      "loss": 1.2803,
+      "step": 900
+    },
+    {
+      "epoch": 0.6111064010173802,
+      "grad_norm": 0.7248262364678507,
+      "learning_rate": 1.3878233240728287e-05,
+      "loss": 1.2675,
+      "step": 901
+    },
+    {
+      "epoch": 0.6117846545146248,
+      "grad_norm": 0.7658558357705719,
+      "learning_rate": 1.383638934208335e-05,
+      "loss": 1.2547,
+      "step": 902
+    },
+    {
+      "epoch": 0.6124629080118694,
+      "grad_norm": 0.7407751363047477,
+      "learning_rate": 1.3794575233425187e-05,
+      "loss": 1.2867,
+      "step": 903
+    },
+    {
+      "epoch": 0.613141161509114,
+      "grad_norm": 0.7305356284556889,
+      "learning_rate": 1.37527911168499e-05,
+      "loss": 1.2639,
+      "step": 904
+    },
+    {
+      "epoch": 0.6138194150063586,
+      "grad_norm": 0.754015236627855,
+      "learning_rate": 1.3711037194308653e-05,
+      "loss": 1.2414,
+      "step": 905
+    },
+    {
+      "epoch": 0.6144976685036032,
+      "grad_norm": 0.7423896758793486,
+      "learning_rate": 1.3669313667606655e-05,
+      "loss": 1.2469,
+      "step": 906
+    },
+    {
+      "epoch": 0.6151759220008478,
+      "grad_norm": 0.7073978592211552,
+      "learning_rate": 1.3627620738402221e-05,
+      "loss": 1.2865,
+      "step": 907
+    },
+    {
+      "epoch": 0.6158541754980924,
+      "grad_norm": 0.7637953954513205,
+      "learning_rate": 1.3585958608205779e-05,
+      "loss": 1.2629,
+      "step": 908
+    },
+    {
+      "epoch": 0.616532428995337,
+      "grad_norm": 0.6891146950788505,
+      "learning_rate": 1.3544327478378891e-05,
+      "loss": 1.2771,
+      "step": 909
+    },
+    {
+      "epoch": 0.6172106824925816,
+      "grad_norm": 0.6944143467859542,
+      "learning_rate": 1.3502727550133306e-05,
+      "loss": 1.2685,
+      "step": 910
+    },
+    {
+      "epoch": 0.6178889359898262,
+      "grad_norm": 0.7664699532195742,
+      "learning_rate": 1.3461159024529942e-05,
+      "loss": 1.2844,
+      "step": 911
+    },
+    {
+      "epoch": 0.6185671894870708,
+      "grad_norm": 0.7201761910665536,
+      "learning_rate": 1.3419622102477967e-05,
+      "loss": 1.2871,
+      "step": 912
+    },
+    {
+      "epoch": 0.6192454429843154,
+      "grad_norm": 0.7287096841573026,
+      "learning_rate": 1.3378116984733791e-05,
+      "loss": 1.2582,
+      "step": 913
+    },
+    {
+      "epoch": 0.61992369648156,
+      "grad_norm": 0.7374024655158494,
+      "learning_rate": 1.3336643871900101e-05,
+      "loss": 1.2479,
+      "step": 914
+    },
+    {
+      "epoch": 0.6206019499788046,
+      "grad_norm": 0.7344509915633016,
+      "learning_rate": 1.3295202964424925e-05,
+      "loss": 1.2704,
+      "step": 915
+    },
+    {
+      "epoch": 0.6212802034760492,
+      "grad_norm": 0.7182428414638428,
+      "learning_rate": 1.3253794462600592e-05,
+      "loss": 1.2547,
+      "step": 916
+    },
+    {
+      "epoch": 0.6219584569732938,
+      "grad_norm": 0.7456646350427812,
+      "learning_rate": 1.3212418566562857e-05,
+      "loss": 1.2791,
+      "step": 917
+    },
+    {
+      "epoch": 0.6226367104705384,
+      "grad_norm": 0.4978577354755168,
+      "learning_rate": 1.3171075476289835e-05,
+      "loss": 1.382,
+      "step": 918
+    },
+    {
+      "epoch": 0.623314963967783,
+      "grad_norm": 0.745274485467238,
+      "learning_rate": 1.3129765391601135e-05,
+      "loss": 1.2466,
+      "step": 919
+    },
+    {
+      "epoch": 0.6239932174650276,
+      "grad_norm": 0.7878519121126696,
+      "learning_rate": 1.3088488512156792e-05,
+      "loss": 1.2795,
+      "step": 920
+    },
+    {
+      "epoch": 0.6246714709622722,
+      "grad_norm": 0.7220277267261397,
+      "learning_rate": 1.3047245037456392e-05,
+      "loss": 1.272,
+      "step": 921
+    },
+    {
+      "epoch": 0.6253497244595168,
+      "grad_norm": 0.7432803867257498,
+      "learning_rate": 1.3006035166838068e-05,
+      "loss": 1.2744,
+      "step": 922
+    },
+    {
+      "epoch": 0.6260279779567614,
+      "grad_norm": 0.7460263032906728,
+      "learning_rate": 1.2964859099477499e-05,
+      "loss": 1.2815,
+      "step": 923
+    },
+    {
+      "epoch": 0.626706231454006,
+      "grad_norm": 0.6981401291287395,
+      "learning_rate": 1.2923717034387035e-05,
+      "loss": 1.2057,
+      "step": 924
+    },
+    {
+      "epoch": 0.6273844849512505,
+      "grad_norm": 0.7180593879124996,
+      "learning_rate": 1.2882609170414646e-05,
+      "loss": 1.2235,
+      "step": 925
+    },
+    {
+      "epoch": 0.6280627384484951,
+      "grad_norm": 0.751874092601359,
+      "learning_rate": 1.2841535706243039e-05,
+      "loss": 1.2681,
+      "step": 926
+    },
+    {
+      "epoch": 0.6287409919457397,
+      "grad_norm": 0.7238837555578427,
+      "learning_rate": 1.280049684038861e-05,
+      "loss": 1.2546,
+      "step": 927
+    },
+    {
+      "epoch": 0.6294192454429843,
+      "grad_norm": 0.7623351431358577,
+      "learning_rate": 1.2759492771200588e-05,
+      "loss": 1.274,
+      "step": 928
+    },
+    {
+      "epoch": 0.6300974989402289,
+      "grad_norm": 0.726521566712189,
+      "learning_rate": 1.2718523696859992e-05,
+      "loss": 1.2556,
+      "step": 929
+    },
+    {
+      "epoch": 0.6307757524374735,
+      "grad_norm": 0.7294458811856235,
+      "learning_rate": 1.2677589815378703e-05,
+      "loss": 1.2611,
+      "step": 930
+    },
+    {
+      "epoch": 0.6314540059347181,
+      "grad_norm": 0.7337933448476249,
+      "learning_rate": 1.2636691324598527e-05,
+      "loss": 1.2638,
+      "step": 931
+    },
+    {
+      "epoch": 0.6321322594319627,
+      "grad_norm": 0.7620793167630296,
+      "learning_rate": 1.2595828422190195e-05,
+      "loss": 1.2631,
+      "step": 932
+    },
+    {
+      "epoch": 0.6328105129292073,
+      "grad_norm": 0.7333252664894931,
+      "learning_rate": 1.2555001305652454e-05,
+      "loss": 1.2744,
+      "step": 933
+    },
+    {
+      "epoch": 0.6334887664264519,
+      "grad_norm": 0.7009821990937899,
+      "learning_rate": 1.2514210172311074e-05,
+      "loss": 1.2504,
+      "step": 934
+    },
+    {
+      "epoch": 0.6341670199236965,
+      "grad_norm": 0.7252465526498413,
+      "learning_rate": 1.247345521931792e-05,
+      "loss": 1.2772,
+      "step": 935
+    },
+    {
+      "epoch": 0.6348452734209411,
+      "grad_norm": 0.7538115830639216,
+      "learning_rate": 1.2432736643649985e-05,
+      "loss": 1.2551,
+      "step": 936
+    },
+    {
+      "epoch": 0.6355235269181857,
+      "grad_norm": 0.7276674431447986,
+      "learning_rate": 1.239205464210845e-05,
+      "loss": 1.2507,
+      "step": 937
+    },
+    {
+      "epoch": 0.6362017804154303,
+      "grad_norm": 0.7249572800476735,
+      "learning_rate": 1.2351409411317725e-05,
+      "loss": 1.2512,
+      "step": 938
+    },
+    {
+      "epoch": 0.6368800339126749,
+      "grad_norm": 0.7200866241176038,
+      "learning_rate": 1.2310801147724484e-05,
+      "loss": 1.2177,
+      "step": 939
+    },
+    {
+      "epoch": 0.6375582874099195,
+      "grad_norm": 0.6966135730154913,
+      "learning_rate": 1.2270230047596757e-05,
+      "loss": 1.238,
+      "step": 940
+    },
+    {
+      "epoch": 0.6382365409071641,
+      "grad_norm": 0.7119468935646069,
+      "learning_rate": 1.2229696307022926e-05,
+      "loss": 1.2271,
+      "step": 941
+    },
+    {
+      "epoch": 0.6389147944044087,
+      "grad_norm": 0.7384160273867104,
+      "learning_rate": 1.218920012191084e-05,
+      "loss": 1.2609,
+      "step": 942
+    },
+    {
+      "epoch": 0.6395930479016533,
+      "grad_norm": 0.7613184802931767,
+      "learning_rate": 1.21487416879868e-05,
+      "loss": 1.2668,
+      "step": 943
+    },
+    {
+      "epoch": 0.6402713013988979,
+      "grad_norm": 0.7124097044569389,
+      "learning_rate": 1.2108321200794672e-05,
+      "loss": 1.2441,
+      "step": 944
+    },
+    {
+      "epoch": 0.6409495548961425,
+      "grad_norm": 0.7095189582515088,
+      "learning_rate": 1.2067938855694919e-05,
+      "loss": 1.2373,
+      "step": 945
+    },
+    {
+      "epoch": 0.6416278083933871,
+      "grad_norm": 0.7271405420859809,
+      "learning_rate": 1.202759484786363e-05,
+      "loss": 1.2449,
+      "step": 946
+    },
+    {
+      "epoch": 0.6423060618906317,
+      "grad_norm": 0.70386838213244,
+      "learning_rate": 1.198728937229163e-05,
+      "loss": 1.2466,
+      "step": 947
+    },
+    {
+      "epoch": 0.6429843153878763,
+      "grad_norm": 0.7311311552571067,
+      "learning_rate": 1.1947022623783495e-05,
+      "loss": 1.236,
+      "step": 948
+    },
+    {
+      "epoch": 0.6436625688851209,
+      "grad_norm": 0.7445659895700161,
+      "learning_rate": 1.1906794796956633e-05,
+      "loss": 1.2718,
+      "step": 949
+    },
+    {
+      "epoch": 0.6443408223823655,
+      "grad_norm": 0.714864727873201,
+      "learning_rate": 1.1866606086240325e-05,
+      "loss": 1.2805,
+      "step": 950
+    },
+    {
+      "epoch": 0.64501907587961,
+      "grad_norm": 0.7349415371328099,
+      "learning_rate": 1.1826456685874801e-05,
+      "loss": 1.2555,
+      "step": 951
+    },
+    {
+      "epoch": 0.6456973293768546,
+      "grad_norm": 0.7234244895672503,
+      "learning_rate": 1.1786346789910316e-05,
+      "loss": 1.2613,
+      "step": 952
+    },
+    {
+      "epoch": 0.6463755828740992,
+      "grad_norm": 0.7179485946292944,
+      "learning_rate": 1.1746276592206147e-05,
+      "loss": 1.2623,
+      "step": 953
+    },
+    {
+      "epoch": 0.6470538363713438,
+      "grad_norm": 0.7391341017957087,
+      "learning_rate": 1.1706246286429752e-05,
+      "loss": 1.2486,
+      "step": 954
+    },
+    {
+      "epoch": 0.6477320898685884,
+      "grad_norm": 0.7150250700985901,
+      "learning_rate": 1.1666256066055739e-05,
+      "loss": 1.2459,
+      "step": 955
+    },
+    {
+      "epoch": 0.648410343365833,
+      "grad_norm": 0.686582064168532,
+      "learning_rate": 1.162630612436501e-05,
+      "loss": 1.2713,
+      "step": 956
+    },
+    {
+      "epoch": 0.6490885968630776,
+      "grad_norm": 0.7075187279569484,
+      "learning_rate": 1.1586396654443773e-05,
+      "loss": 1.223,
+      "step": 957
+    },
+    {
+      "epoch": 0.6497668503603222,
+      "grad_norm": 0.7178501779928439,
+      "learning_rate": 1.1546527849182621e-05,
+      "loss": 1.2227,
+      "step": 958
+    },
+    {
+      "epoch": 0.6504451038575668,
+      "grad_norm": 0.7192989993624956,
+      "learning_rate": 1.1506699901275633e-05,
+      "loss": 1.2598,
+      "step": 959
+    },
+    {
+      "epoch": 0.6511233573548114,
+      "grad_norm": 0.7301835358461147,
+      "learning_rate": 1.1466913003219395e-05,
+      "loss": 1.2332,
+      "step": 960
+    },
+    {
+      "epoch": 0.6518016108520559,
+      "grad_norm": 0.6994674329929395,
+      "learning_rate": 1.1427167347312093e-05,
+      "loss": 1.2699,
+      "step": 961
+    },
+    {
+      "epoch": 0.6524798643493005,
+      "grad_norm": 0.7078327956045938,
+      "learning_rate": 1.1387463125652579e-05,
+      "loss": 1.2414,
+      "step": 962
+    },
+    {
+      "epoch": 0.6531581178465451,
+      "grad_norm": 0.7630351678619136,
+      "learning_rate": 1.1347800530139463e-05,
+      "loss": 1.266,
+      "step": 963
+    },
+    {
+      "epoch": 0.6538363713437897,
+      "grad_norm": 0.7594897772201343,
+      "learning_rate": 1.1308179752470147e-05,
+      "loss": 1.2626,
+      "step": 964
+    },
+    {
+      "epoch": 0.6545146248410343,
+      "grad_norm": 0.6955575097724982,
+      "learning_rate": 1.126860098413993e-05,
+      "loss": 1.2517,
+      "step": 965
+    },
+    {
+      "epoch": 0.6551928783382789,
+      "grad_norm": 0.7326960738706964,
+      "learning_rate": 1.1229064416441055e-05,
+      "loss": 1.2897,
+      "step": 966
+    },
+    {
+      "epoch": 0.6558711318355235,
+      "grad_norm": 0.7312816709206589,
+      "learning_rate": 1.1189570240461825e-05,
+      "loss": 1.2711,
+      "step": 967
+    },
+    {
+      "epoch": 0.6565493853327681,
+      "grad_norm": 0.7707904224347778,
+      "learning_rate": 1.1150118647085653e-05,
+      "loss": 1.2429,
+      "step": 968
+    },
+    {
+      "epoch": 0.6572276388300127,
+      "grad_norm": 0.7181578925761877,
+      "learning_rate": 1.111070982699011e-05,
+      "loss": 1.2603,
+      "step": 969
+    },
+    {
+      "epoch": 0.6579058923272573,
+      "grad_norm": 0.7485309379333241,
+      "learning_rate": 1.1071343970646069e-05,
+      "loss": 1.241,
+      "step": 970
+    },
+    {
+      "epoch": 0.6585841458245019,
+      "grad_norm": 0.7420054340386768,
+      "learning_rate": 1.1032021268316742e-05,
+      "loss": 1.2773,
+      "step": 971
+    },
+    {
+      "epoch": 0.6592623993217465,
+      "grad_norm": 0.7406520244405697,
+      "learning_rate": 1.0992741910056758e-05,
+      "loss": 1.2649,
