diff --git "a/checkpoint-1476/trainer_state.json" "b/checkpoint-1476/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1476/trainer_state.json"
@@ -0,0 +1,10365 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.001354096140826,
+  "eval_steps": 500,
+  "global_step": 1476,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0013540961408259986,
+      "grad_norm": 2.4126635555318527,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 3.1189,
+      "step": 1
+    },
+    {
+      "epoch": 0.002708192281651997,
+      "grad_norm": 2.8141112841232654,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 2.6861,
+      "step": 2
+    },
+    {
+      "epoch": 0.004062288422477996,
+      "grad_norm": 2.6738862039513274,
+      "learning_rate": 7.5e-07,
+      "loss": 3.3382,
+      "step": 3
+    },
+    {
+      "epoch": 0.005416384563303994,
+      "grad_norm": 2.30287728842049,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.4565,
+      "step": 4
+    },
+    {
+      "epoch": 0.006770480704129994,
+      "grad_norm": 42.907191662846444,
+      "learning_rate": 1.25e-06,
+      "loss": 2.5327,
+      "step": 5
+    },
+    {
+      "epoch": 0.008124576844955992,
+      "grad_norm": 2.9404364361574724,
+      "learning_rate": 1.5e-06,
+      "loss": 2.7222,
+      "step": 6
+    },
+    {
+      "epoch": 0.009478672985781991,
+      "grad_norm": 1.9739382530413463,
+      "learning_rate": 1.75e-06,
+      "loss": 2.1649,
+      "step": 7
+    },
+    {
+      "epoch": 0.010832769126607989,
+      "grad_norm": 3.463912464916458,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.9378,
+      "step": 8
+    },
+    {
+      "epoch": 0.012186865267433988,
+      "grad_norm": 2.207345459970691,
+      "learning_rate": 2.25e-06,
+      "loss": 3.6521,
+      "step": 9
+    },
+    {
+      "epoch": 0.013540961408259987,
+      "grad_norm": 2.752714935099265,
+      "learning_rate": 2.5e-06,
+      "loss": 2.5672,
+      "step": 10
+    },
+    {
+      "epoch": 0.014895057549085985,
+      "grad_norm": 2.7722408021236187,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 3.0411,
+      "step": 11
+    },
+    {
+      "epoch": 0.016249153689911984,
+      "grad_norm": 2.9367100387488185,
+      "learning_rate": 3e-06,
+      "loss": 2.8985,
+      "step": 12
+    },
+    {
+      "epoch": 0.017603249830737983,
+      "grad_norm": 2.7601030286888797,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 2.8232,
+      "step": 13
+    },
+    {
+      "epoch": 0.018957345971563982,
+      "grad_norm": 3.184703964873984,
+      "learning_rate": 3.5e-06,
+      "loss": 2.9731,
+      "step": 14
+    },
+    {
+      "epoch": 0.020311442112389978,
+      "grad_norm": 2.2952697179626105,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 2.823,
+      "step": 15
+    },
+    {
+      "epoch": 0.021665538253215978,
+      "grad_norm": 2.4601547818230913,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.7683,
+      "step": 16
+    },
+    {
+      "epoch": 0.023019634394041977,
+      "grad_norm": 2.0169369154885217,
+      "learning_rate": 4.25e-06,
+      "loss": 2.566,
+      "step": 17
+    },
+    {
+      "epoch": 0.024373730534867976,
+      "grad_norm": 1.8874822039928005,
+      "learning_rate": 4.5e-06,
+      "loss": 2.4032,
+      "step": 18
+    },
+    {
+      "epoch": 0.025727826675693975,
+      "grad_norm": 1.4345346790759714,
+      "learning_rate": 4.75e-06,
+      "loss": 2.8099,
+      "step": 19
+    },
+    {
+      "epoch": 0.027081922816519974,
+      "grad_norm": 2.5317082125734025,
+      "learning_rate": 5e-06,
+      "loss": 2.4169,
+      "step": 20
+    },
+    {
+      "epoch": 0.02843601895734597,
+      "grad_norm": 1.4417852513317821,
+      "learning_rate": 5.2500000000000006e-06,
+      "loss": 3.0868,
+      "step": 21
+    },
+    {
+      "epoch": 0.02979011509817197,
+      "grad_norm": 0.9044314800944161,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 2.7374,
+      "step": 22
+    },
+    {
+      "epoch": 0.03114421123899797,
+      "grad_norm": 1.4184446315710213,
+      "learning_rate": 5.75e-06,
+      "loss": 2.5317,
+      "step": 23
+    },
+    {
+      "epoch": 0.03249830737982397,
+      "grad_norm": 1.6900469714019433,
+      "learning_rate": 6e-06,
+      "loss": 2.4404,
+      "step": 24
+    },
+    {
+      "epoch": 0.033852403520649964,
+      "grad_norm": 1.0886842936641719,
+      "learning_rate": 6.25e-06,
+      "loss": 2.8003,
+      "step": 25
+    },
+    {
+      "epoch": 0.035206499661475966,
+      "grad_norm": 1.463135149887311,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.8819,
+      "step": 26
+    },
+    {
+      "epoch": 0.03656059580230196,
+      "grad_norm": 1.1222026842497739,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 2.2671,
+      "step": 27
+    },
+    {
+      "epoch": 0.037914691943127965,
+      "grad_norm": 1.0087919882501093,
+      "learning_rate": 7e-06,
+      "loss": 2.4491,
+      "step": 28
+    },
+    {
+      "epoch": 0.03926878808395396,
+      "grad_norm": 1.2761649114763567,
+      "learning_rate": 7.25e-06,
+      "loss": 2.7121,
+      "step": 29
+    },
+    {
+      "epoch": 0.040622884224779957,
+      "grad_norm": 1.0185872249128933,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 2.6497,
+      "step": 30
+    },
+    {
+      "epoch": 0.04197698036560596,
+      "grad_norm": 1.2789270965641044,
+      "learning_rate": 7.75e-06,
+      "loss": 2.668,
+      "step": 31
+    },
+    {
+      "epoch": 0.043331076506431955,
+      "grad_norm": 1.1115471485822677,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.6767,
+      "step": 32
+    },
+    {
+      "epoch": 0.04468517264725796,
+      "grad_norm": 1.0627476896958792,
+      "learning_rate": 8.25e-06,
+      "loss": 2.6262,
+      "step": 33
+    },
+    {
+      "epoch": 0.046039268788083954,
+      "grad_norm": 1.1212584420019067,
+      "learning_rate": 8.5e-06,
+      "loss": 2.4854,
+      "step": 34
+    },
+    {
+      "epoch": 0.04739336492890995,
+      "grad_norm": 0.714156466912055,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 2.3662,
+      "step": 35
+    },
+    {
+      "epoch": 0.04874746106973595,
+      "grad_norm": 1.4520911693345544,
+      "learning_rate": 9e-06,
+      "loss": 2.7126,
+      "step": 36
+    },
+    {
+      "epoch": 0.05010155721056195,
+      "grad_norm": 1.493660601298817,
+      "learning_rate": 9.250000000000001e-06,
+      "loss": 3.0131,
+      "step": 37
+    },
+    {
+      "epoch": 0.05145565335138795,
+      "grad_norm": 1.2721754658047073,
+      "learning_rate": 9.5e-06,
+      "loss": 3.0784,
+      "step": 38
+    },
+    {
+      "epoch": 0.052809749492213946,
+      "grad_norm": 0.8803541178840371,
+      "learning_rate": 9.75e-06,
+      "loss": 2.1007,
+      "step": 39
+    },
+    {
+      "epoch": 0.05416384563303995,
+      "grad_norm": 2.0655028083700504,
+      "learning_rate": 1e-05,
+      "loss": 2.5182,
+      "step": 40
+    },
+    {
+      "epoch": 0.055517941773865945,
+      "grad_norm": 1.0487372085128044,
+      "learning_rate": 9.999997090241333e-06,
+      "loss": 2.2771,
+      "step": 41
+    },
+    {
+      "epoch": 0.05687203791469194,
+      "grad_norm": 0.8160509238097648,
+      "learning_rate": 9.999988360968714e-06,
+      "loss": 2.2042,
+      "step": 42
+    },
+    {
+      "epoch": 0.05822613405551794,
+      "grad_norm": 0.8864211341294704,
+      "learning_rate": 9.999973812192306e-06,
+      "loss": 2.7951,
+      "step": 43
+    },
+    {
+      "epoch": 0.05958023019634394,
+      "grad_norm": 0.7681880276765591,
+      "learning_rate": 9.99995344392904e-06,
+      "loss": 2.859,
+      "step": 44
+    },
+    {
+      "epoch": 0.06093432633716994,
+      "grad_norm": 1.7239414196388092,
+      "learning_rate": 9.999927256202626e-06,
+      "loss": 2.7866,
+      "step": 45
+    },
+    {
+      "epoch": 0.06228842247799594,
+      "grad_norm": 0.7879562468694528,
+      "learning_rate": 9.999895249043542e-06,
+      "loss": 2.1671,
+      "step": 46
+    },
+    {
+      "epoch": 0.06364251861882193,
+      "grad_norm": 0.8376315627603867,
+      "learning_rate": 9.99985742248904e-06,
+      "loss": 2.6406,
+      "step": 47
+    },
+    {
+      "epoch": 0.06499661475964794,
+      "grad_norm": 0.8839739394970294,
+      "learning_rate": 9.999813776583148e-06,
+      "loss": 2.3163,
+      "step": 48
+    },
+    {
+      "epoch": 0.06635071090047394,
+      "grad_norm": 0.7373151185194805,
+      "learning_rate": 9.999764311376664e-06,
+      "loss": 2.5735,
+      "step": 49
+    },
+    {
+      "epoch": 0.06770480704129993,
+      "grad_norm": 1.1273015474311354,
+      "learning_rate": 9.999709026927162e-06,
+      "loss": 2.2133,
+      "step": 50
+    },
+    {
+      "epoch": 0.06905890318212593,
+      "grad_norm": 0.8478718145728893,
+      "learning_rate": 9.99964792329899e-06,
+      "loss": 2.4075,
+      "step": 51
+    },
+    {
+      "epoch": 0.07041299932295193,
+      "grad_norm": 1.0667576833889847,
+      "learning_rate": 9.999581000563265e-06,
+      "loss": 2.6946,
+      "step": 52
+    },
+    {
+      "epoch": 0.07176709546377792,
+      "grad_norm": 1.0620013368810508,
+      "learning_rate": 9.999508258797876e-06,
+      "loss": 2.6205,
+      "step": 53
+    },
+    {
+      "epoch": 0.07312119160460392,
+      "grad_norm": 0.8072331151341536,
+      "learning_rate": 9.999429698087491e-06,
+      "loss": 2.5132,
+      "step": 54
+    },
+    {
+      "epoch": 0.07447528774542993,
+      "grad_norm": 0.6900996192892075,
+      "learning_rate": 9.999345318523544e-06,
+      "loss": 2.5071,
+      "step": 55
+    },
+    {
+      "epoch": 0.07582938388625593,
+      "grad_norm": 0.7145430930616551,
+      "learning_rate": 9.999255120204248e-06,
+      "loss": 2.6579,
+      "step": 56
+    },
+    {
+      "epoch": 0.07718348002708192,
+      "grad_norm": 0.9813015245466205,
+      "learning_rate": 9.999159103234582e-06,
+      "loss": 2.1667,
+      "step": 57
+    },
+    {
+      "epoch": 0.07853757616790792,
+      "grad_norm": 0.6801613421398072,
+      "learning_rate": 9.999057267726304e-06,
+      "loss": 2.2016,
+      "step": 58
+    },
+    {
+      "epoch": 0.07989167230873392,
+      "grad_norm": 1.340722366857769,
+      "learning_rate": 9.998949613797937e-06,
+      "loss": 2.5416,
+      "step": 59
+    },
+    {
+      "epoch": 0.08124576844955991,
+      "grad_norm": 0.8044848626001605,
+      "learning_rate": 9.998836141574781e-06,
+      "loss": 2.6258,
+      "step": 60
+    },
+    {
+      "epoch": 0.08259986459038592,
+      "grad_norm": 0.9118107600823588,
+      "learning_rate": 9.99871685118891e-06,
+      "loss": 2.5182,
+      "step": 61
+    },
+    {
+      "epoch": 0.08395396073121192,
+      "grad_norm": 0.8103939602292338,
+      "learning_rate": 9.99859174277916e-06,
+      "loss": 2.4597,
+      "step": 62
+    },
+    {
+      "epoch": 0.08530805687203792,
+      "grad_norm": 0.7953248414753887,
+      "learning_rate": 9.99846081649115e-06,
+      "loss": 2.6165,
+      "step": 63
+    },
+    {
+      "epoch": 0.08666215301286391,
+      "grad_norm": 0.6885480399283996,
+      "learning_rate": 9.998324072477266e-06,
+      "loss": 2.3687,
+      "step": 64
+    },
+    {
+      "epoch": 0.08801624915368991,
+      "grad_norm": 0.968337926868628,
+      "learning_rate": 9.99818151089666e-06,
+      "loss": 2.4591,
+      "step": 65
+    },
+    {
+      "epoch": 0.08937034529451592,
+      "grad_norm": 0.7154001784843524,
+      "learning_rate": 9.998033131915266e-06,
+      "loss": 2.2175,
+      "step": 66
+    },
+    {
+      "epoch": 0.0907244414353419,
+      "grad_norm": 1.18262426992356,
+      "learning_rate": 9.997878935705778e-06,
+      "loss": 3.1191,
+      "step": 67
+    },
+    {
+      "epoch": 0.09207853757616791,
+      "grad_norm": 0.7400521396635642,
+      "learning_rate": 9.997718922447669e-06,
+      "loss": 2.5125,
+      "step": 68
+    },
+    {
+      "epoch": 0.09343263371699391,
+      "grad_norm": 0.8148295313823943,
+      "learning_rate": 9.997553092327174e-06,
+      "loss": 2.7714,
+      "step": 69
+    },
+    {
+      "epoch": 0.0947867298578199,
+      "grad_norm": 0.7378483977717124,
+      "learning_rate": 9.997381445537309e-06,
+      "loss": 2.3631,
+      "step": 70
+    },
+    {
+      "epoch": 0.0961408259986459,
+      "grad_norm": 0.6266190246235469,
+      "learning_rate": 9.997203982277852e-06,
+      "loss": 2.4936,
+      "step": 71
+    },
+    {
+      "epoch": 0.0974949221394719,
+      "grad_norm": 1.50407707220557,
+      "learning_rate": 9.997020702755353e-06,
+      "loss": 2.7048,
+      "step": 72
+    },
+    {
+      "epoch": 0.0988490182802979,
+      "grad_norm": 0.8300611098600416,
+      "learning_rate": 9.996831607183132e-06,
+      "loss": 2.7146,
+      "step": 73
+    },
+    {
+      "epoch": 0.1002031144211239,
+      "grad_norm": 0.7552213758897641,
+      "learning_rate": 9.996636695781276e-06,
+      "loss": 2.5399,
+      "step": 74
+    },
+    {
+      "epoch": 0.1015572105619499,
+      "grad_norm": 0.705291949087134,
+      "learning_rate": 9.996435968776646e-06,
+      "loss": 2.6843,
+      "step": 75
+    },
+    {
+      "epoch": 0.1029113067027759,
+      "grad_norm": 0.7673837145200612,
+      "learning_rate": 9.996229426402867e-06,
+      "loss": 2.7562,
+      "step": 76
+    },
+    {
+      "epoch": 0.10426540284360189,
+      "grad_norm": 0.7439456610505197,
+      "learning_rate": 9.996017068900335e-06,
+      "loss": 2.0488,
+      "step": 77
+    },
+    {
+      "epoch": 0.10561949898442789,
+      "grad_norm": 0.8945277132428896,
+      "learning_rate": 9.995798896516215e-06,
+      "loss": 2.48,
+      "step": 78
+    },
+    {
+      "epoch": 0.1069735951252539,
+      "grad_norm": 0.8901395084271732,
+      "learning_rate": 9.995574909504434e-06,
+      "loss": 2.6099,
+      "step": 79
+    },
+    {
+      "epoch": 0.1083276912660799,
+      "grad_norm": 0.824778949464918,
+      "learning_rate": 9.995345108125698e-06,
+      "loss": 2.69,
+      "step": 80
+    },
+    {
+      "epoch": 0.10968178740690589,
+      "grad_norm": 0.8965013093653279,
+      "learning_rate": 9.995109492647467e-06,
+      "loss": 1.8424,
+      "step": 81
+    },
+    {
+      "epoch": 0.11103588354773189,
+      "grad_norm": 1.0115160015541755,
+      "learning_rate": 9.99486806334398e-06,
+      "loss": 2.3693,
+      "step": 82
+    },
+    {
+      "epoch": 0.11238997968855789,
+      "grad_norm": 0.8624086155070334,
+      "learning_rate": 9.994620820496234e-06,
+      "loss": 2.6173,
+      "step": 83
+    },
+    {
+      "epoch": 0.11374407582938388,
+      "grad_norm": 0.7454258690069137,
+      "learning_rate": 9.994367764391998e-06,
+      "loss": 2.6624,
+      "step": 84
+    },
+    {
+      "epoch": 0.11509817197020988,
+      "grad_norm": 1.0680963212316428,
+      "learning_rate": 9.994108895325802e-06,
+      "loss": 2.478,
+      "step": 85
+    },
+    {
+      "epoch": 0.11645226811103589,
+      "grad_norm": 0.8212102801807549,
+      "learning_rate": 9.993844213598949e-06,
+      "loss": 2.0231,
+      "step": 86
+    },
+    {
+      "epoch": 0.11780636425186188,
+      "grad_norm": 1.0684938905029466,
+      "learning_rate": 9.993573719519498e-06,
+      "loss": 2.595,
+      "step": 87
+    },
+    {
+      "epoch": 0.11916046039268788,
+      "grad_norm": 0.9867071296775374,
+      "learning_rate": 9.993297413402282e-06,
+      "loss": 2.3636,
+      "step": 88
+    },
+    {
+      "epoch": 0.12051455653351388,
+      "grad_norm": 1.404157430750734,
+      "learning_rate": 9.993015295568893e-06,
+      "loss": 2.1992,
+      "step": 89
+    },
+    {
+      "epoch": 0.12186865267433988,
+      "grad_norm": 0.8693790763825426,
+      "learning_rate": 9.992727366347688e-06,
+      "loss": 2.7016,
+      "step": 90
+    },
+    {
+      "epoch": 0.12322274881516587,
+      "grad_norm": 0.7501248710359847,
+      "learning_rate": 9.99243362607379e-06,
+      "loss": 3.2797,
+      "step": 91
+    },
+    {
+      "epoch": 0.12457684495599188,
+      "grad_norm": 1.0643003422752406,
+      "learning_rate": 9.992134075089085e-06,
+      "loss": 2.5437,
+      "step": 92
+    },
+    {
+      "epoch": 0.12593094109681788,
+      "grad_norm": 0.9103830863991241,
+      "learning_rate": 9.991828713742218e-06,
+      "loss": 3.0088,
+      "step": 93
+    },
+    {
+      "epoch": 0.12728503723764387,
+      "grad_norm": 1.142388030489728,
+      "learning_rate": 9.991517542388605e-06,
+      "loss": 2.4977,
+      "step": 94
+    },
+    {
+      "epoch": 0.12863913337846988,
+      "grad_norm": 0.9920213403910143,
+      "learning_rate": 9.991200561390417e-06,
+      "loss": 2.5349,
+      "step": 95
+    },
+    {
+      "epoch": 0.12999322951929587,
+      "grad_norm": 0.8981930618864583,
+      "learning_rate": 9.990877771116588e-06,
+      "loss": 2.8531,
+      "step": 96
+    },
+    {
+      "epoch": 0.13134732566012186,
+      "grad_norm": 1.154244133374351,
+      "learning_rate": 9.990549171942817e-06,
+      "loss": 2.5311,
+      "step": 97
+    },
+    {
+      "epoch": 0.13270142180094788,
+      "grad_norm": 0.7466294529414305,
+      "learning_rate": 9.99021476425156e-06,
+      "loss": 2.4112,
+      "step": 98
+    },
+    {
+      "epoch": 0.13405551794177387,
+      "grad_norm": 1.1851681560394296,
+      "learning_rate": 9.989874548432037e-06,
+      "loss": 2.2467,
+      "step": 99
+    },
+    {
+      "epoch": 0.13540961408259986,
+      "grad_norm": 1.295896953150764,
+      "learning_rate": 9.989528524880225e-06,
+      "loss": 2.5688,
+      "step": 100
+    },
+    {
+      "epoch": 0.13676371022342587,
+      "grad_norm": 0.8892799849200678,
+      "learning_rate": 9.989176693998863e-06,
+      "loss": 2.151,
+      "step": 101
+    },
+    {
+      "epoch": 0.13811780636425186,
+      "grad_norm": 1.133636165142649,
+      "learning_rate": 9.988819056197448e-06,
+      "loss": 1.8402,
+      "step": 102
+    },
+    {
+      "epoch": 0.13947190250507785,
+      "grad_norm": 0.8016376414959372,
+      "learning_rate": 9.988455611892237e-06,
+      "loss": 2.3469,
+      "step": 103
+    },
+    {
+      "epoch": 0.14082599864590387,
+      "grad_norm": 0.964428420697917,
+      "learning_rate": 9.98808636150624e-06,
+      "loss": 2.3062,
+      "step": 104
+    },
+    {
+      "epoch": 0.14218009478672985,
+      "grad_norm": 0.7711620898648806,
+      "learning_rate": 9.987711305469232e-06,
+      "loss": 2.7206,
+      "step": 105
+    },
+    {
+      "epoch": 0.14353419092755584,
+      "grad_norm": 0.7778439840699306,
+      "learning_rate": 9.987330444217739e-06,
+      "loss": 2.4899,
+      "step": 106
+    },
+    {
+      "epoch": 0.14488828706838186,
+      "grad_norm": 0.9642875586985667,
+      "learning_rate": 9.986943778195052e-06,
+      "loss": 2.8345,
+      "step": 107
+    },
+    {
+      "epoch": 0.14624238320920785,
+      "grad_norm": 0.8268272296488341,
+      "learning_rate": 9.98655130785121e-06,
+      "loss": 2.5926,
+      "step": 108
+    },
+    {
+      "epoch": 0.14759647935003387,
+      "grad_norm": 0.7977529748864721,
+      "learning_rate": 9.986153033643011e-06,
+      "loss": 1.9835,
+      "step": 109
+    },
+    {
+      "epoch": 0.14895057549085985,
+      "grad_norm": 1.140904527079228,
+      "learning_rate": 9.985748956034007e-06,
+      "loss": 2.2239,
+      "step": 110
+    },
+    {
+      "epoch": 0.15030467163168584,
+      "grad_norm": 1.020890884482194,
+      "learning_rate": 9.985339075494504e-06,
+      "loss": 2.6764,
+      "step": 111
+    },
+    {
+      "epoch": 0.15165876777251186,
+      "grad_norm": 0.8869762247213703,
+      "learning_rate": 9.984923392501567e-06,
+      "loss": 2.2497,
+      "step": 112
+    },
+    {
+      "epoch": 0.15301286391333785,
+      "grad_norm": 0.9590605445485386,
+      "learning_rate": 9.98450190753901e-06,
+      "loss": 2.3991,
+      "step": 113
+    },
+    {
+      "epoch": 0.15436696005416384,
+      "grad_norm": 1.27578790045338,
+      "learning_rate": 9.984074621097397e-06,
+      "loss": 2.3551,
+      "step": 114
+    },
+    {
+      "epoch": 0.15572105619498985,
+      "grad_norm": 1.0374791947621171,
+      "learning_rate": 9.983641533674053e-06,
+      "loss": 2.6919,
+      "step": 115
+    },
+    {
+      "epoch": 0.15707515233581584,
+      "grad_norm": 0.923336386037252,
+      "learning_rate": 9.983202645773049e-06,
+      "loss": 2.6477,
+      "step": 116
+    },
+    {
+      "epoch": 0.15842924847664183,
+      "grad_norm": 0.8526247474365115,
+      "learning_rate": 9.982757957905204e-06,
+      "loss": 2.2264,
+      "step": 117
+    },
+    {
+      "epoch": 0.15978334461746785,
+      "grad_norm": 1.1258963168949678,
+      "learning_rate": 9.982307470588097e-06,
+      "loss": 2.5224,
+      "step": 118
+    },
+    {
+      "epoch": 0.16113744075829384,
+      "grad_norm": 0.7569639815489783,
+      "learning_rate": 9.98185118434605e-06,
+      "loss": 1.8855,
+      "step": 119
+    },
+    {
+      "epoch": 0.16249153689911983,
+      "grad_norm": 1.0355226629532543,
+      "learning_rate": 9.981389099710132e-06,
+      "loss": 3.0085,
+      "step": 120
+    },
+    {
+      "epoch": 0.16384563303994584,
+      "grad_norm": 1.1486309630139306,
+      "learning_rate": 9.980921217218173e-06,
+      "loss": 2.7178,
+      "step": 121
+    },
+    {
+      "epoch": 0.16519972918077183,
+      "grad_norm": 0.7466296304493713,
+      "learning_rate": 9.980447537414736e-06,
+      "loss": 2.251,
+      "step": 122
+    },
+    {
+      "epoch": 0.16655382532159782,
+      "grad_norm": 0.651170263411467,
+      "learning_rate": 9.979968060851144e-06,
+      "loss": 2.1519,
+      "step": 123
+    },
+    {
+      "epoch": 0.16790792146242384,
+      "grad_norm": 1.008261184550054,
+      "learning_rate": 9.979482788085455e-06,
+      "loss": 2.2199,
+      "step": 124
+    },
+    {
+      "epoch": 0.16926201760324983,
+      "grad_norm": 2.416142037739212,
+      "learning_rate": 9.978991719682486e-06,
+      "loss": 2.3314,
+      "step": 125
+    },
+    {
+      "epoch": 0.17061611374407584,
+      "grad_norm": 0.8174937699006882,
+      "learning_rate": 9.97849485621379e-06,
+      "loss": 2.5571,
+      "step": 126
+    },
+    {
+      "epoch": 0.17197020988490183,
+      "grad_norm": 1.0052455897777324,
+      "learning_rate": 9.977992198257668e-06,
+      "loss": 2.5827,
+      "step": 127
+    },
+    {
+      "epoch": 0.17332430602572782,
+      "grad_norm": 0.8663685595604006,
+      "learning_rate": 9.977483746399168e-06,
+      "loss": 2.5957,
+      "step": 128
+    },
+    {
+      "epoch": 0.17467840216655384,
+      "grad_norm": 0.8454778683126596,
+      "learning_rate": 9.976969501230074e-06,
+      "loss": 2.2387,
+      "step": 129
+    },
+    {
+      "epoch": 0.17603249830737983,
+      "grad_norm": 1.1589088141787116,
+      "learning_rate": 9.976449463348924e-06,
+      "loss": 2.721,
+      "step": 130
+    },
+    {
+      "epoch": 0.17738659444820581,
+      "grad_norm": 0.8357485092208772,
+      "learning_rate": 9.975923633360985e-06,
+      "loss": 2.1065,
+      "step": 131
+    },
+    {
+      "epoch": 0.17874069058903183,
+      "grad_norm": 1.0483474542727864,
+      "learning_rate": 9.975392011878278e-06,
+      "loss": 2.5342,
+      "step": 132
+    },
+    {
+      "epoch": 0.18009478672985782,
+      "grad_norm": 0.8310501434875607,
+      "learning_rate": 9.974854599519557e-06,
+      "loss": 2.9813,
+      "step": 133
+    },
+    {
+      "epoch": 0.1814488828706838,
+      "grad_norm": 1.1998035390299338,
+      "learning_rate": 9.974311396910317e-06,
+      "loss": 2.5218,
+      "step": 134
+    },
+    {
+      "epoch": 0.18280297901150983,
+      "grad_norm": 1.1877384820263168,
+      "learning_rate": 9.973762404682795e-06,
+      "loss": 3.0512,
+      "step": 135
+    },
+    {
+      "epoch": 0.18415707515233581,
+      "grad_norm": 0.7431690298915412,
+      "learning_rate": 9.973207623475964e-06,
+      "loss": 1.9473,
+      "step": 136
+    },
+    {
+      "epoch": 0.1855111712931618,
+      "grad_norm": 0.8085958372588746,
+      "learning_rate": 9.972647053935536e-06,
+      "loss": 2.2599,
+      "step": 137
+    },
+    {
+      "epoch": 0.18686526743398782,
+      "grad_norm": 1.0088420757147314,
+      "learning_rate": 9.972080696713962e-06,
+      "loss": 2.3532,
+      "step": 138
+    },
+    {
+      "epoch": 0.1882193635748138,
+      "grad_norm": 0.871056602633167,
+      "learning_rate": 9.971508552470424e-06,
+      "loss": 2.1344,
+      "step": 139
+    },
+    {
+      "epoch": 0.1895734597156398,
+      "grad_norm": 0.9551730724308998,
+      "learning_rate": 9.970930621870843e-06,
+      "loss": 1.7945,
+      "step": 140
+    },
+    {
+      "epoch": 0.1909275558564658,
+      "grad_norm": 0.9099059345338785,
+      "learning_rate": 9.970346905587875e-06,
+      "loss": 2.4697,
+      "step": 141
+    },
+    {
+      "epoch": 0.1922816519972918,
+      "grad_norm": 2.3602640235819288,
+      "learning_rate": 9.969757404300911e-06,
+      "loss": 2.3396,
+      "step": 142
+    },
+    {
+      "epoch": 0.19363574813811782,
+      "grad_norm": 1.0387196911625636,
+      "learning_rate": 9.969162118696072e-06,
+      "loss": 2.2526,
+      "step": 143
+    },
+    {
+      "epoch": 0.1949898442789438,
+      "grad_norm": 1.0660871258756586,
+      "learning_rate": 9.968561049466214e-06,
+      "loss": 2.2518,
+      "step": 144
+    },
+    {
+      "epoch": 0.1963439404197698,
+      "grad_norm": 0.994954020536128,
+      "learning_rate": 9.967954197310922e-06,
+      "loss": 2.1365,
+      "step": 145
+    },
+    {
+      "epoch": 0.1976980365605958,
+      "grad_norm": 1.1172902304571317,
+      "learning_rate": 9.967341562936515e-06,
+      "loss": 2.4633,
+      "step": 146
+    },
+    {
+      "epoch": 0.1990521327014218,
+      "grad_norm": 0.8851246859608983,
+      "learning_rate": 9.966723147056036e-06,
+      "loss": 2.3302,
+      "step": 147
+    },
+    {
+      "epoch": 0.2004062288422478,
+      "grad_norm": 0.9673278398671098,
+      "learning_rate": 9.966098950389268e-06,
+      "loss": 2.3481,
+      "step": 148
+    },
+    {
+      "epoch": 0.2017603249830738,
+      "grad_norm": 1.0923553526656322,
+      "learning_rate": 9.965468973662712e-06,
+      "loss": 2.7291,
+      "step": 149
+    },
+    {
+      "epoch": 0.2031144211238998,
+      "grad_norm": 0.6885203668960329,
+      "learning_rate": 9.9648332176096e-06,
+      "loss": 2.4216,
+      "step": 150
+    },
+    {
+      "epoch": 0.20446851726472579,
+      "grad_norm": 0.8835456349684598,
+      "learning_rate": 9.964191682969891e-06,
+      "loss": 2.5524,
+      "step": 151
+    },
+    {
+      "epoch": 0.2058226134055518,
+      "grad_norm": 1.110246201094945,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 2.7307,
+      "step": 152
+    },
+    {
+      "epoch": 0.2071767095463778,
+      "grad_norm": 0.8335391435494296,
+      "learning_rate": 9.962891280924148e-06,
+      "loss": 2.2845,
+      "step": 153
+    },
+    {
+      "epoch": 0.20853080568720378,
+      "grad_norm": 0.7269532952637685,
+      "learning_rate": 9.962232415031653e-06,
+      "loss": 2.2107,
+      "step": 154
+    },
+    {
+      "epoch": 0.2098849018280298,
+      "grad_norm": 0.9516173733706272,
+      "learning_rate": 9.961567773579645e-06,
+      "loss": 2.549,
+      "step": 155
+    },
+    {
+      "epoch": 0.21123899796885579,
+      "grad_norm": 0.9192964960712486,
+      "learning_rate": 9.960897357341703e-06,
+      "loss": 2.382,
+      "step": 156
+    },
+    {
+      "epoch": 0.21259309410968177,
+      "grad_norm": 0.8697364214950628,
+      "learning_rate": 9.960221167098124e-06,
+      "loss": 2.7404,
+      "step": 157
+    },
+    {
+      "epoch": 0.2139471902505078,
+      "grad_norm": 0.854037317845471,
+      "learning_rate": 9.959539203635931e-06,
+      "loss": 2.2796,
+      "step": 158
+    },
+    {
+      "epoch": 0.21530128639133378,
+      "grad_norm": 9.059153514312463,
+      "learning_rate": 9.958851467748863e-06,
+      "loss": 2.1798,
+      "step": 159
+    },
+    {
+      "epoch": 0.2166553825321598,
+      "grad_norm": 0.9333843627998799,
+      "learning_rate": 9.958157960237376e-06,
+      "loss": 2.3693,
+      "step": 160
+    },
+    {
+      "epoch": 0.21800947867298578,
+      "grad_norm": 1.3475036120106114,
+      "learning_rate": 9.957458681908647e-06,
+      "loss": 2.4024,
+      "step": 161
+    },
+    {
+      "epoch": 0.21936357481381177,
+      "grad_norm": 1.2121402320200159,
+      "learning_rate": 9.956753633576571e-06,
+      "loss": 2.5439,
+      "step": 162
+    },
+    {
+      "epoch": 0.2207176709546378,
+      "grad_norm": 0.7792210046361225,
+      "learning_rate": 9.956042816061752e-06,
+      "loss": 2.0299,
+      "step": 163
+    },
+    {
+      "epoch": 0.22207176709546378,
+      "grad_norm": 0.8226985573354776,
+      "learning_rate": 9.955326230191517e-06,
+      "loss": 2.8253,
+      "step": 164
+    },
+    {
+      "epoch": 0.22342586323628977,
+      "grad_norm": 0.8424020935830455,
+      "learning_rate": 9.9546038767999e-06,
+      "loss": 2.2754,
+      "step": 165
+    },
+    {
+      "epoch": 0.22477995937711578,
+      "grad_norm": 1.1561344660760495,
+      "learning_rate": 9.95387575672765e-06,
+      "loss": 1.9543,
+      "step": 166
+    },
+    {
+      "epoch": 0.22613405551794177,
+      "grad_norm": 0.8810904992087591,
+      "learning_rate": 9.953141870822232e-06,
+      "loss": 2.8316,
+      "step": 167
+    },
+    {
+      "epoch": 0.22748815165876776,
+      "grad_norm": 1.1327457989077157,
+      "learning_rate": 9.952402219937817e-06,
+      "loss": 2.1659,
+      "step": 168
+    },
+    {
+      "epoch": 0.22884224779959378,
+      "grad_norm": 0.847237764337931,
+      "learning_rate": 9.951656804935284e-06,
+      "loss": 2.5104,
+      "step": 169
+    },
+    {
+      "epoch": 0.23019634394041977,
+      "grad_norm": 1.3311662339401327,
+      "learning_rate": 9.950905626682229e-06,
+      "loss": 2.7411,
+      "step": 170
+    },
+    {
+      "epoch": 0.23155044008124576,
+      "grad_norm": 0.8754683872906716,
+      "learning_rate": 9.950148686052948e-06,
+      "loss": 2.2843,
+      "step": 171
+    },
+    {
+      "epoch": 0.23290453622207177,
+      "grad_norm": 1.1984097361442936,
+      "learning_rate": 9.949385983928446e-06,
+      "loss": 2.6191,
+      "step": 172
+    },
+    {
+      "epoch": 0.23425863236289776,
+      "grad_norm": 1.042317105372754,
+      "learning_rate": 9.948617521196438e-06,
+      "loss": 2.5479,
+      "step": 173
+    },
+    {
+      "epoch": 0.23561272850372375,
+      "grad_norm": 0.7787896671463191,
+      "learning_rate": 9.947843298751337e-06,
+      "loss": 2.3346,
+      "step": 174
+    },
+    {
+      "epoch": 0.23696682464454977,
+      "grad_norm": 1.2170235533278835,
+      "learning_rate": 9.947063317494265e-06,
+      "loss": 2.2332,
+      "step": 175
+    },
+    {
+      "epoch": 0.23832092078537576,
+      "grad_norm": 1.0390706514693488,
+      "learning_rate": 9.946277578333045e-06,
+      "loss": 2.5912,
+      "step": 176
+    },
+    {
+      "epoch": 0.23967501692620177,
+      "grad_norm": 1.4816304376017728,
+      "learning_rate": 9.945486082182201e-06,
+      "loss": 2.4719,
+      "step": 177
+    },
+    {
+      "epoch": 0.24102911306702776,
+      "grad_norm": 0.8487826972928669,
+      "learning_rate": 9.944688829962957e-06,
+      "loss": 2.2601,
+      "step": 178
+    },
+    {
+      "epoch": 0.24238320920785375,
+      "grad_norm": 1.0267911713585076,
+      "learning_rate": 9.94388582260324e-06,
+      "loss": 2.4376,
+      "step": 179
+    },
+    {
+      "epoch": 0.24373730534867977,
+      "grad_norm": 0.8526086813518827,
+      "learning_rate": 9.943077061037672e-06,
+      "loss": 2.6301,
+      "step": 180
+    },
+    {
+      "epoch": 0.24509140148950576,
+      "grad_norm": 1.356267698059979,
+      "learning_rate": 9.942262546207572e-06,
+      "loss": 2.723,
+      "step": 181
+    },
+    {
+      "epoch": 0.24644549763033174,
+      "grad_norm": 0.7426455232138849,
+      "learning_rate": 9.94144227906096e-06,
+      "loss": 2.4959,
+      "step": 182
+    },
+    {
+      "epoch": 0.24779959377115776,
+      "grad_norm": 0.8750438781520365,
+      "learning_rate": 9.940616260552545e-06,
+      "loss": 2.3425,
+      "step": 183
+    },
+    {
+      "epoch": 0.24915368991198375,
+      "grad_norm": 0.8785910140656594,
+      "learning_rate": 9.939784491643734e-06,
+      "loss": 2.2364,
+      "step": 184
+    },
+    {
+      "epoch": 0.25050778605280977,
+      "grad_norm": 0.8938618124220896,
+      "learning_rate": 9.938946973302624e-06,
+      "loss": 2.19,
+      "step": 185
+    },
+    {
+      "epoch": 0.25186188219363576,
+      "grad_norm": 1.021044776141451,
+      "learning_rate": 9.938103706504007e-06,
+      "loss": 2.7688,
+      "step": 186
+    },
+    {
+      "epoch": 0.25321597833446174,
+      "grad_norm": 1.2815667789992267,
+      "learning_rate": 9.937254692229363e-06,
+      "loss": 2.1036,
+      "step": 187
+    },
+    {
+      "epoch": 0.25457007447528773,
+      "grad_norm": 0.9597963383831872,
+      "learning_rate": 9.936399931466866e-06,
+      "loss": 2.7931,
+      "step": 188
+    },
+    {
+      "epoch": 0.2559241706161137,
+      "grad_norm": 1.1496492688566942,
+      "learning_rate": 9.935539425211371e-06,
+      "loss": 2.0287,
+      "step": 189
+    },
+    {
+      "epoch": 0.25727826675693977,
+      "grad_norm": 0.9029859331791751,
+      "learning_rate": 9.934673174464426e-06,
+      "loss": 2.555,
+      "step": 190
+    },
+    {
+      "epoch": 0.25863236289776576,
+      "grad_norm": 0.8623354344497336,
+      "learning_rate": 9.933801180234263e-06,
+      "loss": 2.4571,
+      "step": 191
+    },
+    {
+      "epoch": 0.25998645903859174,
+      "grad_norm": 1.172827727876581,
+      "learning_rate": 9.932923443535798e-06,
+      "loss": 2.5339,
+      "step": 192
+    },
+    {
+      "epoch": 0.26134055517941773,
+      "grad_norm": 0.7844413148427782,
+      "learning_rate": 9.932039965390634e-06,
+      "loss": 2.232,
+      "step": 193
+    },
+    {
+      "epoch": 0.2626946513202437,
+      "grad_norm": 2.4674022908729993,
+      "learning_rate": 9.931150746827055e-06,
+      "loss": 2.4686,
+      "step": 194
+    },
+    {
+      "epoch": 0.2640487474610697,
+      "grad_norm": 0.8938856211560369,
+      "learning_rate": 9.930255788880021e-06,
+      "loss": 2.9519,
+      "step": 195
+    },
+    {
+      "epoch": 0.26540284360189575,
+      "grad_norm": 0.9967031966791289,
+      "learning_rate": 9.92935509259118e-06,
+      "loss": 2.3506,
+      "step": 196
+    },
+    {
+      "epoch": 0.26675693974272174,
+      "grad_norm": 0.8181579212619565,
+      "learning_rate": 9.928448659008856e-06,
+      "loss": 2.3992,
+      "step": 197
+    },
+    {
+      "epoch": 0.26811103588354773,
+      "grad_norm": 0.9019033329287899,
+      "learning_rate": 9.927536489188047e-06,
+      "loss": 1.8896,
+      "step": 198
+    },
+    {
+      "epoch": 0.2694651320243737,
+      "grad_norm": 1.196294290686767,
+      "learning_rate": 9.926618584190435e-06,
+      "loss": 2.6578,
+      "step": 199
+    },
+    {
+      "epoch": 0.2708192281651997,
+      "grad_norm": 0.8815838170577491,
+      "learning_rate": 9.925694945084369e-06,
+      "loss": 2.8371,
+      "step": 200
+    },
+    {
+      "epoch": 0.27217332430602575,
+      "grad_norm": 0.933774886972981,
+      "learning_rate": 9.924765572944879e-06,
+      "loss": 2.0347,
+      "step": 201
+    },
+    {
+      "epoch": 0.27352742044685174,
+      "grad_norm": 0.9459867483651119,
+      "learning_rate": 9.923830468853662e-06,
+      "loss": 2.3687,
+      "step": 202
+    },
+    {
+      "epoch": 0.27488151658767773,
+      "grad_norm": 0.8281062486347639,
+      "learning_rate": 9.92288963389909e-06,
+      "loss": 2.1651,
+      "step": 203
+    },
+    {
+      "epoch": 0.2762356127285037,
+      "grad_norm": 0.7955305725702541,
+      "learning_rate": 9.921943069176203e-06,
+      "loss": 2.5368,
+      "step": 204
+    },
+    {
+      "epoch": 0.2775897088693297,
+      "grad_norm": 1.026304391265509,
+      "learning_rate": 9.920990775786712e-06,
+      "loss": 2.9058,
+      "step": 205
+    },
+    {
+      "epoch": 0.2789438050101557,
+      "grad_norm": 0.9971825829470802,
+      "learning_rate": 9.920032754838994e-06,
+      "loss": 2.8694,
+      "step": 206
+    },
+    {
+      "epoch": 0.28029790115098174,
+      "grad_norm": 1.2484687747540484,
+      "learning_rate": 9.919069007448093e-06,
+      "loss": 2.4603,
+      "step": 207
+    },
+    {
+      "epoch": 0.28165199729180773,
+      "grad_norm": 0.864015784119935,
+      "learning_rate": 9.91809953473572e-06,
+      "loss": 2.2217,
+      "step": 208
+    },
+    {
+      "epoch": 0.2830060934326337,
+      "grad_norm": 0.9647460693812894,
+      "learning_rate": 9.917124337830242e-06,
+      "loss": 2.4799,
+      "step": 209
+    },
+    {
+      "epoch": 0.2843601895734597,
+      "grad_norm": 1.0623146158662309,
+      "learning_rate": 9.916143417866702e-06,
+      "loss": 2.3235,
+      "step": 210
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.5128627766738143,
+      "learning_rate": 9.915156775986789e-06,
+      "loss": 2.6927,
+      "step": 211
+    },
+    {
+      "epoch": 0.2870683818551117,
+      "grad_norm": 0.9157505604273191,
+      "learning_rate": 9.914164413338863e-06,
+      "loss": 2.056,
+      "step": 212
+    },
+    {
+      "epoch": 0.28842247799593773,
+      "grad_norm": 2.1672553239870114,
+      "learning_rate": 9.913166331077937e-06,
+      "loss": 2.3637,
+      "step": 213
+    },
+    {
+      "epoch": 0.2897765741367637,
+      "grad_norm": 0.8565503250451203,
+      "learning_rate": 9.912162530365683e-06,
+      "loss": 2.3108,
+      "step": 214
+    },
+    {
+      "epoch": 0.2911306702775897,
+      "grad_norm": 0.8071561412028346,
+      "learning_rate": 9.911153012370427e-06,
+      "loss": 2.4094,
+      "step": 215
+    },
+    {
+      "epoch": 0.2924847664184157,
+      "grad_norm": 0.960350260627165,
+      "learning_rate": 9.910137778267153e-06,
+      "loss": 2.2326,
+      "step": 216
+    },
+    {
+      "epoch": 0.2938388625592417,
+      "grad_norm": 1.1434301374173532,
+      "learning_rate": 9.909116829237492e-06,
+      "loss": 2.1396,
+      "step": 217
+    },
+    {
+      "epoch": 0.29519295870006773,
+      "grad_norm": 1.0512962564460284,
+      "learning_rate": 9.908090166469733e-06,
+      "loss": 2.4862,
+      "step": 218
+    },
+    {
+      "epoch": 0.2965470548408937,
+      "grad_norm": 0.8803334872680652,
+      "learning_rate": 9.90705779115881e-06,
+      "loss": 2.3063,
+      "step": 219
+    },
+    {
+      "epoch": 0.2979011509817197,
+      "grad_norm": 0.7957632137271137,
+      "learning_rate": 9.90601970450631e-06,
+      "loss": 2.1209,
+      "step": 220
+    },
+    {
+      "epoch": 0.2992552471225457,
+      "grad_norm": 1.5559922150186727,
+      "learning_rate": 9.904975907720465e-06,
+      "loss": 2.745,
+      "step": 221
+    },
+    {
+      "epoch": 0.3006093432633717,
+      "grad_norm": 1.0287819888236789,
+      "learning_rate": 9.903926402016153e-06,
+      "loss": 2.3034,
+      "step": 222
+    },
+    {
+      "epoch": 0.3019634394041977,
+      "grad_norm": 0.8755596479469875,
+      "learning_rate": 9.902871188614898e-06,
+      "loss": 2.6008,
+      "step": 223
+    },
+    {
+      "epoch": 0.3033175355450237,
+      "grad_norm": 2.222977688536351,
+      "learning_rate": 9.901810268744868e-06,
+      "loss": 2.5897,
+      "step": 224
+    },
+    {
+      "epoch": 0.3046716316858497,
+      "grad_norm": 0.8294734280823934,
+      "learning_rate": 9.90074364364087e-06,
+      "loss": 2.373,
+      "step": 225
+    },
+    {
+      "epoch": 0.3060257278266757,
+      "grad_norm": 0.9955963385213202,
+      "learning_rate": 9.899671314544352e-06,
+      "loss": 2.8267,
+      "step": 226
+    },
+    {
+      "epoch": 0.3073798239675017,
+      "grad_norm": 0.9863487128858249,
+      "learning_rate": 9.898593282703402e-06,
+      "loss": 2.3585,
+      "step": 227
+    },
+    {
+      "epoch": 0.3087339201083277,
+      "grad_norm": 1.3476414208683485,
+      "learning_rate": 9.897509549372745e-06,
+      "loss": 2.0764,
+      "step": 228
+    },
+    {
+      "epoch": 0.31008801624915366,
+      "grad_norm": 1.1281976478830502,
+      "learning_rate": 9.896420115813741e-06,
+      "loss": 2.1232,
+      "step": 229
+    },
+    {
+      "epoch": 0.3114421123899797,
+      "grad_norm": 0.9905003772016358,
+      "learning_rate": 9.89532498329439e-06,
+      "loss": 2.0276,
+      "step": 230
+    },
+    {
+      "epoch": 0.3127962085308057,
+      "grad_norm": 0.9642354729606564,
+      "learning_rate": 9.894224153089313e-06,
+      "loss": 1.7903,
+      "step": 231
+    },
+    {
+      "epoch": 0.3141503046716317,
+      "grad_norm": 1.8609542881386758,
+      "learning_rate": 9.893117626479778e-06,
+      "loss": 2.6118,
+      "step": 232
+    },
+    {
+      "epoch": 0.3155044008124577,
+      "grad_norm": 1.0931602473722466,
+      "learning_rate": 9.892005404753669e-06,
+      "loss": 2.4775,
+      "step": 233
+    },
+    {
+      "epoch": 0.31685849695328366,
+      "grad_norm": 0.7800560012460497,
+      "learning_rate": 9.890887489205507e-06,
+      "loss": 2.1569,
+      "step": 234
+    },
+    {
+      "epoch": 0.3182125930941097,
+      "grad_norm": 1.150743273577776,
+      "learning_rate": 9.889763881136439e-06,
+      "loss": 2.4256,
+      "step": 235
+    },
+    {
+      "epoch": 0.3195666892349357,
+      "grad_norm": 0.9159836558576258,
+      "learning_rate": 9.888634581854235e-06,
+      "loss": 2.3495,
+      "step": 236
+    },
+    {
+      "epoch": 0.3209207853757617,
+      "grad_norm": 1.0174141433627475,
+      "learning_rate": 9.88749959267329e-06,
+      "loss": 2.8219,
+      "step": 237
+    },
+    {
+      "epoch": 0.3222748815165877,
+      "grad_norm": 0.85962610230906,
+      "learning_rate": 9.886358914914624e-06,
+      "loss": 2.3995,
+      "step": 238
+    },
+    {
+      "epoch": 0.32362897765741366,
+      "grad_norm": 0.8714138576280448,
+      "learning_rate": 9.885212549905874e-06,
+      "loss": 2.1309,
+      "step": 239
+    },
+    {
+      "epoch": 0.32498307379823965,
+      "grad_norm": 0.985120553099594,
+      "learning_rate": 9.884060498981297e-06,
+      "loss": 2.3078,
+      "step": 240
+    },
+    {
+      "epoch": 0.3263371699390657,
+      "grad_norm": 1.1509228640558309,
+      "learning_rate": 9.88290276348177e-06,
+      "loss": 2.762,
+      "step": 241
+    },
+    {
+      "epoch": 0.3276912660798917,
+      "grad_norm": 0.8569907806447795,
+      "learning_rate": 9.881739344754789e-06,
+      "loss": 2.4162,
+      "step": 242
+    },
+    {
+      "epoch": 0.3290453622207177,
+      "grad_norm": 0.8953655946187061,
+      "learning_rate": 9.880570244154455e-06,
+      "loss": 2.1708,
+      "step": 243
+    },
+    {
+      "epoch": 0.33039945836154366,
+      "grad_norm": 2.6172719168994782,
+      "learning_rate": 9.879395463041493e-06,
+      "loss": 2.4244,
+      "step": 244
+    },
+    {
+      "epoch": 0.33175355450236965,
+      "grad_norm": 1.1692362620244634,
+      "learning_rate": 9.87821500278323e-06,
+      "loss": 2.3573,
+      "step": 245
+    },
+    {
+      "epoch": 0.33310765064319564,
+      "grad_norm": 1.3260737783720347,
+      "learning_rate": 9.877028864753614e-06,
+      "loss": 2.2204,
+      "step": 246
+    },
+    {
+      "epoch": 0.3344617467840217,
+      "grad_norm": 1.1284130571617974,
+      "learning_rate": 9.87583705033319e-06,
+      "loss": 2.3806,
+      "step": 247
+    },
+    {
+      "epoch": 0.3358158429248477,
+      "grad_norm": 0.8740876722841778,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 1.8413,
+      "step": 248
+    },
+    {
+      "epoch": 0.33716993906567366,
+      "grad_norm": 0.8380346921284223,
+      "learning_rate": 9.87343639787516e-06,
+      "loss": 2.0977,
+      "step": 249
+    },
+    {
+      "epoch": 0.33852403520649965,
+      "grad_norm": 0.8382850128591168,
+      "learning_rate": 9.87222756263168e-06,
+      "loss": 2.1709,
+      "step": 250
+    },
+    {
+      "epoch": 0.33987813134732564,
+      "grad_norm": 1.1238927174638607,
+      "learning_rate": 9.871013056585646e-06,
+      "loss": 2.4974,
+      "step": 251
+    },
+    {
+      "epoch": 0.3412322274881517,
+      "grad_norm": 0.7904179666008488,
+      "learning_rate": 9.869792881150624e-06,
+      "loss": 2.6544,
+      "step": 252
+    },
+    {
+      "epoch": 0.3425863236289777,
+      "grad_norm": 1.5604245138186315,
+      "learning_rate": 9.868567037746784e-06,
+      "loss": 2.4196,
+      "step": 253
+    },
+    {
+      "epoch": 0.34394041976980366,
+      "grad_norm": 0.9541984067972759,
+      "learning_rate": 9.867335527800887e-06,
+      "loss": 2.437,
+      "step": 254
+    },
+    {
+      "epoch": 0.34529451591062965,
+      "grad_norm": 1.1385820836347318,
+      "learning_rate": 9.866098352746295e-06,
+      "loss": 2.6164,
+      "step": 255
+    },
+    {
+      "epoch": 0.34664861205145564,
+      "grad_norm": 0.8579010274341586,
+      "learning_rate": 9.864855514022955e-06,
+      "loss": 2.0521,
+      "step": 256
+    },
+    {
+      "epoch": 0.34800270819228163,
+      "grad_norm": 1.0456339797858754,
+      "learning_rate": 9.863607013077414e-06,
+      "loss": 1.8313,
+      "step": 257
+    },
+    {
+      "epoch": 0.3493568043331077,
+      "grad_norm": 0.9189984940983889,
+      "learning_rate": 9.862352851362808e-06,
+      "loss": 2.2923,
+      "step": 258
+    },
+    {
+      "epoch": 0.35071090047393366,
+      "grad_norm": 1.0071081750591437,
+      "learning_rate": 9.861093030338859e-06,
+      "loss": 2.5982,
+      "step": 259
+    },
+    {
+      "epoch": 0.35206499661475965,
+      "grad_norm": 0.7791457829988662,
+      "learning_rate": 9.859827551471877e-06,
+      "loss": 2.1629,
+      "step": 260
+    },
+    {
+      "epoch": 0.35341909275558564,
+      "grad_norm": 1.3508700131079088,
+      "learning_rate": 9.858556416234755e-06,
+      "loss": 2.3283,
+      "step": 261
+    },
+    {
+      "epoch": 0.35477318889641163,
+      "grad_norm": 1.4779336761646127,
+      "learning_rate": 9.857279626106975e-06,
+      "loss": 2.1788,
+      "step": 262
+    },
+    {
+      "epoch": 0.3561272850372376,
+      "grad_norm": 0.8287354033035986,
+      "learning_rate": 9.855997182574598e-06,
+      "loss": 2.4326,
+      "step": 263
+    },
+    {
+      "epoch": 0.35748138117806366,
+      "grad_norm": 1.2376776693679805,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 1.9501,
+      "step": 264
+    },
+    {
+      "epoch": 0.35883547731888965,
+      "grad_norm": 0.9140324089635932,
+      "learning_rate": 9.853415341273185e-06,
+      "loss": 2.1236,
+      "step": 265
+    },
+    {
+      "epoch": 0.36018957345971564,
+      "grad_norm": 0.8399357799885816,
+      "learning_rate": 9.852115946509163e-06,
+      "loss": 2.2648,
+      "step": 266
+    },
+    {
+      "epoch": 0.36154366960054163,
+      "grad_norm": 1.0405854607468001,
+      "learning_rate": 9.85081090435057e-06,
+      "loss": 2.2106,
+      "step": 267
+    },
+    {
+      "epoch": 0.3628977657413676,
+      "grad_norm": 0.8825320659362774,
+      "learning_rate": 9.849500216316346e-06,
+      "loss": 1.9139,
+      "step": 268
+    },
+    {
+      "epoch": 0.36425186188219366,
+      "grad_norm": 0.8884433268338788,
+      "learning_rate": 9.848183883932003e-06,
+      "loss": 2.3441,
+      "step": 269
+    },
+    {
+      "epoch": 0.36560595802301965,
+      "grad_norm": 0.8144408969478175,
+      "learning_rate": 9.846861908729628e-06,
+      "loss": 2.7794,
+      "step": 270
+    },
+    {
+      "epoch": 0.36696005416384564,
+      "grad_norm": 3.0643775465082093,
+      "learning_rate": 9.845534292247872e-06,
+      "loss": 2.6927,
+      "step": 271
+    },
+    {
+      "epoch": 0.36831415030467163,
+      "grad_norm": 1.0918166740808275,
+      "learning_rate": 9.844201036031952e-06,
+      "loss": 2.6845,
+      "step": 272
+    },
+    {
+      "epoch": 0.3696682464454976,
+      "grad_norm": 1.0782629441989242,
+      "learning_rate": 9.84286214163365e-06,
+      "loss": 2.1323,
+      "step": 273
+    },
+    {
+      "epoch": 0.3710223425863236,
+      "grad_norm": 0.899216159697703,
+      "learning_rate": 9.841517610611309e-06,
+      "loss": 2.5939,
+      "step": 274
+    },
+    {
+      "epoch": 0.37237643872714965,
+      "grad_norm": 0.8784697558034725,
+      "learning_rate": 9.840167444529834e-06,
+      "loss": 2.0398,
+      "step": 275
+    },
+    {
+      "epoch": 0.37373053486797564,
+      "grad_norm": 0.7893703178184768,
+      "learning_rate": 9.838811644960686e-06,
+      "loss": 2.3489,
+      "step": 276
+    },
+    {
+      "epoch": 0.37508463100880163,
+      "grad_norm": 1.5759025892565637,
+      "learning_rate": 9.837450213481888e-06,
+      "loss": 2.1618,
+      "step": 277
+    },
+    {
+      "epoch": 0.3764387271496276,
+      "grad_norm": 0.8915508527852146,
+      "learning_rate": 9.836083151678014e-06,
+      "loss": 2.2966,
+      "step": 278
+    },
+    {
+      "epoch": 0.3777928232904536,
+      "grad_norm": 0.9656457496381649,
+      "learning_rate": 9.834710461140191e-06,
+      "loss": 2.2487,
+      "step": 279
+    },
+    {
+      "epoch": 0.3791469194312796,
+      "grad_norm": 1.0512896665744822,
+      "learning_rate": 9.833332143466099e-06,
+      "loss": 2.4065,
+      "step": 280
+    },
+    {
+      "epoch": 0.38050101557210564,
+      "grad_norm": 1.0658387523528818,
+      "learning_rate": 9.831948200259966e-06,
+      "loss": 2.4469,
+      "step": 281
+    },
+    {
+      "epoch": 0.3818551117129316,
+      "grad_norm": 1.117757922229749,
+      "learning_rate": 9.830558633132568e-06,
+      "loss": 2.2469,
+      "step": 282
+    },
+    {
+      "epoch": 0.3832092078537576,
+      "grad_norm": 0.8846882799705557,
+      "learning_rate": 9.82916344370123e-06,
+      "loss": 1.8672,
+      "step": 283
+    },
+    {
+      "epoch": 0.3845633039945836,
+      "grad_norm": 0.8829678982284991,
+      "learning_rate": 9.827762633589813e-06,
+      "loss": 2.3709,
+      "step": 284
+    },
+    {
+      "epoch": 0.3859174001354096,
+      "grad_norm": 1.5559611510853741,
+      "learning_rate": 9.826356204428726e-06,
+      "loss": 2.531,
+      "step": 285
+    },
+    {
+      "epoch": 0.38727149627623564,
+      "grad_norm": 0.8832988192797496,
+      "learning_rate": 9.82494415785492e-06,
+      "loss": 2.0908,
+      "step": 286
+    },
+    {
+      "epoch": 0.3886255924170616,
+      "grad_norm": 0.893313058823839,
+      "learning_rate": 9.82352649551188e-06,
+      "loss": 2.2141,
+      "step": 287
+    },
+    {
+      "epoch": 0.3899796885578876,
+      "grad_norm": 1.915840114647375,
+      "learning_rate": 9.822103219049625e-06,
+      "loss": 2.3759,
+      "step": 288
+    },
+    {
+      "epoch": 0.3913337846987136,
+      "grad_norm": 1.091811159406276,
+      "learning_rate": 9.820674330124716e-06,
+      "loss": 1.949,
+      "step": 289
+    },
+    {