+      "step": 972
+    },
+    {
+      "epoch": 0.6599406528189911,
+      "grad_norm": 0.7497464155716638,
+      "learning_rate": 1.0953506085711258e-05,
+      "loss": 1.251,
+      "step": 973
+    },
+    {
+      "epoch": 0.6606189063162357,
+      "grad_norm": 0.7075855670188584,
+      "learning_rate": 1.0914313984914987e-05,
+      "loss": 1.2554,
+      "step": 974
+    },
+    {
+      "epoch": 0.6612971598134803,
+      "grad_norm": 0.7163704059152383,
+      "learning_rate": 1.0875165797091371e-05,
+      "loss": 1.2594,
+      "step": 975
+    },
+    {
+      "epoch": 0.6619754133107248,
+      "grad_norm": 0.7107842817416975,
+      "learning_rate": 1.0836061711451552e-05,
+      "loss": 1.2553,
+      "step": 976
+    },
+    {
+      "epoch": 0.6626536668079694,
+      "grad_norm": 0.7121994112451051,
+      "learning_rate": 1.0797001916993572e-05,
+      "loss": 1.2851,
+      "step": 977
+    },
+    {
+      "epoch": 0.663331920305214,
+      "grad_norm": 0.742016146380806,
+      "learning_rate": 1.0757986602501365e-05,
+      "loss": 1.2729,
+      "step": 978
+    },
+    {
+      "epoch": 0.6640101738024586,
+      "grad_norm": 0.7614468214071379,
+      "learning_rate": 1.071901595654392e-05,
+      "loss": 1.2791,
+      "step": 979
+    },
+    {
+      "epoch": 0.6646884272997032,
+      "grad_norm": 0.7137224832937834,
+      "learning_rate": 1.0680090167474284e-05,
+      "loss": 1.2646,
+      "step": 980
+    },
+    {
+      "epoch": 0.6653666807969478,
+      "grad_norm": 0.6944784698203637,
+      "learning_rate": 1.0641209423428745e-05,
+      "loss": 1.2663,
+      "step": 981
+    },
+    {
+      "epoch": 0.6660449342941924,
+      "grad_norm": 0.6926775709241877,
+      "learning_rate": 1.0602373912325872e-05,
+      "loss": 1.2559,
+      "step": 982
+    },
+    {
+      "epoch": 0.666723187791437,
+      "grad_norm": 0.7121621198396063,
+      "learning_rate": 1.0563583821865598e-05,
+      "loss": 1.259,
+      "step": 983
+    },
+    {
+      "epoch": 0.6674014412886816,
+      "grad_norm": 0.7492579217654587,
+      "learning_rate": 1.0524839339528344e-05,
+      "loss": 1.2615,
+      "step": 984
+    },
+    {
+      "epoch": 0.6680796947859262,
+      "grad_norm": 0.7234426728331795,
+      "learning_rate": 1.0486140652574077e-05,
+      "loss": 1.2353,
+      "step": 985
+    },
+    {
+      "epoch": 0.6687579482831708,
+      "grad_norm": 0.6977853252293624,
+      "learning_rate": 1.0447487948041454e-05,
+      "loss": 1.2321,
+      "step": 986
+    },
+    {
+      "epoch": 0.6694362017804154,
+      "grad_norm": 0.7180470239654396,
+      "learning_rate": 1.040888141274688e-05,
+      "loss": 1.2575,
+      "step": 987
+    },
+    {
+      "epoch": 0.67011445527766,
+      "grad_norm": 0.7166912234194147,
+      "learning_rate": 1.0370321233283587e-05,
+      "loss": 1.2592,
+      "step": 988
+    },
+    {
+      "epoch": 0.6707927087749046,
+      "grad_norm": 0.719471786130956,
+      "learning_rate": 1.0331807596020804e-05,
+      "loss": 1.2263,
+      "step": 989
+    },
+    {
+      "epoch": 0.6714709622721492,
+      "grad_norm": 0.7153381227740333,
+      "learning_rate": 1.029334068710279e-05,
+      "loss": 1.257,
+      "step": 990
+    },
+    {
+      "epoch": 0.6721492157693938,
+      "grad_norm": 0.7211376167286312,
+      "learning_rate": 1.0254920692447946e-05,
+      "loss": 1.2396,
+      "step": 991
+    },
+    {
+      "epoch": 0.6728274692666384,
+      "grad_norm": 0.7353230173357311,
+      "learning_rate": 1.0216547797747935e-05,
+      "loss": 1.2579,
+      "step": 992
+    },
+    {
+      "epoch": 0.673505722763883,
+      "grad_norm": 0.7749549540065295,
+      "learning_rate": 1.01782221884668e-05,
+      "loss": 1.2472,
+      "step": 993
+    },
+    {
+      "epoch": 0.6741839762611276,
+      "grad_norm": 0.7052582586968418,
+      "learning_rate": 1.013994404984001e-05,
+      "loss": 1.222,
+      "step": 994
+    },
+    {
+      "epoch": 0.6748622297583722,
+      "grad_norm": 0.7238807278644459,
+      "learning_rate": 1.0101713566873612e-05,
+      "loss": 1.2622,
+      "step": 995
+    },
+    {
+      "epoch": 0.6755404832556168,
+      "grad_norm": 0.7224437920546443,
+      "learning_rate": 1.006353092434332e-05,
+      "loss": 1.2417,
+      "step": 996
+    },
+    {
+      "epoch": 0.6762187367528614,
+      "grad_norm": 0.7030898940390986,
+      "learning_rate": 1.002539630679364e-05,
+      "loss": 1.2487,
+      "step": 997
+    },
+    {
+      "epoch": 0.676896990250106,
+      "grad_norm": 0.7063999825585107,
+      "learning_rate": 9.987309898536946e-06,
+      "loss": 1.2402,
+      "step": 998
+    },
+    {
+      "epoch": 0.6775752437473506,
+      "grad_norm": 0.7216994485791334,
+      "learning_rate": 9.949271883652605e-06,
+      "loss": 1.2535,
+      "step": 999
+    },
+    {
+      "epoch": 0.6782534972445952,
+      "grad_norm": 0.715290491246088,
+      "learning_rate": 9.911282445986115e-06,
+      "loss": 1.2452,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6789317507418398,
+      "grad_norm": 0.7235774612430038,
+      "learning_rate": 9.87334176914816e-06,
+      "loss": 1.2694,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6796100042390844,
+      "grad_norm": 0.7822215823589889,
+      "learning_rate": 9.835450036513772e-06,
+      "loss": 1.2491,
+      "step": 1002
+    },
+    {
+      "epoch": 0.680288257736329,
+      "grad_norm": 0.7435176319997631,
+      "learning_rate": 9.797607431221405e-06,
+      "loss": 1.2358,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6809665112335735,
+      "grad_norm": 0.7402778365799259,
+      "learning_rate": 9.759814136172097e-06,
+      "loss": 1.2333,
+      "step": 1004
+    },
+    {
+      "epoch": 0.6816447647308181,
+      "grad_norm": 0.6991177876465452,
+      "learning_rate": 9.722070334028557e-06,
+      "loss": 1.2448,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6823230182280627,
+      "grad_norm": 0.7390069128689151,
+      "learning_rate": 9.684376207214252e-06,
+      "loss": 1.2667,
+      "step": 1006
+    },
+    {
+      "epoch": 0.6830012717253073,
+      "grad_norm": 0.744668014688226,
+      "learning_rate": 9.646731937912596e-06,
+      "loss": 1.238,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6836795252225519,
+      "grad_norm": 0.6847769198917258,
+      "learning_rate": 9.609137708066007e-06,
+      "loss": 1.249,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6843577787197965,
+      "grad_norm": 0.7011560538142221,
+      "learning_rate": 9.571593699375082e-06,
+      "loss": 1.2518,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6850360322170411,
+      "grad_norm": 0.7296332469643578,
+      "learning_rate": 9.534100093297637e-06,
+      "loss": 1.2667,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 0.7564106571457987,
+      "learning_rate": 9.49665707104793e-06,
+      "loss": 1.2803,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6863925392115303,
+      "grad_norm": 0.7497025167281187,
+      "learning_rate": 9.459264813595736e-06,
+      "loss": 1.2676,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6870707927087749,
+      "grad_norm": 0.6994523711051618,
+      "learning_rate": 9.421923501665426e-06,
+      "loss": 1.2505,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6877490462060195,
+      "grad_norm": 0.7215420284145486,
+      "learning_rate": 9.384633315735197e-06,
+      "loss": 1.2526,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6884272997032641,
+      "grad_norm": 0.7182576007658268,
+      "learning_rate": 9.3473944360361e-06,
+      "loss": 1.2344,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6891055532005087,
+      "grad_norm": 0.7424567342468971,
+      "learning_rate": 9.310207042551258e-06,
+      "loss": 1.2352,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6897838066977533,
+      "grad_norm": 0.7086078062679702,
+      "learning_rate": 9.273071315014897e-06,
+      "loss": 1.2633,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6904620601949979,
+      "grad_norm": 0.7444779504149994,
+      "learning_rate": 9.235987432911567e-06,
+      "loss": 1.2711,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6911403136922425,
+      "grad_norm": 0.77022707965804,
+      "learning_rate": 9.198955575475241e-06,
+      "loss": 1.2317,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6918185671894871,
+      "grad_norm": 0.7276754899754154,
+      "learning_rate": 9.161975921688427e-06,
+      "loss": 1.2791,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6924968206867317,
+      "grad_norm": 0.7026188417040135,
+      "learning_rate": 9.12504865028133e-06,
+      "loss": 1.2493,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6931750741839763,
+      "grad_norm": 0.6926211613260367,
+      "learning_rate": 9.088173939730971e-06,
+      "loss": 1.2467,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6938533276812209,
+      "grad_norm": 0.7017132659341797,
+      "learning_rate": 9.051351968260362e-06,
+      "loss": 1.2537,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6945315811784655,
+      "grad_norm": 0.7157201337759609,
+      "learning_rate": 9.014582913837588e-06,
+      "loss": 1.2521,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6952098346757101,
+      "grad_norm": 0.7397450250982527,
+      "learning_rate": 8.977866954174983e-06,
+      "loss": 1.2667,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6958880881729547,
+      "grad_norm": 0.7381070889903316,
+      "learning_rate": 8.941204266728259e-06,
+      "loss": 1.2799,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6965663416701993,
+      "grad_norm": 0.471275683966,
+      "learning_rate": 8.904595028695673e-06,
+      "loss": 1.3761,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6972445951674439,
+      "grad_norm": 0.7477049967173038,
+      "learning_rate": 8.868039417017128e-06,
+      "loss": 1.264,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6979228486646885,
+      "grad_norm": 0.7439285336253352,
+      "learning_rate": 8.831537608373337e-06,
+      "loss": 1.2387,
+      "step": 1029
+    },
+    {
+      "epoch": 0.698601102161933,
+      "grad_norm": 0.7372188855937621,
+      "learning_rate": 8.795089779185e-06,
+      "loss": 1.2242,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6992793556591776,
+      "grad_norm": 0.7108484812069134,
+      "learning_rate": 8.758696105611895e-06,
+      "loss": 1.2376,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6999576091564222,
+      "grad_norm": 0.7112819048115543,
+      "learning_rate": 8.72235676355207e-06,
+      "loss": 1.2577,
+      "step": 1032
+    },
+    {
+      "epoch": 0.7006358626536668,
+      "grad_norm": 0.46441117142961635,
+      "learning_rate": 8.68607192864096e-06,
+      "loss": 1.3853,
+      "step": 1033
+    },
+    {
+      "epoch": 0.7013141161509114,
+      "grad_norm": 0.704005301317096,
+      "learning_rate": 8.64984177625059e-06,
+      "loss": 1.2523,
+      "step": 1034
+    },
+    {
+      "epoch": 0.701992369648156,
+      "grad_norm": 0.6760654966910298,
+      "learning_rate": 8.613666481488658e-06,
+      "loss": 1.2579,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7026706231454006,
+      "grad_norm": 0.6859875086712081,
+      "learning_rate": 8.577546219197734e-06,
+      "loss": 1.203,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7033488766426452,
+      "grad_norm": 0.7250776709559372,
+      "learning_rate": 8.541481163954426e-06,
+      "loss": 1.2512,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7040271301398898,
+      "grad_norm": 0.6828939714058161,
+      "learning_rate": 8.505471490068487e-06,
+      "loss": 1.2535,
+      "step": 1038
+    },
+    {
+      "epoch": 0.7047053836371344,
+      "grad_norm": 0.7247935847952882,
+      "learning_rate": 8.46951737158201e-06,
+      "loss": 1.2648,
+      "step": 1039
+    },
+    {
+      "epoch": 0.705383637134379,
+      "grad_norm": 0.7062728261435427,
+      "learning_rate": 8.433618982268575e-06,
+      "loss": 1.2427,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7060618906316236,
+      "grad_norm": 0.708341700820802,
+      "learning_rate": 8.397776495632424e-06,
+      "loss": 1.2364,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7067401441288682,
+      "grad_norm": 0.7057742843892745,
+      "learning_rate": 8.361990084907609e-06,
+      "loss": 1.2597,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7074183976261128,
+      "grad_norm": 0.7134634886760335,
+      "learning_rate": 8.326259923057129e-06,
+      "loss": 1.2567,
+      "step": 1043
+    },
+    {
+      "epoch": 0.7080966511233574,
+      "grad_norm": 0.7112068813359304,
+      "learning_rate": 8.290586182772153e-06,
+      "loss": 1.2588,
+      "step": 1044
+    },
+    {
+      "epoch": 0.708774904620602,
+      "grad_norm": 0.7166082576786501,