+      "epoch": 0.3926878808395396,
+      "grad_norm": 0.9101585142251976,
+      "learning_rate": 9.819239830400238e-06,
+      "loss": 2.2837,
+      "step": 290
+    },
+    {
+      "epoch": 0.3940419769803656,
+      "grad_norm": 0.8669554630795423,
+      "learning_rate": 9.81779972154581e-06,
+      "loss": 2.3568,
+      "step": 291
+    },
+    {
+      "epoch": 0.3953960731211916,
+      "grad_norm": 1.0812216653348674,
+      "learning_rate": 9.816354005237583e-06,
+      "loss": 2.5594,
+      "step": 292
+    },
+    {
+      "epoch": 0.3967501692620176,
+      "grad_norm": 1.0135729926732555,
+      "learning_rate": 9.814902683158227e-06,
+      "loss": 2.2677,
+      "step": 293
+    },
+    {
+      "epoch": 0.3981042654028436,
+      "grad_norm": 1.1354280196245004,
+      "learning_rate": 9.813445756996946e-06,
+      "loss": 2.1235,
+      "step": 294
+    },
+    {
+      "epoch": 0.3994583615436696,
+      "grad_norm": 0.8872371772513353,
+      "learning_rate": 9.811983228449457e-06,
+      "loss": 2.6214,
+      "step": 295
+    },
+    {
+      "epoch": 0.4008124576844956,
+      "grad_norm": 1.198158875138794,
+      "learning_rate": 9.810515099218004e-06,
+      "loss": 2.5482,
+      "step": 296
+    },
+    {
+      "epoch": 0.40216655382532157,
+      "grad_norm": 1.122073024930223,
+      "learning_rate": 9.809041371011347e-06,
+      "loss": 2.3414,
+      "step": 297
+    },
+    {
+      "epoch": 0.4035206499661476,
+      "grad_norm": 0.7450105063153143,
+      "learning_rate": 9.807562045544764e-06,
+      "loss": 2.5729,
+      "step": 298
+    },
+    {
+      "epoch": 0.4048747461069736,
+      "grad_norm": 1.096232419286508,
+      "learning_rate": 9.806077124540045e-06,
+      "loss": 2.5012,
+      "step": 299
+    },
+    {
+      "epoch": 0.4062288422477996,
+      "grad_norm": 1.0045118062203406,
+      "learning_rate": 9.804586609725499e-06,
+      "loss": 2.0684,
+      "step": 300
+    },
+    {
+      "epoch": 0.4075829383886256,
+      "grad_norm": 1.0368766866449026,
+      "learning_rate": 9.803090502835938e-06,
+      "loss": 2.1287,
+      "step": 301
+    },
+    {
+      "epoch": 0.40893703452945157,
+      "grad_norm": 1.0208588144634514,
+      "learning_rate": 9.801588805612685e-06,
+      "loss": 2.3341,
+      "step": 302
+    },
+    {
+      "epoch": 0.4102911306702776,
+      "grad_norm": 1.0454209454651813,
+      "learning_rate": 9.800081519803575e-06,
+      "loss": 2.088,
+      "step": 303
+    },
+    {
+      "epoch": 0.4116452268111036,
+      "grad_norm": 1.0546734132886375,
+      "learning_rate": 9.798568647162939e-06,
+      "loss": 2.3576,
+      "step": 304
+    },
+    {
+      "epoch": 0.4129993229519296,
+      "grad_norm": 1.1121493752708365,
+      "learning_rate": 9.797050189451615e-06,
+      "loss": 2.1073,
+      "step": 305
+    },
+    {
+      "epoch": 0.4143534190927556,
+      "grad_norm": 1.1189378911389287,
+      "learning_rate": 9.795526148436945e-06,
+      "loss": 2.308,
+      "step": 306
+    },
+    {
+      "epoch": 0.41570751523358157,
+      "grad_norm": 0.9899429575476786,
+      "learning_rate": 9.793996525892762e-06,
+      "loss": 2.2555,
+      "step": 307
+    },
+    {
+      "epoch": 0.41706161137440756,
+      "grad_norm": 1.0479984863056457,
+      "learning_rate": 9.7924613235994e-06,
+      "loss": 2.3882,
+      "step": 308
+    },
+    {
+      "epoch": 0.4184157075152336,
+      "grad_norm": 0.8437095271965467,
+      "learning_rate": 9.790920543343686e-06,
+      "loss": 2.8099,
+      "step": 309
+    },
+    {
+      "epoch": 0.4197698036560596,
+      "grad_norm": 1.1267645343143333,
+      "learning_rate": 9.78937418691894e-06,
+      "loss": 2.377,
+      "step": 310
+    },
+    {
+      "epoch": 0.4211238997968856,
+      "grad_norm": 0.954265542677354,
+      "learning_rate": 9.787822256124972e-06,
+      "loss": 2.2119,
+      "step": 311
+    },
+    {
+      "epoch": 0.42247799593771157,
+      "grad_norm": 1.0897928699498936,
+      "learning_rate": 9.78626475276808e-06,
+      "loss": 2.5392,
+      "step": 312
+    },
+    {
+      "epoch": 0.42383209207853756,
+      "grad_norm": 0.8557954442168854,
+      "learning_rate": 9.784701678661045e-06,
+      "loss": 2.0243,
+      "step": 313
+    },
+    {
+      "epoch": 0.42518618821936355,
+      "grad_norm": 1.007998713387866,
+      "learning_rate": 9.783133035623136e-06,
+      "loss": 2.0593,
+      "step": 314
+    },
+    {
+      "epoch": 0.4265402843601896,
+      "grad_norm": 0.8245127331943697,
+      "learning_rate": 9.781558825480104e-06,
+      "loss": 1.9544,
+      "step": 315
+    },
+    {
+      "epoch": 0.4278943805010156,
+      "grad_norm": 1.1343475642376186,
+      "learning_rate": 9.779979050064174e-06,
+      "loss": 2.1917,
+      "step": 316
+    },
+    {
+      "epoch": 0.42924847664184157,
+      "grad_norm": 1.1336908348453985,
+      "learning_rate": 9.778393711214054e-06,
+      "loss": 2.3615,
+      "step": 317
+    },
+    {
+      "epoch": 0.43060257278266756,
+      "grad_norm": 1.0074004622651764,
+      "learning_rate": 9.776802810774924e-06,
+      "loss": 2.4642,
+      "step": 318
+    },
+    {
+      "epoch": 0.43195666892349355,
+      "grad_norm": 0.869836917179775,
+      "learning_rate": 9.77520635059844e-06,
+      "loss": 2.2773,
+      "step": 319
+    },
+    {
+      "epoch": 0.4333107650643196,
+      "grad_norm": 0.8822131547317252,
+      "learning_rate": 9.77360433254273e-06,
+      "loss": 2.5491,
+      "step": 320
+    },
+    {
+      "epoch": 0.4346648612051456,
+      "grad_norm": 0.9871053514468092,
+      "learning_rate": 9.771996758472381e-06,
+      "loss": 2.5086,
+      "step": 321
+    },
+    {
+      "epoch": 0.43601895734597157,
+      "grad_norm": 0.8810949134669234,
+      "learning_rate": 9.770383630258463e-06,
+      "loss": 2.4492,
+      "step": 322
+    },
+    {
+      "epoch": 0.43737305348679756,
+      "grad_norm": 0.9634447042117443,
+      "learning_rate": 9.768764949778495e-06,
+      "loss": 2.4698,
+      "step": 323
+    },
+    {
+      "epoch": 0.43872714962762355,
+      "grad_norm": 0.9578841465826307,
+      "learning_rate": 9.767140718916467e-06,
+      "loss": 2.8468,
+      "step": 324
+    },
+    {
+      "epoch": 0.44008124576844954,
+      "grad_norm": 1.045773660121023,
+      "learning_rate": 9.765510939562827e-06,
+      "loss": 2.4331,
+      "step": 325
+    },
+    {
+      "epoch": 0.4414353419092756,
+      "grad_norm": 1.5236244191611785,
+      "learning_rate": 9.763875613614482e-06,
+      "loss": 2.4025,
+      "step": 326
+    },
+    {
+      "epoch": 0.44278943805010157,
+      "grad_norm": 0.9193628457370401,
+      "learning_rate": 9.762234742974793e-06,
+      "loss": 2.2136,
+      "step": 327
+    },
+    {
+      "epoch": 0.44414353419092756,
+      "grad_norm": 0.8949925268143644,
+      "learning_rate": 9.76058832955357e-06,
+      "loss": 2.4088,
+      "step": 328
+    },
+    {
+      "epoch": 0.44549763033175355,
+      "grad_norm": 2.0556356286018413,
+      "learning_rate": 9.758936375267087e-06,
+      "loss": 2.1351,
+      "step": 329
+    },
+    {
+      "epoch": 0.44685172647257954,
+      "grad_norm": 1.0444826822145854,
+      "learning_rate": 9.757278882038056e-06,
+      "loss": 2.7799,
+      "step": 330
+    },
+    {
+      "epoch": 0.4482058226134055,
+      "grad_norm": 1.0824437177931403,
+      "learning_rate": 9.755615851795639e-06,
+      "loss": 1.8668,
+      "step": 331
+    },
+    {
+      "epoch": 0.44955991875423157,
+      "grad_norm": 1.1111629995250478,
+      "learning_rate": 9.753947286475442e-06,
+      "loss": 2.754,
+      "step": 332
+    },
+    {
+      "epoch": 0.45091401489505756,
+      "grad_norm": 0.9573737568401282,
+      "learning_rate": 9.752273188019514e-06,
+      "loss": 2.5329,
+      "step": 333
+    },
+    {
+      "epoch": 0.45226811103588355,
+      "grad_norm": 1.0357238977124108,
+      "learning_rate": 9.750593558376347e-06,
+      "loss": 2.2007,
+      "step": 334
+    },
+    {
+      "epoch": 0.45362220717670954,
+      "grad_norm": 0.9095487712099384,
+      "learning_rate": 9.748908399500863e-06,
+      "loss": 1.8825,
+      "step": 335
+    },
+    {
+      "epoch": 0.4549763033175355,
+      "grad_norm": 1.7445295348550816,
+      "learning_rate": 9.747217713354428e-06,
+      "loss": 2.5305,
+      "step": 336
+    },
+    {
+      "epoch": 0.45633039945836157,
+      "grad_norm": 0.9070558170854458,
+      "learning_rate": 9.745521501904835e-06,
+      "loss": 2.1951,
+      "step": 337
+    },
+    {
+      "epoch": 0.45768449559918756,
+      "grad_norm": 0.8995516521874707,
+      "learning_rate": 9.743819767126312e-06,
+      "loss": 2.2137,
+      "step": 338
+    },
+    {
+      "epoch": 0.45903859174001355,
+      "grad_norm": 1.9426471289696312,
+      "learning_rate": 9.742112510999516e-06,
+      "loss": 2.3716,
+      "step": 339
+    },
+    {
+      "epoch": 0.46039268788083954,
+      "grad_norm": 0.9725576998378328,
+      "learning_rate": 9.740399735511524e-06,
+      "loss": 2.6405,
+      "step": 340
+    },
+    {
+      "epoch": 0.4617467840216655,
+      "grad_norm": 0.8305802783581133,
+      "learning_rate": 9.738681442655842e-06,
+      "loss": 2.2231,
+      "step": 341
+    },
+    {
+      "epoch": 0.4631008801624915,
+      "grad_norm": 0.9734526439933455,
+      "learning_rate": 9.736957634432398e-06,
+      "loss": 2.3041,
+      "step": 342
+    },
+    {
+      "epoch": 0.46445497630331756,
+      "grad_norm": 0.824808147309629,
+      "learning_rate": 9.73522831284754e-06,
+      "loss": 2.3024,
+      "step": 343
+    },
+    {
+      "epoch": 0.46580907244414355,
+      "grad_norm": 0.9599600802177057,
+      "learning_rate": 9.733493479914031e-06,
+      "loss": 2.3563,
+      "step": 344
+    },
+    {
+      "epoch": 0.46716316858496953,
+      "grad_norm": 0.9340063834780592,
+      "learning_rate": 9.731753137651047e-06,
+      "loss": 2.1237,
+      "step": 345
+    },
+    {
+      "epoch": 0.4685172647257955,
+      "grad_norm": 1.562533249030556,
+      "learning_rate": 9.730007288084178e-06,
+      "loss": 2.3575,
+      "step": 346
+    },
+    {
+      "epoch": 0.4698713608666215,
+      "grad_norm": 0.9490993069469829,
+      "learning_rate": 9.728255933245428e-06,
+      "loss": 2.8822,
+      "step": 347
+    },
+    {
+      "epoch": 0.4712254570074475,
+      "grad_norm": 0.9566724872274933,
+      "learning_rate": 9.726499075173201e-06,
+      "loss": 2.3402,
+      "step": 348
+    },
+    {
+      "epoch": 0.47257955314827355,
+      "grad_norm": 1.008513385617003,
+      "learning_rate": 9.724736715912313e-06,
+      "loss": 2.1406,
+      "step": 349
+    },
+    {
+      "epoch": 0.47393364928909953,
+      "grad_norm": 1.1766610712644368,
+      "learning_rate": 9.72296885751398e-06,
+      "loss": 2.1934,
+      "step": 350
+    },
+    {
+      "epoch": 0.4752877454299255,
+      "grad_norm": 0.9325244223791722,
+      "learning_rate": 9.721195502035817e-06,
+      "loss": 1.9835,
+      "step": 351
+    },
+    {
+      "epoch": 0.4766418415707515,
+      "grad_norm": 0.837427966240756,
+      "learning_rate": 9.719416651541839e-06,
+      "loss": 2.481,
+      "step": 352
+    },
+    {
+      "epoch": 0.4779959377115775,
+      "grad_norm": 1.02007044942659,
+      "learning_rate": 9.717632308102455e-06,
+      "loss": 2.7148,
+      "step": 353
+    },
+    {
+      "epoch": 0.47935003385240355,
+      "grad_norm": 0.959116878107154,
+      "learning_rate": 9.715842473794472e-06,
+      "loss": 2.4738,
+      "step": 354
+    },
+    {
+      "epoch": 0.48070412999322953,
+      "grad_norm": 1.0102624905029922,
+      "learning_rate": 9.714047150701082e-06,
+      "loss": 2.0945,
+      "step": 355
+    },
+    {
+      "epoch": 0.4820582261340555,
+      "grad_norm": 2.0984222479912438,
+      "learning_rate": 9.712246340911866e-06,
+      "loss": 2.1697,
+      "step": 356
+    },
+    {
+      "epoch": 0.4834123222748815,
+      "grad_norm": 0.917396475775467,
+      "learning_rate": 9.710440046522797e-06,
+      "loss": 2.2475,
+      "step": 357
+    },
+    {
+      "epoch": 0.4847664184157075,
+      "grad_norm": 1.005361407668732,
+      "learning_rate": 9.708628269636224e-06,
+      "loss": 2.2307,
+      "step": 358
+    },
+    {
+      "epoch": 0.4861205145565335,
+      "grad_norm": 1.0309046462532379,
+      "learning_rate": 9.706811012360882e-06,
+      "loss": 2.4654,
+      "step": 359
+    },
+    {
+      "epoch": 0.48747461069735953,
+      "grad_norm": 3.0659420771722834,
+      "learning_rate": 9.704988276811883e-06,
+      "loss": 2.7097,
+      "step": 360
+    },
+    {
+      "epoch": 0.4888287068381855,
+      "grad_norm": 1.0772251854315056,
+      "learning_rate": 9.703160065110716e-06,
+      "loss": 2.3728,
+      "step": 361
+    },
+    {
+      "epoch": 0.4901828029790115,
+      "grad_norm": 1.1656105776873809,
+      "learning_rate": 9.701326379385238e-06,
+      "loss": 2.6172,
+      "step": 362
+    },
+    {
+      "epoch": 0.4915368991198375,
+      "grad_norm": 1.0293153712278362,
+      "learning_rate": 9.699487221769687e-06,
+      "loss": 2.427,
+      "step": 363
+    },
+    {
+      "epoch": 0.4928909952606635,
+      "grad_norm": 1.1554858971025437,
+      "learning_rate": 9.697642594404666e-06,
+      "loss": 2.3854,
+      "step": 364
+    },
+    {
+      "epoch": 0.4942450914014895,
+      "grad_norm": 1.1552799937404215,
+      "learning_rate": 9.69579249943714e-06,
+      "loss": 2.2642,
+      "step": 365
+    },
+    {
+      "epoch": 0.4955991875423155,
+      "grad_norm": 1.3024770839893858,
+      "learning_rate": 9.693936939020441e-06,
+      "loss": 2.7555,
+      "step": 366
+    },
+    {
+      "epoch": 0.4969532836831415,
+      "grad_norm": 0.9139562471712885,
+      "learning_rate": 9.692075915314265e-06,
+      "loss": 1.9849,
+      "step": 367
+    },
+    {
+      "epoch": 0.4983073798239675,
+      "grad_norm": 1.1108979731718802,
+      "learning_rate": 9.69020943048466e-06,
+      "loss": 2.7987,
+      "step": 368
+    },
+    {
+      "epoch": 0.4996614759647935,
+      "grad_norm": 1.043554360812148,
+      "learning_rate": 9.688337486704038e-06,
+      "loss": 2.379,
+      "step": 369
+    },
+    {
+      "epoch": 0.5010155721056195,
+      "grad_norm": 0.8482819857067955,
+      "learning_rate": 9.686460086151159e-06,
+      "loss": 2.3031,
+      "step": 370
+    },
+    {
+      "epoch": 0.5023696682464455,
+      "grad_norm": 0.882893031313843,
+      "learning_rate": 9.684577231011134e-06,
+      "loss": 1.9398,
+      "step": 371
+    },
+    {
+      "epoch": 0.5037237643872715,
+      "grad_norm": 0.8900569461037698,
+      "learning_rate": 9.68268892347543e-06,
+      "loss": 2.3579,
+      "step": 372
+    },
+    {
+      "epoch": 0.5050778605280974,
+      "grad_norm": 1.027440419577199,
+      "learning_rate": 9.680795165741849e-06,
+      "loss": 1.8953,
+      "step": 373
+    },
+    {
+      "epoch": 0.5064319566689235,
+      "grad_norm": 1.2020524441224305,
+      "learning_rate": 9.678895960014545e-06,
+      "loss": 2.6735,
+      "step": 374
+    },
+    {
+      "epoch": 0.5077860528097495,
+      "grad_norm": 1.6359343554936903,
+      "learning_rate": 9.676991308504012e-06,
+      "loss": 2.5313,
+      "step": 375
+    },
+    {
+      "epoch": 0.5091401489505755,
+      "grad_norm": 0.9318914822552815,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 2.7319,
+      "step": 376
+    },
+    {
+      "epoch": 0.5104942450914015,
+      "grad_norm": 1.0629414040228036,
+      "learning_rate": 9.673165677006906e-06,
+      "loss": 2.5041,
+      "step": 377
+    },
+    {
+      "epoch": 0.5118483412322274,
+      "grad_norm": 0.9015543787336042,
+      "learning_rate": 9.671244701472999e-06,
+      "loss": 2.2052,
+      "step": 378
+    },
+    {
+      "epoch": 0.5132024373730535,
+      "grad_norm": 0.9603399577043747,
+      "learning_rate": 9.669318289061191e-06,
+      "loss": 2.5069,
+      "step": 379
+    },
+    {
+      "epoch": 0.5145565335138795,
+      "grad_norm": 0.8421778244290412,
+      "learning_rate": 9.667386442013634e-06,
+      "loss": 2.6313,
+      "step": 380
+    },
+    {
+      "epoch": 0.5159106296547055,
+      "grad_norm": 0.8356138567382542,
+      "learning_rate": 9.665449162578814e-06,
+      "loss": 2.0145,
+      "step": 381
+    },
+    {
+      "epoch": 0.5172647257955315,
+      "grad_norm": 0.9669755721462949,
+      "learning_rate": 9.663506453011538e-06,
+      "loss": 2.2227,
+      "step": 382
+    },
+    {
+      "epoch": 0.5186188219363574,
+      "grad_norm": 1.163959225455655,
+      "learning_rate": 9.66155831557293e-06,
+      "loss": 2.5775,
+      "step": 383
+    },
+    {
+      "epoch": 0.5199729180771835,
+      "grad_norm": 0.9362192144637581,
+      "learning_rate": 9.659604752530434e-06,
+      "loss": 2.2014,
+      "step": 384
+    },
+    {
+      "epoch": 0.5213270142180095,
+      "grad_norm": 0.9892445662855336,
+      "learning_rate": 9.657645766157813e-06,
+      "loss": 2.4195,
+      "step": 385
+    },
+    {
+      "epoch": 0.5226811103588355,
+      "grad_norm": 0.8722826858897498,
+      "learning_rate": 9.655681358735134e-06,
+      "loss": 2.6966,
+      "step": 386
+    },
+    {
+      "epoch": 0.5240352064996615,
+      "grad_norm": 1.0196349369430946,
+      "learning_rate": 9.653711532548778e-06,
+      "loss": 2.4237,
+      "step": 387
+    },
+    {
+      "epoch": 0.5253893026404874,
+      "grad_norm": 1.0011773469402157,
+      "learning_rate": 9.651736289891434e-06,
+      "loss": 2.3765,
+      "step": 388
+    },
+    {
+      "epoch": 0.5267433987813135,
+      "grad_norm": 0.9843650150083161,
+      "learning_rate": 9.649755633062092e-06,
+      "loss": 2.4698,
+      "step": 389
+    },
+    {
+      "epoch": 0.5280974949221394,
+      "grad_norm": 1.2199752234970775,
+      "learning_rate": 9.647769564366048e-06,
+      "loss": 2.2025,
+      "step": 390
+    },
+    {
+      "epoch": 0.5294515910629655,
+      "grad_norm": 1.048526738243433,
+      "learning_rate": 9.645778086114892e-06,
+      "loss": 2.5315,
+      "step": 391
+    },
+    {
+      "epoch": 0.5308056872037915,
+      "grad_norm": 1.0079058862169128,
+      "learning_rate": 9.643781200626512e-06,
+      "loss": 2.2688,
+      "step": 392
+    },
+    {
+      "epoch": 0.5321597833446174,
+      "grad_norm": 0.9175841000628026,
+      "learning_rate": 9.641778910225093e-06,
+      "loss": 2.4924,
+      "step": 393
+    },
+    {
+      "epoch": 0.5335138794854435,
+      "grad_norm": 0.9394126148321758,
+      "learning_rate": 9.639771217241104e-06,
+      "loss": 2.6604,
+      "step": 394
+    },
+    {
+      "epoch": 0.5348679756262694,
+      "grad_norm": 1.8536643574976448,
+      "learning_rate": 9.637758124011307e-06,
+      "loss": 2.3891,
+      "step": 395
+    },
+    {
+      "epoch": 0.5362220717670955,
+      "grad_norm": 1.353118661391877,
+      "learning_rate": 9.63573963287875e-06,
+      "loss": 2.2486,
+      "step": 396
+    },
+    {
+      "epoch": 0.5375761679079215,
+      "grad_norm": 0.9114437873284958,
+      "learning_rate": 9.633715746192762e-06,
+      "loss": 2.9009,
+      "step": 397
+    },
+    {
+      "epoch": 0.5389302640487474,
+      "grad_norm": 0.9178272536878976,
+      "learning_rate": 9.631686466308947e-06,
+      "loss": 2.6545,
+      "step": 398
+    },
+    {
+      "epoch": 0.5402843601895735,
+      "grad_norm": 0.8349824532422451,
+      "learning_rate": 9.629651795589197e-06,
+      "loss": 2.3238,
+      "step": 399
+    },
+    {
+      "epoch": 0.5416384563303994,
+      "grad_norm": 0.8002054931156397,
+      "learning_rate": 9.627611736401668e-06,
+      "loss": 2.1503,
+      "step": 400
+    },
+    {
+      "epoch": 0.5429925524712255,
+      "grad_norm": 1.014043339956385,
+      "learning_rate": 9.625566291120794e-06,
+      "loss": 2.3366,
+      "step": 401
+    },
+    {
+      "epoch": 0.5443466486120515,
+      "grad_norm": 0.9753524140226457,
+      "learning_rate": 9.623515462127276e-06,
+      "loss": 2.4777,
+      "step": 402
+    },
+    {
+      "epoch": 0.5457007447528774,
+      "grad_norm": 0.9904758899407581,
+      "learning_rate": 9.621459251808078e-06,
+      "loss": 2.5355,
+      "step": 403
+    },
+    {
+      "epoch": 0.5470548408937035,
+      "grad_norm": 0.9176434471710123,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 2.1775,
+      "step": 404
+    },
+    {
+      "epoch": 0.5484089370345294,
+      "grad_norm": 0.9994849139704428,
+      "learning_rate": 9.617330696771834e-06,
+      "loss": 2.3618,
+      "step": 405
+    },
+    {
+      "epoch": 0.5497630331753555,
+      "grad_norm": 0.9718579619606619,
+      "learning_rate": 9.615258356860027e-06,
+      "loss": 2.065,
+      "step": 406
+    },
+    {
+      "epoch": 0.5511171293161814,
+      "grad_norm": 0.8890137365695405,
+      "learning_rate": 9.613180645233014e-06,
+      "loss": 2.153,
+      "step": 407
+    },
+    {
+      "epoch": 0.5524712254570074,
+      "grad_norm": 1.001837019128747,
+      "learning_rate": 9.611097564309054e-06,
+      "loss": 2.4168,
+      "step": 408
+    },
+    {
+      "epoch": 0.5538253215978335,
+      "grad_norm": 1.205118639472601,
+      "learning_rate": 9.609009116512648e-06,
+      "loss": 2.5229,
+      "step": 409
+    },
+    {
+      "epoch": 0.5551794177386594,
+      "grad_norm": 0.9757578479246173,
+      "learning_rate": 9.60691530427455e-06,
+      "loss": 2.1174,
+      "step": 410
+    },
+    {
+      "epoch": 0.5565335138794855,
+      "grad_norm": 0.9985946290929901,
+      "learning_rate": 9.60481613003176e-06,
+      "loss": 1.9395,
+      "step": 411
+    },
+    {
+      "epoch": 0.5578876100203114,
+      "grad_norm": 0.9058862928790223,
+      "learning_rate": 9.602711596227507e-06,
+      "loss": 2.1988,
+      "step": 412
+    },
+    {
+      "epoch": 0.5592417061611374,
+      "grad_norm": 0.9620888369560685,
+      "learning_rate": 9.600601705311267e-06,
+      "loss": 2.3826,
+      "step": 413
+    },
+    {
+      "epoch": 0.5605958023019635,
+      "grad_norm": 1.0478032599035778,
+      "learning_rate": 9.598486459738751e-06,
+      "loss": 2.4528,
+      "step": 414
+    },
+    {
+      "epoch": 0.5619498984427894,
+      "grad_norm": 0.8838781940692847,
+      "learning_rate": 9.5963658619719e-06,
+      "loss": 2.2238,
+      "step": 415
+    },
+    {
+      "epoch": 0.5633039945836155,
+      "grad_norm": 1.044653451464992,
+      "learning_rate": 9.594239914478886e-06,
+      "loss": 2.8379,
+      "step": 416
+    },
+    {
+      "epoch": 0.5646580907244414,
+      "grad_norm": 0.9143098376517602,
+      "learning_rate": 9.592108619734107e-06,
+      "loss": 2.3621,
+      "step": 417
+    },
+    {
+      "epoch": 0.5660121868652674,
+      "grad_norm": 1.3056702719267006,
+      "learning_rate": 9.58997198021818e-06,
+      "loss": 2.3495,
+      "step": 418
+    },
+    {
+      "epoch": 0.5673662830060935,
+      "grad_norm": 1.2114169289052528,
+      "learning_rate": 9.587829998417953e-06,
+      "loss": 2.6471,
+      "step": 419
+    },
+    {
+      "epoch": 0.5687203791469194,
+      "grad_norm": 1.046324944660913,
+      "learning_rate": 9.58568267682648e-06,
+      "loss": 2.402,
+      "step": 420
+    },
+    {
+      "epoch": 0.5700744752877455,
+      "grad_norm": 0.9349583810327691,
+      "learning_rate": 9.58353001794304e-06,
+      "loss": 2.2225,
+      "step": 421
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 1.481772765240176,
+      "learning_rate": 9.581372024273121e-06,
+      "loss": 1.915,
+      "step": 422
+    },
+    {
+      "epoch": 0.5727826675693974,
+      "grad_norm": 0.928476736332317,
+      "learning_rate": 9.579208698328419e-06,
+      "loss": 1.8008,
+      "step": 423
+    },
+    {
+      "epoch": 0.5741367637102234,
+      "grad_norm": 0.9260270264242221,
+      "learning_rate": 9.577040042626832e-06,
+      "loss": 2.417,
+      "step": 424
+    },
+    {
+      "epoch": 0.5754908598510494,
+      "grad_norm": 1.1064706208462298,
+      "learning_rate": 9.574866059692471e-06,
+      "loss": 2.5265,
+      "step": 425
+    },
+    {
+      "epoch": 0.5768449559918755,
+      "grad_norm": 2.139248004116943,
+      "learning_rate": 9.57268675205564e-06,
+      "loss": 2.1862,
+      "step": 426
+    },
+    {
+      "epoch": 0.5781990521327014,
+      "grad_norm": 1.2117707663910726,
+      "learning_rate": 9.570502122252844e-06,
+      "loss": 2.8615,
+      "step": 427
+    },
+    {
+      "epoch": 0.5795531482735274,
+      "grad_norm": 0.8678663266783263,
+      "learning_rate": 9.568312172826779e-06,
+      "loss": 2.2196,
+      "step": 428
+    },
+    {
+      "epoch": 0.5809072444143534,
+      "grad_norm": 1.183740118318461,
+      "learning_rate": 9.566116906326336e-06,
+      "loss": 2.7205,
+      "step": 429
+    },
+    {
+      "epoch": 0.5822613405551794,
+      "grad_norm": 0.9774576381702014,
+      "learning_rate": 9.563916325306595e-06,
+      "loss": 2.2203,
+      "step": 430
+    },
+    {
+      "epoch": 0.5836154366960055,
+      "grad_norm": 0.9824933411377383,
+      "learning_rate": 9.561710432328817e-06,
+      "loss": 2.7149,
+      "step": 431
+    },
+    {
+      "epoch": 0.5849695328368314,
+      "grad_norm": 1.8285370131236038,
+      "learning_rate": 9.55949922996045e-06,
+      "loss": 2.1549,
+      "step": 432
+    },
+    {
+      "epoch": 0.5863236289776574,
+      "grad_norm": 0.92698096294728,
+      "learning_rate": 9.55728272077512e-06,
+      "loss": 2.0743,
+      "step": 433
+    },
+    {
+      "epoch": 0.5876777251184834,
+      "grad_norm": 1.0344108156736416,
+      "learning_rate": 9.555060907352632e-06,
+      "loss": 2.3762,
+      "step": 434
+    },
+    {
+      "epoch": 0.5890318212593094,
+      "grad_norm": 1.0045554696436734,
+      "learning_rate": 9.552833792278957e-06,
+      "loss": 2.1954,
+      "step": 435
+    },
+    {
+      "epoch": 0.5903859174001355,
+      "grad_norm": 0.7967159103822394,
+      "learning_rate": 9.550601378146246e-06,
+      "loss": 2.0138,
+      "step": 436
+    },
+    {
+      "epoch": 0.5917400135409614,
+      "grad_norm": 0.8111896941287193,
+      "learning_rate": 9.54836366755281e-06,
+      "loss": 2.3701,
+      "step": 437
+    },
+    {
+      "epoch": 0.5930941096817874,
+      "grad_norm": 0.9726524616144652,
+      "learning_rate": 9.546120663103134e-06,
+      "loss": 1.9217,
+      "step": 438
+    },
+    {
+      "epoch": 0.5944482058226134,
+      "grad_norm": 1.0256397319731523,
+      "learning_rate": 9.543872367407854e-06,
+      "loss": 2.0282,
+      "step": 439
+    },
+    {
+      "epoch": 0.5958023019634394,
+      "grad_norm": 1.0079882867447063,
+      "learning_rate": 9.54161878308377e-06,
+      "loss": 2.3361,
+      "step": 440
+    },
+    {
+      "epoch": 0.5971563981042654,
+      "grad_norm": 0.9800767836635148,
+      "learning_rate": 9.539359912753839e-06,
+      "loss": 2.6659,
+      "step": 441
+    },
+    {
+      "epoch": 0.5985104942450914,
+      "grad_norm": 1.0649583722690574,
+      "learning_rate": 9.537095759047163e-06,
+      "loss": 2.4143,
+      "step": 442
+    },
+    {
+      "epoch": 0.5998645903859174,
+      "grad_norm": 1.174278203502222,
+      "learning_rate": 9.534826324599002e-06,
+      "loss": 2.22,
+      "step": 443
+    },
+    {
+      "epoch": 0.6012186865267434,
+      "grad_norm": 1.3010142344912385,
+      "learning_rate": 9.53255161205076e-06,
+      "loss": 2.1634,
+      "step": 444
+    },
+    {
+      "epoch": 0.6025727826675694,
+      "grad_norm": 0.8409212163726476,
+      "learning_rate": 9.530271624049979e-06,
+      "loss": 1.8782,
+      "step": 445
+    },
+    {
+      "epoch": 0.6039268788083954,
+      "grad_norm": 0.8027451978851314,
+      "learning_rate": 9.527986363250348e-06,
+      "loss": 2.203,
+      "step": 446
+    },
+    {
+      "epoch": 0.6052809749492214,
+      "grad_norm": 0.9039555297703499,
+      "learning_rate": 9.525695832311688e-06,
+      "loss": 2.3687,
+      "step": 447
+    },
+    {
+      "epoch": 0.6066350710900474,
+      "grad_norm": 0.8912812069256285,
+      "learning_rate": 9.523400033899957e-06,
+      "loss": 2.2258,
+      "step": 448
+    },
+    {
+      "epoch": 0.6079891672308734,
+      "grad_norm": 0.8488642026117442,
+      "learning_rate": 9.52109897068724e-06,
+      "loss": 2.4719,
+      "step": 449
+    },
+    {
+      "epoch": 0.6093432633716994,
+      "grad_norm": 0.9189612986590668,
+      "learning_rate": 9.518792645351757e-06,
+      "loss": 2.1323,
+      "step": 450
+    },
+    {
+      "epoch": 0.6106973595125254,
+      "grad_norm": 1.0241173909729102,
+      "learning_rate": 9.516481060577847e-06,
+      "loss": 2.3125,
+      "step": 451
+    },
+    {
+      "epoch": 0.6120514556533514,
+      "grad_norm": 0.9700124960446453,
+      "learning_rate": 9.514164219055969e-06,
+      "loss": 2.1262,
+      "step": 452
+    },
+    {
+      "epoch": 0.6134055517941774,
+      "grad_norm": 0.9865209216124129,
+      "learning_rate": 9.511842123482703e-06,
+      "loss": 2.0346,
+      "step": 453
+    },
+    {
+      "epoch": 0.6147596479350034,
+      "grad_norm": 0.8339628050854324,
+      "learning_rate": 9.509514776560747e-06,
+      "loss": 2.2793,
+      "step": 454
+    },
+    {
+      "epoch": 0.6161137440758294,
+      "grad_norm": 1.2124630572926367,
+      "learning_rate": 9.507182180998906e-06,
+      "loss": 2.1059,
+      "step": 455
+    },
+    {
+      "epoch": 0.6174678402166554,
+      "grad_norm": 1.1523483755668846,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 2.1094,
+      "step": 456
+    },
+    {
+      "epoch": 0.6188219363574814,
+      "grad_norm": 0.9250806368586671,
+      "learning_rate": 9.50250125482134e-06,
+      "loss": 2.6985,
+      "step": 457
+    },
+    {
+      "epoch": 0.6201760324983073,
+      "grad_norm": 1.2028362769677976,
+      "learning_rate": 9.500152929653764e-06,
+      "loss": 2.3153,
+      "step": 458
+    },
+    {
+      "epoch": 0.6215301286391334,
+      "grad_norm": 1.071969908607831,
+      "learning_rate": 9.497799366742586e-06,
+      "loss": 2.0759,
+      "step": 459
+    },
+    {
+      "epoch": 0.6228842247799594,
+      "grad_norm": 1.3535021074481768,
+      "learning_rate": 9.49544056882713e-06,
+      "loss": 2.2538,
+      "step": 460
+    },
+    {
+      "epoch": 0.6242383209207854,
+      "grad_norm": 1.1103088277032604,
+      "learning_rate": 9.49307653865281e-06,
+      "loss": 2.0217,
+      "step": 461
+    },
+    {
+      "epoch": 0.6255924170616114,
+      "grad_norm": 0.9163278193375946,
+      "learning_rate": 9.490707278971127e-06,
+      "loss": 2.2543,
+      "step": 462
+    },
+    {
+      "epoch": 0.6269465132024373,
+      "grad_norm": 0.953261634167317,
+      "learning_rate": 9.488332792539672e-06,
+      "loss": 2.6539,
+      "step": 463
+    },
+    {
+      "epoch": 0.6283006093432634,
+      "grad_norm": 1.029379230332492,
+      "learning_rate": 9.485953082122116e-06,
+      "loss": 2.3507,
+      "step": 464
+    },
+    {
+      "epoch": 0.6296547054840894,
+      "grad_norm": 0.9618296595544104,
+      "learning_rate": 9.483568150488215e-06,
+      "loss": 2.4448,
+      "step": 465
+    },
+    {
+      "epoch": 0.6310088016249153,
+      "grad_norm": 0.9594228312002971,
+      "learning_rate": 9.481178000413796e-06,
+      "loss": 2.6108,
+      "step": 466
+    },
+    {
+      "epoch": 0.6323628977657414,
+      "grad_norm": 0.9624426030456921,
+      "learning_rate": 9.478782634680765e-06,
+      "loss": 2.3803,
+      "step": 467
+    },
+    {
+      "epoch": 0.6337169939065673,
+      "grad_norm": 0.9808684308680445,
+      "learning_rate": 9.476382056077097e-06,
+      "loss": 2.5205,
+      "step": 468
+    },
+    {
+      "epoch": 0.6350710900473934,
+      "grad_norm": 1.0917634456153018,
+      "learning_rate": 9.473976267396831e-06,
+      "loss": 2.008,
+      "step": 469
+    },
+    {
+      "epoch": 0.6364251861882194,
+      "grad_norm": 1.85759777161883,
+      "learning_rate": 9.471565271440075e-06,
+      "loss": 1.9461,
+      "step": 470
+    },
+    {
+      "epoch": 0.6377792823290453,
+      "grad_norm": 0.9263896064370456,
+      "learning_rate": 9.469149071012996e-06,
+      "loss": 2.1459,
+      "step": 471
+    },
+    {
+      "epoch": 0.6391333784698714,
+      "grad_norm": 1.2831155066680264,
+      "learning_rate": 9.466727668927817e-06,
+      "loss": 2.5968,
+      "step": 472
+    },
+    {
+      "epoch": 0.6404874746106973,
+      "grad_norm": 2.5926011027880405,
+      "learning_rate": 9.464301068002815e-06,
+      "loss": 2.8532,
+      "step": 473
+    },
+    {
+      "epoch": 0.6418415707515234,
+      "grad_norm": 1.3517479249404691,
+      "learning_rate": 9.461869271062322e-06,
+      "loss": 2.5085,
+      "step": 474
+    },
+    {
+      "epoch": 0.6431956668923493,
+      "grad_norm": 1.0653079693612253,
+      "learning_rate": 9.459432280936714e-06,
+      "loss": 1.926,
+      "step": 475
+    },
+    {
+      "epoch": 0.6445497630331753,
+      "grad_norm": 1.0222358313347504,
+      "learning_rate": 9.456990100462411e-06,
+      "loss": 2.0835,
+      "step": 476
+    },
+    {
+      "epoch": 0.6459038591740014,
+      "grad_norm": 0.9472201990170999,
+      "learning_rate": 9.454542732481876e-06,
+      "loss": 2.5298,
+      "step": 477
+    },
+    {
+      "epoch": 0.6472579553148273,
+      "grad_norm": 1.05236931626042,
+      "learning_rate": 9.452090179843609e-06,
+      "loss": 2.5688,
+      "step": 478
+    },
+    {
+      "epoch": 0.6486120514556534,
+      "grad_norm": 0.9205333314410238,
+      "learning_rate": 9.449632445402146e-06,
+      "loss": 2.2376,
+      "step": 479
+    },
+    {
+      "epoch": 0.6499661475964793,
+      "grad_norm": 0.936047746739389,
+      "learning_rate": 9.44716953201805e-06,
+      "loss": 2.2657,
+      "step": 480
+    },
+    {
+      "epoch": 0.6513202437373053,
+      "grad_norm": 0.984012799628733,
+      "learning_rate": 9.444701442557917e-06,
+      "loss": 2.2747,
+      "step": 481
+    },
+    {
+      "epoch": 0.6526743398781314,
+      "grad_norm": 1.1605171743997218,
+      "learning_rate": 9.442228179894362e-06,
+      "loss": 2.5819,
+      "step": 482
+    },
+    {
+      "epoch": 0.6540284360189573,
+      "grad_norm": 1.0367078942789454,
+      "learning_rate": 9.439749746906027e-06,
+      "loss": 2.3326,
+      "step": 483
+    },
+    {
+      "epoch": 0.6553825321597834,
+      "grad_norm": 0.8511526169640604,
+      "learning_rate": 9.437266146477567e-06,
+      "loss": 2.3916,
+      "step": 484
+    },
+    {
+      "epoch": 0.6567366283006093,
+      "grad_norm": 0.9546785433848596,
+      "learning_rate": 9.434777381499654e-06,
+      "loss": 2.2932,
+      "step": 485
+    },
+    {
+      "epoch": 0.6580907244414353,
+      "grad_norm": 0.897715373597665,
+      "learning_rate": 9.43228345486897e-06,
+      "loss": 2.2329,
+      "step": 486
+    },
+    {
+      "epoch": 0.6594448205822614,
+      "grad_norm": 1.1576802918228548,
+      "learning_rate": 9.429784369488205e-06,
+      "loss": 2.5906,
+      "step": 487
+    },
+    {
+      "epoch": 0.6607989167230873,
+      "grad_norm": 1.2669978365090186,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 2.0969,
+      "step": 488
+    },
+    {
+      "epoch": 0.6621530128639134,
+      "grad_norm": 1.116706571792338,
+      "learning_rate": 9.424770734117206e-06,
+      "loss": 2.0978,
+      "step": 489
+    },
+    {
+      "epoch": 0.6635071090047393,
+      "grad_norm": 1.0731484382258585,
+      "learning_rate": 9.42225618996236e-06,
+      "loss": 2.2606,
+      "step": 490
+    },
+    {
+      "epoch": 0.6648612051455653,
+      "grad_norm": 1.0217610153976007,
+      "learning_rate": 9.419736498728203e-06,
+      "loss": 2.0359,
+      "step": 491
+    },
+    {
+      "epoch": 0.6662153012863913,
+      "grad_norm": 0.9470105904564678,
+      "learning_rate": 9.417211663347407e-06,
+      "loss": 2.3133,
+      "step": 492
+    },
+    {
+      "epoch": 0.6675693974272173,
+      "grad_norm": 1.1496175064197292,
+      "learning_rate": 9.414681686758645e-06,
+      "loss": 2.6914,
+      "step": 493
+    },
+    {
+      "epoch": 0.6689234935680434,
+      "grad_norm": 1.1088711952773571,
+      "learning_rate": 9.412146571906556e-06,
+      "loss": 2.1629,
+      "step": 494
+    },
+    {
+      "epoch": 0.6702775897088693,
+      "grad_norm": 2.4048301880979315,
+      "learning_rate": 9.409606321741776e-06,
+      "loss": 2.3511,
+      "step": 495
+    },
+    {
+      "epoch": 0.6716316858496953,
+      "grad_norm": 1.0855961625720172,
+      "learning_rate": 9.407060939220907e-06,
+      "loss": 3.0198,
+      "step": 496
+    },
+    {
+      "epoch": 0.6729857819905213,
+      "grad_norm": 0.831939027060162,
+      "learning_rate": 9.404510427306533e-06,
+      "loss": 2.0721,
+      "step": 497
+    },
+    {
+      "epoch": 0.6743398781313473,
+      "grad_norm": 0.9189030221620978,
+      "learning_rate": 9.401954788967199e-06,
+      "loss": 2.1533,
+      "step": 498
+    },
+    {
+      "epoch": 0.6756939742721734,
+      "grad_norm": 1.1215682593702503,
+      "learning_rate": 9.39939402717742e-06,
+      "loss": 2.3017,
+      "step": 499
+    },
+    {
+      "epoch": 0.6770480704129993,
+      "grad_norm": 0.9997121388139094,
+      "learning_rate": 9.396828144917682e-06,
+      "loss": 2.1092,
+      "step": 500
+    },
+    {
+      "epoch": 0.6784021665538253,
+      "grad_norm": 1.0419306713261767,
+      "learning_rate": 9.39425714517442e-06,
+      "loss": 2.3772,
+      "step": 501
+    },
+    {
+      "epoch": 0.6797562626946513,
+      "grad_norm": 0.8546730463318373,
+      "learning_rate": 9.391681030940031e-06,
+      "loss": 2.4692,
+      "step": 502
+    },
+    {
+      "epoch": 0.6811103588354773,
+      "grad_norm": 1.036666833583794,
+      "learning_rate": 9.389099805212862e-06,
+      "loss": 2.4967,
+      "step": 503
+    },
+    {
+      "epoch": 0.6824644549763034,
+      "grad_norm": 0.9028903653709767,
+      "learning_rate": 9.38651347099721e-06,
+      "loss": 2.2658,
+      "step": 504
+    },
+    {
+      "epoch": 0.6838185511171293,
+      "grad_norm": 1.004993430273087,
+      "learning_rate": 9.38392203130332e-06,
+      "loss": 2.2413,
+      "step": 505
+    },
+    {
+      "epoch": 0.6851726472579553,
+      "grad_norm": 4.244116799684234,
+      "learning_rate": 9.38132548914738e-06,
+      "loss": 2.2353,
+      "step": 506
+    },
+    {
+      "epoch": 0.6865267433987813,
+      "grad_norm": 2.0623005376684516,
+      "learning_rate": 9.37872384755151e-06,
+      "loss": 2.3859,
+      "step": 507
+    },
+    {
+      "epoch": 0.6878808395396073,
+      "grad_norm": 1.099884816816885,
+      "learning_rate": 9.376117109543769e-06,
+      "loss": 1.8825,
+      "step": 508
+    },
+    {
+      "epoch": 0.6892349356804333,
+      "grad_norm": 0.9736281444679528,
+      "learning_rate": 9.373505278158152e-06,
+      "loss": 2.4243,
+      "step": 509
+    },
+    {
+      "epoch": 0.6905890318212593,
+      "grad_norm": 0.9265991736695152,
+      "learning_rate": 9.370888356434577e-06,
+      "loss": 2.5581,
+      "step": 510
+    },
+    {
+      "epoch": 0.6919431279620853,
+      "grad_norm": 0.9742740668751364,
+      "learning_rate": 9.368266347418891e-06,
+      "loss": 2.2436,
+      "step": 511
+    },
+    {
+      "epoch": 0.6932972241029113,
+      "grad_norm": 0.903250624410521,
+      "learning_rate": 9.365639254162855e-06,
+      "loss": 2.3704,
+      "step": 512
+    },
+    {
+      "epoch": 0.6946513202437373,
+      "grad_norm": 1.011884451867896,
+      "learning_rate": 9.363007079724153e-06,
+      "loss": 2.3554,
+      "step": 513
+    },
+    {
+      "epoch": 0.6960054163845633,
+      "grad_norm": 2.1268078439425504,
+      "learning_rate": 9.360369827166385e-06,
+      "loss": 2.2531,
+      "step": 514
+    },
+    {
+      "epoch": 0.6973595125253893,
+      "grad_norm": 0.9226076560386388,
+      "learning_rate": 9.357727499559055e-06,
+      "loss": 2.1536,
+      "step": 515
+    },
+    {
+      "epoch": 0.6987136086662153,
+      "grad_norm": 0.9235717043788523,
+      "learning_rate": 9.355080099977579e-06,
+      "loss": 2.5795,
+      "step": 516
+    },
+    {
+      "epoch": 0.7000677048070413,
+      "grad_norm": 0.997624756104421,
+      "learning_rate": 9.352427631503274e-06,
+      "loss": 2.4086,
+      "step": 517
+    },
+    {
+      "epoch": 0.7014218009478673,
+      "grad_norm": 1.0325935042936567,
+      "learning_rate": 9.349770097223356e-06,
+      "loss": 2.1905,
+      "step": 518
+    },
+    {
+      "epoch": 0.7027758970886933,
+      "grad_norm": 0.9779488073579523,
+      "learning_rate": 9.347107500230941e-06,
+      "loss": 2.1902,
+      "step": 519
+    },
+    {
+      "epoch": 0.7041299932295193,
+      "grad_norm": 0.9071035111230149,
+      "learning_rate": 9.344439843625034e-06,
+      "loss": 2.4966,
+      "step": 520
+    },
+    {
+      "epoch": 0.7054840893703453,
+      "grad_norm": 0.9294082146423078,
+      "learning_rate": 9.341767130510529e-06,
+      "loss": 2.1556,
+      "step": 521
+    },
+    {
+      "epoch": 0.7068381855111713,
+      "grad_norm": 0.8010254757576656,
+      "learning_rate": 9.339089363998206e-06,
+      "loss": 2.2359,
+      "step": 522
+    },
+    {
+      "epoch": 0.7081922816519973,
+      "grad_norm": 1.2492381753400523,
+      "learning_rate": 9.336406547204726e-06,
+      "loss": 2.0768,
+      "step": 523
+    },
+    {
+      "epoch": 0.7095463777928233,
+      "grad_norm": 0.960396750578361,
+      "learning_rate": 9.333718683252631e-06,
+      "loss": 2.2373,
+      "step": 524
+    },
+    {
+      "epoch": 0.7109004739336493,
+      "grad_norm": 0.9089783184513304,
+      "learning_rate": 9.331025775270335e-06,
+      "loss": 2.0008,
+      "step": 525
+    },
+    {
+      "epoch": 0.7122545700744752,
+      "grad_norm": 1.0902851866790608,
+      "learning_rate": 9.32832782639212e-06,
+      "loss": 2.452,
+      "step": 526
+    },
+    {
+      "epoch": 0.7136086662153013,
+      "grad_norm": 1.0819110431428438,
+      "learning_rate": 9.325624839758142e-06,
+      "loss": 2.1797,
+      "step": 527
+    },
+    {
+      "epoch": 0.7149627623561273,
+      "grad_norm": 1.035728859841243,
+      "learning_rate": 9.322916818514414e-06,
+      "loss": 2.0016,
+      "step": 528
+    },
+    {
+      "epoch": 0.7163168584969533,
+      "grad_norm": 0.8429686871803489,
+      "learning_rate": 9.32020376581281e-06,
+      "loss": 2.1271,
+      "step": 529
+    },
+    {
+      "epoch": 0.7176709546377793,
+      "grad_norm": 0.948124830150651,
+      "learning_rate": 9.317485684811065e-06,
+      "loss": 2.463,
+      "step": 530
+    },
+    {
+      "epoch": 0.7190250507786052,
+      "grad_norm": 1.0966777201484916,
+      "learning_rate": 9.31476257867276e-06,
+      "loss": 2.7078,
+      "step": 531
+    },
+    {
+      "epoch": 0.7203791469194313,
+      "grad_norm": 0.9601018625399497,
+      "learning_rate": 9.312034450567331e-06,
+      "loss": 2.0091,
+      "step": 532
+    },
+    {
+      "epoch": 0.7217332430602573,
+      "grad_norm": 1.0195440951721928,
+      "learning_rate": 9.309301303670053e-06,
+      "loss": 2.2595,
+      "step": 533
+    },
+    {
+      "epoch": 0.7230873392010833,
+      "grad_norm": 1.0365088177530009,
+      "learning_rate": 9.306563141162046e-06,
+      "loss": 2.5509,
+      "step": 534
+    },
+    {
+      "epoch": 0.7244414353419093,
+      "grad_norm": 0.89557097021405,
+      "learning_rate": 9.303819966230265e-06,
+      "loss": 2.184,
+      "step": 535
+    },
+    {
+      "epoch": 0.7257955314827352,
+      "grad_norm": 1.1201396043943863,
+      "learning_rate": 9.301071782067504e-06,
+      "loss": 2.2719,
+      "step": 536
+    },
+    {
+      "epoch": 0.7271496276235613,
+      "grad_norm": 1.0416986531898793,
+      "learning_rate": 9.298318591872381e-06,
+      "loss": 2.9307,
+      "step": 537
+    },
+    {
+      "epoch": 0.7285037237643873,
+      "grad_norm": 0.8874369002318888,
+      "learning_rate": 9.295560398849348e-06,
+      "loss": 2.0185,
+      "step": 538
+    },
+    {
+      "epoch": 0.7298578199052133,
+      "grad_norm": 1.3062496312023244,
+      "learning_rate": 9.29279720620867e-06,
+      "loss": 2.4988,
+      "step": 539
+    },
+    {
+      "epoch": 0.7312119160460393,
+      "grad_norm": 0.8723686506256735,
+      "learning_rate": 9.290029017166439e-06,
+      "loss": 2.1358,
+      "step": 540
+    },
+    {
+      "epoch": 0.7325660121868652,
+      "grad_norm": 0.9558509794048232,
+      "learning_rate": 9.287255834944563e-06,
+      "loss": 2.5911,
+      "step": 541
+    },
+    {
+      "epoch": 0.7339201083276913,
+      "grad_norm": 0.9765924363071334,
+      "learning_rate": 9.284477662770753e-06,
+      "loss": 2.2083,
+      "step": 542
+    },
+    {
+      "epoch": 0.7352742044685172,
+      "grad_norm": 0.9439430068504877,
+      "learning_rate": 9.281694503878536e-06,
+      "loss": 2.2064,
+      "step": 543
+    },
+    {
+      "epoch": 0.7366283006093433,
+      "grad_norm": 1.1103666583629448,
+      "learning_rate": 9.278906361507238e-06,
+      "loss": 2.2118,
+      "step": 544
+    },
+    {
+      "epoch": 0.7379823967501693,
+      "grad_norm": 1.015972947443489,
+      "learning_rate": 9.276113238901992e-06,
+      "loss": 2.3865,
+      "step": 545
+    },
+    {
+      "epoch": 0.7393364928909952,
+      "grad_norm": 0.9787500482208973,
+      "learning_rate": 9.273315139313719e-06,
+      "loss": 2.0949,
+      "step": 546
+    },
+    {
+      "epoch": 0.7406905890318213,
+      "grad_norm": 1.1247301478380771,
+      "learning_rate": 9.270512065999139e-06,
+      "loss": 1.9341,
+      "step": 547
+    },
+    {
+      "epoch": 0.7420446851726472,
+      "grad_norm": 0.8630459962648358,
+      "learning_rate": 9.267704022220758e-06,
+      "loss": 2.1934,
+      "step": 548
+    },
+    {
+      "epoch": 0.7433987813134733,
+      "grad_norm": 0.9456846452275267,
+      "learning_rate": 9.264891011246867e-06,
+      "loss": 2.1091,
+      "step": 549
+    },
+    {
+      "epoch": 0.7447528774542993,
+      "grad_norm": 0.9901864770446673,
+      "learning_rate": 9.26207303635154e-06,
+      "loss": 2.0778,
+      "step": 550
+    },
+    {
+      "epoch": 0.7461069735951252,
+      "grad_norm": 0.9659375715716257,
+      "learning_rate": 9.25925010081463e-06,
+      "loss": 2.2679,
+      "step": 551
+    },
+    {
+      "epoch": 0.7474610697359513,
+      "grad_norm": 0.9764027265044928,
+      "learning_rate": 9.256422207921757e-06,
+      "loss": 2.7468,
+      "step": 552
+    },
+    {
+      "epoch": 0.7488151658767772,
+      "grad_norm": 0.899162453540954,
+      "learning_rate": 9.25358936096432e-06,
+      "loss": 2.312,
+      "step": 553
+    },
+    {
+      "epoch": 0.7501692620176033,
+      "grad_norm": 0.7884101704032132,
+      "learning_rate": 9.250751563239473e-06,
+      "loss": 1.8254,
+      "step": 554
+    },
+    {
+      "epoch": 0.7515233581584293,
+      "grad_norm": 1.0492022449068705,
+      "learning_rate": 9.247908818050146e-06,
+      "loss": 2.6885,
+      "step": 555
+    },
+    {
+      "epoch": 0.7528774542992552,
+      "grad_norm": 1.0133767522172044,
+      "learning_rate": 9.245061128705017e-06,
+      "loss": 2.2732,
+      "step": 556
+    },
+    {
+      "epoch": 0.7542315504400813,
+      "grad_norm": 1.1740684334910383,
+      "learning_rate": 9.242208498518522e-06,
+      "loss": 2.4037,
+      "step": 557
+    },
+    {
+      "epoch": 0.7555856465809072,
+      "grad_norm": 1.0034773672404191,
+      "learning_rate": 9.239350930810843e-06,
+      "loss": 2.6555,
+      "step": 558
+    },
+    {
+      "epoch": 0.7569397427217333,
+      "grad_norm": 1.0559371991027349,
+      "learning_rate": 9.236488428907919e-06,
+      "loss": 1.8506,
+      "step": 559
+    },
+    {
+      "epoch": 0.7582938388625592,
+      "grad_norm": 0.9422190702574703,
+      "learning_rate": 9.233620996141421e-06,
+      "loss": 2.393,
+      "step": 560
+    },
+    {
+      "epoch": 0.7596479350033852,
+      "grad_norm": 1.701435588173234,
+      "learning_rate": 9.230748635848768e-06,
+      "loss": 2.4996,
+      "step": 561
+    },
+    {
+      "epoch": 0.7610020311442113,
+      "grad_norm": 0.9848241894053293,
+      "learning_rate": 9.227871351373108e-06,
+      "loss": 2.2563,
+      "step": 562
+    },
+    {
+      "epoch": 0.7623561272850372,
+      "grad_norm": 0.9551626201609839,
+      "learning_rate": 9.224989146063322e-06,
+      "loss": 2.3359,
+      "step": 563
+    },
+    {
+      "epoch": 0.7637102234258633,
+      "grad_norm": 1.0501196818584786,
+      "learning_rate": 9.22210202327402e-06,
+      "loss": 2.3484,
+      "step": 564
+    },
+    {
+      "epoch": 0.7650643195666892,
+      "grad_norm": 1.0118693595069355,
+      "learning_rate": 9.219209986365533e-06,
+      "loss": 2.0198,
+      "step": 565
+    },
+    {
+      "epoch": 0.7664184157075152,
+      "grad_norm": 1.000596790716306,
+      "learning_rate": 9.216313038703914e-06,
+      "loss": 2.4147,
+      "step": 566
+    },
+    {
+      "epoch": 0.7677725118483413,
+      "grad_norm": 1.0925850682201712,
+      "learning_rate": 9.21341118366093e-06,
+      "loss": 2.0607,
+      "step": 567
+    },
+    {
+      "epoch": 0.7691266079891672,
+      "grad_norm": 0.8195614498636685,
+      "learning_rate": 9.21050442461406e-06,
+      "loss": 2.3161,
+      "step": 568
+    },
+    {
+      "epoch": 0.7704807041299933,
+      "grad_norm": 1.4095142340461897,
+      "learning_rate": 9.207592764946491e-06,
+      "loss": 2.3281,
+      "step": 569
+    },
+    {
+      "epoch": 0.7718348002708192,
+      "grad_norm": 1.0305488353359658,
+      "learning_rate": 9.204676208047112e-06,
+      "loss": 2.209,
+      "step": 570
+    },
+    {
+      "epoch": 0.7731888964116452,
+      "grad_norm": 0.9457457073106245,
+      "learning_rate": 9.201754757310518e-06,
+      "loss": 1.9644,
+      "step": 571
+    },
+    {
+      "epoch": 0.7745429925524713,
+      "grad_norm": 0.8798589503356994,
+      "learning_rate": 9.198828416136991e-06,
+      "loss": 2.5588,
+      "step": 572
+    },
+    {
+      "epoch": 0.7758970886932972,
+      "grad_norm": 1.1021679000741063,
+      "learning_rate": 9.195897187932513e-06,
+      "loss": 2.2556,
+      "step": 573
+    },
+    {
+      "epoch": 0.7772511848341233,
+      "grad_norm": 1.2881391770879362,
+      "learning_rate": 9.192961076108748e-06,
+      "loss": 2.2786,
+      "step": 574
+    },
+    {
+      "epoch": 0.7786052809749492,
+      "grad_norm": 0.8079890380239451,
+      "learning_rate": 9.190020084083048e-06,
+      "loss": 1.9979,
+      "step": 575
+    },
+    {
+      "epoch": 0.7799593771157752,
+      "grad_norm": 2.3299606691457706,
+      "learning_rate": 9.187074215278444e-06,
+      "loss": 2.3944,
+      "step": 576
+    },
+    {
+      "epoch": 0.7813134732566012,
+      "grad_norm": 0.8875887654620506,
+      "learning_rate": 9.184123473123643e-06,
+      "loss": 3.159,
+      "step": 577
+    },
+    {
+      "epoch": 0.7826675693974272,
+      "grad_norm": 1.2732756612102565,
+      "learning_rate": 9.181167861053024e-06,
+      "loss": 2.5586,
+      "step": 578