+      "learning_rate": 8.254969036471133e-06,
+      "loss": 1.2684,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7094531581178466,
+      "grad_norm": 0.6774303776112416,
+      "learning_rate": 8.21940865629901e-06,
+      "loss": 1.2518,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7101314116150912,
+      "grad_norm": 0.7369725642504088,
+      "learning_rate": 8.183905214126331e-06,
+      "loss": 1.2812,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7108096651123358,
+      "grad_norm": 0.7125161557801593,
+      "learning_rate": 8.148458881548478e-06,
+      "loss": 1.2579,
+      "step": 1048
+    },
+    {
+      "epoch": 0.7114879186095804,
+      "grad_norm": 0.7123133162717256,
+      "learning_rate": 8.11306982988481e-06,
+      "loss": 1.2381,
+      "step": 1049
+    },
+    {
+      "epoch": 0.712166172106825,
+      "grad_norm": 0.4556155958071684,
+      "learning_rate": 8.077738230177823e-06,
+      "loss": 1.3769,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7128444256040696,
+      "grad_norm": 0.7296551165413036,
+      "learning_rate": 8.042464253192339e-06,
+      "loss": 1.2557,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7135226791013141,
+      "grad_norm": 0.7153326806368623,
+      "learning_rate": 8.007248069414673e-06,
+      "loss": 1.2453,
+      "step": 1052
+    },
+    {
+      "epoch": 0.7142009325985587,
+      "grad_norm": 0.7370663314394464,
+      "learning_rate": 7.972089849051834e-06,
+      "loss": 1.2612,
+      "step": 1053
+    },
+    {
+      "epoch": 0.7148791860958033,
+      "grad_norm": 0.7071174488237575,
+      "learning_rate": 7.936989762030662e-06,
+      "loss": 1.2661,
+      "step": 1054
+    },
+    {
+      "epoch": 0.7155574395930479,
+      "grad_norm": 0.6903000911952577,
+      "learning_rate": 7.901947977997031e-06,
+      "loss": 1.2708,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7162356930902924,
+      "grad_norm": 0.6865849284089116,
+      "learning_rate": 7.866964666315018e-06,
+      "loss": 1.2527,
+      "step": 1056
+    },
+    {
+      "epoch": 0.716913946587537,
+      "grad_norm": 0.7297435263907307,
+      "learning_rate": 7.832039996066118e-06,
+      "loss": 1.2346,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7175922000847816,
+      "grad_norm": 0.7267674081261121,
+      "learning_rate": 7.797174136048373e-06,
+      "loss": 1.2684,
+      "step": 1058
+    },
+    {
+      "epoch": 0.7182704535820262,
+      "grad_norm": 0.770600917229959,
+      "learning_rate": 7.762367254775584e-06,
+      "loss": 1.27,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7189487070792708,
+      "grad_norm": 0.6989971112730319,
+      "learning_rate": 7.727619520476522e-06,
+      "loss": 1.2327,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7196269605765154,
+      "grad_norm": 0.6845992496825357,
+      "learning_rate": 7.692931101094055e-06,
+      "loss": 1.2393,
+      "step": 1061
+    },
+    {
+      "epoch": 0.72030521407376,
+      "grad_norm": 0.6972088539223147,
+      "learning_rate": 7.658302164284393e-06,
+      "loss": 1.2649,
+      "step": 1062
+    },
+    {
+      "epoch": 0.7209834675710046,
+      "grad_norm": 0.6925122480528125,
+      "learning_rate": 7.623732877416232e-06,
+      "loss": 1.2515,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7216617210682492,
+      "grad_norm": 0.7035579922671705,
+      "learning_rate": 7.589223407570006e-06,
+      "loss": 1.2495,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7223399745654938,
+      "grad_norm": 0.4509807827721255,
+      "learning_rate": 7.554773921537002e-06,
+      "loss": 1.3852,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7230182280627384,
+      "grad_norm": 0.748892704277171,
+      "learning_rate": 7.520384585818608e-06,
+      "loss": 1.267,
+      "step": 1066
+    },
+    {
+      "epoch": 0.723696481559983,
+      "grad_norm": 0.7108147391195934,
+      "learning_rate": 7.486055566625503e-06,
+      "loss": 1.2619,
+      "step": 1067
+    },
+    {
+      "epoch": 0.7243747350572276,
+      "grad_norm": 0.7086880409021801,
+      "learning_rate": 7.451787029876829e-06,
+      "loss": 1.2643,
+      "step": 1068
+    },
+    {
+      "epoch": 0.7250529885544722,
+      "grad_norm": 0.6977432184670711,
+      "learning_rate": 7.4175791411994114e-06,
+      "loss": 1.2146,
+      "step": 1069
+    },
+    {
+      "epoch": 0.7257312420517168,
+      "grad_norm": 0.6687009426082059,
+      "learning_rate": 7.383432065926943e-06,
+      "loss": 1.2536,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7264094955489614,
+      "grad_norm": 0.6879116674692699,
+      "learning_rate": 7.349345969099211e-06,
+      "loss": 1.2608,
+      "step": 1071
+    },
+    {
+      "epoch": 0.727087749046206,
+      "grad_norm": 0.7102026991932625,
+      "learning_rate": 7.315321015461263e-06,
+      "loss": 1.2416,
+      "step": 1072
+    },
+    {
+      "epoch": 0.7277660025434506,
+      "grad_norm": 0.7263982096199556,
+      "learning_rate": 7.281357369462632e-06,
+      "loss": 1.2463,
+      "step": 1073
+    },
+    {
+      "epoch": 0.7284442560406952,
+      "grad_norm": 0.7199407337871239,
+      "learning_rate": 7.247455195256552e-06,
+      "loss": 1.2308,
+      "step": 1074
+    },
+    {
+      "epoch": 0.7291225095379398,
+      "grad_norm": 0.6968257341434265,
+      "learning_rate": 7.21361465669914e-06,
+      "loss": 1.2689,
+      "step": 1075
+    },
+    {
+      "epoch": 0.7298007630351844,
+      "grad_norm": 0.7107034728723381,
+      "learning_rate": 7.179835917348614e-06,
+      "loss": 1.2153,
+      "step": 1076
+    },
+    {
+      "epoch": 0.730479016532429,
+      "grad_norm": 0.7479033842548566,
+      "learning_rate": 7.146119140464501e-06,
+      "loss": 1.2619,
+      "step": 1077
+    },
+    {
+      "epoch": 0.7311572700296736,
+      "grad_norm": 0.6977861461512729,
+      "learning_rate": 7.112464489006865e-06,
+      "loss": 1.2417,
+      "step": 1078
+    },
+    {
+      "epoch": 0.7318355235269182,
+      "grad_norm": 0.6592519646440719,
+      "learning_rate": 7.078872125635507e-06,
+      "loss": 1.2182,
+      "step": 1079
+    },
+    {
+      "epoch": 0.7325137770241628,
+      "grad_norm": 0.4570588318163805,
+      "learning_rate": 7.045342212709146e-06,
+      "loss": 1.3656,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7331920305214074,
+      "grad_norm": 0.7473703638144439,
+      "learning_rate": 7.0118749122847e-06,
+      "loss": 1.2813,
+      "step": 1081
+    },
+    {
+      "epoch": 0.733870284018652,
+      "grad_norm": 0.7159820456458268,
+      "learning_rate": 6.978470386116445e-06,
+      "loss": 1.2572,
+      "step": 1082
+    },
+    {
+      "epoch": 0.7345485375158965,
+      "grad_norm": 0.694451155338397,
+      "learning_rate": 6.945128795655283e-06,
+      "loss": 1.247,
+      "step": 1083
+    },
+    {
+      "epoch": 0.7352267910131411,
+      "grad_norm": 0.6686217234195255,
+      "learning_rate": 6.911850302047893e-06,
+      "loss": 1.2527,
+      "step": 1084
+    },
+    {
+      "epoch": 0.7359050445103857,
+      "grad_norm": 0.7154649690871586,
+      "learning_rate": 6.878635066136032e-06,
+      "loss": 1.2548,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7365832980076303,
+      "grad_norm": 0.7025619787414288,
+      "learning_rate": 6.845483248455711e-06,
+      "loss": 1.2515,
+      "step": 1086
+    },
+    {
+      "epoch": 0.7372615515048749,
+      "grad_norm": 0.6904241313849837,
+      "learning_rate": 6.812395009236416e-06,
+      "loss": 1.2605,
+      "step": 1087
+    },
+    {
+      "epoch": 0.7379398050021195,
+      "grad_norm": 0.7067524261458777,
+      "learning_rate": 6.779370508400356e-06,
+      "loss": 1.2371,
+      "step": 1088
+    },
+    {
+      "epoch": 0.7386180584993641,
+      "grad_norm": 0.6998978069695649,
+      "learning_rate": 6.7464099055616635e-06,
+      "loss": 1.2396,
+      "step": 1089
+    },
+    {
+      "epoch": 0.7392963119966087,
+      "grad_norm": 0.7127235331064138,
+      "learning_rate": 6.713513360025667e-06,
+      "loss": 1.2826,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7399745654938533,
+      "grad_norm": 0.709084315722468,
+      "learning_rate": 6.680681030788072e-06,
+      "loss": 1.2685,
+      "step": 1091
+    },
+    {
+      "epoch": 0.7406528189910979,
+      "grad_norm": 0.7439738397680048,
+      "learning_rate": 6.6479130765342185e-06,
+      "loss": 1.2548,
+      "step": 1092
+    },
+    {
+      "epoch": 0.7413310724883425,
+      "grad_norm": 0.6913571968581045,
+      "learning_rate": 6.615209655638299e-06,
+      "loss": 1.2527,
+      "step": 1093
+    },
+    {
+      "epoch": 0.7420093259855871,
+      "grad_norm": 0.6695385106248662,
+      "learning_rate": 6.582570926162628e-06,
+      "loss": 1.245,
+      "step": 1094
+    },
+    {
+      "epoch": 0.7426875794828317,
+      "grad_norm": 0.7061064658406417,
+      "learning_rate": 6.549997045856835e-06,
+      "loss": 1.2422,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7433658329800763,
+      "grad_norm": 0.7014131205678311,
+      "learning_rate": 6.517488172157113e-06,
+      "loss": 1.2609,
+      "step": 1096
+    },
+    {
+      "epoch": 0.7440440864773209,
+      "grad_norm": 0.6878508712453276,
+      "learning_rate": 6.485044462185492e-06,
+      "loss": 1.2718,
+      "step": 1097
+    },
+    {
+      "epoch": 0.7447223399745655,
+      "grad_norm": 0.6677191346667652,
+      "learning_rate": 6.452666072749028e-06,
+      "loss": 1.2492,
+      "step": 1098
+    },
+    {
+      "epoch": 0.7454005934718101,
+      "grad_norm": 0.7090961144596659,
+      "learning_rate": 6.4203531603390765e-06,
+      "loss": 1.2373,
+      "step": 1099
+    },
+    {
+      "epoch": 0.7460788469690547,
+      "grad_norm": 0.45255204902887586,
+      "learning_rate": 6.3881058811305264e-06,
+      "loss": 1.3588,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7467571004662993,
+      "grad_norm": 0.7135393080533664,
+      "learning_rate": 6.355924390981061e-06,
+      "loss": 1.2277,
+      "step": 1101
+    },
+    {
+      "epoch": 0.7474353539635439,
+      "grad_norm": 0.4458247664684505,
+      "learning_rate": 6.323808845430379e-06,
+      "loss": 1.377,
+      "step": 1102
+    },
+    {
+      "epoch": 0.7481136074607885,
+      "grad_norm": 0.7185164012176813,
+      "learning_rate": 6.29175939969945e-06,
+      "loss": 1.2373,
+      "step": 1103
+    },
+    {
+      "epoch": 0.7487918609580331,
+      "grad_norm": 0.6982441752973358,
+      "learning_rate": 6.259776208689796e-06,
+      "loss": 1.2612,
+      "step": 1104
+    },
+    {
+      "epoch": 0.7494701144552777,
+      "grad_norm": 0.6953929763573126,
+      "learning_rate": 6.227859426982688e-06,
+      "loss": 1.2782,
+      "step": 1105
+    },
+    {
+      "epoch": 0.7501483679525223,
+      "grad_norm": 0.697754814965526,
+      "learning_rate": 6.196009208838438e-06,
+      "loss": 1.2536,
+      "step": 1106
+    },
+    {
+      "epoch": 0.7508266214497669,
+      "grad_norm": 0.6748304634062791,
+      "learning_rate": 6.164225708195642e-06,
+      "loss": 1.2523,
+      "step": 1107
+    },
+    {
+      "epoch": 0.7515048749470115,
+      "grad_norm": 0.6782055822317078,
+      "learning_rate": 6.132509078670437e-06,
+      "loss": 1.239,
+      "step": 1108
+    },
+    {
+      "epoch": 0.752183128444256,
+      "grad_norm": 0.6808116233647222,
+      "learning_rate": 6.100859473555776e-06,
+      "loss": 1.2442,
+      "step": 1109
+    },
+    {
+      "epoch": 0.7528613819415007,
+      "grad_norm": 0.7036659853403976,
+      "learning_rate": 6.069277045820625e-06,
+      "loss": 1.2323,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7535396354387452,
+      "grad_norm": 0.6828322290009959,
+      "learning_rate": 6.037761948109318e-06,
+      "loss": 1.2374,
+      "step": 1111
+    },
+    {
+      "epoch": 0.7542178889359898,
+      "grad_norm": 0.713945854741471,
+      "learning_rate": 6.006314332740735e-06,
+      "loss": 1.2603,
+      "step": 1112
+    },
+    {
+      "epoch": 0.7548961424332344,
+      "grad_norm": 0.7020795096881565,
+      "learning_rate": 5.9749343517076155e-06,
+      "loss": 1.2399,
+      "step": 1113
+    },
+    {
+      "epoch": 0.755574395930479,
+      "grad_norm": 0.6998554795750904,
+      "learning_rate": 5.943622156675799e-06,
+      "loss": 1.2494,
+      "step": 1114
+    },
+    {
+      "epoch": 0.7562526494277236,
+      "grad_norm": 0.7122693984226326,
+      "learning_rate": 5.91237789898351e-06,
+      "loss": 1.2505,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7569309029249682,
+      "grad_norm": 0.7314913158727478,
+      "learning_rate": 5.881201729640629e-06,
+      "loss": 1.2491,
+      "step": 1116
+    },
+    {
+      "epoch": 0.7576091564222128,
+      "grad_norm": 0.6878164452463644,