+    },
+    {
+      "epoch": 0.7840216655382533,
+      "grad_norm": 1.1186141048074447,
+      "learning_rate": 9.178207382506634e-06,
+      "loss": 2.3656,
+      "step": 579
+    },
+    {
+      "epoch": 0.7853757616790792,
+      "grad_norm": 1.0038802401880105,
+      "learning_rate": 9.175242040930185e-06,
+      "loss": 2.3621,
+      "step": 580
+    },
+    {
+      "epoch": 0.7867298578199052,
+      "grad_norm": 1.0084194487112716,
+      "learning_rate": 9.172271839775046e-06,
+      "loss": 2.2325,
+      "step": 581
+    },
+    {
+      "epoch": 0.7880839539607312,
+      "grad_norm": 0.8495420286871743,
+      "learning_rate": 9.169296782498249e-06,
+      "loss": 2.4249,
+      "step": 582
+    },
+    {
+      "epoch": 0.7894380501015572,
+      "grad_norm": 1.0108960715157491,
+      "learning_rate": 9.166316872562467e-06,
+      "loss": 2.0571,
+      "step": 583
+    },
+    {
+      "epoch": 0.7907921462423833,
+      "grad_norm": 0.7599417123654713,
+      "learning_rate": 9.163332113436031e-06,
+      "loss": 2.2401,
+      "step": 584
+    },
+    {
+      "epoch": 0.7921462423832092,
+      "grad_norm": 0.9153720931911937,
+      "learning_rate": 9.160342508592916e-06,
+      "loss": 2.0296,
+      "step": 585
+    },
+    {
+      "epoch": 0.7935003385240352,
+      "grad_norm": 1.0606631317988464,
+      "learning_rate": 9.157348061512728e-06,
+      "loss": 1.995,
+      "step": 586
+    },
+    {
+      "epoch": 0.7948544346648612,
+      "grad_norm": 0.9696771933921041,
+      "learning_rate": 9.154348775680714e-06,
+      "loss": 2.4553,
+      "step": 587
+    },
+    {
+      "epoch": 0.7962085308056872,
+      "grad_norm": 1.1359820431460632,
+      "learning_rate": 9.151344654587758e-06,
+      "loss": 2.6147,
+      "step": 588
+    },
+    {
+      "epoch": 0.7975626269465133,
+      "grad_norm": 1.1620764175753133,
+      "learning_rate": 9.148335701730363e-06,
+      "loss": 2.1694,
+      "step": 589
+    },
+    {
+      "epoch": 0.7989167230873392,
+      "grad_norm": 1.0906791903175577,
+      "learning_rate": 9.145321920610662e-06,
+      "loss": 2.0159,
+      "step": 590
+    },
+    {
+      "epoch": 0.8002708192281652,
+      "grad_norm": 0.9262503779509312,
+      "learning_rate": 9.142303314736405e-06,
+      "loss": 1.8704,
+      "step": 591
+    },
+    {
+      "epoch": 0.8016249153689912,
+      "grad_norm": 0.8897378574716819,
+      "learning_rate": 9.139279887620955e-06,
+      "loss": 2.2212,
+      "step": 592
+    },
+    {
+      "epoch": 0.8029790115098172,
+      "grad_norm": 0.8785520958211444,
+      "learning_rate": 9.136251642783294e-06,
+      "loss": 2.2422,
+      "step": 593
+    },
+    {
+      "epoch": 0.8043331076506431,
+      "grad_norm": 1.4708082564367666,
+      "learning_rate": 9.133218583748002e-06,
+      "loss": 2.4264,
+      "step": 594
+    },
+    {
+      "epoch": 0.8056872037914692,
+      "grad_norm": 0.9127946818898369,
+      "learning_rate": 9.130180714045271e-06,
+      "loss": 1.985,
+      "step": 595
+    },
+    {
+      "epoch": 0.8070412999322952,
+      "grad_norm": 1.0273656011649834,
+      "learning_rate": 9.127138037210884e-06,
+      "loss": 2.1752,
+      "step": 596
+    },
+    {
+      "epoch": 0.8083953960731212,
+      "grad_norm": 1.0442611846851202,
+      "learning_rate": 9.124090556786227e-06,
+      "loss": 2.2373,
+      "step": 597
+    },
+    {
+      "epoch": 0.8097494922139472,
+      "grad_norm": 1.0020569110815791,
+      "learning_rate": 9.12103827631827e-06,
+      "loss": 2.6107,
+      "step": 598
+    },
+    {
+      "epoch": 0.8111035883547731,
+      "grad_norm": 1.0433877333186745,
+      "learning_rate": 9.117981199359575e-06,
+      "loss": 2.2095,
+      "step": 599
+    },
+    {
+      "epoch": 0.8124576844955992,
+      "grad_norm": 0.9172575390929391,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 2.0521,
+      "step": 600
+    },
+    {
+      "epoch": 0.8138117806364252,
+      "grad_norm": 1.0426202271128402,
+      "learning_rate": 9.111852670208116e-06,
+      "loss": 1.8817,
+      "step": 601
+    },
+    {
+      "epoch": 0.8151658767772512,
+      "grad_norm": 1.1301209307772506,
+      "learning_rate": 9.108781225148369e-06,
+      "loss": 3.0088,
+      "step": 602
+    },
+    {
+      "epoch": 0.8165199729180772,
+      "grad_norm": 0.9133026858806589,
+      "learning_rate": 9.105704997863907e-06,
+      "loss": 2.5472,
+      "step": 603
+    },
+    {
+      "epoch": 0.8178740690589031,
+      "grad_norm": 1.1108413215075008,
+      "learning_rate": 9.102623991935163e-06,
+      "loss": 2.5752,
+      "step": 604
+    },
+    {
+      "epoch": 0.8192281651997292,
+      "grad_norm": 0.9702100994035456,
+      "learning_rate": 9.099538210948128e-06,
+      "loss": 2.4404,
+      "step": 605
+    },
+    {
+      "epoch": 0.8205822613405552,
+      "grad_norm": 0.9485690500451401,
+      "learning_rate": 9.096447658494357e-06,
+      "loss": 2.2257,
+      "step": 606
+    },
+    {
+      "epoch": 0.8219363574813812,
+      "grad_norm": 26.292288625318744,
+      "learning_rate": 9.09335233817095e-06,
+      "loss": 2.676,
+      "step": 607
+    },
+    {
+      "epoch": 0.8232904536222072,
+      "grad_norm": 0.9757213752439725,
+      "learning_rate": 9.090252253580565e-06,
+      "loss": 2.083,
+      "step": 608
+    },
+    {
+      "epoch": 0.8246445497630331,
+      "grad_norm": 0.977754402583003,
+      "learning_rate": 9.087147408331399e-06,
+      "loss": 2.396,
+      "step": 609
+    },
+    {
+      "epoch": 0.8259986459038592,
+      "grad_norm": 1.1224622683564631,
+      "learning_rate": 9.084037806037194e-06,
+      "loss": 2.5127,
+      "step": 610
+    },
+    {
+      "epoch": 0.8273527420446851,
+      "grad_norm": 1.0702372697619176,
+      "learning_rate": 9.080923450317226e-06,
+      "loss": 2.44,
+      "step": 611
+    },
+    {
+      "epoch": 0.8287068381855112,
+      "grad_norm": 0.9854940413106023,
+      "learning_rate": 9.077804344796302e-06,
+      "loss": 2.3179,
+      "step": 612
+    },
+    {
+      "epoch": 0.8300609343263372,
+      "grad_norm": 1.0317420348048132,
+      "learning_rate": 9.074680493104764e-06,
+      "loss": 2.2255,
+      "step": 613
+    },
+    {
+      "epoch": 0.8314150304671631,
+      "grad_norm": 1.2553199368667127,
+      "learning_rate": 9.071551898878471e-06,
+      "loss": 2.196,
+      "step": 614
+    },
+    {
+      "epoch": 0.8327691266079892,
+      "grad_norm": 0.9428355967192141,
+      "learning_rate": 9.068418565758805e-06,
+      "loss": 2.4896,
+      "step": 615
+    },
+    {
+      "epoch": 0.8341232227488151,
+      "grad_norm": 0.919605019517011,
+      "learning_rate": 9.065280497392663e-06,
+      "loss": 2.724,
+      "step": 616
+    },
+    {
+      "epoch": 0.8354773188896412,
+      "grad_norm": 0.9006367820269072,
+      "learning_rate": 9.062137697432457e-06,
+      "loss": 2.3021,
+      "step": 617
+    },
+    {
+      "epoch": 0.8368314150304672,
+      "grad_norm": 0.9254413019481625,
+      "learning_rate": 9.058990169536098e-06,
+      "loss": 2.8054,
+      "step": 618
+    },
+    {
+      "epoch": 0.8381855111712931,
+      "grad_norm": 0.9723010342216781,
+      "learning_rate": 9.055837917367006e-06,
+      "loss": 2.1242,
+      "step": 619
+    },
+    {
+      "epoch": 0.8395396073121192,
+      "grad_norm": 0.9646095993807202,
+      "learning_rate": 9.052680944594101e-06,
+      "loss": 2.7603,
+      "step": 620
+    },
+    {
+      "epoch": 0.8408937034529451,
+      "grad_norm": 1.3152197929562175,
+      "learning_rate": 9.049519254891793e-06,
+      "loss": 2.1762,
+      "step": 621
+    },
+    {
+      "epoch": 0.8422477995937712,
+      "grad_norm": 0.8097257743614227,
+      "learning_rate": 9.046352851939981e-06,
+      "loss": 2.4746,
+      "step": 622
+    },
+    {
+      "epoch": 0.8436018957345972,
+      "grad_norm": 0.9993019751259021,
+      "learning_rate": 9.043181739424054e-06,
+      "loss": 1.8778,
+      "step": 623
+    },
+    {
+      "epoch": 0.8449559918754231,
+      "grad_norm": 1.1394511401717762,
+      "learning_rate": 9.040005921034884e-06,
+      "loss": 2.2527,
+      "step": 624
+    },
+    {
+      "epoch": 0.8463100880162492,
+      "grad_norm": 1.0167098829317125,
+      "learning_rate": 9.036825400468814e-06,
+      "loss": 2.4564,
+      "step": 625
+    },
+    {
+      "epoch": 0.8476641841570751,
+      "grad_norm": 1.0799560174241793,
+      "learning_rate": 9.033640181427661e-06,
+      "loss": 2.1033,
+      "step": 626
+    },
+    {
+      "epoch": 0.8490182802979012,
+      "grad_norm": 1.174120607908624,
+      "learning_rate": 9.030450267618717e-06,
+      "loss": 2.4757,
+      "step": 627
+    },
+    {
+      "epoch": 0.8503723764387271,
+      "grad_norm": 1.050685050196375,
+      "learning_rate": 9.02725566275473e-06,
+      "loss": 2.238,
+      "step": 628
+    },
+    {
+      "epoch": 0.8517264725795531,
+      "grad_norm": 0.9338038071047584,
+      "learning_rate": 9.024056370553916e-06,
+      "loss": 2.1313,
+      "step": 629
+    },
+    {
+      "epoch": 0.8530805687203792,
+      "grad_norm": 1.156893096752152,
+      "learning_rate": 9.020852394739936e-06,
+      "loss": 2.2162,
+      "step": 630
+    },
+    {
+      "epoch": 0.8544346648612051,
+      "grad_norm": 1.2811232350478499,
+      "learning_rate": 9.017643739041914e-06,
+      "loss": 2.9062,
+      "step": 631
+    },
+    {
+      "epoch": 0.8557887610020312,
+      "grad_norm": 1.3380924542089319,
+      "learning_rate": 9.014430407194413e-06,
+      "loss": 2.4905,
+      "step": 632
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 1.068909763409773,
+      "learning_rate": 9.011212402937441e-06,
+      "loss": 1.973,
+      "step": 633
+    },
+    {
+      "epoch": 0.8584969532836831,
+      "grad_norm": 1.8932248755263266,
+      "learning_rate": 9.007989730016446e-06,
+      "loss": 2.4409,
+      "step": 634
+    },
+    {
+      "epoch": 0.8598510494245092,
+      "grad_norm": 1.2891934792560118,
+      "learning_rate": 9.004762392182307e-06,
+      "loss": 2.3235,
+      "step": 635
+    },
+    {
+      "epoch": 0.8612051455653351,
+      "grad_norm": 0.858242619324268,
+      "learning_rate": 9.001530393191334e-06,
+      "loss": 1.9865,
+      "step": 636
+    },
+    {
+      "epoch": 0.8625592417061612,
+      "grad_norm": 1.1164311194466396,
+      "learning_rate": 8.998293736805262e-06,
+      "loss": 2.3765,
+      "step": 637
+    },
+    {
+      "epoch": 0.8639133378469871,
+      "grad_norm": 1.2157891571025403,
+      "learning_rate": 8.995052426791247e-06,
+      "loss": 2.9348,
+      "step": 638
+    },
+    {
+      "epoch": 0.8652674339878131,
+      "grad_norm": 1.0909914707099782,
+      "learning_rate": 8.991806466921858e-06,
+      "loss": 2.8517,
+      "step": 639
+    },
+    {
+      "epoch": 0.8666215301286392,
+      "grad_norm": 0.9001967266754044,
+      "learning_rate": 8.988555860975082e-06,
+      "loss": 2.3269,
+      "step": 640
+    },
+    {
+      "epoch": 0.8679756262694651,
+      "grad_norm": 1.0105961392957552,
+      "learning_rate": 8.98530061273431e-06,
+      "loss": 2.101,
+      "step": 641
+    },
+    {
+      "epoch": 0.8693297224102912,
+      "grad_norm": 0.9677003303068539,
+      "learning_rate": 8.982040725988337e-06,
+      "loss": 2.2758,
+      "step": 642
+    },
+    {
+      "epoch": 0.8706838185511171,
+      "grad_norm": 0.9813548892015744,
+      "learning_rate": 8.978776204531354e-06,
+      "loss": 2.33,
+      "step": 643
+    },
+    {
+      "epoch": 0.8720379146919431,
+      "grad_norm": 1.0150274086140143,
+      "learning_rate": 8.975507052162954e-06,
+      "loss": 2.3066,
+      "step": 644
+    },
+    {
+      "epoch": 0.8733920108327691,
+      "grad_norm": 1.0301765250355548,
+      "learning_rate": 8.972233272688107e-06,
+      "loss": 2.1587,
+      "step": 645
+    },
+    {
+      "epoch": 0.8747461069735951,
+      "grad_norm": 1.1970855668426001,
+      "learning_rate": 8.968954869917183e-06,
+      "loss": 2.3845,
+      "step": 646
+    },
+    {
+      "epoch": 0.8761002031144212,
+      "grad_norm": 1.1034962442803717,
+      "learning_rate": 8.965671847665925e-06,
+      "loss": 1.7817,
+      "step": 647
+    },
+    {
+      "epoch": 0.8774542992552471,
+      "grad_norm": 0.9455749614353804,
+      "learning_rate": 8.962384209755453e-06,
+      "loss": 1.8517,
+      "step": 648
+    },
+    {
+      "epoch": 0.8788083953960731,
+      "grad_norm": 0.9934755324282731,
+      "learning_rate": 8.95909196001226e-06,
+      "loss": 2.2541,
+      "step": 649
+    },
+    {
+      "epoch": 0.8801624915368991,
+      "grad_norm": 1.0284578656567072,
+      "learning_rate": 8.955795102268206e-06,
+      "loss": 2.3968,
+      "step": 650
+    },
+    {
+      "epoch": 0.8815165876777251,
+      "grad_norm": 1.037663714131099,
+      "learning_rate": 8.952493640360518e-06,
+      "loss": 2.6138,
+      "step": 651
+    },
+    {
+      "epoch": 0.8828706838185512,
+      "grad_norm": 1.0926666615727456,
+      "learning_rate": 8.949187578131777e-06,
+      "loss": 1.9031,
+      "step": 652
+    },
+    {
+      "epoch": 0.8842247799593771,
+      "grad_norm": 1.110544847466399,
+      "learning_rate": 8.945876919429922e-06,
+      "loss": 2.2067,
+      "step": 653
+    },
+    {
+      "epoch": 0.8855788761002031,
+      "grad_norm": 1.0868074118477926,
+      "learning_rate": 8.942561668108236e-06,
+      "loss": 2.3532,
+      "step": 654
+    },
+    {
+      "epoch": 0.8869329722410291,
+      "grad_norm": 1.1283691827694036,
+      "learning_rate": 8.939241828025356e-06,
+      "loss": 2.2511,
+      "step": 655
+    },
+    {
+      "epoch": 0.8882870683818551,
+      "grad_norm": 1.9053437775194237,
+      "learning_rate": 8.935917403045251e-06,
+      "loss": 2.1626,
+      "step": 656
+    },
+    {
+      "epoch": 0.8896411645226812,
+      "grad_norm": 1.2141944808281238,
+      "learning_rate": 8.932588397037236e-06,
+      "loss": 2.3409,
+      "step": 657
+    },
+    {
+      "epoch": 0.8909952606635071,
+      "grad_norm": 1.0922208185194722,
+      "learning_rate": 8.92925481387595e-06,
+      "loss": 1.7841,
+      "step": 658
+    },
+    {
+      "epoch": 0.8923493568043331,
+      "grad_norm": 1.5140227295690092,
+      "learning_rate": 8.92591665744136e-06,
+      "loss": 2.369,
+      "step": 659
+    },
+    {
+      "epoch": 0.8937034529451591,
+      "grad_norm": 0.8526913110142397,
+      "learning_rate": 8.922573931618763e-06,
+      "loss": 2.0121,
+      "step": 660
+    },
+    {
+      "epoch": 0.8950575490859851,
+      "grad_norm": 1.043736998424973,
+      "learning_rate": 8.919226640298763e-06,
+      "loss": 2.2168,
+      "step": 661
+    },
+    {
+      "epoch": 0.896411645226811,
+      "grad_norm": 0.8649236144416748,
+      "learning_rate": 8.915874787377289e-06,
+      "loss": 1.9414,
+      "step": 662
+    },
+    {
+      "epoch": 0.8977657413676371,
+      "grad_norm": 1.0840827090139065,
+      "learning_rate": 8.912518376755572e-06,
+      "loss": 3.087,
+      "step": 663
+    },
+    {
+      "epoch": 0.8991198375084631,
+      "grad_norm": 1.1561825732631927,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 2.1006,
+      "step": 664
+    },
+    {
+      "epoch": 0.9004739336492891,
+      "grad_norm": 0.9793291395752391,
+      "learning_rate": 8.905791898042861e-06,
+      "loss": 1.85,
+      "step": 665
+    },
+    {
+      "epoch": 0.9018280297901151,
+      "grad_norm": 1.2185211503560396,
+      "learning_rate": 8.902421837780839e-06,
+      "loss": 2.7844,
+      "step": 666
+    },
+    {
+      "epoch": 0.903182125930941,
+      "grad_norm": 1.0236959874012588,
+      "learning_rate": 8.89904723547651e-06,
+      "loss": 2.1372,
+      "step": 667
+    },
+    {
+      "epoch": 0.9045362220717671,
+      "grad_norm": 1.0598101921330767,
+      "learning_rate": 8.895668095057584e-06,
+      "loss": 2.413,
+      "step": 668
+    },
+    {
+      "epoch": 0.9058903182125931,
+      "grad_norm": 0.9973400634976877,
+      "learning_rate": 8.892284420457054e-06,
+      "loss": 2.1688,
+      "step": 669
+    },
+    {
+      "epoch": 0.9072444143534191,
+      "grad_norm": 1.11854939708825,
+      "learning_rate": 8.888896215613192e-06,
+      "loss": 2.3154,
+      "step": 670
+    },
+    {
+      "epoch": 0.9085985104942451,
+      "grad_norm": 1.0425526725686658,
+      "learning_rate": 8.885503484469539e-06,
+      "loss": 2.2115,
+      "step": 671
+    },
+    {
+      "epoch": 0.909952606635071,
+      "grad_norm": 0.9579083361963218,
+      "learning_rate": 8.88210623097491e-06,
+      "loss": 2.0552,
+      "step": 672
+    },
+    {
+      "epoch": 0.9113067027758971,
+      "grad_norm": 0.8237407553812768,
+      "learning_rate": 8.878704459083376e-06,
+      "loss": 2.8482,
+      "step": 673
+    },
+    {
+      "epoch": 0.9126607989167231,
+      "grad_norm": 0.8200385146933777,
+      "learning_rate": 8.875298172754274e-06,
+      "loss": 2.1437,
+      "step": 674
+    },
+    {
+      "epoch": 0.9140148950575491,
+      "grad_norm": 1.2074180626236117,
+      "learning_rate": 8.871887375952192e-06,
+      "loss": 2.0688,
+      "step": 675
+    },
+    {
+      "epoch": 0.9153689911983751,
+      "grad_norm": 1.2749052909022498,
+      "learning_rate": 8.868472072646968e-06,
+      "loss": 2.4633,
+      "step": 676
+    },
+    {
+      "epoch": 0.916723087339201,
+      "grad_norm": 1.219617129030152,
+      "learning_rate": 8.865052266813686e-06,
+      "loss": 2.3246,
+      "step": 677
+    },
+    {
+      "epoch": 0.9180771834800271,
+      "grad_norm": 0.981406917677342,
+      "learning_rate": 8.861627962432669e-06,
+      "loss": 2.2168,
+      "step": 678
+    },
+    {
+      "epoch": 0.919431279620853,
+      "grad_norm": 1.1941352888187047,
+      "learning_rate": 8.858199163489476e-06,
+      "loss": 2.0688,
+      "step": 679
+    },
+    {
+      "epoch": 0.9207853757616791,
+      "grad_norm": 1.0302188684262572,
+      "learning_rate": 8.854765873974898e-06,
+      "loss": 2.6023,
+      "step": 680
+    },
+    {
+      "epoch": 0.9221394719025051,
+      "grad_norm": 1.0315351443360778,
+      "learning_rate": 8.851328097884956e-06,
+      "loss": 2.4473,
+      "step": 681
+    },
+    {
+      "epoch": 0.923493568043331,
+      "grad_norm": 0.9270825061496434,
+      "learning_rate": 8.847885839220884e-06,
+      "loss": 2.3396,
+      "step": 682
+    },
+    {
+      "epoch": 0.9248476641841571,
+      "grad_norm": 1.5668051937578615,
+      "learning_rate": 8.844439101989145e-06,
+      "loss": 2.1618,
+      "step": 683
+    },
+    {
+      "epoch": 0.926201760324983,
+      "grad_norm": 1.063281721298216,
+      "learning_rate": 8.840987890201404e-06,
+      "loss": 2.0825,
+      "step": 684
+    },
+    {
+      "epoch": 0.9275558564658091,
+      "grad_norm": 5.49994743188365,
+      "learning_rate": 8.837532207874539e-06,
+      "loss": 2.3203,
+      "step": 685
+    },
+    {
+      "epoch": 0.9289099526066351,
+      "grad_norm": 0.805384496899143,
+      "learning_rate": 8.834072059030631e-06,
+      "loss": 2.2733,
+      "step": 686
+    },
+    {
+      "epoch": 0.930264048747461,
+      "grad_norm": 1.2279223754931932,
+      "learning_rate": 8.83060744769696e-06,
+      "loss": 2.519,
+      "step": 687
+    },
+    {
+      "epoch": 0.9316181448882871,
+      "grad_norm": 0.9224586328427765,
+      "learning_rate": 8.827138377905999e-06,
+      "loss": 2.8326,
+      "step": 688
+    },
+    {
+      "epoch": 0.932972241029113,
+      "grad_norm": 1.6089966176184487,
+      "learning_rate": 8.823664853695408e-06,
+      "loss": 1.6853,
+      "step": 689
+    },
+    {
+      "epoch": 0.9343263371699391,
+      "grad_norm": 0.8870008681490367,
+      "learning_rate": 8.820186879108038e-06,
+      "loss": 2.4482,
+      "step": 690
+    },
+    {
+      "epoch": 0.9356804333107651,
+      "grad_norm": 0.91606230626226,
+      "learning_rate": 8.816704458191913e-06,
+      "loss": 2.4619,
+      "step": 691
+    },
+    {
+      "epoch": 0.937034529451591,
+      "grad_norm": 1.103378118700351,
+      "learning_rate": 8.813217595000234e-06,
+      "loss": 2.0875,
+      "step": 692
+    },
+    {
+      "epoch": 0.9383886255924171,
+      "grad_norm": 1.1782658973751619,
+      "learning_rate": 8.809726293591376e-06,
+      "loss": 2.2522,
+      "step": 693
+    },
+    {
+      "epoch": 0.939742721733243,
+      "grad_norm": 3.1082108827531383,
+      "learning_rate": 8.806230558028874e-06,
+      "loss": 2.4234,
+      "step": 694
+    },
+    {
+      "epoch": 0.9410968178740691,
+      "grad_norm": 1.2771263022311399,
+      "learning_rate": 8.80273039238143e-06,
+      "loss": 2.2953,
+      "step": 695
+    },
+    {
+      "epoch": 0.942450914014895,
+      "grad_norm": 1.0851177832663736,
+      "learning_rate": 8.799225800722895e-06,
+      "loss": 2.1023,
+      "step": 696
+    },
+    {
+      "epoch": 0.943805010155721,
+      "grad_norm": 0.8982930726085163,
+      "learning_rate": 8.795716787132278e-06,
+      "loss": 2.261,
+      "step": 697
+    },
+    {
+      "epoch": 0.9451591062965471,
+      "grad_norm": 0.9710609235588211,
+      "learning_rate": 8.792203355693731e-06,
+      "loss": 2.0593,
+      "step": 698
+    },
+    {
+      "epoch": 0.946513202437373,
+      "grad_norm": 1.1589767335158037,
+      "learning_rate": 8.788685510496549e-06,
+      "loss": 2.2502,
+      "step": 699
+    },
+    {
+      "epoch": 0.9478672985781991,
+      "grad_norm": 0.923602460813735,
+      "learning_rate": 8.785163255635166e-06,
+      "loss": 2.1052,
+      "step": 700
+    },
+    {
+      "epoch": 0.949221394719025,
+      "grad_norm": 1.1331497754597915,
+      "learning_rate": 8.781636595209145e-06,
+      "loss": 2.2342,
+      "step": 701
+    },
+    {
+      "epoch": 0.950575490859851,
+      "grad_norm": 1.0911649916589345,
+      "learning_rate": 8.778105533323177e-06,
+      "loss": 2.5234,
+      "step": 702
+    },
+    {
+      "epoch": 0.9519295870006771,
+      "grad_norm": 0.7939604108123635,
+      "learning_rate": 8.77457007408708e-06,
+      "loss": 2.3947,
+      "step": 703
+    },
+    {
+      "epoch": 0.953283683141503,
+      "grad_norm": 0.9952541607703701,
+      "learning_rate": 8.771030221615786e-06,
+      "loss": 2.4793,
+      "step": 704
+    },
+    {
+      "epoch": 0.9546377792823291,
+      "grad_norm": 1.0107073531709874,
+      "learning_rate": 8.767485980029342e-06,
+      "loss": 2.0951,
+      "step": 705
+    },
+    {
+      "epoch": 0.955991875423155,
+      "grad_norm": 1.0100357247296368,
+      "learning_rate": 8.763937353452902e-06,
+      "loss": 2.1011,
+      "step": 706
+    },
+    {
+      "epoch": 0.957345971563981,
+      "grad_norm": 0.9804384454423593,
+      "learning_rate": 8.760384346016726e-06,
+      "loss": 1.7807,
+      "step": 707
+    },
+    {
+      "epoch": 0.9587000677048071,
+      "grad_norm": 1.948525190001482,
+      "learning_rate": 8.756826961856171e-06,
+      "loss": 2.2514,
+      "step": 708
+    },
+    {
+      "epoch": 0.960054163845633,
+      "grad_norm": 2.325650876165679,
+      "learning_rate": 8.75326520511169e-06,
+      "loss": 2.43,
+      "step": 709
+    },
+    {
+      "epoch": 0.9614082599864591,
+      "grad_norm": 1.4411906872847189,
+      "learning_rate": 8.74969907992882e-06,
+      "loss": 2.5396,
+      "step": 710
+    },
+    {
+      "epoch": 0.962762356127285,
+      "grad_norm": 1.1588565975941216,
+      "learning_rate": 8.746128590458191e-06,
+      "loss": 2.2853,
+      "step": 711
+    },
+    {
+      "epoch": 0.964116452268111,
+      "grad_norm": 1.018506381733888,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 2.3352,
+      "step": 712
+    },
+    {
+      "epoch": 0.965470548408937,
+      "grad_norm": 0.9885672968116678,
+      "learning_rate": 8.738974535281545e-06,
+      "loss": 2.3807,
+      "step": 713
+    },
+    {
+      "epoch": 0.966824644549763,
+      "grad_norm": 0.9360638976400335,
+      "learning_rate": 8.735390977902159e-06,
+      "loss": 2.522,
+      "step": 714
+    },
+    {
+      "epoch": 0.9681787406905891,
+      "grad_norm": 1.0138074790076992,
+      "learning_rate": 8.731803072888262e-06,
+      "loss": 2.0687,
+      "step": 715
+    },
+    {
+      "epoch": 0.969532836831415,
+      "grad_norm": 0.8593670335614861,
+      "learning_rate": 8.728210824415829e-06,
+      "loss": 2.3171,
+      "step": 716
+    },
+    {
+      "epoch": 0.970886932972241,
+      "grad_norm": 0.8782947340827698,
+      "learning_rate": 8.724614236665889e-06,
+      "loss": 2.1826,
+      "step": 717
+    },
+    {
+      "epoch": 0.972241029113067,
+      "grad_norm": 1.041877641264968,
+      "learning_rate": 8.721013313824527e-06,
+      "loss": 2.5999,
+      "step": 718
+    },
+    {
+      "epoch": 0.973595125253893,
+      "grad_norm": 1.0184265790630844,
+      "learning_rate": 8.717408060082865e-06,
+      "loss": 2.3649,
+      "step": 719
+    },
+    {
+      "epoch": 0.9749492213947191,
+      "grad_norm": 0.8805171555606713,
+      "learning_rate": 8.713798479637073e-06,
+      "loss": 2.5529,
+      "step": 720
+    },
+    {
+      "epoch": 0.976303317535545,
+      "grad_norm": 0.8250466080220131,
+      "learning_rate": 8.710184576688353e-06,
+      "loss": 2.1707,
+      "step": 721
+    },
+    {
+      "epoch": 0.977657413676371,
+      "grad_norm": 0.9735738962932305,
+      "learning_rate": 8.70656635544294e-06,
+      "loss": 1.9474,
+      "step": 722
+    },
+    {
+      "epoch": 0.979011509817197,
+      "grad_norm": 0.9574755177161404,
+      "learning_rate": 8.702943820112094e-06,