+      "learning_rate": 5.850093799327914e-06,
+      "loss": 1.2419,
+      "step": 1117
+    },
+    {
+      "epoch": 0.7582874099194574,
+      "grad_norm": 0.7070919917450872,
+      "learning_rate": 5.81905425839635e-06,
+      "loss": 1.2535,
+      "step": 1118
+    },
+    {
+      "epoch": 0.758965663416702,
+      "grad_norm": 0.4473660675198994,
+      "learning_rate": 5.788083256866357e-06,
+      "loss": 1.3981,
+      "step": 1119
+    },
+    {
+      "epoch": 0.7596439169139466,
+      "grad_norm": 0.6899640026759862,
+      "learning_rate": 5.757180944427115e-06,
+      "loss": 1.223,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7603221704111912,
+      "grad_norm": 0.7150514267374936,
+      "learning_rate": 5.7263474704357715e-06,
+      "loss": 1.2355,
+      "step": 1121
+    },
+    {
+      "epoch": 0.7610004239084358,
+      "grad_norm": 0.7389898022126541,
+      "learning_rate": 5.6955829839168165e-06,
+      "loss": 1.2777,
+      "step": 1122
+    },
+    {
+      "epoch": 0.7616786774056804,
+      "grad_norm": 0.6796754729134707,
+      "learning_rate": 5.664887633561269e-06,
+      "loss": 1.2135,
+      "step": 1123
+    },
+    {
+      "epoch": 0.762356930902925,
+      "grad_norm": 0.6900893470393483,
+      "learning_rate": 5.63426156772603e-06,
+      "loss": 1.2354,
+      "step": 1124
+    },
+    {
+      "epoch": 0.7630351844001696,
+      "grad_norm": 0.7012314507491941,
+      "learning_rate": 5.603704934433107e-06,
+      "loss": 1.2196,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7637134378974142,
+      "grad_norm": 0.7056609640562619,
+      "learning_rate": 5.573217881368936e-06,
+      "loss": 1.2807,
+      "step": 1126
+    },
+    {
+      "epoch": 0.7643916913946588,
+      "grad_norm": 0.6769326414922961,
+      "learning_rate": 5.54280055588367e-06,
+      "loss": 1.2424,
+      "step": 1127
+    },
+    {
+      "epoch": 0.7650699448919034,
+      "grad_norm": 0.6758653545849139,
+      "learning_rate": 5.5124531049904385e-06,
+      "loss": 1.2471,
+      "step": 1128
+    },
+    {
+      "epoch": 0.765748198389148,
+      "grad_norm": 0.6630169402173475,
+      "learning_rate": 5.4821756753646584e-06,
+      "loss": 1.2248,
+      "step": 1129
+    },
+    {
+      "epoch": 0.7664264518863926,
+      "grad_norm": 0.6716307852815654,
+      "learning_rate": 5.451968413343309e-06,
+      "loss": 1.2381,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7671047053836372,
+      "grad_norm": 0.6890962976180042,
+      "learning_rate": 5.421831464924263e-06,
+      "loss": 1.2713,
+      "step": 1131
+    },
+    {
+      "epoch": 0.7677829588808818,
+      "grad_norm": 0.6962004345825891,
+      "learning_rate": 5.3917649757655275e-06,
+      "loss": 1.2561,
+      "step": 1132
+    },
+    {
+      "epoch": 0.7684612123781264,
+      "grad_norm": 0.7023013942987391,
+      "learning_rate": 5.361769091184566e-06,
+      "loss": 1.22,
+      "step": 1133
+    },
+    {
+      "epoch": 0.769139465875371,
+      "grad_norm": 0.6767598844140867,
+      "learning_rate": 5.3318439561576186e-06,
+      "loss": 1.2404,
+      "step": 1134
+    },
+    {
+      "epoch": 0.7698177193726156,
+      "grad_norm": 0.6771636851434627,
+      "learning_rate": 5.301989715318954e-06,
+      "loss": 1.2369,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7704959728698602,
+      "grad_norm": 0.6971245037688651,
+      "learning_rate": 5.272206512960205e-06,
+      "loss": 1.2324,
+      "step": 1136
+    },
+    {
+      "epoch": 0.7711742263671048,
+      "grad_norm": 0.6765112279690619,
+      "learning_rate": 5.242494493029655e-06,
+      "loss": 1.2217,
+      "step": 1137
+    },
+    {
+      "epoch": 0.7718524798643494,
+      "grad_norm": 0.6737178423734921,
+      "learning_rate": 5.212853799131566e-06,
+      "loss": 1.2588,
+      "step": 1138
+    },
+    {
+      "epoch": 0.772530733361594,
+      "grad_norm": 0.6798938652668959,
+      "learning_rate": 5.183284574525444e-06,
+      "loss": 1.2555,
+      "step": 1139
+    },
+    {
+      "epoch": 0.7732089868588385,
+      "grad_norm": 0.4647417772652952,
+      "learning_rate": 5.1537869621253774e-06,
+      "loss": 1.4124,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7738872403560831,
+      "grad_norm": 0.7075653241795719,
+      "learning_rate": 5.124361104499349e-06,
+      "loss": 1.2511,
+      "step": 1141
+    },
+    {
+      "epoch": 0.7745654938533277,
+      "grad_norm": 0.7135593274232547,
+      "learning_rate": 5.095007143868522e-06,
+      "loss": 1.2313,
+      "step": 1142
+    },
+    {
+      "epoch": 0.7752437473505722,
+      "grad_norm": 0.4479664062052932,
+      "learning_rate": 5.065725222106574e-06,
+      "loss": 1.3455,
+      "step": 1143
+    },
+    {
+      "epoch": 0.7759220008478168,
+      "grad_norm": 0.6998180607890737,
+      "learning_rate": 5.036515480738995e-06,
+      "loss": 1.2638,
+      "step": 1144
+    },
+    {
+      "epoch": 0.7766002543450614,
+      "grad_norm": 0.6780873565049669,
+      "learning_rate": 5.007378060942425e-06,
+      "loss": 1.2585,
+      "step": 1145
+    },
+    {
+      "epoch": 0.777278507842306,
+      "grad_norm": 0.6951550032542391,
+      "learning_rate": 4.978313103543964e-06,
+      "loss": 1.2609,
+      "step": 1146
+    },
+    {
+      "epoch": 0.7779567613395506,
+      "grad_norm": 0.43274514832271704,
+      "learning_rate": 4.949320749020454e-06,
+      "loss": 1.353,
+      "step": 1147
+    },
+    {
+      "epoch": 0.7786350148367952,
+      "grad_norm": 0.6959484335867067,
+      "learning_rate": 4.920401137497872e-06,
+      "loss": 1.226,
+      "step": 1148
+    },
+    {
+      "epoch": 0.7793132683340398,
+      "grad_norm": 0.6818736987116557,
+      "learning_rate": 4.891554408750585e-06,
+      "loss": 1.261,
+      "step": 1149
+    },
+    {
+      "epoch": 0.7799915218312844,
+      "grad_norm": 0.6928424892107953,
+      "learning_rate": 4.862780702200729e-06,
+      "loss": 1.2267,
+      "step": 1150
+    },
+    {
+      "epoch": 0.780669775328529,
+      "grad_norm": 0.6754834335451715,
+      "learning_rate": 4.8340801569174735e-06,
+      "loss": 1.2191,
+      "step": 1151
+    },
+    {
+      "epoch": 0.7813480288257736,
+      "grad_norm": 0.7004613308457371,
+      "learning_rate": 4.805452911616417e-06,
+      "loss": 1.2375,
+      "step": 1152
+    },
+    {
+      "epoch": 0.7820262823230182,
+      "grad_norm": 0.6806930381395735,
+      "learning_rate": 4.77689910465887e-06,
+      "loss": 1.229,
+      "step": 1153
+    },
+    {
+      "epoch": 0.7827045358202628,
+      "grad_norm": 0.6774321121177972,
+      "learning_rate": 4.748418874051195e-06,
+      "loss": 1.2545,
+      "step": 1154
+    },
+    {
+      "epoch": 0.7833827893175074,
+      "grad_norm": 0.6948812717558227,
+      "learning_rate": 4.720012357444162e-06,
+      "loss": 1.2469,
+      "step": 1155
+    },
+    {
+      "epoch": 0.784061042814752,
+      "grad_norm": 0.6889011025050227,
+      "learning_rate": 4.691679692132247e-06,
+      "loss": 1.226,
+      "step": 1156
+    },
+    {
+      "epoch": 0.7847392963119966,
+      "grad_norm": 0.6679127571883847,
+      "learning_rate": 4.663421015053016e-06,
+      "loss": 1.2602,
+      "step": 1157
+    },
+    {
+      "epoch": 0.7854175498092412,
+      "grad_norm": 0.6579398726491081,
+      "learning_rate": 4.63523646278639e-06,
+      "loss": 1.2256,
+      "step": 1158
+    },
+    {
+      "epoch": 0.7860958033064858,
+      "grad_norm": 0.6688306932896388,
+      "learning_rate": 4.607126171554075e-06,
+      "loss": 1.2536,
+      "step": 1159
+    },
+    {
+      "epoch": 0.7867740568037304,
+      "grad_norm": 0.7004512005802537,
+      "learning_rate": 4.579090277218825e-06,
+      "loss": 1.2502,
+      "step": 1160
+    },
+    {
+      "epoch": 0.787452310300975,
+      "grad_norm": 0.6694251736984261,
+      "learning_rate": 4.5511289152838444e-06,
+      "loss": 1.2204,
+      "step": 1161
+    },
+    {
+      "epoch": 0.7881305637982196,
+      "grad_norm": 0.6843283642295125,
+      "learning_rate": 4.523242220892092e-06,
+      "loss": 1.2483,
+      "step": 1162
+    },
+    {
+      "epoch": 0.7888088172954641,
+      "grad_norm": 0.7067471312697848,
+      "learning_rate": 4.495430328825639e-06,
+      "loss": 1.253,
+      "step": 1163
+    },
+    {
+      "epoch": 0.7894870707927087,
+      "grad_norm": 0.6817423205506041,
+      "learning_rate": 4.46769337350504e-06,
+      "loss": 1.235,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7901653242899533,
+      "grad_norm": 0.7247217035269676,
+      "learning_rate": 4.440031488988647e-06,
+      "loss": 1.2366,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7908435777871979,
+      "grad_norm": 0.6579826200974558,
+      "learning_rate": 4.412444808971994e-06,
+      "loss": 1.2285,
+      "step": 1166
+    },
+    {
+      "epoch": 0.7915218312844425,
+      "grad_norm": 0.667181720898605,
+      "learning_rate": 4.384933466787116e-06,
+      "loss": 1.2345,
+      "step": 1167
+    },
+    {
+      "epoch": 0.7922000847816871,
+      "grad_norm": 0.6899309881903906,
+      "learning_rate": 4.357497595401954e-06,
+      "loss": 1.2662,
+      "step": 1168
+    },
+    {
+      "epoch": 0.7928783382789317,
+      "grad_norm": 0.7080472474113639,
+      "learning_rate": 4.330137327419656e-06,
+      "loss": 1.2547,
+      "step": 1169
+    },
+    {
+      "epoch": 0.7935565917761763,
+      "grad_norm": 0.6877701187705579,
+      "learning_rate": 4.302852795077976e-06,
+      "loss": 1.2547,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7942348452734209,
+      "grad_norm": 0.7248362233029715,
+      "learning_rate": 4.275644130248629e-06,
+      "loss": 1.2631,
+      "step": 1171
+    },
+    {
+      "epoch": 0.7949130987706655,
+      "grad_norm": 0.6786411729582461,
+      "learning_rate": 4.248511464436629e-06,
+      "loss": 1.2471,
+      "step": 1172
+    },
+    {
+      "epoch": 0.7955913522679101,
+      "grad_norm": 0.6759310563754385,
+      "learning_rate": 4.221454928779687e-06,
+      "loss": 1.225,
+      "step": 1173
+    },
+    {
+      "epoch": 0.7962696057651547,
+      "grad_norm": 0.6934587765306394,
+      "learning_rate": 4.1944746540475465e-06,
+      "loss": 1.2548,
+      "step": 1174
+    },
+    {
+      "epoch": 0.7969478592623993,
+      "grad_norm": 0.6521853026195695,
+      "learning_rate": 4.167570770641387e-06,
+      "loss": 1.2171,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7976261127596439,
+      "grad_norm": 0.695071649092735,
+      "learning_rate": 4.140743408593158e-06,
+      "loss": 1.2618,
+      "step": 1176
+    },
+    {
+      "epoch": 0.7983043662568885,
+      "grad_norm": 0.6946974156403914,
+      "learning_rate": 4.113992697564959e-06,
+      "loss": 1.2202,
+      "step": 1177
+    },
+    {
+      "epoch": 0.7989826197541331,
+      "grad_norm": 0.6875018646078283,
+      "learning_rate": 4.0873187668484444e-06,
+      "loss": 1.2475,
+      "step": 1178
+    },
+    {
+      "epoch": 0.7996608732513777,
+      "grad_norm": 0.6837968878391059,
+      "learning_rate": 4.060721745364153e-06,
+      "loss": 1.2521,
+      "step": 1179
+    },
+    {
+      "epoch": 0.8003391267486223,
+      "grad_norm": 0.6999303464068111,
+      "learning_rate": 4.0342017616609095e-06,
+      "loss": 1.2452,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8010173802458669,
+      "grad_norm": 0.6700378318277015,
+      "learning_rate": 4.007758943915197e-06,
+      "loss": 1.2421,
+      "step": 1181
+    },
+    {
+      "epoch": 0.8016956337431115,
+      "grad_norm": 0.6818739644010207,
+      "learning_rate": 3.981393419930555e-06,
+      "loss": 1.2235,
+      "step": 1182
+    },
+    {
+      "epoch": 0.8023738872403561,
+      "grad_norm": 0.6793140252226033,
+      "learning_rate": 3.955105317136929e-06,
+      "loss": 1.2663,
+      "step": 1183
+    },
+    {
+      "epoch": 0.8030521407376007,
+      "grad_norm": 0.6875039701252489,
+      "learning_rate": 3.928894762590076e-06,
+      "loss": 1.2192,
+      "step": 1184
+    },
+    {
+      "epoch": 0.8037303942348453,
+      "grad_norm": 0.711408992755773,
+      "learning_rate": 3.902761882970958e-06,
+      "loss": 1.2628,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8044086477320899,
+      "grad_norm": 0.7195784285174972,
+      "learning_rate": 3.8767068045850975e-06,
+      "loss": 1.2556,
+      "step": 1186
+    },
+    {
+      "epoch": 0.8050869012293345,
+      "grad_norm": 0.7028743882976222,
+      "learning_rate": 3.850729653362018e-06,
+      "loss": 1.238,
+      "step": 1187
+    },
+    {
+      "epoch": 0.8057651547265791,
+      "grad_norm": 0.7195821566351972,
+      "learning_rate": 3.824830554854566e-06,
+      "loss": 1.2312,
+      "step": 1188
+    },
+    {
+      "epoch": 0.8064434082238237,
+      "grad_norm": 0.7077679764250616,
+      "learning_rate": 3.7990096342383775e-06,