+      "loss": 2.0991,
+      "step": 723
+    },
+    {
+      "epoch": 0.980365605958023,
+      "grad_norm": 1.1008036232420575,
+      "learning_rate": 8.699316974912097e-06,
+      "loss": 2.6044,
+      "step": 724
+    },
+    {
+      "epoch": 0.9817197020988491,
+      "grad_norm": 1.0106221199017615,
+      "learning_rate": 8.695685824064246e-06,
+      "loss": 2.4405,
+      "step": 725
+    },
+    {
+      "epoch": 0.983073798239675,
+      "grad_norm": 1.167558559334278,
+      "learning_rate": 8.692050371794849e-06,
+      "loss": 2.6635,
+      "step": 726
+    },
+    {
+      "epoch": 0.984427894380501,
+      "grad_norm": 0.8542048524860696,
+      "learning_rate": 8.688410622335222e-06,
+      "loss": 2.5169,
+      "step": 727
+    },
+    {
+      "epoch": 0.985781990521327,
+      "grad_norm": 1.0044160146409207,
+      "learning_rate": 8.684766579921684e-06,
+      "loss": 2.477,
+      "step": 728
+    },
+    {
+      "epoch": 0.987136086662153,
+      "grad_norm": 0.9438656631341891,
+      "learning_rate": 8.681118248795548e-06,
+      "loss": 1.9811,
+      "step": 729
+    },
+    {
+      "epoch": 0.988490182802979,
+      "grad_norm": 0.9095360753452296,
+      "learning_rate": 8.677465633203117e-06,
+      "loss": 2.3501,
+      "step": 730
+    },
+    {
+      "epoch": 0.989844278943805,
+      "grad_norm": 1.5751129180562193,
+      "learning_rate": 8.673808737395686e-06,
+      "loss": 1.9711,
+      "step": 731
+    },
+    {
+      "epoch": 0.991198375084631,
+      "grad_norm": 1.0244041655385774,
+      "learning_rate": 8.670147565629526e-06,
+      "loss": 2.1411,
+      "step": 732
+    },
+    {
+      "epoch": 0.992552471225457,
+      "grad_norm": 1.0999564192877611,
+      "learning_rate": 8.66648212216589e-06,
+      "loss": 2.7566,
+      "step": 733
+    },
+    {
+      "epoch": 0.993906567366283,
+      "grad_norm": 1.0783864346910368,
+      "learning_rate": 8.662812411270997e-06,
+      "loss": 2.2272,
+      "step": 734
+    },
+    {
+      "epoch": 0.995260663507109,
+      "grad_norm": 1.011628347758097,
+      "learning_rate": 8.65913843721604e-06,
+      "loss": 2.5275,
+      "step": 735
+    },
+    {
+      "epoch": 0.996614759647935,
+      "grad_norm": 1.2637326655889927,
+      "learning_rate": 8.655460204277167e-06,
+      "loss": 2.7871,
+      "step": 736
+    },
+    {
+      "epoch": 0.997968855788761,
+      "grad_norm": 0.8940421539736993,
+      "learning_rate": 8.651777716735488e-06,
+      "loss": 2.8055,
+      "step": 737
+    },
+    {
+      "epoch": 0.999322951929587,
+      "grad_norm": 0.8894453146981818,
+      "learning_rate": 8.648090978877063e-06,
+      "loss": 2.2385,
+      "step": 738
+    },
+    {
+      "epoch": 1.000677048070413,
+      "grad_norm": 0.9588436493567126,
+      "learning_rate": 8.644399994992898e-06,
+      "loss": 2.1316,
+      "step": 739
+    },
+    {
+      "epoch": 1.001354096140826,
+      "grad_norm": 1.042596861537762,
+      "learning_rate": 8.640704769378943e-06,
+      "loss": 2.1595,
+      "step": 740
+    },
+    {
+      "epoch": 1.002708192281652,
+      "grad_norm": 0.9542700464860409,
+      "learning_rate": 8.637005306336083e-06,
+      "loss": 2.2683,
+      "step": 741
+    },
+    {
+      "epoch": 1.004062288422478,
+      "grad_norm": 1.0792626697240038,
+      "learning_rate": 8.633301610170136e-06,
+      "loss": 1.9535,
+      "step": 742
+    },
+    {
+      "epoch": 1.005416384563304,
+      "grad_norm": 0.9813632090639985,
+      "learning_rate": 8.629593685191846e-06,
+      "loss": 2.3908,
+      "step": 743
+    },
+    {
+      "epoch": 1.00677048070413,
+      "grad_norm": 1.0359954057575655,
+      "learning_rate": 8.625881535716883e-06,
+      "loss": 2.6186,
+      "step": 744
+    },
+    {
+      "epoch": 1.008124576844956,
+      "grad_norm": 0.9759453555493578,
+      "learning_rate": 8.622165166065828e-06,
+      "loss": 1.926,
+      "step": 745
+    },
+    {
+      "epoch": 1.009478672985782,
+      "grad_norm": 1.086383812605352,
+      "learning_rate": 8.618444580564175e-06,
+      "loss": 2.4117,
+      "step": 746
+    },
+    {
+      "epoch": 1.010832769126608,
+      "grad_norm": 1.066485208921513,
+      "learning_rate": 8.61471978354233e-06,
+      "loss": 2.2121,
+      "step": 747
+    },
+    {
+      "epoch": 1.012186865267434,
+      "grad_norm": 1.0075256921761013,
+      "learning_rate": 8.610990779335594e-06,
+      "loss": 2.2686,
+      "step": 748
+    },
+    {
+      "epoch": 1.01354096140826,
+      "grad_norm": 4.05838919324133,
+      "learning_rate": 8.60725757228417e-06,
+      "loss": 2.3371,
+      "step": 749
+    },
+    {
+      "epoch": 1.014895057549086,
+      "grad_norm": 1.0649638809266402,
+      "learning_rate": 8.603520166733153e-06,
+      "loss": 2.2376,
+      "step": 750
+    },
+    {
+      "epoch": 1.016249153689912,
+      "grad_norm": 0.9312921856108541,
+      "learning_rate": 8.599778567032515e-06,
+      "loss": 2.2602,
+      "step": 751
+    },
+    {
+      "epoch": 1.017603249830738,
+      "grad_norm": 1.013555848121877,
+      "learning_rate": 8.596032777537124e-06,
+      "loss": 2.2348,
+      "step": 752
+    },
+    {
+      "epoch": 1.018957345971564,
+      "grad_norm": 1.0949907267977028,
+      "learning_rate": 8.592282802606713e-06,
+      "loss": 2.2799,
+      "step": 753
+    },
+    {
+      "epoch": 1.02031144211239,
+      "grad_norm": 0.8061510705350392,
+      "learning_rate": 8.588528646605893e-06,
+      "loss": 2.1358,
+      "step": 754
+    },
+    {
+      "epoch": 1.021665538253216,
+      "grad_norm": 0.8912985567283312,
+      "learning_rate": 8.584770313904138e-06,
+      "loss": 2.1013,
+      "step": 755
+    },
+    {
+      "epoch": 1.0230196343940419,
+      "grad_norm": 0.938063703434926,
+      "learning_rate": 8.581007808875786e-06,
+      "loss": 2.4889,
+      "step": 756
+    },
+    {
+      "epoch": 1.024373730534868,
+      "grad_norm": 0.7577265350168257,
+      "learning_rate": 8.577241135900026e-06,
+      "loss": 1.8219,
+      "step": 757
+    },
+    {
+      "epoch": 1.025727826675694,
+      "grad_norm": 1.0606203403633938,
+      "learning_rate": 8.573470299360906e-06,
+      "loss": 2.4437,
+      "step": 758
+    },
+    {
+      "epoch": 1.02708192281652,
+      "grad_norm": 1.0551285967576851,
+      "learning_rate": 8.569695303647313e-06,
+      "loss": 2.1422,
+      "step": 759
+    },
+    {
+      "epoch": 1.028436018957346,
+      "grad_norm": 0.9409375423377818,
+      "learning_rate": 8.565916153152982e-06,
+      "loss": 2.1203,
+      "step": 760
+    },
+    {
+      "epoch": 1.0297901150981719,
+      "grad_norm": 1.1003263863457742,
+      "learning_rate": 8.562132852276474e-06,
+      "loss": 2.338,
+      "step": 761
+    },
+    {
+      "epoch": 1.031144211238998,
+      "grad_norm": 0.9457106296289286,
+      "learning_rate": 8.558345405421189e-06,
+      "loss": 1.9891,
+      "step": 762
+    },
+    {
+      "epoch": 1.032498307379824,
+      "grad_norm": 1.0094972443672454,
+      "learning_rate": 8.554553816995347e-06,
+      "loss": 2.064,
+      "step": 763
+    },
+    {
+      "epoch": 1.03385240352065,
+      "grad_norm": 0.9931533657371708,
+      "learning_rate": 8.550758091411993e-06,
+      "loss": 2.5872,
+      "step": 764
+    },
+    {
+      "epoch": 1.035206499661476,
+      "grad_norm": 1.00970787542275,
+      "learning_rate": 8.546958233088986e-06,
+      "loss": 1.6308,
+      "step": 765
+    },
+    {
+      "epoch": 1.0365605958023019,
+      "grad_norm": 0.9046067132912955,
+      "learning_rate": 8.543154246448992e-06,
+      "loss": 2.4595,
+      "step": 766
+    },
+    {
+      "epoch": 1.037914691943128,
+      "grad_norm": 0.8869062626220933,
+      "learning_rate": 8.539346135919485e-06,
+      "loss": 2.2769,
+      "step": 767
+    },
+    {
+      "epoch": 1.039268788083954,
+      "grad_norm": 1.1369978829634468,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 2.2687,
+      "step": 768
+    },
+    {
+      "epoch": 1.04062288422478,
+      "grad_norm": 1.000069084621953,
+      "learning_rate": 8.531717560925819e-06,
+      "loss": 1.9722,
+      "step": 769
+    },
+    {
+      "epoch": 1.041976980365606,
+      "grad_norm": 1.1369770696046089,
+      "learning_rate": 8.527897105340587e-06,
+      "loss": 2.3523,
+      "step": 770
+    },
+    {
+      "epoch": 1.0433310765064319,
+      "grad_norm": 1.2290321379198472,
+      "learning_rate": 8.52407254362368e-06,
+      "loss": 2.2146,
+      "step": 771
+    },
+    {
+      "epoch": 1.044685172647258,
+      "grad_norm": 1.0959412542423825,
+      "learning_rate": 8.52024388022652e-06,
+      "loss": 2.6636,
+      "step": 772
+    },
+    {
+      "epoch": 1.046039268788084,
+      "grad_norm": 1.0258500561432276,
+      "learning_rate": 8.516411119605303e-06,
+      "loss": 2.8278,
+      "step": 773
+    },
+    {
+      "epoch": 1.04739336492891,
+      "grad_norm": 0.9866371420929005,
+      "learning_rate": 8.512574266220991e-06,
+      "loss": 2.5915,
+      "step": 774
+    },
+    {
+      "epoch": 1.048747461069736,
+      "grad_norm": 1.0612020165548255,
+      "learning_rate": 8.50873332453931e-06,
+      "loss": 2.4831,
+      "step": 775
+    },
+    {
+      "epoch": 1.0501015572105619,
+      "grad_norm": 0.9966103250455697,
+      "learning_rate": 8.504888299030748e-06,
+      "loss": 2.1428,
+      "step": 776
+    },
+    {
+      "epoch": 1.051455653351388,
+      "grad_norm": 1.0897590636471566,
+      "learning_rate": 8.501039194170542e-06,
+      "loss": 2.0724,
+      "step": 777
+    },
+    {
+      "epoch": 1.052809749492214,
+      "grad_norm": 1.0670474685447335,
+      "learning_rate": 8.497186014438677e-06,
+      "loss": 1.9695,
+      "step": 778
+    },
+    {
+      "epoch": 1.05416384563304,
+      "grad_norm": 1.1907785256874888,
+      "learning_rate": 8.493328764319886e-06,
+      "loss": 2.4858,
+      "step": 779
+    },
+    {
+      "epoch": 1.055517941773866,
+      "grad_norm": 1.2893351808110196,
+      "learning_rate": 8.489467448303633e-06,
+      "loss": 2.7964,
+      "step": 780
+    },
+    {
+      "epoch": 1.0568720379146919,
+      "grad_norm": 1.0302272518222444,
+      "learning_rate": 8.485602070884118e-06,
+      "loss": 2.3342,
+      "step": 781
+    },
+    {
+      "epoch": 1.058226134055518,
+      "grad_norm": 1.0206240865882117,
+      "learning_rate": 8.481732636560266e-06,
+      "loss": 2.4793,
+      "step": 782
+    },
+    {
+      "epoch": 1.059580230196344,
+      "grad_norm": 1.0121397248564201,
+      "learning_rate": 8.477859149835726e-06,
+      "loss": 1.8307,
+      "step": 783
+    },
+    {
+      "epoch": 1.06093432633717,
+      "grad_norm": 1.0578580561059918,
+      "learning_rate": 8.473981615218863e-06,
+      "loss": 2.1949,
+      "step": 784
+    },
+    {
+      "epoch": 1.062288422477996,
+      "grad_norm": 0.9969690177718763,
+      "learning_rate": 8.470100037222751e-06,
+      "loss": 2.1533,
+      "step": 785
+    },
+    {
+      "epoch": 1.0636425186188219,
+      "grad_norm": 1.1212678356984322,
+      "learning_rate": 8.466214420365176e-06,
+      "loss": 2.1861,
+      "step": 786
+    },
+    {
+      "epoch": 1.064996614759648,
+      "grad_norm": 0.9923066032834624,
+      "learning_rate": 8.462324769168618e-06,
+      "loss": 2.4372,
+      "step": 787
+    },
+    {
+      "epoch": 1.066350710900474,
+      "grad_norm": 1.2404978604054542,
+      "learning_rate": 8.458431088160255e-06,
+      "loss": 2.0584,
+      "step": 788
+    },
+    {
+      "epoch": 1.0677048070413,
+      "grad_norm": 1.060465239616255,
+      "learning_rate": 8.454533381871957e-06,
+      "loss": 2.3261,
+      "step": 789
+    },
+    {
+      "epoch": 1.069058903182126,
+      "grad_norm": 1.0780439600876397,
+      "learning_rate": 8.450631654840279e-06,
+      "loss": 2.2541,
+      "step": 790
+    },
+    {
+      "epoch": 1.0704129993229519,
+      "grad_norm": 1.0543355817090738,
+      "learning_rate": 8.446725911606451e-06,
+      "loss": 2.3729,
+      "step": 791
+    },
+    {
+      "epoch": 1.071767095463778,
+      "grad_norm": 1.5929978362842498,
+      "learning_rate": 8.442816156716386e-06,
+      "loss": 2.2915,
+      "step": 792
+    },
+    {
+      "epoch": 1.073121191604604,
+      "grad_norm": 0.8559195523501567,
+      "learning_rate": 8.438902394720659e-06,
+      "loss": 1.9731,
+      "step": 793
+    },
+    {
+      "epoch": 1.07447528774543,
+      "grad_norm": 1.0220594117408883,
+      "learning_rate": 8.43498463017451e-06,
+      "loss": 2.0958,
+      "step": 794
+    },
+    {
+      "epoch": 1.0758293838862558,
+      "grad_norm": 1.0326806004121811,
+      "learning_rate": 8.431062867637838e-06,
+      "loss": 2.0738,
+      "step": 795
+    },
+    {
+      "epoch": 1.0771834800270819,
+      "grad_norm": 1.18276220665135,
+      "learning_rate": 8.4271371116752e-06,
+      "loss": 2.3479,
+      "step": 796
+    },
+    {
+      "epoch": 1.078537576167908,
+      "grad_norm": 1.1320006680638006,
+      "learning_rate": 8.423207366855794e-06,
+      "loss": 1.8838,
+      "step": 797
+    },
+    {
+      "epoch": 1.079891672308734,
+      "grad_norm": 0.8967800856440162,
+      "learning_rate": 8.419273637753463e-06,
+      "loss": 2.2114,
+      "step": 798
+    },
+    {
+      "epoch": 1.08124576844956,
+      "grad_norm": 0.9935120415012008,
+      "learning_rate": 8.415335928946691e-06,
+      "loss": 2.3093,
+      "step": 799
+    },
+    {
+      "epoch": 1.0825998645903858,
+      "grad_norm": 0.9813682594289855,
+      "learning_rate": 8.411394245018589e-06,
+      "loss": 2.2393,
+      "step": 800
+    },
+    {
+      "epoch": 1.0839539607312119,
+      "grad_norm": 1.0470052038748332,
+      "learning_rate": 8.407448590556895e-06,
+      "loss": 2.0396,
+      "step": 801
+    },
+    {
+      "epoch": 1.085308056872038,
+      "grad_norm": 1.0097113780081433,
+      "learning_rate": 8.403498970153973e-06,
+      "loss": 2.1874,
+      "step": 802
+    },
+    {
+      "epoch": 1.086662153012864,
+      "grad_norm": 0.9573725041864002,
+      "learning_rate": 8.399545388406798e-06,
+      "loss": 2.2734,
+      "step": 803
+    },
+    {
+      "epoch": 1.08801624915369,
+      "grad_norm": 0.9578847774313853,
+      "learning_rate": 8.39558784991696e-06,
+      "loss": 2.186,
+      "step": 804
+    },
+    {
+      "epoch": 1.0893703452945158,
+      "grad_norm": 0.9151868078460803,
+      "learning_rate": 8.391626359290648e-06,
+      "loss": 2.0554,
+      "step": 805
+    },
+    {
+      "epoch": 1.0907244414353419,
+      "grad_norm": 0.9630038514154003,
+      "learning_rate": 8.387660921138657e-06,
+      "loss": 1.6792,
+      "step": 806
+    },
+    {
+      "epoch": 1.092078537576168,
+      "grad_norm": 1.0284639983718127,
+      "learning_rate": 8.383691540076372e-06,
+      "loss": 2.795,
+      "step": 807
+    },
+    {
+      "epoch": 1.093432633716994,
+      "grad_norm": 1.0684029705952454,
+      "learning_rate": 8.379718220723772e-06,
+      "loss": 2.2904,
+      "step": 808
+    },
+    {
+      "epoch": 1.09478672985782,
+      "grad_norm": 1.0034018245502179,
+      "learning_rate": 8.375740967705418e-06,
+      "loss": 2.0063,
+      "step": 809
+    },
+    {
+      "epoch": 1.0961408259986458,
+      "grad_norm": 0.9850203635464737,
+      "learning_rate": 8.371759785650444e-06,
+      "loss": 2.0104,
+      "step": 810
+    },
+    {
+      "epoch": 1.0974949221394719,
+      "grad_norm": 1.0272152976204258,
+      "learning_rate": 8.367774679192566e-06,
+      "loss": 2.4944,
+      "step": 811
+    },
+    {
+      "epoch": 1.098849018280298,
+      "grad_norm": 1.014673449942826,
+      "learning_rate": 8.363785652970062e-06,
+      "loss": 2.4625,
+      "step": 812
+    },
+    {
+      "epoch": 1.100203114421124,
+      "grad_norm": 0.9983178702745071,
+      "learning_rate": 8.359792711625773e-06,
+      "loss": 2.6321,
+      "step": 813
+    },
+    {
+      "epoch": 1.10155721056195,
+      "grad_norm": 1.2478151045690562,
+      "learning_rate": 8.355795859807097e-06,
+      "loss": 2.6152,
+      "step": 814
+    },
+    {
+      "epoch": 1.1029113067027758,
+      "grad_norm": 1.1326333593128552,
+      "learning_rate": 8.351795102165983e-06,
+      "loss": 2.3933,
+      "step": 815
+    },
+    {
+      "epoch": 1.1042654028436019,
+      "grad_norm": 1.2156960926171398,
+      "learning_rate": 8.34779044335893e-06,
+      "loss": 2.5298,
+      "step": 816
+    },
+    {
+      "epoch": 1.105619498984428,
+      "grad_norm": 1.0345252180165674,
+      "learning_rate": 8.343781888046971e-06,
+      "loss": 2.0308,
+      "step": 817
+    },
+    {
+      "epoch": 1.106973595125254,
+      "grad_norm": 0.8846702899770009,
+      "learning_rate": 8.339769440895678e-06,
+      "loss": 2.0716,
+      "step": 818
+    },
+    {
+      "epoch": 1.10832769126608,
+      "grad_norm": 0.9633796785097113,
+      "learning_rate": 8.335753106575153e-06,
+      "loss": 1.9232,
+      "step": 819
+    },
+    {
+      "epoch": 1.1096817874069058,
+      "grad_norm": 1.0958725343390576,
+      "learning_rate": 8.331732889760021e-06,
+      "loss": 2.2685,
+      "step": 820
+    },
+    {
+      "epoch": 1.1110358835477319,
+      "grad_norm": 0.9434748502922011,
+      "learning_rate": 8.327708795129426e-06,
+      "loss": 2.5492,
+      "step": 821
+    },
+    {
+      "epoch": 1.112389979688558,
+      "grad_norm": 1.0218749260607791,
+      "learning_rate": 8.323680827367027e-06,
+      "loss": 2.1566,
+      "step": 822
+    },
+    {
+      "epoch": 1.113744075829384,
+      "grad_norm": 0.9991712465528062,
+      "learning_rate": 8.319648991160987e-06,
+      "loss": 1.9646,
+      "step": 823
+    },
+    {
+      "epoch": 1.1150981719702098,
+      "grad_norm": 0.7872947693815001,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 2.1754,
+      "step": 824
+    },
+    {
+      "epoch": 1.1164522681110358,
+      "grad_norm": 1.0951406212655457,
+      "learning_rate": 8.311573732193162e-06,
+      "loss": 2.3643,
+      "step": 825
+    },
+    {
+      "epoch": 1.1178063642518619,
+      "grad_norm": 1.162631966134576,
+      "learning_rate": 8.307530318830195e-06,
+      "loss": 2.447,
+      "step": 826
+    },
+    {
+      "epoch": 1.119160460392688,
+      "grad_norm": 1.1057256649020668,
+      "learning_rate": 8.303483055821222e-06,
+      "loss": 2.037,
+      "step": 827
+    },
+    {
+      "epoch": 1.120514556533514,
+      "grad_norm": 1.3428986372890304,
+      "learning_rate": 8.299431947876868e-06,
+      "loss": 2.3625,
+      "step": 828
+    },
+    {
+      "epoch": 1.12186865267434,
+      "grad_norm": 1.1140355194533162,
+      "learning_rate": 8.29537699971223e-06,
+      "loss": 2.6255,
+      "step": 829
+    },
+    {
+      "epoch": 1.1232227488151658,
+      "grad_norm": 1.3177560768420005,
+      "learning_rate": 8.291318216046874e-06,
+      "loss": 2.193,
+      "step": 830
+    },
+    {
+      "epoch": 1.1245768449559919,
+      "grad_norm": 1.0528043127642381,
+      "learning_rate": 8.287255601604834e-06,
+      "loss": 1.8949,
+      "step": 831
+    },
+    {
+      "epoch": 1.125930941096818,
+      "grad_norm": 1.1208525533921565,
+      "learning_rate": 8.283189161114602e-06,
+      "loss": 2.8527,
+      "step": 832
+    },
+    {
+      "epoch": 1.127285037237644,
+      "grad_norm": 1.0730370677327041,
+      "learning_rate": 8.279118899309121e-06,
+      "loss": 2.4503,
+      "step": 833
+    },
+    {
+      "epoch": 1.1286391333784698,
+      "grad_norm": 0.9202040373418652,
+      "learning_rate": 8.275044820925784e-06,
+      "loss": 2.066,
+      "step": 834
+    },
+    {
+      "epoch": 1.1299932295192958,
+      "grad_norm": 1.0579688082270045,
+      "learning_rate": 8.270966930706424e-06,
+      "loss": 2.9167,
+      "step": 835
+    },
+    {
+      "epoch": 1.1313473256601219,
+      "grad_norm": 1.0848914875005917,
+      "learning_rate": 8.266885233397312e-06,
+      "loss": 2.8931,
+      "step": 836
+    },
+    {
+      "epoch": 1.132701421800948,
+      "grad_norm": 0.9167254802739708,
+      "learning_rate": 8.262799733749147e-06,
+      "loss": 2.5057,
+      "step": 837
+    },
+    {
+      "epoch": 1.134055517941774,
+      "grad_norm": 1.066012346340753,
+      "learning_rate": 8.25871043651706e-06,
+      "loss": 2.3063,
+      "step": 838
+    },
+    {
+      "epoch": 1.1354096140825998,
+      "grad_norm": 1.0335819197341622,
+      "learning_rate": 8.254617346460598e-06,
+      "loss": 2.333,
+      "step": 839
+    },
+    {
+      "epoch": 1.1367637102234258,
+      "grad_norm": 1.2740330227860488,
+      "learning_rate": 8.250520468343722e-06,
+      "loss": 2.1914,
+      "step": 840
+    },
+    {
+      "epoch": 1.1381178063642519,
+      "grad_norm": 0.9819332138988384,
+      "learning_rate": 8.2464198069348e-06,
+      "loss": 2.2113,
+      "step": 841
+    },
+    {
+      "epoch": 1.139471902505078,
+      "grad_norm": 0.876882108695887,
+      "learning_rate": 8.24231536700661e-06,
+      "loss": 2.0397,
+      "step": 842
+    },
+    {
+      "epoch": 1.140825998645904,
+      "grad_norm": 1.0295962020138234,
+      "learning_rate": 8.238207153336322e-06,
+      "loss": 2.0076,
+      "step": 843
+    },
+    {
+      "epoch": 1.1421800947867298,
+      "grad_norm": 0.934240502035731,
+      "learning_rate": 8.2340951707055e-06,
+      "loss": 2.1013,
+      "step": 844
+    },
+    {
+      "epoch": 1.1435341909275558,
+      "grad_norm": 1.0551844052623185,
+      "learning_rate": 8.229979423900095e-06,
+      "loss": 2.0243,
+      "step": 845
+    },
+    {
+      "epoch": 1.1448882870683819,
+      "grad_norm": 1.0967095459364826,
+      "learning_rate": 8.22585991771044e-06,
+      "loss": 2.1247,
+      "step": 846
+    },
+    {
+      "epoch": 1.146242383209208,
+      "grad_norm": 1.752846560905714,
+      "learning_rate": 8.22173665693124e-06,
+      "loss": 2.0735,
+      "step": 847
+    },
+    {
+      "epoch": 1.147596479350034,
+      "grad_norm": 1.3734604707116838,
+      "learning_rate": 8.217609646361574e-06,
+      "loss": 1.8235,
+      "step": 848
+    },
+    {
+      "epoch": 1.1489505754908598,
+      "grad_norm": 1.1360548028045079,
+      "learning_rate": 8.213478890804884e-06,
+      "loss": 2.4021,
+      "step": 849
+    },
+    {
+      "epoch": 1.1503046716316858,
+      "grad_norm": 1.0221199810893757,
+      "learning_rate": 8.209344395068972e-06,
+      "loss": 1.9002,
+      "step": 850
+    },
+    {
+      "epoch": 1.1516587677725119,
+      "grad_norm": 0.8652009856618762,
+      "learning_rate": 8.20520616396599e-06,
+      "loss": 1.6506,
+      "step": 851
+    },
+    {
+      "epoch": 1.153012863913338,
+      "grad_norm": 0.8365704134586867,
+      "learning_rate": 8.20106420231244e-06,
+      "loss": 2.4525,
+      "step": 852
+    },
+    {
+      "epoch": 1.1543669600541637,
+      "grad_norm": 1.1184140556213078,
+      "learning_rate": 8.196918514929166e-06,
+      "loss": 1.9014,
+      "step": 853
+    },
+    {
+      "epoch": 1.1557210561949898,
+      "grad_norm": 1.0082190304094705,
+      "learning_rate": 8.192769106641349e-06,
+      "loss": 2.3114,
+      "step": 854
+    },
+    {
+      "epoch": 1.1570751523358158,
+      "grad_norm": 1.2193028904213525,
+      "learning_rate": 8.188615982278496e-06,
+      "loss": 2.6133,
+      "step": 855
+    },
+    {
+      "epoch": 1.1584292484766419,
+      "grad_norm": 1.361443418854842,
+      "learning_rate": 8.184459146674447e-06,
+      "loss": 2.4022,
+      "step": 856
+    },
+    {
+      "epoch": 1.159783344617468,
+      "grad_norm": 1.1918066000555871,
+      "learning_rate": 8.180298604667354e-06,
+      "loss": 2.4705,
+      "step": 857
+    },
+    {
+      "epoch": 1.161137440758294,
+      "grad_norm": 1.061895871052199,
+      "learning_rate": 8.17613436109969e-06,
+      "loss": 2.2618,
+      "step": 858
+    },
+    {
+      "epoch": 1.1624915368991198,
+      "grad_norm": 1.107285284892013,
+      "learning_rate": 8.171966420818227e-06,
+      "loss": 2.2433,
+      "step": 859
+    },
+    {
+      "epoch": 1.1638456330399458,
+      "grad_norm": 1.1870517898017734,
+      "learning_rate": 8.167794788674051e-06,
+      "loss": 2.2089,
+      "step": 860
+    },
+    {
+      "epoch": 1.1651997291807719,
+      "grad_norm": 0.9482677548752911,
+      "learning_rate": 8.163619469522537e-06,
+      "loss": 2.2702,
+      "step": 861
+    },
+    {
+      "epoch": 1.166553825321598,
+      "grad_norm": 0.9617964781357382,
+      "learning_rate": 8.159440468223352e-06,
+      "loss": 1.9168,
+      "step": 862
+    },
+    {
+      "epoch": 1.1679079214624237,
+      "grad_norm": 0.9780264000997955,
+      "learning_rate": 8.155257789640452e-06,
+      "loss": 2.1226,
+      "step": 863
+    },
+    {
+      "epoch": 1.1692620176032498,
+      "grad_norm": 1.0030932897097675,
+      "learning_rate": 8.15107143864207e-06,
+      "loss": 2.2428,
+      "step": 864
+    },
+    {
+      "epoch": 1.1706161137440758,
+      "grad_norm": 1.0659315727721794,
+      "learning_rate": 8.146881420100715e-06,
+      "loss": 2.3583,
+      "step": 865
+    },
+    {
+      "epoch": 1.1719702098849019,
+      "grad_norm": 1.1995965464770133,
+      "learning_rate": 8.142687738893161e-06,
+      "loss": 2.1148,
+      "step": 866
+    },
+    {
+      "epoch": 1.173324306025728,
+      "grad_norm": 1.0100674325958858,
+      "learning_rate": 8.138490399900454e-06,
+      "loss": 1.7792,
+      "step": 867
+    },
+    {
+      "epoch": 1.174678402166554,
+      "grad_norm": 1.1528480305040367,
+      "learning_rate": 8.134289408007887e-06,