+      "loss": 1.2487,
+      "step": 1189
+    },
+    {
+      "epoch": 0.8071216617210683,
+      "grad_norm": 0.6877957096422861,
+      "learning_rate": 3.773267016311215e-06,
+      "loss": 1.2407,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8077999152183128,
+      "grad_norm": 0.7134703340862909,
+      "learning_rate": 3.7476028254924115e-06,
+      "loss": 1.2524,
+      "step": 1191
+    },
+    {
+      "epoch": 0.8084781687155574,
+      "grad_norm": 0.65709662877186,
+      "learning_rate": 3.7220171858222264e-06,
+      "loss": 1.214,
+      "step": 1192
+    },
+    {
+      "epoch": 0.809156422212802,
+      "grad_norm": 0.6768556017907369,
+      "learning_rate": 3.6965102209612667e-06,
+      "loss": 1.241,
+      "step": 1193
+    },
+    {
+      "epoch": 0.8098346757100466,
+      "grad_norm": 0.6839758361669105,
+      "learning_rate": 3.6710820541899097e-06,
+      "loss": 1.2783,
+      "step": 1194
+    },
+    {
+      "epoch": 0.8105129292072912,
+      "grad_norm": 0.7094232562572493,
+      "learning_rate": 3.645732808407647e-06,
+      "loss": 1.2558,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8111911827045358,
+      "grad_norm": 0.6969915627207982,
+      "learning_rate": 3.6204626061325666e-06,
+      "loss": 1.2748,
+      "step": 1196
+    },
+    {
+      "epoch": 0.8118694362017804,
+      "grad_norm": 0.6971469387166436,
+      "learning_rate": 3.595271569500698e-06,
+      "loss": 1.2325,
+      "step": 1197
+    },
+    {
+      "epoch": 0.812547689699025,
+      "grad_norm": 0.6627792415071245,
+      "learning_rate": 3.570159820265464e-06,
+      "loss": 1.2179,
+      "step": 1198
+    },
+    {
+      "epoch": 0.8132259431962696,
+      "grad_norm": 0.6920530698070828,
+      "learning_rate": 3.545127479797068e-06,
+      "loss": 1.2169,
+      "step": 1199
+    },
+    {
+      "epoch": 0.8139041966935142,
+      "grad_norm": 0.7002031808124952,
+      "learning_rate": 3.520174669081904e-06,
+      "loss": 1.2683,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8145824501907588,
+      "grad_norm": 0.6637127831585506,
+      "learning_rate": 3.4953015087220043e-06,
+      "loss": 1.2278,
+      "step": 1201
+    },
+    {
+      "epoch": 0.8152607036880034,
+      "grad_norm": 0.6803390774543788,
+      "learning_rate": 3.4705081189344214e-06,
+      "loss": 1.236,
+      "step": 1202
+    },
+    {
+      "epoch": 0.815938957185248,
+      "grad_norm": 0.7062394197270683,
+      "learning_rate": 3.4457946195506576e-06,
+      "loss": 1.2286,
+      "step": 1203
+    },
+    {
+      "epoch": 0.8166172106824926,
+      "grad_norm": 0.7383625500117024,
+      "learning_rate": 3.421161130016093e-06,
+      "loss": 1.2587,
+      "step": 1204
+    },
+    {
+      "epoch": 0.8172954641797372,
+      "grad_norm": 0.6688377168685776,
+      "learning_rate": 3.3966077693894106e-06,
+      "loss": 1.237,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8179737176769818,
+      "grad_norm": 0.7132285176191728,
+      "learning_rate": 3.3721346563420033e-06,
+      "loss": 1.2448,
+      "step": 1206
+    },
+    {
+      "epoch": 0.8186519711742264,
+      "grad_norm": 0.6680523424443547,
+      "learning_rate": 3.3477419091574092e-06,
+      "loss": 1.2498,
+      "step": 1207
+    },
+    {
+      "epoch": 0.819330224671471,
+      "grad_norm": 0.6905435032148992,
+      "learning_rate": 3.3234296457307625e-06,
+      "loss": 1.2649,
+      "step": 1208
+    },
+    {
+      "epoch": 0.8200084781687156,
+      "grad_norm": 0.6926944873494554,
+      "learning_rate": 3.2991979835681788e-06,
+      "loss": 1.2572,
+      "step": 1209
+    },
+    {
+      "epoch": 0.8206867316659602,
+      "grad_norm": 0.6847897982257526,
+      "learning_rate": 3.2750470397862232e-06,
+      "loss": 1.233,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8213649851632048,
+      "grad_norm": 0.6737318091933269,
+      "learning_rate": 3.2509769311113227e-06,
+      "loss": 1.2455,
+      "step": 1211
+    },
+    {
+      "epoch": 0.8220432386604494,
+      "grad_norm": 0.6670064588219545,
+      "learning_rate": 3.226987773879233e-06,
+      "loss": 1.2395,
+      "step": 1212
+    },
+    {
+      "epoch": 0.822721492157694,
+      "grad_norm": 0.6800698653690069,
+      "learning_rate": 3.2030796840344335e-06,
+      "loss": 1.2046,
+      "step": 1213
+    },
+    {
+      "epoch": 0.8233997456549386,
+      "grad_norm": 0.6954571148384878,
+      "learning_rate": 3.1792527771295934e-06,
+      "loss": 1.2555,
+      "step": 1214
+    },
+    {
+      "epoch": 0.8240779991521832,
+      "grad_norm": 0.6800480785017151,
+      "learning_rate": 3.1555071683250183e-06,
+      "loss": 1.2576,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8247562526494278,
+      "grad_norm": 0.6731537475624323,
+      "learning_rate": 3.1318429723880705e-06,
+      "loss": 1.2664,
+      "step": 1216
+    },
+    {
+      "epoch": 0.8254345061466724,
+      "grad_norm": 0.6744093289438378,
+      "learning_rate": 3.1082603036926363e-06,
+      "loss": 1.2681,
+      "step": 1217
+    },
+    {
+      "epoch": 0.826112759643917,
+      "grad_norm": 0.66118151158709,
+      "learning_rate": 3.0847592762185563e-06,
+      "loss": 1.2482,
+      "step": 1218
+    },
+    {
+      "epoch": 0.8267910131411615,
+      "grad_norm": 0.7773444601932658,
+      "learning_rate": 3.061340003551092e-06,
+      "loss": 1.2329,
+      "step": 1219
+    },
+    {
+      "epoch": 0.8274692666384061,
+      "grad_norm": 0.6699375883984999,
+      "learning_rate": 3.038002598880363e-06,
+      "loss": 1.2248,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8281475201356507,
+      "grad_norm": 0.6754455023957769,
+      "learning_rate": 3.014747175000794e-06,
+      "loss": 1.2342,
+      "step": 1221
+    },
+    {
+      "epoch": 0.8288257736328953,
+      "grad_norm": 0.6738655402613765,
+      "learning_rate": 2.9915738443106e-06,
+      "loss": 1.2322,
+      "step": 1222
+    },
+    {
+      "epoch": 0.8295040271301399,
+      "grad_norm": 0.6775994450297527,
+      "learning_rate": 2.9684827188112054e-06,
+      "loss": 1.2306,
+      "step": 1223
+    },
+    {
+      "epoch": 0.8301822806273845,
+      "grad_norm": 0.6793157607402639,
+      "learning_rate": 2.9454739101067376e-06,
+      "loss": 1.2551,
+      "step": 1224
+    },
+    {
+      "epoch": 0.8308605341246291,
+      "grad_norm": 0.7046333525864542,
+      "learning_rate": 2.9225475294034434e-06,
+      "loss": 1.2199,
+      "step": 1225
+    },
+    {
+      "epoch": 0.8315387876218737,
+      "grad_norm": 0.6956025857176882,
+      "learning_rate": 2.8997036875092056e-06,
+      "loss": 1.2423,
+      "step": 1226
+    },
+    {
+      "epoch": 0.8322170411191183,
+      "grad_norm": 0.6932938866887467,
+      "learning_rate": 2.8769424948329617e-06,
+      "loss": 1.2414,
+      "step": 1227
+    },
+    {
+      "epoch": 0.8328952946163629,
+      "grad_norm": 0.6903105564714394,
+      "learning_rate": 2.8542640613842043e-06,
+      "loss": 1.2612,
+      "step": 1228
+    },
+    {
+      "epoch": 0.8335735481136075,
+      "grad_norm": 0.6640271856515038,
+      "learning_rate": 2.8316684967724216e-06,
+      "loss": 1.2215,
+      "step": 1229
+    },
+    {
+      "epoch": 0.8342518016108521,
+      "grad_norm": 0.6924264638696419,
+      "learning_rate": 2.8091559102065757e-06,
+      "loss": 1.2574,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8349300551080967,
+      "grad_norm": 0.6688585447783828,
+      "learning_rate": 2.7867264104946e-06,
+      "loss": 1.2621,
+      "step": 1231
+    },
+    {
+      "epoch": 0.8356083086053413,
+      "grad_norm": 0.6990612602692031,
+      "learning_rate": 2.764380106042832e-06,
+      "loss": 1.2473,
+      "step": 1232
+    },
+    {
+      "epoch": 0.8362865621025859,
+      "grad_norm": 0.6605622117162815,
+      "learning_rate": 2.7421171048555174e-06,
+      "loss": 1.2119,
+      "step": 1233
+    },
+    {
+      "epoch": 0.8369648155998305,
+      "grad_norm": 0.6798097456514123,
+      "learning_rate": 2.7199375145342723e-06,
+      "loss": 1.248,
+      "step": 1234
+    },
+    {
+      "epoch": 0.837643069097075,
+      "grad_norm": 0.6845646072497431,
+      "learning_rate": 2.6978414422775913e-06,
+      "loss": 1.249,
+      "step": 1235
+    },
+    {
+      "epoch": 0.8383213225943196,
+      "grad_norm": 0.4316045622329302,
+      "learning_rate": 2.6758289948802873e-06,
+      "loss": 1.3416,
+      "step": 1236
+    },
+    {
+      "epoch": 0.8389995760915642,
+      "grad_norm": 0.6920372772638658,
+      "learning_rate": 2.653900278733006e-06,
+      "loss": 1.2551,
+      "step": 1237
+    },
+    {
+      "epoch": 0.8396778295888088,
+      "grad_norm": 0.7074558805325211,
+      "learning_rate": 2.632055399821707e-06,
+      "loss": 1.2667,
+      "step": 1238
+    },
+    {
+      "epoch": 0.8403560830860534,
+      "grad_norm": 0.6764498414633098,
+      "learning_rate": 2.610294463727141e-06,
+      "loss": 1.2472,
+      "step": 1239
+    },
+    {
+      "epoch": 0.841034336583298,
+      "grad_norm": 0.6665093041129415,
+      "learning_rate": 2.588617575624346e-06,
+      "loss": 1.2397,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8417125900805426,
+      "grad_norm": 0.6719411187283858,
+      "learning_rate": 2.5670248402821416e-06,
+      "loss": 1.2681,
+      "step": 1241
+    },
+    {
+      "epoch": 0.8423908435777872,
+      "grad_norm": 0.6633310055649266,
+      "learning_rate": 2.545516362062623e-06,
+      "loss": 1.2174,
+      "step": 1242
+    },
+    {
+      "epoch": 0.8430690970750317,
+      "grad_norm": 0.7009012164351331,
+      "learning_rate": 2.5240922449206485e-06,
+      "loss": 1.2499,
+      "step": 1243
+    },
+    {
+      "epoch": 0.8437473505722763,
+      "grad_norm": 0.65658740687541,
+      "learning_rate": 2.5027525924033393e-06,
+      "loss": 1.179,
+      "step": 1244
+    },
+    {
+      "epoch": 0.8444256040695209,
+      "grad_norm": 0.4369524964249072,
+      "learning_rate": 2.4814975076495928e-06,
+      "loss": 1.4163,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8451038575667655,
+      "grad_norm": 0.6493205856586579,
+      "learning_rate": 2.460327093389563e-06,
+      "loss": 1.2481,
+      "step": 1246
+    },
+    {
+      "epoch": 0.8457821110640101,
+      "grad_norm": 0.6521636141722069,
+      "learning_rate": 2.4392414519441766e-06,
+      "loss": 1.2141,
+      "step": 1247
+    },
+    {
+      "epoch": 0.8464603645612547,
+      "grad_norm": 0.6986484333798653,
+      "learning_rate": 2.4182406852246353e-06,
+      "loss": 1.2385,
+      "step": 1248
+    },
+    {
+      "epoch": 0.8471386180584993,
+      "grad_norm": 0.6773567173483765,
+      "learning_rate": 2.3973248947319337e-06,
+      "loss": 1.2637,
+      "step": 1249
+    },
+    {
+      "epoch": 0.8478168715557439,
+      "grad_norm": 0.7222705415612313,
+      "learning_rate": 2.3764941815563456e-06,
+      "loss": 1.2504,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8484951250529885,
+      "grad_norm": 0.663347159603606,
+      "learning_rate": 2.355748646376952e-06,
+      "loss": 1.2924,
+      "step": 1251
+    },
+    {
+      "epoch": 0.8491733785502331,
+      "grad_norm": 0.672155101613964,
+      "learning_rate": 2.3350883894611574e-06,
+      "loss": 1.2404,
+      "step": 1252
+    },
+    {
+      "epoch": 0.8498516320474777,
+      "grad_norm": 0.6724349931683334,
+      "learning_rate": 2.314513510664196e-06,
+      "loss": 1.2531,
+      "step": 1253
+    },
+    {
+      "epoch": 0.8505298855447223,
+      "grad_norm": 0.654967406341385,
+      "learning_rate": 2.2940241094286475e-06,
+      "loss": 1.2368,
+      "step": 1254
+    },
+    {
+      "epoch": 0.8512081390419669,
+      "grad_norm": 0.7050572649979452,
+      "learning_rate": 2.2736202847839616e-06,
+      "loss": 1.2503,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8518863925392115,
+      "grad_norm": 0.6593550166493525,
+      "learning_rate": 2.2533021353459917e-06,
+      "loss": 1.2496,
+      "step": 1256
+    },
+    {
+      "epoch": 0.8525646460364561,
+      "grad_norm": 0.6797192175460736,
+      "learning_rate": 2.233069759316491e-06,
+      "loss": 1.2683,
+      "step": 1257
+    },
+    {
+      "epoch": 0.8532428995337007,
+      "grad_norm": 0.6594627663012708,
+      "learning_rate": 2.212923254482653e-06,
+      "loss": 1.2374,
+      "step": 1258
+    },
+    {
+      "epoch": 0.8539211530309453,
+      "grad_norm": 0.6790072404571235,
+      "learning_rate": 2.1928627182166527e-06,
+      "loss": 1.2256,
+      "step": 1259
+    },
+    {
+      "epoch": 0.8545994065281899,
+      "grad_norm": 0.6531785829339144,
+      "learning_rate": 2.17288824747514e-06,
+      "loss": 1.246,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8552776600254345,
+      "grad_norm": 0.6697866159646256,