+      "loss": 2.2195,
+      "step": 868
+    },
+    {
+      "epoch": 1.1760324983073798,
+      "grad_norm": 0.9508957577380588,
+      "learning_rate": 8.130084768105011e-06,
+      "loss": 1.5612,
+      "step": 869
+    },
+    {
+      "epoch": 1.1773865944482058,
+      "grad_norm": 1.0003595585872833,
+      "learning_rate": 8.12587648508562e-06,
+      "loss": 1.8605,
+      "step": 870
+    },
+    {
+      "epoch": 1.1787406905890319,
+      "grad_norm": 1.1420205399905372,
+      "learning_rate": 8.12166456384775e-06,
+      "loss": 2.0476,
+      "step": 871
+    },
+    {
+      "epoch": 1.180094786729858,
+      "grad_norm": 1.091147689337014,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 2.238,
+      "step": 872
+    },
+    {
+      "epoch": 1.1814488828706837,
+      "grad_norm": 1.1033755210296698,
+      "learning_rate": 8.113229826329876e-06,
+      "loss": 2.3103,
+      "step": 873
+    },
+    {
+      "epoch": 1.1828029790115098,
+      "grad_norm": 0.9843912490480584,
+      "learning_rate": 8.109007019867097e-06,
+      "loss": 2.2559,
+      "step": 874
+    },
+    {
+      "epoch": 1.1841570751523358,
+      "grad_norm": 1.0569456356202054,
+      "learning_rate": 8.104780594820265e-06,
+      "loss": 2.0617,
+      "step": 875
+    },
+    {
+      "epoch": 1.1855111712931619,
+      "grad_norm": 0.9431626809735268,
+      "learning_rate": 8.100550556108534e-06,
+      "loss": 2.216,
+      "step": 876
+    },
+    {
+      "epoch": 1.186865267433988,
+      "grad_norm": 0.9619228085042159,
+      "learning_rate": 8.09631690865526e-06,
+      "loss": 2.6777,
+      "step": 877
+    },
+    {
+      "epoch": 1.1882193635748137,
+      "grad_norm": 1.1921389359070207,
+      "learning_rate": 8.092079657388e-06,
+      "loss": 2.1743,
+      "step": 878
+    },
+    {
+      "epoch": 1.1895734597156398,
+      "grad_norm": 0.9898902262989339,
+      "learning_rate": 8.087838807238506e-06,
+      "loss": 2.0959,
+      "step": 879
+    },
+    {
+      "epoch": 1.1909275558564658,
+      "grad_norm": 1.143265112227052,
+      "learning_rate": 8.083594363142717e-06,
+      "loss": 1.7081,
+      "step": 880
+    },
+    {
+      "epoch": 1.1922816519972919,
+      "grad_norm": 1.036462321572088,
+      "learning_rate": 8.079346330040757e-06,
+      "loss": 1.7609,
+      "step": 881
+    },
+    {
+      "epoch": 1.193635748138118,
+      "grad_norm": 1.0246337259282456,
+      "learning_rate": 8.075094712876925e-06,
+      "loss": 2.4473,
+      "step": 882
+    },
+    {
+      "epoch": 1.1949898442789437,
+      "grad_norm": 1.1209881301108675,
+      "learning_rate": 8.070839516599695e-06,
+      "loss": 1.7249,
+      "step": 883
+    },
+    {
+      "epoch": 1.1963439404197698,
+      "grad_norm": 1.128942962423783,
+      "learning_rate": 8.066580746161705e-06,
+      "loss": 2.8118,
+      "step": 884
+    },
+    {
+      "epoch": 1.1976980365605958,
+      "grad_norm": 1.1416681649771092,
+      "learning_rate": 8.062318406519751e-06,
+      "loss": 2.3983,
+      "step": 885
+    },
+    {
+      "epoch": 1.1990521327014219,
+      "grad_norm": 1.1873758496607623,
+      "learning_rate": 8.058052502634786e-06,
+      "loss": 1.9744,
+      "step": 886
+    },
+    {
+      "epoch": 1.200406228842248,
+      "grad_norm": 1.0805673821547641,
+      "learning_rate": 8.053783039471909e-06,
+      "loss": 2.4546,
+      "step": 887
+    },
+    {
+      "epoch": 1.2017603249830737,
+      "grad_norm": 1.0642882838498282,
+      "learning_rate": 8.049510022000365e-06,
+      "loss": 2.063,
+      "step": 888
+    },
+    {
+      "epoch": 1.2031144211238998,
+      "grad_norm": 1.0624621525792766,
+      "learning_rate": 8.045233455193531e-06,
+      "loss": 2.1409,
+      "step": 889
+    },
+    {
+      "epoch": 1.2044685172647258,
+      "grad_norm": 1.2346691816252744,
+      "learning_rate": 8.040953344028921e-06,
+      "loss": 2.0547,
+      "step": 890
+    },
+    {
+      "epoch": 1.2058226134055519,
+      "grad_norm": 0.9557989229824211,
+      "learning_rate": 8.03666969348817e-06,
+      "loss": 1.9712,
+      "step": 891
+    },
+    {
+      "epoch": 1.2071767095463777,
+      "grad_norm": 0.8441961698009585,
+      "learning_rate": 8.032382508557033e-06,
+      "loss": 2.2598,
+      "step": 892
+    },
+    {
+      "epoch": 1.2085308056872037,
+      "grad_norm": 1.009678392621935,
+      "learning_rate": 8.02809179422538e-06,
+      "loss": 2.2036,
+      "step": 893
+    },
+    {
+      "epoch": 1.2098849018280298,
+      "grad_norm": 1.4057498442542968,
+      "learning_rate": 8.023797555487188e-06,
+      "loss": 2.648,
+      "step": 894
+    },
+    {
+      "epoch": 1.2112389979688558,
+      "grad_norm": 1.1266529635992393,
+      "learning_rate": 8.019499797340537e-06,
+      "loss": 2.0414,
+      "step": 895
+    },
+    {
+      "epoch": 1.2125930941096819,
+      "grad_norm": 1.2136121598603382,
+      "learning_rate": 8.015198524787603e-06,
+      "loss": 2.2459,
+      "step": 896
+    },
+    {
+      "epoch": 1.213947190250508,
+      "grad_norm": 1.1213074117332114,
+      "learning_rate": 8.01089374283465e-06,
+      "loss": 2.2054,
+      "step": 897
+    },
+    {
+      "epoch": 1.2153012863913337,
+      "grad_norm": 1.0818726596445376,
+      "learning_rate": 8.00658545649203e-06,
+      "loss": 2.2397,
+      "step": 898
+    },
+    {
+      "epoch": 1.2166553825321598,
+      "grad_norm": 1.0597814871253108,
+      "learning_rate": 8.002273670774172e-06,
+      "loss": 2.4344,
+      "step": 899
+    },
+    {
+      "epoch": 1.2180094786729858,
+      "grad_norm": 1.0180557577675222,
+      "learning_rate": 7.99795839069958e-06,
+      "loss": 1.9542,
+      "step": 900
+    },
+    {
+      "epoch": 1.2193635748138119,
+      "grad_norm": 1.0732743540144358,
+      "learning_rate": 7.99363962129082e-06,
+      "loss": 1.9053,
+      "step": 901
+    },
+    {
+      "epoch": 1.2207176709546377,
+      "grad_norm": 1.1464961982270008,
+      "learning_rate": 7.989317367574528e-06,
+      "loss": 2.2012,
+      "step": 902
+    },
+    {
+      "epoch": 1.2220717670954637,
+      "grad_norm": 0.990300248509241,
+      "learning_rate": 7.984991634581384e-06,
+      "loss": 1.9006,
+      "step": 903
+    },
+    {
+      "epoch": 1.2234258632362898,
+      "grad_norm": 1.0378186156083944,
+      "learning_rate": 7.980662427346127e-06,
+      "loss": 1.8861,
+      "step": 904
+    },
+    {
+      "epoch": 1.2247799593771158,
+      "grad_norm": 4.430236914366399,
+      "learning_rate": 7.976329750907536e-06,
+      "loss": 2.1467,
+      "step": 905
+    },
+    {
+      "epoch": 1.2261340555179419,
+      "grad_norm": 1.0098877057349103,
+      "learning_rate": 7.971993610308428e-06,
+      "loss": 2.0393,
+      "step": 906
+    },
+    {
+      "epoch": 1.2274881516587677,
+      "grad_norm": 1.0949603281546505,
+      "learning_rate": 7.967654010595653e-06,
+      "loss": 2.2995,
+      "step": 907
+    },
+    {
+      "epoch": 1.2288422477995937,
+      "grad_norm": 1.1108897890970795,
+      "learning_rate": 7.963310956820085e-06,
+      "loss": 2.2542,
+      "step": 908
+    },
+    {
+      "epoch": 1.2301963439404198,
+      "grad_norm": 1.1882474635276283,
+      "learning_rate": 7.95896445403662e-06,
+      "loss": 1.9627,
+      "step": 909
+    },
+    {
+      "epoch": 1.2315504400812458,
+      "grad_norm": 1.0695557890668739,
+      "learning_rate": 7.954614507304166e-06,
+      "loss": 2.8768,
+      "step": 910
+    },
+    {
+      "epoch": 1.2329045362220719,
+      "grad_norm": 1.1617487541356466,
+      "learning_rate": 7.950261121685642e-06,
+      "loss": 2.1828,
+      "step": 911
+    },
+    {
+      "epoch": 1.2342586323628977,
+      "grad_norm": 1.1384341697301983,
+      "learning_rate": 7.945904302247968e-06,
+      "loss": 2.3497,
+      "step": 912
+    },
+    {
+      "epoch": 1.2356127285037237,
+      "grad_norm": 1.0784446459158834,
+      "learning_rate": 7.941544054062065e-06,
+      "loss": 2.2683,
+      "step": 913
+    },
+    {
+      "epoch": 1.2369668246445498,
+      "grad_norm": 1.038286976843538,
+      "learning_rate": 7.937180382202836e-06,
+      "loss": 2.2668,
+      "step": 914
+    },
+    {
+      "epoch": 1.2383209207853758,
+      "grad_norm": 1.0029564115724932,
+      "learning_rate": 7.932813291749177e-06,
+      "loss": 2.0261,
+      "step": 915
+    },
+    {
+      "epoch": 1.2396750169262019,
+      "grad_norm": 1.0333541710726606,
+      "learning_rate": 7.928442787783958e-06,
+      "loss": 2.2212,
+      "step": 916
+    },
+    {
+      "epoch": 1.2410291130670277,
+      "grad_norm": 1.03717078747682,
+      "learning_rate": 7.924068875394024e-06,
+      "loss": 2.0631,
+      "step": 917
+    },
+    {
+      "epoch": 1.2423832092078537,
+      "grad_norm": 0.9341937079507571,
+      "learning_rate": 7.919691559670188e-06,
+      "loss": 2.0421,
+      "step": 918
+    },
+    {
+      "epoch": 1.2437373053486798,
+      "grad_norm": 1.1684039509398019,
+      "learning_rate": 7.915310845707223e-06,
+      "loss": 2.0931,
+      "step": 919
+    },
+    {
+      "epoch": 1.2450914014895058,
+      "grad_norm": 0.9994376608594767,
+      "learning_rate": 7.910926738603855e-06,
+      "loss": 2.6169,
+      "step": 920
+    },
+    {
+      "epoch": 1.2464454976303316,
+      "grad_norm": 1.3781109917383612,
+      "learning_rate": 7.906539243462763e-06,
+      "loss": 2.8901,
+      "step": 921
+    },
+    {
+      "epoch": 1.2477995937711577,
+      "grad_norm": 1.1519867240702086,
+      "learning_rate": 7.902148365390567e-06,
+      "loss": 2.3781,
+      "step": 922
+    },
+    {
+      "epoch": 1.2491536899119837,
+      "grad_norm": 1.2921377062772406,
+      "learning_rate": 7.897754109497826e-06,
+      "loss": 2.4671,
+      "step": 923
+    },
+    {
+      "epoch": 1.2505077860528098,
+      "grad_norm": 1.0832508824791298,
+      "learning_rate": 7.89335648089903e-06,
+      "loss": 2.3355,
+      "step": 924
+    },
+    {
+      "epoch": 1.2518618821936358,
+      "grad_norm": 1.156819119690463,
+      "learning_rate": 7.888955484712592e-06,
+      "loss": 2.6164,
+      "step": 925
+    },
+    {
+      "epoch": 1.2532159783344619,
+      "grad_norm": 1.0295051950872909,
+      "learning_rate": 7.88455112606085e-06,
+      "loss": 1.9817,
+      "step": 926
+    },
+    {
+      "epoch": 1.2545700744752877,
+      "grad_norm": 0.9867097354242217,
+      "learning_rate": 7.880143410070051e-06,
+      "loss": 1.9863,
+      "step": 927
+    },
+    {
+      "epoch": 1.2559241706161137,
+      "grad_norm": 1.1107662231325113,
+      "learning_rate": 7.875732341870349e-06,
+      "loss": 2.0024,
+      "step": 928
+    },
+    {
+      "epoch": 1.2572782667569398,
+      "grad_norm": 0.9832002800266002,
+      "learning_rate": 7.871317926595804e-06,
+      "loss": 2.1088,
+      "step": 929
+    },
+    {
+      "epoch": 1.2586323628977658,
+      "grad_norm": 0.9422090734118035,
+      "learning_rate": 7.866900169384368e-06,
+      "loss": 2.4123,
+      "step": 930
+    },
+    {
+      "epoch": 1.2599864590385916,
+      "grad_norm": 1.0997264185351083,
+      "learning_rate": 7.862479075377884e-06,
+      "loss": 2.2158,
+      "step": 931
+    },
+    {
+      "epoch": 1.2613405551794177,
+      "grad_norm": 1.1188615261512294,
+      "learning_rate": 7.85805464972208e-06,
+      "loss": 2.3005,
+      "step": 932
+    },
+    {
+      "epoch": 1.2626946513202437,
+      "grad_norm": 1.0299930093784284,
+      "learning_rate": 7.85362689756656e-06,
+      "loss": 2.2651,
+      "step": 933
+    },
+    {
+      "epoch": 1.2640487474610698,
+      "grad_norm": 0.9937695521744077,
+      "learning_rate": 7.849195824064797e-06,
+      "loss": 2.2419,
+      "step": 934
+    },
+    {
+      "epoch": 1.2654028436018958,
+      "grad_norm": 1.0175704885653492,
+      "learning_rate": 7.844761434374138e-06,
+      "loss": 2.2974,
+      "step": 935
+    },
+    {
+      "epoch": 1.2667569397427219,
+      "grad_norm": 1.1723623677027686,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 2.0875,
+      "step": 936
+    },
+    {
+      "epoch": 1.2681110358835477,
+      "grad_norm": 1.2528863590890928,
+      "learning_rate": 7.835882727074779e-06,
+      "loss": 2.1546,
+      "step": 937
+    },
+    {
+      "epoch": 1.2694651320243737,
+      "grad_norm": 1.3110354157785997,
+      "learning_rate": 7.831438419800042e-06,
+      "loss": 2.1202,
+      "step": 938
+    },
+    {
+      "epoch": 1.2708192281651998,
+      "grad_norm": 1.0304392874864414,
+      "learning_rate": 7.826990817004307e-06,
+      "loss": 2.1668,
+      "step": 939
+    },
+    {
+      "epoch": 1.2721733243060258,
+      "grad_norm": 1.0116042176431888,
+      "learning_rate": 7.82253992386416e-06,
+      "loss": 2.2464,
+      "step": 940
+    },
+    {
+      "epoch": 1.2735274204468516,
+      "grad_norm": 0.9962700643735357,
+      "learning_rate": 7.818085745560009e-06,
+      "loss": 2.2486,
+      "step": 941
+    },
+    {
+      "epoch": 1.2748815165876777,
+      "grad_norm": 1.7303323063471459,
+      "learning_rate": 7.813628287276087e-06,
+      "loss": 2.2405,
+      "step": 942
+    },
+    {
+      "epoch": 1.2762356127285037,
+      "grad_norm": 1.0526531855129573,
+      "learning_rate": 7.809167554200446e-06,
+      "loss": 2.4385,
+      "step": 943
+    },
+    {
+      "epoch": 1.2775897088693298,
+      "grad_norm": 0.9488843491039597,
+      "learning_rate": 7.804703551524948e-06,
+      "loss": 2.0235,
+      "step": 944
+    },
+    {
+      "epoch": 1.2789438050101558,
+      "grad_norm": 1.0271145865465217,
+      "learning_rate": 7.800236284445262e-06,
+      "loss": 1.9179,
+      "step": 945
+    },
+    {
+      "epoch": 1.2802979011509819,
+      "grad_norm": 1.0978694506737823,
+      "learning_rate": 7.795765758160855e-06,
+      "loss": 2.6674,
+      "step": 946
+    },
+    {
+      "epoch": 1.2816519972918077,
+      "grad_norm": 0.9144971550877312,
+      "learning_rate": 7.791291977874989e-06,
+      "loss": 2.1267,
+      "step": 947
+    },
+    {
+      "epoch": 1.2830060934326337,
+      "grad_norm": 1.0316859906377736,
+      "learning_rate": 7.78681494879471e-06,
+      "loss": 2.2941,
+      "step": 948
+    },
+    {
+      "epoch": 1.2843601895734598,
+      "grad_norm": 1.2313895009261298,
+      "learning_rate": 7.78233467613085e-06,
+      "loss": 2.1893,
+      "step": 949
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 1.158158610918687,
+      "learning_rate": 7.777851165098012e-06,
+      "loss": 1.8157,
+      "step": 950
+    },
+    {
+      "epoch": 1.2870683818551116,
+      "grad_norm": 1.0349810102864025,
+      "learning_rate": 7.773364420914572e-06,
+      "loss": 2.4442,
+      "step": 951
+    },
+    {
+      "epoch": 1.2884224779959377,
+      "grad_norm": 1.1215212355798856,
+      "learning_rate": 7.768874448802665e-06,
+      "loss": 2.1174,
+      "step": 952
+    },
+    {
+      "epoch": 1.2897765741367637,
+      "grad_norm": 0.9965099934668875,
+      "learning_rate": 7.764381253988189e-06,
+      "loss": 2.5512,
+      "step": 953
+    },
+    {
+      "epoch": 1.2911306702775898,
+      "grad_norm": 1.104679475403435,
+      "learning_rate": 7.759884841700785e-06,
+      "loss": 2.0502,
+      "step": 954
+    },
+    {
+      "epoch": 1.2924847664184158,
+      "grad_norm": 1.660982051919665,
+      "learning_rate": 7.755385217173845e-06,
+      "loss": 2.4249,
+      "step": 955
+    },
+    {
+      "epoch": 1.2938388625592416,
+      "grad_norm": 1.0755785420417305,
+      "learning_rate": 7.750882385644495e-06,
+      "loss": 2.3396,
+      "step": 956
+    },
+    {
+      "epoch": 1.2951929587000677,
+      "grad_norm": 1.0925433109915934,
+      "learning_rate": 7.746376352353599e-06,
+      "loss": 2.249,
+      "step": 957
+    },
+    {
+      "epoch": 1.2965470548408937,
+      "grad_norm": 0.9542648024212192,
+      "learning_rate": 7.741867122545746e-06,
+      "loss": 2.2574,
+      "step": 958
+    },
+    {
+      "epoch": 1.2979011509817198,
+      "grad_norm": 1.1252306171349105,
+      "learning_rate": 7.73735470146924e-06,
+      "loss": 2.1923,
+      "step": 959
+    },
+    {
+      "epoch": 1.2992552471225456,
+      "grad_norm": 1.0754233683086076,
+      "learning_rate": 7.732839094376106e-06,
+      "loss": 2.0116,
+      "step": 960
+    },
+    {
+      "epoch": 1.3006093432633716,
+      "grad_norm": 1.1306744896780743,
+      "learning_rate": 7.728320306522074e-06,
+      "loss": 2.3124,
+      "step": 961
+    },
+    {
+      "epoch": 1.3019634394041977,
+      "grad_norm": 1.072223164700719,
+      "learning_rate": 7.723798343166578e-06,
+      "loss": 2.0477,
+      "step": 962
+    },
+    {
+      "epoch": 1.3033175355450237,
+      "grad_norm": 1.107055290773056,
+      "learning_rate": 7.719273209572745e-06,
+      "loss": 2.029,
+      "step": 963
+    },
+    {
+      "epoch": 1.3046716316858498,
+      "grad_norm": 1.0396566849531226,
+      "learning_rate": 7.714744911007395e-06,
+      "loss": 2.1424,
+      "step": 964
+    },
+    {
+      "epoch": 1.3060257278266758,
+      "grad_norm": 1.1282128940500473,
+      "learning_rate": 7.710213452741028e-06,
+      "loss": 2.3644,
+      "step": 965
+    },
+    {
+      "epoch": 1.3073798239675016,
+      "grad_norm": 1.0660837047209504,
+      "learning_rate": 7.70567884004783e-06,
+      "loss": 2.3568,
+      "step": 966
+    },
+    {
+      "epoch": 1.3087339201083277,
+      "grad_norm": 1.0098446863045805,
+      "learning_rate": 7.701141078205648e-06,
+      "loss": 2.2343,
+      "step": 967
+    },
+    {
+      "epoch": 1.3100880162491537,
+      "grad_norm": 9.794773052164214,
+      "learning_rate": 7.696600172495997e-06,
+      "loss": 2.7452,
+      "step": 968
+    },
+    {
+      "epoch": 1.3114421123899798,
+      "grad_norm": 1.3318037186474474,
+      "learning_rate": 7.692056128204056e-06,
+      "loss": 1.8471,
+      "step": 969
+    },
+    {
+      "epoch": 1.3127962085308056,
+      "grad_norm": 1.151203349471892,
+      "learning_rate": 7.687508950618654e-06,
+      "loss": 2.6144,
+      "step": 970
+    },
+    {
+      "epoch": 1.3141503046716316,
+      "grad_norm": 1.0959293878977527,
+      "learning_rate": 7.682958645032265e-06,
+      "loss": 2.22,
+      "step": 971
+    },
+    {
+      "epoch": 1.3155044008124577,
+      "grad_norm": 1.0050631329121955,
+      "learning_rate": 7.678405216741008e-06,
+      "loss": 2.4293,
+      "step": 972
+    },
+    {
+      "epoch": 1.3168584969532837,
+      "grad_norm": 1.0613207475015503,
+      "learning_rate": 7.67384867104463e-06,
+      "loss": 1.9871,
+      "step": 973
+    },
+    {
+      "epoch": 1.3182125930941098,
+      "grad_norm": 1.1967438926432024,
+      "learning_rate": 7.669289013246512e-06,
+      "loss": 2.085,
+      "step": 974
+    },
+    {
+      "epoch": 1.3195666892349358,
+      "grad_norm": 1.2081442687520356,
+      "learning_rate": 7.664726248653658e-06,
+      "loss": 2.1026,
+      "step": 975
+    },
+    {
+      "epoch": 1.3209207853757616,
+      "grad_norm": 1.117288478243434,
+      "learning_rate": 7.660160382576683e-06,
+      "loss": 2.274,
+      "step": 976
+    },
+    {
+      "epoch": 1.3222748815165877,
+      "grad_norm": 1.146272553819504,
+      "learning_rate": 7.655591420329816e-06,
+      "loss": 2.3935,
+      "step": 977
+    },
+    {
+      "epoch": 1.3236289776574137,
+      "grad_norm": 1.658633099303849,
+      "learning_rate": 7.651019367230886e-06,
+      "loss": 2.0573,
+      "step": 978
+    },
+    {
+      "epoch": 1.3249830737982395,
+      "grad_norm": 1.4532457972038215,
+      "learning_rate": 7.646444228601321e-06,
+      "loss": 2.0818,
+      "step": 979
+    },
+    {
+      "epoch": 1.3263371699390656,
+      "grad_norm": 0.9729335641505342,
+      "learning_rate": 7.641866009766145e-06,
+      "loss": 2.1297,
+      "step": 980
+    },
+    {
+      "epoch": 1.3276912660798916,
+      "grad_norm": 1.0010902233959411,
+      "learning_rate": 7.637284716053959e-06,
+      "loss": 1.9135,
+      "step": 981
+    },
+    {
+      "epoch": 1.3290453622207177,
+      "grad_norm": 1.1644542386857344,
+      "learning_rate": 7.63270035279695e-06,
+      "loss": 2.0995,
+      "step": 982
+    },
+    {
+      "epoch": 1.3303994583615437,
+      "grad_norm": 1.17699731802077,
+      "learning_rate": 7.628112925330867e-06,
+      "loss": 2.3395,
+      "step": 983
+    },
+    {
+      "epoch": 1.3317535545023698,
+      "grad_norm": 1.9239341747345802,
+      "learning_rate": 7.62352243899504e-06,
+      "loss": 2.2552,
+      "step": 984
+    },
+    {
+      "epoch": 1.3331076506431956,
+      "grad_norm": 1.104296574087454,
+      "learning_rate": 7.6189288991323505e-06,
+      "loss": 2.12,
+      "step": 985
+    },
+    {
+      "epoch": 1.3344617467840216,
+      "grad_norm": 1.1563231154836133,
+      "learning_rate": 7.614332311089234e-06,
+      "loss": 2.4421,
+      "step": 986
+    },
+    {
+      "epoch": 1.3358158429248477,
+      "grad_norm": 1.1224897324302643,
+      "learning_rate": 7.609732680215676e-06,
+      "loss": 2.5472,
+      "step": 987
+    },
+    {
+      "epoch": 1.3371699390656737,
+      "grad_norm": 1.0798698576819643,
+      "learning_rate": 7.605130011865201e-06,
+      "loss": 1.9996,
+      "step": 988
+    },
+    {
+      "epoch": 1.3385240352064995,
+      "grad_norm": 1.1427247534168647,
+      "learning_rate": 7.600524311394873e-06,
+      "loss": 2.4247,
+      "step": 989
+    },
+    {
+      "epoch": 1.3398781313473256,
+      "grad_norm": 1.1436870392125549,
+      "learning_rate": 7.595915584165283e-06,
+      "loss": 2.2447,
+      "step": 990
+    },
+    {
+      "epoch": 1.3412322274881516,
+      "grad_norm": 1.0123472207553952,
+      "learning_rate": 7.591303835540544e-06,
+      "loss": 2.5301,
+      "step": 991
+    },
+    {
+      "epoch": 1.3425863236289777,
+      "grad_norm": 1.2308969840782176,
+      "learning_rate": 7.586689070888284e-06,
+      "loss": 2.4497,
+      "step": 992
+    },
+    {
+      "epoch": 1.3439404197698037,
+      "grad_norm": 1.2549119339933823,
+      "learning_rate": 7.582071295579647e-06,
+      "loss": 2.4945,
+      "step": 993
+    },
+    {
+      "epoch": 1.3452945159106298,
+      "grad_norm": 1.1551978030637418,
+      "learning_rate": 7.577450514989275e-06,
+      "loss": 1.7533,
+      "step": 994
+    },
+    {
+      "epoch": 1.3466486120514556,
+      "grad_norm": 0.8995145500568735,
+      "learning_rate": 7.572826734495313e-06,
+      "loss": 2.0904,
+      "step": 995
+    },
+    {
+      "epoch": 1.3480027081922816,
+      "grad_norm": 1.0456560935840085,
+      "learning_rate": 7.568199959479393e-06,
+      "loss": 2.3237,
+      "step": 996
+    },
+    {
+      "epoch": 1.3493568043331077,
+      "grad_norm": 1.1047521602796329,
+      "learning_rate": 7.563570195326635e-06,
+      "loss": 2.2112,
+      "step": 997
+    },
+    {
+      "epoch": 1.3507109004739337,
+      "grad_norm": 1.0312134506506243,
+      "learning_rate": 7.558937447425638e-06,
+      "loss": 1.9692,
+      "step": 998
+    },
+    {
+      "epoch": 1.3520649966147595,
+      "grad_norm": 1.114956621453039,
+      "learning_rate": 7.5543017211684745e-06,
+      "loss": 2.5477,
+      "step": 999
+    },
+    {
+      "epoch": 1.3534190927555856,
+      "grad_norm": 1.0810952077309708,
+      "learning_rate": 7.5496630219506805e-06,
+      "loss": 2.5975,
+      "step": 1000
+    },
+    {
+      "epoch": 1.3547731888964116,
+      "grad_norm": 1.0751552768256476,
+      "learning_rate": 7.545021355171253e-06,
+      "loss": 2.4863,
+      "step": 1001
+    },
+    {
+      "epoch": 1.3561272850372377,
+      "grad_norm": 1.1285166971489493,
+      "learning_rate": 7.540376726232648e-06,
+      "loss": 2.1198,
+      "step": 1002
+    },
+    {
+      "epoch": 1.3574813811780637,
+      "grad_norm": 1.3021965465577348,
+      "learning_rate": 7.535729140540762e-06,
+      "loss": 2.5985,
+      "step": 1003
+    },
+    {
+      "epoch": 1.3588354773188898,
+      "grad_norm": 1.1282925324454272,
+      "learning_rate": 7.531078603504938e-06,
+      "loss": 2.4989,
+      "step": 1004
+    },
+    {
+      "epoch": 1.3601895734597156,
+      "grad_norm": 1.1221756439884651,
+      "learning_rate": 7.52642512053795e-06,
+      "loss": 2.2481,
+      "step": 1005
+    },
+    {
+      "epoch": 1.3615436696005416,
+      "grad_norm": 1.057396172107965,
+      "learning_rate": 7.521768697056004e-06,
+      "loss": 2.181,
+      "step": 1006
+    },
+    {
+      "epoch": 1.3628977657413677,
+      "grad_norm": 1.0186231893170619,
+      "learning_rate": 7.517109338478729e-06,
+      "loss": 2.1148,
+      "step": 1007
+    },
+    {
+      "epoch": 1.3642518618821937,
+      "grad_norm": 0.996322134726628,
+      "learning_rate": 7.512447050229166e-06,
+      "loss": 2.0567,
+      "step": 1008
+    },
+    {
+      "epoch": 1.3656059580230195,
+      "grad_norm": 1.0628848521017902,
+      "learning_rate": 7.507781837733771e-06,
+      "loss": 2.0321,
+      "step": 1009
+    },
+    {
+      "epoch": 1.3669600541638456,
+      "grad_norm": 1.2154305322289178,
+      "learning_rate": 7.503113706422398e-06,
+      "loss": 2.6109,
+      "step": 1010
+    },
+    {
+      "epoch": 1.3683141503046716,
+      "grad_norm": 1.2116773535315877,
+      "learning_rate": 7.498442661728305e-06,
+      "loss": 2.2339,
+      "step": 1011
+    },
+    {
+      "epoch": 1.3696682464454977,
+      "grad_norm": 1.0098078280775096,
+      "learning_rate": 7.493768709088133e-06,
+      "loss": 2.9048,
+      "step": 1012
+    },
+    {
+      "epoch": 1.3710223425863237,