+      "learning_rate": 2.1529999387988164e-06,
+      "loss": 1.2297,
+      "step": 1261
+    },
+    {
+      "epoch": 0.8559559135226791,
+      "grad_norm": 0.6883522445090433,
+      "learning_rate": 2.1331978883119175e-06,
+      "loss": 1.2578,
+      "step": 1262
+    },
+    {
+      "epoch": 0.8566341670199237,
+      "grad_norm": 0.6489524920906364,
+      "learning_rate": 2.113482191721801e-06,
+      "loss": 1.2351,
+      "step": 1263
+    },
+    {
+      "epoch": 0.8573124205171683,
+      "grad_norm": 0.6689527801097199,
+      "learning_rate": 2.0938529443184395e-06,
+      "loss": 1.2589,
+      "step": 1264
+    },
+    {
+      "epoch": 0.8579906740144129,
+      "grad_norm": 0.6636152186845068,
+      "learning_rate": 2.0743102409739956e-06,
+      "loss": 1.2717,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8586689275116575,
+      "grad_norm": 0.643077165344515,
+      "learning_rate": 2.0548541761423335e-06,
+      "loss": 1.2472,
+      "step": 1266
+    },
+    {
+      "epoch": 0.8593471810089021,
+      "grad_norm": 0.6733627311390646,
+      "learning_rate": 2.0354848438585793e-06,
+      "loss": 1.2518,
+      "step": 1267
+    },
+    {
+      "epoch": 0.8600254345061467,
+      "grad_norm": 0.6500042790933303,
+      "learning_rate": 2.0162023377386684e-06,
+      "loss": 1.2392,
+      "step": 1268
+    },
+    {
+      "epoch": 0.8607036880033913,
+      "grad_norm": 0.6711783818853033,
+      "learning_rate": 1.9970067509788828e-06,
+      "loss": 1.2434,
+      "step": 1269
+    },
+    {
+      "epoch": 0.8613819415006359,
+      "grad_norm": 0.6878334385995437,
+      "learning_rate": 1.977898176355404e-06,
+      "loss": 1.2482,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8620601949978804,
+      "grad_norm": 0.6732797757524392,
+      "learning_rate": 1.9588767062238666e-06,
+      "loss": 1.2258,
+      "step": 1271
+    },
+    {
+      "epoch": 0.862738448495125,
+      "grad_norm": 0.690590706308697,
+      "learning_rate": 1.939942432518922e-06,
+      "loss": 1.2454,
+      "step": 1272
+    },
+    {
+      "epoch": 0.8634167019923696,
+      "grad_norm": 0.6742559709870798,
+      "learning_rate": 1.921095446753767e-06,
+      "loss": 1.2544,
+      "step": 1273
+    },
+    {
+      "epoch": 0.8640949554896142,
+      "grad_norm": 0.6812447470627214,
+      "learning_rate": 1.9023358400197267e-06,
+      "loss": 1.2221,
+      "step": 1274
+    },
+    {
+      "epoch": 0.8647732089868588,
+      "grad_norm": 0.6651274724851445,
+      "learning_rate": 1.8836637029858073e-06,
+      "loss": 1.2365,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8654514624841034,
+      "grad_norm": 0.6968195379849065,
+      "learning_rate": 1.8650791258982525e-06,
+      "loss": 1.2246,
+      "step": 1276
+    },
+    {
+      "epoch": 0.866129715981348,
+      "grad_norm": 0.6887124005022628,
+      "learning_rate": 1.8465821985801113e-06,
+      "loss": 1.2464,
+      "step": 1277
+    },
+    {
+      "epoch": 0.8668079694785926,
+      "grad_norm": 0.677103488304023,
+      "learning_rate": 1.8281730104308027e-06,
+      "loss": 1.2849,
+      "step": 1278
+    },
+    {
+      "epoch": 0.8674862229758372,
+      "grad_norm": 0.6806202048462158,
+      "learning_rate": 1.809851650425689e-06,
+      "loss": 1.2361,
+      "step": 1279
+    },
+    {
+      "epoch": 0.8681644764730818,
+      "grad_norm": 0.6594204417492138,
+      "learning_rate": 1.7916182071156352e-06,
+      "loss": 1.2087,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8688427299703264,
+      "grad_norm": 0.6775538501405443,
+      "learning_rate": 1.7734727686265896e-06,
+      "loss": 1.2521,
+      "step": 1281
+    },
+    {
+      "epoch": 0.869520983467571,
+      "grad_norm": 0.6732775057867245,
+      "learning_rate": 1.7554154226591591e-06,
+      "loss": 1.2443,
+      "step": 1282
+    },
+    {
+      "epoch": 0.8701992369648156,
+      "grad_norm": 0.670136546211293,
+      "learning_rate": 1.7374462564881734e-06,
+      "loss": 1.2578,
+      "step": 1283
+    },
+    {
+      "epoch": 0.8708774904620602,
+      "grad_norm": 0.6825310488125489,
+      "learning_rate": 1.7195653569622806e-06,
+      "loss": 1.2209,
+      "step": 1284
+    },
+    {
+      "epoch": 0.8715557439593048,
+      "grad_norm": 0.6779073695972031,
+      "learning_rate": 1.7017728105035037e-06,
+      "loss": 1.2423,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8722339974565494,
+      "grad_norm": 0.7023270067247336,
+      "learning_rate": 1.684068703106858e-06,
+      "loss": 1.2305,
+      "step": 1286
+    },
+    {
+      "epoch": 0.872912250953794,
+      "grad_norm": 0.6567892230707757,
+      "learning_rate": 1.666453120339897e-06,
+      "loss": 1.2317,
+      "step": 1287
+    },
+    {
+      "epoch": 0.8735905044510386,
+      "grad_norm": 0.6761140088046478,
+      "learning_rate": 1.6489261473423246e-06,
+      "loss": 1.2294,
+      "step": 1288
+    },
+    {
+      "epoch": 0.8742687579482832,
+      "grad_norm": 0.6897471789058747,
+      "learning_rate": 1.6314878688255742e-06,
+      "loss": 1.2455,
+      "step": 1289
+    },
+    {
+      "epoch": 0.8749470114455278,
+      "grad_norm": 0.6552396693987875,
+      "learning_rate": 1.6141383690724e-06,
+      "loss": 1.2306,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8756252649427724,
+      "grad_norm": 0.6606878888166857,
+      "learning_rate": 1.596877731936477e-06,
+      "loss": 1.2483,
+      "step": 1291
+    },
+    {
+      "epoch": 0.876303518440017,
+      "grad_norm": 0.6849908384639455,
+      "learning_rate": 1.579706040841973e-06,
+      "loss": 1.2554,
+      "step": 1292
+    },
+    {
+      "epoch": 0.8769817719372616,
+      "grad_norm": 0.6602215845249478,
+      "learning_rate": 1.5626233787831791e-06,
+      "loss": 1.2384,
+      "step": 1293
+    },
+    {
+      "epoch": 0.8776600254345062,
+      "grad_norm": 0.6611842345331809,
+      "learning_rate": 1.54562982832408e-06,
+      "loss": 1.237,
+      "step": 1294
+    },
+    {
+      "epoch": 0.8783382789317508,
+      "grad_norm": 0.6656283658878175,
+      "learning_rate": 1.5287254715979672e-06,
+      "loss": 1.2548,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8790165324289954,
+      "grad_norm": 0.6829406117992067,
+      "learning_rate": 1.5119103903070476e-06,
+      "loss": 1.2509,
+      "step": 1296
+    },
+    {
+      "epoch": 0.87969478592624,
+      "grad_norm": 0.43142912689574664,
+      "learning_rate": 1.4951846657220336e-06,
+      "loss": 1.3619,
+      "step": 1297
+    },
+    {
+      "epoch": 0.8803730394234845,
+      "grad_norm": 0.6636443493647105,
+      "learning_rate": 1.4785483786817678e-06,
+      "loss": 1.2532,
+      "step": 1298
+    },
+    {
+      "epoch": 0.8810512929207291,
+      "grad_norm": 0.6409186280150971,
+      "learning_rate": 1.462001609592807e-06,
+      "loss": 1.2439,
+      "step": 1299
+    },
+    {
+      "epoch": 0.8817295464179737,
+      "grad_norm": 0.6567127599805216,
+      "learning_rate": 1.4455444384290652e-06,
+      "loss": 1.2565,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8824077999152183,
+      "grad_norm": 0.6586436635118995,
+      "learning_rate": 1.429176944731403e-06,
+      "loss": 1.2469,
+      "step": 1301
+    },
+    {
+      "epoch": 0.8830860534124629,
+      "grad_norm": 0.701738944563103,
+      "learning_rate": 1.412899207607259e-06,
+      "loss": 1.2422,
+      "step": 1302
+    },
+    {
+      "epoch": 0.8837643069097075,
+      "grad_norm": 0.447188260094171,
+      "learning_rate": 1.3967113057302495e-06,
+      "loss": 1.3799,
+      "step": 1303
+    },
+    {
+      "epoch": 0.8844425604069521,
+      "grad_norm": 0.6751666808079548,
+      "learning_rate": 1.3806133173398028e-06,
+      "loss": 1.2505,
+      "step": 1304
+    },
+    {
+      "epoch": 0.8851208139041967,
+      "grad_norm": 0.6724348276625257,
+      "learning_rate": 1.3646053202407861e-06,
+      "loss": 1.2424,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8857990674014413,
+      "grad_norm": 0.7038026791021831,
+      "learning_rate": 1.3486873918031096e-06,
+      "loss": 1.248,
+      "step": 1306
+    },
+    {
+      "epoch": 0.8864773208986859,
+      "grad_norm": 0.6493663173977285,
+      "learning_rate": 1.332859608961361e-06,
+      "loss": 1.2563,
+      "step": 1307
+    },
+    {
+      "epoch": 0.8871555743959305,
+      "grad_norm": 0.6761645748668302,
+      "learning_rate": 1.3171220482144452e-06,
+      "loss": 1.2366,
+      "step": 1308
+    },
+    {
+      "epoch": 0.8878338278931751,
+      "grad_norm": 0.6543856339684542,
+      "learning_rate": 1.301474785625203e-06,
+      "loss": 1.228,
+      "step": 1309
+    },
+    {
+      "epoch": 0.8885120813904197,
+      "grad_norm": 0.6736409903221909,
+      "learning_rate": 1.2859178968200437e-06,
+      "loss": 1.2555,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8891903348876643,
+      "grad_norm": 0.7204894887302865,
+      "learning_rate": 1.2704514569885773e-06,
+      "loss": 1.2738,
+      "step": 1311
+    },
+    {
+      "epoch": 0.8898685883849089,
+      "grad_norm": 0.6585100428910327,
+      "learning_rate": 1.255075540883266e-06,
+      "loss": 1.2529,
+      "step": 1312
+    },
+    {
+      "epoch": 0.8905468418821535,
+      "grad_norm": 0.6988673581133306,
+      "learning_rate": 1.2397902228190483e-06,
+      "loss": 1.232,
+      "step": 1313
+    },
+    {
+      "epoch": 0.8912250953793981,
+      "grad_norm": 0.6808638632254461,
+      "learning_rate": 1.2245955766729757e-06,
+      "loss": 1.2578,
+      "step": 1314
+    },
+    {
+      "epoch": 0.8919033488766427,
+      "grad_norm": 0.6931340674170225,
+      "learning_rate": 1.2094916758838715e-06,
+      "loss": 1.2652,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8925816023738873,
+      "grad_norm": 0.6728575515822607,
+      "learning_rate": 1.194478593451973e-06,
+      "loss": 1.2484,
+      "step": 1316
+    },
+    {
+      "epoch": 0.8932598558711319,
+      "grad_norm": 0.6808867615629325,
+      "learning_rate": 1.1795564019385642e-06,
+      "loss": 1.2242,
+      "step": 1317
+    },
+    {
+      "epoch": 0.8939381093683765,
+      "grad_norm": 0.6881827380698176,
+      "learning_rate": 1.1647251734656352e-06,
+      "loss": 1.2757,
+      "step": 1318
+    },
+    {
+      "epoch": 0.8946163628656211,
+      "grad_norm": 0.6730892890394782,
+      "learning_rate": 1.1499849797155438e-06,
+      "loss": 1.2477,
+      "step": 1319
+    },
+    {
+      "epoch": 0.8952946163628657,
+      "grad_norm": 0.6617266640968748,
+      "learning_rate": 1.1353358919306468e-06,
+      "loss": 1.2306,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8959728698601103,
+      "grad_norm": 0.6551174490404256,
+      "learning_rate": 1.1207779809129748e-06,
+      "loss": 1.246,
+      "step": 1321
+    },
+    {
+      "epoch": 0.8966511233573549,
+      "grad_norm": 0.6541381635552229,
+      "learning_rate": 1.1063113170238715e-06,
+      "loss": 1.2171,
+      "step": 1322
+    },
+    {
+      "epoch": 0.8973293768545995,
+      "grad_norm": 0.6624929688825112,
+      "learning_rate": 1.0919359701836818e-06,
+      "loss": 1.2357,
+      "step": 1323
+    },
+    {
+      "epoch": 0.898007630351844,
+      "grad_norm": 0.658560451189581,
+      "learning_rate": 1.0776520098713838e-06,
+      "loss": 1.2348,
+      "step": 1324
+    },
+    {
+      "epoch": 0.8986858838490887,
+      "grad_norm": 0.655584899768396,
+      "learning_rate": 1.063459505124267e-06,
+      "loss": 1.2534,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8993641373463331,
+      "grad_norm": 0.6523259218703032,
+      "learning_rate": 1.0493585245376048e-06,
+      "loss": 1.2545,
+      "step": 1326
+    },
+    {
+      "epoch": 0.9000423908435777,
+      "grad_norm": 0.6712031142934327,
+      "learning_rate": 1.0353491362643054e-06,
+      "loss": 1.2311,
+      "step": 1327
+    },
+    {
+      "epoch": 0.9007206443408223,
+      "grad_norm": 0.6929628155813327,
+      "learning_rate": 1.0214314080146082e-06,
+      "loss": 1.2531,
+      "step": 1328
+    },
+    {
+      "epoch": 0.9013988978380669,
+      "grad_norm": 0.6701710022956457,
+      "learning_rate": 1.0076054070557163e-06,
+      "loss": 1.2605,
+      "step": 1329
+    },
+    {
+      "epoch": 0.9020771513353115,
+      "grad_norm": 0.6677084081978719,
+      "learning_rate": 9.938712002115226e-07,
+      "loss": 1.2472,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9027554048325561,
+      "grad_norm": 0.6549213183466188,
+      "learning_rate": 9.802288538622417e-07,
+      "loss": 1.2471,
+      "step": 1331
+    },
+    {
+      "epoch": 0.9034336583298007,
+      "grad_norm": 0.6498290628304042,
+      "learning_rate": 9.666784339441216e-07,
+      "loss": 1.2187,
+      "step": 1332
+    },
+    {
+      "epoch": 0.9041119118270453,
+      "grad_norm": 0.6607792472753872,
+      "learning_rate": 9.532200059490959e-07,
+      "loss": 1.2299,
+      "step": 1333
+    },
+    {
+      "epoch": 0.9047901653242899,
+      "grad_norm": 0.6807448721243032,
+      "learning_rate": 9.398536349244947e-07,
+      "loss": 1.2378,
+      "step": 1334
+    },
+    {
+      "epoch": 0.9054684188215345,
+      "grad_norm": 0.6834284278293586,
+      "learning_rate": 9.265793854727189e-07,
+      "loss": 1.2641,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9061466723187791,
+      "grad_norm": 0.6561570607501847,
+      "learning_rate": 9.133973217509106e-07,
+      "loss": 1.2491,
+      "step": 1336
+    },
+    {
+      "epoch": 0.9068249258160237,
+      "grad_norm": 0.673378051921662,
+      "learning_rate": 9.003075074706791e-07,
+      "loss": 1.2394,
+      "step": 1337
+    },
+    {
+      "epoch": 0.9075031793132683,
+      "grad_norm": 0.6858403613997888,
+      "learning_rate": 8.873100058977613e-07,
+      "loss": 1.2538,
+      "step": 1338
+    },
+    {
+      "epoch": 0.9081814328105129,
+      "grad_norm": 0.673381363207631,
+      "learning_rate": 8.744048798517402e-07,
+      "loss": 1.238,
+      "step": 1339
+    },
+    {
+      "epoch": 0.9088596863077575,
+      "grad_norm": 0.6637418848607853,
+      "learning_rate": 8.615921917057069e-07,
+      "loss": 1.2361,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9095379398050021,
+      "grad_norm": 0.6637814922635127,
+      "learning_rate": 8.488720033860032e-07,
+      "loss": 1.2446,
+      "step": 1341
+    },
+    {
+      "epoch": 0.9102161933022467,
+      "grad_norm": 0.6602838507106655,
+      "learning_rate": 8.362443763718953e-07,
+      "loss": 1.2599,
+      "step": 1342
+    },
+    {
+      "epoch": 0.9108944467994913,
+      "grad_norm": 0.659284695620839,
+      "learning_rate": 8.237093716952737e-07,
+      "loss": 1.2695,
+      "step": 1343
+    },
+    {
+      "epoch": 0.9115727002967359,
+      "grad_norm": 0.665982122184112,
+      "learning_rate": 8.11267049940374e-07,
+      "loss": 1.2202,
+      "step": 1344
+    },
+    {
+      "epoch": 0.9122509537939805,
+      "grad_norm": 0.6393656129125219,
+      "learning_rate": 7.989174712434677e-07,
+      "loss": 1.2217,
+      "step": 1345
+    },
+    {
+      "epoch": 0.9129292072912251,
+      "grad_norm": 0.6707636295089805,
+      "learning_rate": 7.86660695292596e-07,
+      "loss": 1.2275,
+      "step": 1346
+    },
+    {
+      "epoch": 0.9136074607884697,
+      "grad_norm": 0.6638358184981306,
+      "learning_rate": 7.744967813272475e-07,
+      "loss": 1.2489,
+      "step": 1347
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 0.6627214654844239,
+      "learning_rate": 7.624257881380992e-07,
+      "loss": 1.2488,
+      "step": 1348
+    },
+    {
+      "epoch": 0.9149639677829589,
+      "grad_norm": 0.6739167341497851,
+      "learning_rate": 7.504477740667271e-07,
+      "loss": 1.2447,
+      "step": 1349
+    },
+    {
+      "epoch": 0.9156422212802034,
+      "grad_norm": 0.6683330924152061,
+      "learning_rate": 7.385627970053088e-07,
+      "loss": 1.2374,
+      "step": 1350
+    },
+    {
+      "epoch": 0.916320474777448,
+      "grad_norm": 0.6592964271700535,
+      "learning_rate": 7.267709143963663e-07,
+      "loss": 1.2669,
+      "step": 1351
+    },
+    {
+      "epoch": 0.9169987282746926,
+      "grad_norm": 0.6843140708067933,
+      "learning_rate": 7.150721832324659e-07,
+      "loss": 1.246,
+      "step": 1352
+    },
+    {
+      "epoch": 0.9176769817719372,
+      "grad_norm": 0.6661572430839988,
+      "learning_rate": 7.034666600559647e-07,
+      "loss": 1.2383,
+      "step": 1353
+    },
+    {
+      "epoch": 0.9183552352691818,
+      "grad_norm": 0.665393427409558,
+      "learning_rate": 6.919544009587231e-07,
+      "loss": 1.283,
+      "step": 1354
+    },
+    {
+      "epoch": 0.9190334887664264,
+      "grad_norm": 0.669387001203815,
+      "learning_rate": 6.805354615818305e-07,
+      "loss": 1.2452,
+      "step": 1355
+    },
+    {
+      "epoch": 0.919711742263671,
+      "grad_norm": 0.43055144276040425,
+      "learning_rate": 6.692098971153549e-07,
+      "loss": 1.3544,
+      "step": 1356
+    },
+    {
+      "epoch": 0.9203899957609156,
+      "grad_norm": 0.6871511026380561,
+      "learning_rate": 6.579777622980565e-07,
+      "loss": 1.2243,
+      "step": 1357
+    },
+    {
+      "epoch": 0.9210682492581602,
+      "grad_norm": 0.6739005611348086,
+      "learning_rate": 6.468391114171302e-07,
+      "loss": 1.2756,
+      "step": 1358
+    },
+    {
+      "epoch": 0.9217465027554048,
+      "grad_norm": 0.6744433756104877,
+      "learning_rate": 6.357939983079453e-07,
+      "loss": 1.2444,
+      "step": 1359
+    },
+    {
+      "epoch": 0.9224247562526494,
+      "grad_norm": 0.6729612932904839,
+      "learning_rate": 6.248424763537886e-07,
+      "loss": 1.256,
+      "step": 1360
+    },
+    {
+      "epoch": 0.923103009749894,
+      "grad_norm": 0.6829166299546027,
+      "learning_rate": 6.139845984855974e-07,
+      "loss": 1.2518,
+      "step": 1361
+    },
+    {
+      "epoch": 0.9237812632471386,
+      "grad_norm": 0.6709060858281704,
+      "learning_rate": 6.032204171817068e-07,
+      "loss": 1.221,
+      "step": 1362
+    },
+    {
+      "epoch": 0.9244595167443832,
+      "grad_norm": 0.7078854183369467,
+      "learning_rate": 5.92549984467603e-07,
+      "loss": 1.274,
+      "step": 1363
+    },
+    {
+      "epoch": 0.9251377702416278,
+      "grad_norm": 0.693268884844903,
+      "learning_rate": 5.819733519156589e-07,
+      "loss": 1.2452,
+      "step": 1364
+    },
+    {
+      "epoch": 0.9258160237388724,
+      "grad_norm": 0.6758341411358181,
+      "learning_rate": 5.714905706448992e-07,
+      "loss": 1.2596,
+      "step": 1365
+    },
+    {
+      "epoch": 0.926494277236117,
+      "grad_norm": 0.6676882929450928,
+      "learning_rate": 5.611016913207379e-07,
+      "loss": 1.2596,
+      "step": 1366
+    },
+    {
+      "epoch": 0.9271725307333616,
+      "grad_norm": 0.6921240376093872,
+      "learning_rate": 5.508067641547521e-07,
+      "loss": 1.2504,
+      "step": 1367
+    },
+    {
+      "epoch": 0.9278507842306062,
+      "grad_norm": 0.6905148881597328,
+      "learning_rate": 5.406058389044178e-07,
+      "loss": 1.2266,
+      "step": 1368
+    },
+    {
+      "epoch": 0.9285290377278508,
+      "grad_norm": 0.6879011599996939,
+      "learning_rate": 5.30498964872892e-07,
+      "loss": 1.2458,
+      "step": 1369
+    },
+    {
+      "epoch": 0.9292072912250954,
+      "grad_norm": 0.6757393596742982,
+      "learning_rate": 5.204861909087511e-07,
+      "loss": 1.2417,
+      "step": 1370
+    },
+    {
+      "epoch": 0.92988554472234,
+      "grad_norm": 0.6926775176816642,
+      "learning_rate": 5.105675654057752e-07,
+      "loss": 1.2674,
+      "step": 1371
+    },
+    {
+      "epoch": 0.9305637982195846,
+      "grad_norm": 0.6881075927445028,
+      "learning_rate": 5.007431363027082e-07,
+      "loss": 1.2648,
+      "step": 1372
+    },
+    {
+      "epoch": 0.9312420517168292,
+      "grad_norm": 0.644620430770547,
+      "learning_rate": 4.91012951083012e-07,
+      "loss": 1.2202,
+      "step": 1373
+    },
+    {
+      "epoch": 0.9319203052140738,
+      "grad_norm": 0.6765738092957386,
+      "learning_rate": 4.813770567746589e-07,
+      "loss": 1.2685,
+      "step": 1374
+    },
+    {
+      "epoch": 0.9325985587113184,
+      "grad_norm": 0.6759248428214282,
+      "learning_rate": 4.718354999498864e-07,
+      "loss": 1.2147,
+      "step": 1375
+    },
+    {
+      "epoch": 0.933276812208563,
+      "grad_norm": 0.4172155603814433,
+      "learning_rate": 4.6238832672499177e-07,
+      "loss": 1.3648,
+      "step": 1376
+    },
+    {
+      "epoch": 0.9339550657058076,
+      "grad_norm": 0.4602491839575769,
+      "learning_rate": 4.5303558276007744e-07,
+      "loss": 1.4068,
+      "step": 1377
+    },
+    {
+      "epoch": 0.9346333192030521,
+      "grad_norm": 0.6755778465641865,
+      "learning_rate": 4.4377731325887075e-07,
+      "loss": 1.2283,
+      "step": 1378
+    },
+    {
+      "epoch": 0.9353115727002967,
+      "grad_norm": 0.6721429198493364,
+      "learning_rate": 4.3461356296847333e-07,
+      "loss": 1.2553,
+      "step": 1379
+    },
+    {
+      "epoch": 0.9359898261975413,
+      "grad_norm": 0.6699419980408854,
+      "learning_rate": 4.2554437617915666e-07,
+      "loss": 1.2443,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9366680796947859,
+      "grad_norm": 0.6749882416152699,
+      "learning_rate": 4.165697967241511e-07,
+      "loss": 1.2262,
+      "step": 1381
+    },
+    {
+      "epoch": 0.9373463331920305,
+      "grad_norm": 0.6847443257768054,
+      "learning_rate": 4.076898679794261e-07,
+      "loss": 1.2253,
+      "step": 1382
+    },
+    {
+      "epoch": 0.9380245866892751,
+      "grad_norm": 0.6727955333188005,
+      "learning_rate": 3.989046328634927e-07,
+      "loss": 1.2594,
+      "step": 1383
+    },
+    {
+      "epoch": 0.9387028401865197,
+      "grad_norm": 0.6620805374528571,
+      "learning_rate": 3.902141338371768e-07,
+      "loss": 1.2142,
+      "step": 1384
+    },
+    {
+      "epoch": 0.9393810936837643,
+      "grad_norm": 0.6480772047300294,
+      "learning_rate": 3.8161841290343503e-07,
+      "loss": 1.2418,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9400593471810089,
+      "grad_norm": 0.6835792051059321,
+      "learning_rate": 3.7311751160713947e-07,
+      "loss": 1.2387,
+      "step": 1386
+    },
+    {
+      "epoch": 0.9407376006782535,
+      "grad_norm": 0.6649418879644043,
+      "learning_rate": 3.6471147103487756e-07,
+      "loss": 1.2339,
+      "step": 1387
+    },
+    {
+      "epoch": 0.9414158541754981,
+      "grad_norm": 0.6842421143137674,
+      "learning_rate": 3.564003318147569e-07,
+      "loss": 1.226,
+      "step": 1388
+    },
+    {
+      "epoch": 0.9420941076727427,
+      "grad_norm": 0.6811128860613901,
+      "learning_rate": 3.481841341162073e-07,
+      "loss": 1.2558,
+      "step": 1389
+    },
+    {
+      "epoch": 0.9427723611699873,
+      "grad_norm": 0.6613640812358834,
+      "learning_rate": 3.400629176497905e-07,
+      "loss": 1.2507,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9434506146672319,
+      "grad_norm": 0.6722408592326058,
+      "learning_rate": 3.320367216669973e-07,
+      "loss": 1.2396,
+      "step": 1391
+    },
+    {
+      "epoch": 0.9441288681644765,
+      "grad_norm": 0.6624694671185586,
+      "learning_rate": 3.2410558496006825e-07,
+      "loss": 1.2334,
+      "step": 1392
+    },
+    {
+      "epoch": 0.9448071216617211,
+      "grad_norm": 0.7137402370876521,
+      "learning_rate": 3.16269545861807e-07,
+      "loss": 1.2496,
+      "step": 1393
+    },
+    {
+      "epoch": 0.9454853751589657,
+      "grad_norm": 0.6449845768103436,
+      "learning_rate": 3.085286422453893e-07,
+      "loss": 1.2404,
+      "step": 1394
+    },
+    {
+      "epoch": 0.9461636286562103,
+      "grad_norm": 0.6749794928160756,
+      "learning_rate": 3.008829115241763e-07,
+      "loss": 1.2394,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9468418821534549,
+      "grad_norm": 0.6526640710965668,
+      "learning_rate": 2.9333239065154397e-07,
+      "loss": 1.2561,
+      "step": 1396
+    },
+    {
+      "epoch": 0.9475201356506995,
+      "grad_norm": 0.6605969261232847,
+      "learning_rate": 2.858771161206986e-07,
+      "loss": 1.2233,
+      "step": 1397
+    },
+    {
+      "epoch": 0.9481983891479441,
+      "grad_norm": 0.656943414990126,
+      "learning_rate": 2.785171239645057e-07,
+      "loss": 1.2334,
+      "step": 1398
+    },
+    {
+      "epoch": 0.9488766426451887,
+      "grad_norm": 0.6734397241519186,
+      "learning_rate": 2.712524497553037e-07,
+      "loss": 1.2722,
+      "step": 1399
+    },
+    {
+      "epoch": 0.9495548961424333,
+      "grad_norm": 0.6663530577389908,
+      "learning_rate": 2.640831286047441e-07,
+      "loss": 1.2379,
+      "step": 1400
+    },
+    {
+      "epoch": 0.9502331496396779,
+      "grad_norm": 0.6673776018022044,
+      "learning_rate": 2.570091951636178e-07,
+      "loss": 1.2541,
+      "step": 1401
+    },
+    {
+      "epoch": 0.9509114031369225,
+      "grad_norm": 0.6653923384339829,
+      "learning_rate": 2.5003068362168927e-07,
+      "loss": 1.2273,
+      "step": 1402
+    },
+    {
+      "epoch": 0.9515896566341671,
+      "grad_norm": 0.6837729865567328,
+      "learning_rate": 2.431476277075251e-07,
+      "loss": 1.2423,
+      "step": 1403
+    },
+    {
+      "epoch": 0.9522679101314117,
+      "grad_norm": 0.6680668422360202,
+      "learning_rate": 2.3636006068833872e-07,
+      "loss": 1.2451,
+      "step": 1404
+    },
+    {
+      "epoch": 0.9529461636286563,
+      "grad_norm": 0.6799977513226729,
+      "learning_rate": 2.2966801536982387e-07,
+      "loss": 1.239,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9536244171259008,
+      "grad_norm": 0.6777577598433704,
+      "learning_rate": 2.2307152409600575e-07,
+      "loss": 1.2522,
+      "step": 1406
+    },
+    {
+      "epoch": 0.9543026706231454,
+      "grad_norm": 0.6606560971096702,
+      "learning_rate": 2.1657061874906126e-07,
+      "loss": 1.2405,
+      "step": 1407
+    },
+    {
+      "epoch": 0.95498092412039,
+      "grad_norm": 0.6469835351521775,
+      "learning_rate": 2.1016533074919687e-07,
+      "loss": 1.229,
+      "step": 1408
+    },
+    {
+      "epoch": 0.9556591776176346,
+      "grad_norm": 0.6958824972510786,
+      "learning_rate": 2.0385569105447532e-07,
+      "loss": 1.2687,
+      "step": 1409
+    },
+    {
+      "epoch": 0.9563374311148792,
+      "grad_norm": 0.6740777877893196,
+      "learning_rate": 1.9764173016067145e-07,
+      "loss": 1.2389,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9570156846121238,
+      "grad_norm": 0.6736611531923851,
+      "learning_rate": 1.915234781011255e-07,
+      "loss": 1.2415,
+      "step": 1411
+    },
+    {
+      "epoch": 0.9576939381093684,
+      "grad_norm": 0.6883611656465358,
+      "learning_rate": 1.8550096444659216e-07,
+      "loss": 1.2424,
+      "step": 1412
+    },
+    {
+      "epoch": 0.958372191606613,
+      "grad_norm": 0.6827067228567522,
+      "learning_rate": 1.7957421830511412e-07,
+      "loss": 1.233,
+      "step": 1413
+    },
+    {
+      "epoch": 0.9590504451038576,
+      "grad_norm": 0.6784984890612891,
+      "learning_rate": 1.7374326832185983e-07,
+      "loss": 1.2531,
+      "step": 1414
+    },
+    {
+      "epoch": 0.9597286986011022,
+      "grad_norm": 0.6633612107331113,
+      "learning_rate": 1.6800814267899923e-07,
+      "loss": 1.2398,
+      "step": 1415
+    },
+    {
+      "epoch": 0.9604069520983468,
+      "grad_norm": 0.6488222570693454,
+      "learning_rate": 1.6236886909556603e-07,
+      "loss": 1.2664,
+      "step": 1416
+    },
+    {
+      "epoch": 0.9610852055955914,
+      "grad_norm": 0.4189268521205959,
+      "learning_rate": 1.5682547482731791e-07,
+      "loss": 1.3749,
+      "step": 1417
+    },
+    {
+      "epoch": 0.9617634590928359,
+      "grad_norm": 0.6917389003668754,
+      "learning_rate": 1.5137798666660765e-07,
+      "loss": 1.2585,
+      "step": 1418
+    },
+    {
+      "epoch": 0.9624417125900805,
+      "grad_norm": 0.672451082602383,
+      "learning_rate": 1.4602643094225876e-07,
+      "loss": 1.2517,
+      "step": 1419
+    },
+    {
+      "epoch": 0.9631199660873251,
+      "grad_norm": 0.6504255430195721,
+      "learning_rate": 1.4077083351942799e-07,
+      "loss": 1.2365,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9637982195845697,
+      "grad_norm": 0.6850370736126643,
+      "learning_rate": 1.3561121979949188e-07,
+      "loss": 1.2489,
+      "step": 1421
+    },
+    {
+      "epoch": 0.9644764730818143,
+      "grad_norm": 0.66162990708649,
+      "learning_rate": 1.3054761471991362e-07,
+      "loss": 1.2442,
+      "step": 1422
+    },
+    {
+      "epoch": 0.9651547265790589,
+      "grad_norm": 0.6636753229495839,
+      "learning_rate": 1.2558004275412983e-07,
+      "loss": 1.2332,
+      "step": 1423
+    },
+    {
+      "epoch": 0.9658329800763035,
+      "grad_norm": 0.6868071837214885,
+      "learning_rate": 1.207085279114284e-07,
+      "loss": 1.2472,
+      "step": 1424
+    },
+    {
+      "epoch": 0.9665112335735481,
+      "grad_norm": 0.6876603336287529,
+      "learning_rate": 1.1593309373683304e-07,
+      "loss": 1.272,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9671894870707927,
+      "grad_norm": 0.6512671334666834,
+      "learning_rate": 1.1125376331099002e-07,
+      "loss": 1.2344,
+      "step": 1426
+    },
+    {
+      "epoch": 0.9678677405680373,
+      "grad_norm": 0.6715940763143174,
+      "learning_rate": 1.066705592500572e-07,
+      "loss": 1.2247,
+      "step": 1427
+    },
+    {
+      "epoch": 0.9685459940652819,
+      "grad_norm": 0.6754499531697719,
+      "learning_rate": 1.0218350370559515e-07,
+      "loss": 1.2597,
+      "step": 1428
+    },
+    {
+      "epoch": 0.9692242475625265,
+      "grad_norm": 0.6524907799318351,
+      "learning_rate": 9.779261836445841e-08,
+      "loss": 1.2156,
+      "step": 1429
+    },
+    {
+      "epoch": 0.969902501059771,
+      "grad_norm": 0.6792557779960546,
+      "learning_rate": 9.34979244486911e-08,
+      "loss": 1.3041,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9705807545570156,
+      "grad_norm": 0.6823631877198995,
+      "learning_rate": 8.929944271542035e-08,
+      "loss": 1.24,
+      "step": 1431
+    },
+    {
+      "epoch": 0.9712590080542602,
+      "grad_norm": 0.6846773472067208,
+      "learning_rate": 8.51971934567697e-08,
+      "loss": 1.2623,
+      "step": 1432
+    },
+    {
+      "epoch": 0.9719372615515048,
+      "grad_norm": 0.6467976465127085,
+      "learning_rate": 8.119119649974361e-08,
+      "loss": 1.2613,
+      "step": 1433
+    },
+    {
+      "epoch": 0.9726155150487494,
+      "grad_norm": 0.6575189203134979,
+      "learning_rate": 7.728147120614093e-08,
+      "loss": 1.2143,
+      "step": 1434
+    },
+    {
+      "epoch": 0.973293768545994,
+      "grad_norm": 0.6665055528564646,
+      "learning_rate": 7.346803647246381e-08,
+      "loss": 1.2492,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9739720220432386,
+      "grad_norm": 0.6486777079132526,
+      "learning_rate": 6.975091072981777e-08,
+      "loss": 1.2271,
+      "step": 1436
+    },
+    {
+      "epoch": 0.9746502755404832,
+      "grad_norm": 0.7034061981240023,
+      "learning_rate": 6.613011194382957e-08,
+      "loss": 1.2742,
+      "step": 1437
+    },
+    {
+      "epoch": 0.9753285290377278,
+      "grad_norm": 0.6686002236944355,
+      "learning_rate": 6.260565761455616e-08,
+      "loss": 1.2429,
+      "step": 1438
+    },
+    {
+      "epoch": 0.9760067825349724,
+      "grad_norm": 0.6556116978595136,
+      "learning_rate": 5.917756477640702e-08,
+      "loss": 1.2385,
+      "step": 1439
+    },
+    {
+      "epoch": 0.976685036032217,
+      "grad_norm": 0.6554439946819614,
+      "learning_rate": 5.584584999805076e-08,
+      "loss": 1.239,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9773632895294616,
+      "grad_norm": 0.6575403679333993,
+      "learning_rate": 5.2610529382346456e-08,
+      "loss": 1.2633,
+      "step": 1441
+    },
+    {
+      "epoch": 0.9780415430267062,
+      "grad_norm": 0.649925623318637,
+      "learning_rate": 4.947161856625693e-08,
+      "loss": 1.2244,
+      "step": 1442
+    },
+    {
+      "epoch": 0.9787197965239508,
+      "grad_norm": 0.6605273078555803,
+      "learning_rate": 4.642913272077776e-08,
+      "loss": 1.2355,
+      "step": 1443
+    },
+    {
+      "epoch": 0.9793980500211954,
+      "grad_norm": 0.6821874686800649,
+      "learning_rate": 4.3483086550863974e-08,
+      "loss": 1.2527,
+      "step": 1444
+    },
+    {
+      "epoch": 0.98007630351844,
+      "grad_norm": 0.6766151832758971,
+      "learning_rate": 4.063349429535679e-08,
+      "loss": 1.2384,
+      "step": 1445
+    },
+    {
+      "epoch": 0.9807545570156846,
+      "grad_norm": 0.6881276161064582,
+      "learning_rate": 3.788036972691922e-08,
+      "loss": 1.246,
+      "step": 1446
+    },
+    {
+      "epoch": 0.9814328105129292,
+      "grad_norm": 0.6700454329845072,
+      "learning_rate": 3.522372615195835e-08,
+      "loss": 1.2342,
+      "step": 1447
+    },
+    {
+      "epoch": 0.9821110640101738,
+      "grad_norm": 0.6545685722297931,
+      "learning_rate": 3.2663576410576495e-08,
+      "loss": 1.2287,
+      "step": 1448
+    },
+    {
+      "epoch": 0.9827893175074184,
+      "grad_norm": 0.6947327345532218,
+      "learning_rate": 3.0199932876500136e-08,
+      "loss": 1.2424,
+      "step": 1449
+    },
+    {
+      "epoch": 0.983467571004663,
+      "grad_norm": 0.6500645029727457,
+      "learning_rate": 2.7832807457019995e-08,
+      "loss": 1.2511,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9841458245019076,
+      "grad_norm": 0.6973448708269708,
+      "learning_rate": 2.5562211592937703e-08,
+      "loss": 1.2365,
+      "step": 1451
+    },
+    {
+      "epoch": 0.9848240779991522,
+      "grad_norm": 0.6555455445351246,
+      "learning_rate": 2.3388156258501436e-08,
+      "loss": 1.22,
+      "step": 1452
+    },
+    {
+      "epoch": 0.9855023314963968,
+      "grad_norm": 0.6778273501171838,
+      "learning_rate": 2.1310651961368167e-08,
+      "loss": 1.2449,
+      "step": 1453
+    },
+    {
+      "epoch": 0.9861805849936414,
+      "grad_norm": 0.4290647050125529,
+      "learning_rate": 1.9329708742537035e-08,
+      "loss": 1.3346,
+      "step": 1454
+    },
+    {
+      "epoch": 0.986858838490886,
+      "grad_norm": 0.6651972834538079,
+      "learning_rate": 1.744533617631161e-08,
+      "loss": 1.2721,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9875370919881306,
+      "grad_norm": 0.6320225606844798,
+      "learning_rate": 1.5657543370248828e-08,
+      "loss": 1.2154,
+      "step": 1456
+    },
+    {
+      "epoch": 0.9882153454853752,
+      "grad_norm": 0.6817640902595111,
+      "learning_rate": 1.3966338965114567e-08,
+      "loss": 1.2607,
+      "step": 1457
+    },
+    {
+      "epoch": 0.9888935989826197,
+      "grad_norm": 0.6667804231582753,
+      "learning_rate": 1.237173113484591e-08,
+      "loss": 1.2037,
+      "step": 1458
+    },
+    {
+      "epoch": 0.9895718524798643,
+      "grad_norm": 0.6537258264830468,
+      "learning_rate": 1.0873727586506733e-08,
+      "loss": 1.2153,
+      "step": 1459
+    },
+    {
+      "epoch": 0.9902501059771089,
+      "grad_norm": 0.6598848171719655,
+      "learning_rate": 9.472335560254398e-09,
+      "loss": 1.2209,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9909283594743535,
+      "grad_norm": 0.6886054526615797,
+      "learning_rate": 8.167561829299786e-09,
+      "loss": 1.2512,
+      "step": 1461
+    },
+    {
+      "epoch": 0.9916066129715981,
+      "grad_norm": 0.6664574716833053,
+      "learning_rate": 6.9594126998828726e-09,
+      "loss": 1.2156,
+      "step": 1462
+    },
+    {
+      "epoch": 0.9922848664688427,
+      "grad_norm": 0.6751343990247555,
+      "learning_rate": 5.847894011234978e-09,
+      "loss": 1.2391,
+      "step": 1463
+    },
+    {
+      "epoch": 0.9929631199660873,
+      "grad_norm": 0.6850150824771514,
+      "learning_rate": 4.833011135549903e-09,
+      "loss": 1.235,
+      "step": 1464
+    },
+    {
+      "epoch": 0.9936413734633319,
+      "grad_norm": 0.6680179651217075,
+      "learning_rate": 3.914768977966166e-09,
+      "loss": 1.2249,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9943196269605765,
+      "grad_norm": 0.669886714999622,
+      "learning_rate": 3.093171976533693e-09,
+      "loss": 1.2441,
+      "step": 1466
+    },
+    {
+      "epoch": 0.9949978804578211,
+      "grad_norm": 0.6783995175305425,
+      "learning_rate": 2.36822410219828e-09,
+      "loss": 1.2329,
+      "step": 1467
+    },
+    {
+      "epoch": 0.9956761339550657,
+      "grad_norm": 0.6487853495244439,
+      "learning_rate": 1.7399288587816032e-09,
+      "loss": 1.2061,
+      "step": 1468
+    },
+    {
+      "epoch": 0.9963543874523103,
+      "grad_norm": 0.6655272487747175,
+      "learning_rate": 1.2082892829634596e-09,
+      "loss": 1.2523,
+      "step": 1469
+    },
+    {
+      "epoch": 0.9970326409495549,
+      "grad_norm": 0.6756617448317891,
+      "learning_rate": 7.733079442617808e-10,
+      "loss": 1.2319,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9977108944467995,
+      "grad_norm": 0.6582339993988024,
+      "learning_rate": 4.349869450370747e-10,
+      "loss": 1.2534,
+      "step": 1471
+    },
+    {
+      "epoch": 0.9983891479440441,
+      "grad_norm": 0.6616448969554425,
+      "learning_rate": 1.933279204568983e-10,
+      "loss": 1.2107,
+      "step": 1472
+    },
+    {
+      "epoch": 0.9990674014412887,
+      "grad_norm": 0.6720706810300792,
+      "learning_rate": 4.8332038513621e-11,
+      "loss": 1.2564,
+      "step": 1473
+    },
+    {
+      "epoch": 0.9997456549385333,
+      "grad_norm": 0.5783213650920862,
+      "learning_rate": 0.0,
+      "loss": 1.2774,
+      "step": 1474
+    },
+    {
+      "epoch": 0.9997456549385333,
+      "step": 1474,
+      "total_flos": 7384218508197888.0,
+      "train_loss": 1.3047956960980895,
+      "train_runtime": 17436.1201,
+      "train_samples_per_second": 86.585,
+      "train_steps_per_second": 0.085
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1474,
+  "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": 7384218508197888.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}