+      "grad_norm": 0.9136677144647644,
+      "learning_rate": 7.489091853941914e-06,
+      "loss": 1.5463,
+      "step": 1013
+    },
+    {
+      "epoch": 1.3723764387271498,
+      "grad_norm": 1.075788414941271,
+      "learning_rate": 7.4844121017330575e-06,
+      "loss": 2.0282,
+      "step": 1014
+    },
+    {
+      "epoch": 1.3737305348679756,
+      "grad_norm": 0.9532034628399626,
+      "learning_rate": 7.4797294579083405e-06,
+      "loss": 2.0474,
+      "step": 1015
+    },
+    {
+      "epoch": 1.3750846310088016,
+      "grad_norm": 1.0957243116565307,
+      "learning_rate": 7.475043927917908e-06,
+      "loss": 2.1565,
+      "step": 1016
+    },
+    {
+      "epoch": 1.3764387271496277,
+      "grad_norm": 1.0904772853908138,
+      "learning_rate": 7.470355517215267e-06,
+      "loss": 2.3966,
+      "step": 1017
+    },
+    {
+      "epoch": 1.3777928232904535,
+      "grad_norm": 1.3101240056799253,
+      "learning_rate": 7.4656642312572734e-06,
+      "loss": 2.1686,
+      "step": 1018
+    },
+    {
+      "epoch": 1.3791469194312795,
+      "grad_norm": 1.7456227009220926,
+      "learning_rate": 7.460970075504132e-06,
+      "loss": 2.2421,
+      "step": 1019
+    },
+    {
+      "epoch": 1.3805010155721056,
+      "grad_norm": 1.1602149134117044,
+      "learning_rate": 7.4562730554193875e-06,
+      "loss": 2.5002,
+      "step": 1020
+    },
+    {
+      "epoch": 1.3818551117129316,
+      "grad_norm": 1.292010568667381,
+      "learning_rate": 7.451573176469916e-06,
+      "loss": 2.4289,
+      "step": 1021
+    },
+    {
+      "epoch": 1.3832092078537577,
+      "grad_norm": 1.135185428169095,
+      "learning_rate": 7.446870444125926e-06,
+      "loss": 2.5667,
+      "step": 1022
+    },
+    {
+      "epoch": 1.3845633039945837,
+      "grad_norm": 1.0225658545428182,
+      "learning_rate": 7.442164863860939e-06,
+      "loss": 2.0494,
+      "step": 1023
+    },
+    {
+      "epoch": 1.3859174001354095,
+      "grad_norm": 1.057697531883615,
+      "learning_rate": 7.4374564411518e-06,
+      "loss": 1.9321,
+      "step": 1024
+    },
+    {
+      "epoch": 1.3872714962762356,
+      "grad_norm": 0.9672332524393876,
+      "learning_rate": 7.432745181478659e-06,
+      "loss": 2.3441,
+      "step": 1025
+    },
+    {
+      "epoch": 1.3886255924170616,
+      "grad_norm": 1.1227113972766776,
+      "learning_rate": 7.428031090324966e-06,
+      "loss": 3.3247,
+      "step": 1026
+    },
+    {
+      "epoch": 1.3899796885578877,
+      "grad_norm": 1.158636683545892,
+      "learning_rate": 7.423314173177467e-06,
+      "loss": 2.4228,
+      "step": 1027
+    },
+    {
+      "epoch": 1.3913337846987135,
+      "grad_norm": 1.050811409742003,
+      "learning_rate": 7.4185944355261996e-06,
+      "loss": 2.2984,
+      "step": 1028
+    },
+    {
+      "epoch": 1.3926878808395395,
+      "grad_norm": 1.130180934126097,
+      "learning_rate": 7.413871882864483e-06,
+      "loss": 2.0937,
+      "step": 1029
+    },
+    {
+      "epoch": 1.3940419769803656,
+      "grad_norm": 1.1323994964140216,
+      "learning_rate": 7.4091465206889115e-06,
+      "loss": 1.8611,
+      "step": 1030
+    },
+    {
+      "epoch": 1.3953960731211916,
+      "grad_norm": 1.1272097248462323,
+      "learning_rate": 7.404418354499352e-06,
+      "loss": 2.3052,
+      "step": 1031
+    },
+    {
+      "epoch": 1.3967501692620177,
+      "grad_norm": 1.0357729628233883,
+      "learning_rate": 7.399687389798933e-06,
+      "loss": 1.8565,
+      "step": 1032
+    },
+    {
+      "epoch": 1.3981042654028437,
+      "grad_norm": 1.0211169637817514,
+      "learning_rate": 7.39495363209404e-06,
+      "loss": 2.1472,
+      "step": 1033
+    },
+    {
+      "epoch": 1.3994583615436695,
+      "grad_norm": 1.0575088136735646,
+      "learning_rate": 7.390217086894309e-06,
+      "loss": 1.925,
+      "step": 1034
+    },
+    {
+      "epoch": 1.4008124576844956,
+      "grad_norm": 1.153744768232039,
+      "learning_rate": 7.385477759712625e-06,
+      "loss": 2.6485,
+      "step": 1035
+    },
+    {
+      "epoch": 1.4021665538253216,
+      "grad_norm": 1.1427109111587352,
+      "learning_rate": 7.380735656065103e-06,
+      "loss": 2.0497,
+      "step": 1036
+    },
+    {
+      "epoch": 1.4035206499661477,
+      "grad_norm": 1.1900700043389405,
+      "learning_rate": 7.375990781471098e-06,
+      "loss": 2.3127,
+      "step": 1037
+    },
+    {
+      "epoch": 1.4048747461069735,
+      "grad_norm": 1.1628543548305619,
+      "learning_rate": 7.371243141453185e-06,
+      "loss": 2.0708,
+      "step": 1038
+    },
+    {
+      "epoch": 1.4062288422477995,
+      "grad_norm": 1.1557498507007162,
+      "learning_rate": 7.366492741537156e-06,
+      "loss": 2.0788,
+      "step": 1039
+    },
+    {
+      "epoch": 1.4075829383886256,
+      "grad_norm": 1.0074214592668214,
+      "learning_rate": 7.361739587252019e-06,
+      "loss": 2.2388,
+      "step": 1040
+    },
+    {
+      "epoch": 1.4089370345294516,
+      "grad_norm": 0.9850303377628948,
+      "learning_rate": 7.3569836841299905e-06,
+      "loss": 2.3422,
+      "step": 1041
+    },
+    {
+      "epoch": 1.4102911306702777,
+      "grad_norm": 0.9362862055958489,
+      "learning_rate": 7.352225037706476e-06,
+      "loss": 1.9123,
+      "step": 1042
+    },
+    {
+      "epoch": 1.4116452268111037,
+      "grad_norm": 1.0225754131079985,
+      "learning_rate": 7.347463653520086e-06,
+      "loss": 2.1162,
+      "step": 1043
+    },
+    {
+      "epoch": 1.4129993229519295,
+      "grad_norm": 1.1046568630708087,
+      "learning_rate": 7.34269953711261e-06,
+      "loss": 2.744,
+      "step": 1044
+    },
+    {
+      "epoch": 1.4143534190927556,
+      "grad_norm": 1.0769083803347521,
+      "learning_rate": 7.337932694029018e-06,
+      "loss": 2.1198,
+      "step": 1045
+    },
+    {
+      "epoch": 1.4157075152335816,
+      "grad_norm": 0.9800626580167843,
+      "learning_rate": 7.3331631298174596e-06,
+      "loss": 2.3196,
+      "step": 1046
+    },
+    {
+      "epoch": 1.4170616113744074,
+      "grad_norm": 1.0549618215499794,
+      "learning_rate": 7.3283908500292434e-06,
+      "loss": 2.1395,
+      "step": 1047
+    },
+    {
+      "epoch": 1.4184157075152335,
+      "grad_norm": 1.060574227209818,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 2.8704,
+      "step": 1048
+    },
+    {
+      "epoch": 1.4197698036560595,
+      "grad_norm": 1.0781456745775502,
+      "learning_rate": 7.3188381659438865e-06,
+      "loss": 2.2806,
+      "step": 1049
+    },
+    {
+      "epoch": 1.4211238997968856,
+      "grad_norm": 1.551162578778134,
+      "learning_rate": 7.314057772765148e-06,
+      "loss": 2.116,
+      "step": 1050
+    },
+    {
+      "epoch": 1.4224779959377116,
+      "grad_norm": 1.3756005806034488,
+      "learning_rate": 7.309274686246547e-06,
+      "loss": 3.2393,
+      "step": 1051
+    },
+    {
+      "epoch": 1.4238320920785377,
+      "grad_norm": 1.0635108173294194,
+      "learning_rate": 7.304488911955129e-06,
+      "loss": 2.2153,
+      "step": 1052
+    },
+    {
+      "epoch": 1.4251861882193635,
+      "grad_norm": 1.1523212588317262,
+      "learning_rate": 7.299700455461076e-06,
+      "loss": 1.8101,
+      "step": 1053
+    },
+    {
+      "epoch": 1.4265402843601895,
+      "grad_norm": 1.1351217638799846,
+      "learning_rate": 7.294909322337689e-06,
+      "loss": 2.459,
+      "step": 1054
+    },
+    {
+      "epoch": 1.4278943805010156,
+      "grad_norm": 1.113321093812555,
+      "learning_rate": 7.290115518161385e-06,
+      "loss": 1.9065,
+      "step": 1055
+    },
+    {
+      "epoch": 1.4292484766418416,
+      "grad_norm": 1.1651609776327243,
+      "learning_rate": 7.28531904851169e-06,
+      "loss": 2.2955,
+      "step": 1056
+    },
+    {
+      "epoch": 1.4306025727826674,
+      "grad_norm": 0.9391837640644604,
+      "learning_rate": 7.280519918971228e-06,
+      "loss": 2.5399,
+      "step": 1057
+    },
+    {
+      "epoch": 1.4319566689234935,
+      "grad_norm": 1.1481164415158314,
+      "learning_rate": 7.275718135125726e-06,
+      "loss": 2.3987,
+      "step": 1058
+    },
+    {
+      "epoch": 1.4333107650643195,
+      "grad_norm": 0.9819604190905024,
+      "learning_rate": 7.270913702563995e-06,
+      "loss": 2.0484,
+      "step": 1059
+    },
+    {
+      "epoch": 1.4346648612051456,
+      "grad_norm": 1.241655388978992,
+      "learning_rate": 7.266106626877933e-06,
+      "loss": 2.1864,
+      "step": 1060
+    },
+    {
+      "epoch": 1.4360189573459716,
+      "grad_norm": 1.0531564756435357,
+      "learning_rate": 7.26129691366251e-06,
+      "loss": 2.1386,
+      "step": 1061
+    },
+    {
+      "epoch": 1.4373730534867977,
+      "grad_norm": 1.1218122300065656,
+      "learning_rate": 7.256484568515769e-06,
+      "loss": 2.4484,
+      "step": 1062
+    },
+    {
+      "epoch": 1.4387271496276235,
+      "grad_norm": 1.0001136323518225,
+      "learning_rate": 7.251669597038814e-06,
+      "loss": 2.2824,
+      "step": 1063
+    },
+    {
+      "epoch": 1.4400812457684495,
+      "grad_norm": 1.2896307069641786,
+      "learning_rate": 7.246852004835807e-06,
+      "loss": 2.4539,
+      "step": 1064
+    },
+    {
+      "epoch": 1.4414353419092756,
+      "grad_norm": 1.0510167874878298,
+      "learning_rate": 7.242031797513962e-06,
+      "loss": 2.6035,
+      "step": 1065
+    },
+    {
+      "epoch": 1.4427894380501016,
+      "grad_norm": 1.2155395455915952,
+      "learning_rate": 7.2372089806835335e-06,
+      "loss": 1.9087,
+      "step": 1066
+    },
+    {
+      "epoch": 1.4441435341909274,
+      "grad_norm": 1.1199078701755178,
+      "learning_rate": 7.232383559957815e-06,
+      "loss": 2.1759,
+      "step": 1067
+    },
+    {
+      "epoch": 1.4454976303317535,
+      "grad_norm": 1.159792012820624,
+      "learning_rate": 7.227555540953131e-06,
+      "loss": 2.4314,
+      "step": 1068
+    },
+    {
+      "epoch": 1.4468517264725795,
+      "grad_norm": 1.0730576607550595,
+      "learning_rate": 7.22272492928883e-06,
+      "loss": 2.3359,
+      "step": 1069
+    },
+    {
+      "epoch": 1.4482058226134056,
+      "grad_norm": 0.8222048650029028,
+      "learning_rate": 7.217891730587276e-06,
+      "loss": 2.2215,
+      "step": 1070
+    },
+    {
+      "epoch": 1.4495599187542316,
+      "grad_norm": 1.0285016809620628,
+      "learning_rate": 7.2130559504738464e-06,
+      "loss": 2.239,
+      "step": 1071
+    },
+    {
+      "epoch": 1.4509140148950577,
+      "grad_norm": 1.0693497499216194,
+      "learning_rate": 7.2082175945769226e-06,
+      "loss": 2.2506,
+      "step": 1072
+    },
+    {
+      "epoch": 1.4522681110358835,
+      "grad_norm": 1.1167979192807023,
+      "learning_rate": 7.203376668527884e-06,
+      "loss": 2.2712,
+      "step": 1073
+    },
+    {
+      "epoch": 1.4536222071767095,
+      "grad_norm": 1.1307443613901718,
+      "learning_rate": 7.198533177961102e-06,
+      "loss": 2.0017,
+      "step": 1074
+    },
+    {
+      "epoch": 1.4549763033175356,
+      "grad_norm": 1.0564264494899962,
+      "learning_rate": 7.19368712851393e-06,
+      "loss": 2.1056,
+      "step": 1075
+    },
+    {
+      "epoch": 1.4563303994583616,
+      "grad_norm": 1.1101530483846054,
+      "learning_rate": 7.188838525826702e-06,
+      "loss": 2.1903,
+      "step": 1076
+    },
+    {
+      "epoch": 1.4576844955991874,
+      "grad_norm": 1.0103673641252706,
+      "learning_rate": 7.183987375542726e-06,
+      "loss": 2.382,
+      "step": 1077
+    },
+    {
+      "epoch": 1.4590385917400135,
+      "grad_norm": 1.253353521477901,
+      "learning_rate": 7.17913368330827e-06,
+      "loss": 2.0939,
+      "step": 1078
+    },
+    {
+      "epoch": 1.4603926878808395,
+      "grad_norm": 1.2508634701536092,
+      "learning_rate": 7.174277454772566e-06,
+      "loss": 1.9324,
+      "step": 1079
+    },
+    {
+      "epoch": 1.4617467840216656,
+      "grad_norm": 8.122448301970948,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 2.1723,
+      "step": 1080
+    },
+    {
+      "epoch": 1.4631008801624916,
+      "grad_norm": 1.3165876873084112,
+      "learning_rate": 7.164557411409075e-06,
+      "loss": 2.0454,
+      "step": 1081
+    },
+    {
+      "epoch": 1.4644549763033177,
+      "grad_norm": 1.3433865812537615,
+      "learning_rate": 7.1596936078944825e-06,
+      "loss": 1.5743,
+      "step": 1082
+    },
+    {
+      "epoch": 1.4658090724441435,
+      "grad_norm": 1.1708156611040208,
+      "learning_rate": 7.154827290705012e-06,
+      "loss": 2.7672,
+      "step": 1083
+    },
+    {
+      "epoch": 1.4671631685849695,
+      "grad_norm": 1.0739066488240625,
+      "learning_rate": 7.1499584655045866e-06,
+      "loss": 2.0953,
+      "step": 1084
+    },
+    {
+      "epoch": 1.4685172647257956,
+      "grad_norm": 1.0810850204219578,
+      "learning_rate": 7.145087137960047e-06,
+      "loss": 2.6921,
+      "step": 1085
+    },
+    {
+      "epoch": 1.4698713608666214,
+      "grad_norm": 0.9939405297389377,
+      "learning_rate": 7.14021331374115e-06,
+      "loss": 2.0588,
+      "step": 1086
+    },
+    {
+      "epoch": 1.4712254570074474,
+      "grad_norm": 0.9226185898131262,
+      "learning_rate": 7.135336998520557e-06,
+      "loss": 1.9489,
+      "step": 1087
+    },
+    {
+      "epoch": 1.4725795531482735,
+      "grad_norm": 1.0783455175484244,
+      "learning_rate": 7.130458197973828e-06,
+      "loss": 1.8894,
+      "step": 1088
+    },
+    {
+      "epoch": 1.4739336492890995,
+      "grad_norm": 1.0986213365125383,
+      "learning_rate": 7.125576917779414e-06,
+      "loss": 2.3981,
+      "step": 1089
+    },
+    {
+      "epoch": 1.4752877454299256,
+      "grad_norm": 0.9907037990576693,
+      "learning_rate": 7.120693163618656e-06,
+      "loss": 1.9927,
+      "step": 1090
+    },
+    {
+      "epoch": 1.4766418415707516,
+      "grad_norm": 1.067381598945353,
+      "learning_rate": 7.115806941175771e-06,
+      "loss": 1.9368,
+      "step": 1091
+    },
+    {
+      "epoch": 1.4779959377115774,
+      "grad_norm": 1.00952047194142,
+      "learning_rate": 7.1109182561378505e-06,
+      "loss": 2.1691,
+      "step": 1092
+    },
+    {
+      "epoch": 1.4793500338524035,
+      "grad_norm": 1.107779870373698,
+      "learning_rate": 7.106027114194856e-06,
+      "loss": 1.8634,
+      "step": 1093
+    },
+    {
+      "epoch": 1.4807041299932295,
+      "grad_norm": 1.1243485392179005,
+      "learning_rate": 7.101133521039597e-06,
+      "loss": 1.9522,
+      "step": 1094
+    },
+    {
+      "epoch": 1.4820582261340556,
+      "grad_norm": 1.0572503465978025,
+      "learning_rate": 7.09623748236775e-06,
+      "loss": 1.8363,
+      "step": 1095
+    },
+    {
+      "epoch": 1.4834123222748814,
+      "grad_norm": 1.0216562069583985,
+      "learning_rate": 7.091339003877826e-06,
+      "loss": 1.8552,
+      "step": 1096
+    },
+    {
+      "epoch": 1.4847664184157074,
+      "grad_norm": 1.089956344974216,
+      "learning_rate": 7.086438091271186e-06,
+      "loss": 2.5221,
+      "step": 1097
+    },
+    {
+      "epoch": 1.4861205145565335,
+      "grad_norm": 1.11408247722042,
+      "learning_rate": 7.0815347502520185e-06,
+      "loss": 2.2585,
+      "step": 1098
+    },
+    {
+      "epoch": 1.4874746106973595,
+      "grad_norm": 0.9710069262269226,
+      "learning_rate": 7.076628986527335e-06,
+      "loss": 2.3226,
+      "step": 1099
+    },
+    {
+      "epoch": 1.4888287068381856,
+      "grad_norm": 1.0917833338553558,
+      "learning_rate": 7.0717208058069755e-06,
+      "loss": 1.9066,
+      "step": 1100
+    },
+    {
+      "epoch": 1.4901828029790116,
+      "grad_norm": 1.1161624453210208,
+      "learning_rate": 7.066810213803586e-06,
+      "loss": 2.1559,
+      "step": 1101
+    },
+    {
+      "epoch": 1.4915368991198374,
+      "grad_norm": 1.16178506916721,
+      "learning_rate": 7.061897216232622e-06,
+      "loss": 2.3097,
+      "step": 1102
+    },
+    {
+      "epoch": 1.4928909952606635,
+      "grad_norm": 1.0384328202163011,
+      "learning_rate": 7.05698181881234e-06,
+      "loss": 1.938,
+      "step": 1103
+    },
+    {
+      "epoch": 1.4942450914014895,
+      "grad_norm": 1.216771459951468,
+      "learning_rate": 7.052064027263785e-06,
+      "loss": 2.4431,
+      "step": 1104
+    },
+    {
+      "epoch": 1.4955991875423156,
+      "grad_norm": 1.3088163468638734,
+      "learning_rate": 7.047143847310794e-06,
+      "loss": 2.0962,
+      "step": 1105
+    },
+    {
+      "epoch": 1.4969532836831414,
+      "grad_norm": 1.1011041050321808,
+      "learning_rate": 7.042221284679982e-06,
+      "loss": 1.9349,
+      "step": 1106
+    },
+    {
+      "epoch": 1.4983073798239674,
+      "grad_norm": 1.2350163332118804,
+      "learning_rate": 7.037296345100734e-06,
+      "loss": 1.8952,
+      "step": 1107
+    },
+    {
+      "epoch": 1.4996614759647935,
+      "grad_norm": 1.0028965498141413,
+      "learning_rate": 7.032369034305209e-06,
+      "loss": 2.4324,
+      "step": 1108
+    },
+    {
+      "epoch": 1.5010155721056195,
+      "grad_norm": 1.0240481324400763,
+      "learning_rate": 7.027439358028315e-06,
+      "loss": 2.4041,
+      "step": 1109
+    },
+    {
+      "epoch": 1.5023696682464456,
+      "grad_norm": 1.2289451784867027,
+      "learning_rate": 7.022507322007723e-06,
+      "loss": 2.296,
+      "step": 1110
+    },
+    {
+      "epoch": 1.5037237643872716,
+      "grad_norm": 1.1846045097053997,
+      "learning_rate": 7.017572931983846e-06,
+      "loss": 1.9807,
+      "step": 1111
+    },
+    {
+      "epoch": 1.5050778605280974,
+      "grad_norm": 1.3679366822185766,
+      "learning_rate": 7.012636193699838e-06,
+      "loss": 2.3165,
+      "step": 1112
+    },
+    {
+      "epoch": 1.5064319566689235,
+      "grad_norm": 1.0735093643875895,
+      "learning_rate": 7.007697112901586e-06,
+      "loss": 2.0637,
+      "step": 1113
+    },
+    {
+      "epoch": 1.5077860528097495,
+      "grad_norm": 0.9805492134387425,
+      "learning_rate": 7.002755695337703e-06,
+      "loss": 2.2236,
+      "step": 1114
+    },
+    {
+      "epoch": 1.5091401489505754,
+      "grad_norm": 1.1540910897325873,
+      "learning_rate": 6.997811946759522e-06,
+      "loss": 2.5638,
+      "step": 1115
+    },
+    {
+      "epoch": 1.5104942450914014,
+      "grad_norm": 1.045564752122656,
+      "learning_rate": 6.9928658729210885e-06,
+      "loss": 2.5629,
+      "step": 1116
+    },
+    {
+      "epoch": 1.5118483412322274,
+      "grad_norm": 1.5327917333436447,
+      "learning_rate": 6.987917479579156e-06,
+      "loss": 2.3958,
+      "step": 1117
+    },
+    {
+      "epoch": 1.5132024373730535,
+      "grad_norm": 1.0061412022965626,
+      "learning_rate": 6.982966772493176e-06,
+      "loss": 2.3869,
+      "step": 1118
+    },
+    {
+      "epoch": 1.5145565335138795,
+      "grad_norm": 1.143443700093151,
+      "learning_rate": 6.978013757425295e-06,
+      "loss": 2.2255,
+      "step": 1119
+    },
+    {
+      "epoch": 1.5159106296547056,
+      "grad_norm": 1.076563769412944,
+      "learning_rate": 6.973058440140341e-06,
+      "loss": 2.0117,
+      "step": 1120
+    },
+    {
+      "epoch": 1.5172647257955316,
+      "grad_norm": 0.9932180957280397,
+      "learning_rate": 6.96810082640583e-06,
+      "loss": 2.4315,
+      "step": 1121
+    },
+    {
+      "epoch": 1.5186188219363574,
+      "grad_norm": 1.7614630897566435,
+      "learning_rate": 6.963140921991941e-06,
+      "loss": 1.8154,
+      "step": 1122
+    },
+    {
+      "epoch": 1.5199729180771835,
+      "grad_norm": 1.0435661710662918,
+      "learning_rate": 6.958178732671527e-06,
+      "loss": 2.1161,
+      "step": 1123
+    },
+    {
+      "epoch": 1.5213270142180095,
+      "grad_norm": 1.034141937866288,
+      "learning_rate": 6.953214264220095e-06,
+      "loss": 2.2153,
+      "step": 1124
+    },
+    {
+      "epoch": 1.5226811103588354,
+      "grad_norm": 1.1333203921847073,
+      "learning_rate": 6.948247522415811e-06,
+      "loss": 2.1679,
+      "step": 1125
+    },
+    {
+      "epoch": 1.5240352064996614,
+      "grad_norm": 1.1523266319281296,
+      "learning_rate": 6.943278513039477e-06,
+      "loss": 2.0316,
+      "step": 1126
+    },
+    {
+      "epoch": 1.5253893026404874,
+      "grad_norm": 1.1175733614955354,
+      "learning_rate": 6.938307241874547e-06,
+      "loss": 1.9331,
+      "step": 1127
+    },
+    {
+      "epoch": 1.5267433987813135,
+      "grad_norm": 1.0816050768685381,
+      "learning_rate": 6.933333714707094e-06,
+      "loss": 2.1337,
+      "step": 1128
+    },
+    {
+      "epoch": 1.5280974949221395,
+      "grad_norm": 0.9555568730660431,
+      "learning_rate": 6.928357937325829e-06,
+      "loss": 2.1897,
+      "step": 1129
+    },
+    {
+      "epoch": 1.5294515910629656,
+      "grad_norm": 1.1330718777414148,
+      "learning_rate": 6.923379915522075e-06,
+      "loss": 2.5013,
+      "step": 1130
+    },
+    {
+      "epoch": 1.5308056872037916,
+      "grad_norm": 1.4860432492374922,
+      "learning_rate": 6.918399655089768e-06,
+      "loss": 2.2661,
+      "step": 1131
+    },
+    {
+      "epoch": 1.5321597833446174,
+      "grad_norm": 1.0608341510421102,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 2.6459,
+      "step": 1132
+    },
+    {
+      "epoch": 1.5335138794854435,
+      "grad_norm": 1.2535263483596926,
+      "learning_rate": 6.908432441528262e-06,
+      "loss": 2.5183,
+      "step": 1133
+    },
+    {
+      "epoch": 1.5348679756262693,
+      "grad_norm": 1.1881324730918719,
+      "learning_rate": 6.903445499999939e-06,
+      "loss": 1.9834,
+      "step": 1134
+    },
+    {
+      "epoch": 1.5362220717670954,
+      "grad_norm": 0.9936992242467968,
+      "learning_rate": 6.8984563430448e-06,
+      "loss": 2.185,
+      "step": 1135
+    },
+    {
+      "epoch": 1.5375761679079214,
+      "grad_norm": 0.9856859764312084,
+      "learning_rate": 6.893464976469739e-06,
+      "loss": 2.413,
+      "step": 1136
+    },
+    {
+      "epoch": 1.5389302640487474,
+      "grad_norm": 0.9728057769285352,
+      "learning_rate": 6.888471406084227e-06,
+      "loss": 2.2291,
+      "step": 1137
+    },
+    {
+      "epoch": 1.5402843601895735,
+      "grad_norm": 1.9592487450519758,
+      "learning_rate": 6.883475637700298e-06,
+      "loss": 2.2332,
+      "step": 1138
+    },
+    {
+      "epoch": 1.5416384563303995,
+      "grad_norm": 2.989310700259001,
+      "learning_rate": 6.8784776771325426e-06,
+      "loss": 2.5173,
+      "step": 1139
+    },
+    {
+      "epoch": 1.5429925524712256,
+      "grad_norm": 1.0753052749288403,
+      "learning_rate": 6.873477530198107e-06,
+      "loss": 2.2734,
+      "step": 1140
+    },
+    {
+      "epoch": 1.5443466486120516,
+      "grad_norm": 1.1348199790348215,
+      "learning_rate": 6.868475202716677e-06,
+      "loss": 3.0435,
+      "step": 1141
+    },
+    {
+      "epoch": 1.5457007447528774,
+      "grad_norm": 0.9253248443470898,
+      "learning_rate": 6.863470700510479e-06,
+      "loss": 2.2282,
+      "step": 1142
+    },
+    {
+      "epoch": 1.5470548408937035,
+      "grad_norm": 1.0212577294991076,
+      "learning_rate": 6.858464029404272e-06,
+      "loss": 1.9986,
+      "step": 1143
+    },
+    {
+      "epoch": 1.5484089370345293,
+      "grad_norm": 1.0119821954044654,
+      "learning_rate": 6.8534551952253395e-06,
+      "loss": 2.236,
+      "step": 1144
+    },
+    {
+      "epoch": 1.5497630331753554,
+      "grad_norm": 0.9376658731281714,
+      "learning_rate": 6.848444203803476e-06,
+      "loss": 1.9798,
+      "step": 1145
+    },
+    {
+      "epoch": 1.5511171293161814,
+      "grad_norm": 1.0518223950174321,
+      "learning_rate": 6.843431060970995e-06,
+      "loss": 2.2833,
+      "step": 1146
+    },
+    {
+      "epoch": 1.5524712254570074,
+      "grad_norm": 1.6688430719575653,
+      "learning_rate": 6.838415772562711e-06,
+      "loss": 2.2911,
+      "step": 1147
+    },
+    {
+      "epoch": 1.5538253215978335,
+      "grad_norm": 2.211386740791148,
+      "learning_rate": 6.833398344415933e-06,
+      "loss": 2.6092,
+      "step": 1148
+    },
+    {
+      "epoch": 1.5551794177386595,
+      "grad_norm": 1.302197054959585,
+      "learning_rate": 6.8283787823704685e-06,
+      "loss": 2.1764,
+      "step": 1149
+    },
+    {
+      "epoch": 1.5565335138794856,
+      "grad_norm": 1.365025659237606,
+      "learning_rate": 6.823357092268596e-06,
+      "loss": 1.9187,
+      "step": 1150
+    },
+    {
+      "epoch": 1.5578876100203114,
+      "grad_norm": 0.9995629551418774,
+      "learning_rate": 6.8183332799550836e-06,
+      "loss": 2.0254,
+      "step": 1151
+    },
+    {
+      "epoch": 1.5592417061611374,
+      "grad_norm": 1.091795853304648,
+      "learning_rate": 6.813307351277161e-06,
+      "loss": 2.5035,
+      "step": 1152
+    },
+    {
+      "epoch": 1.5605958023019635,
+      "grad_norm": 0.9113959325465628,
+      "learning_rate": 6.808279312084525e-06,
+      "loss": 2.2859,
+      "step": 1153
+    },
+    {
+      "epoch": 1.5619498984427893,
+      "grad_norm": 1.105743242638601,
+      "learning_rate": 6.803249168229329e-06,
+      "loss": 1.9762,
+      "step": 1154
+    },
+    {
+      "epoch": 1.5633039945836154,
+      "grad_norm": 1.3089749040803054,
+      "learning_rate": 6.798216925566171e-06,
+      "loss": 2.245,
+      "step": 1155
+    },
+    {
+      "epoch": 1.5646580907244414,
+      "grad_norm": 1.4155040937471588,
+      "learning_rate": 6.7931825899521e-06,
+      "loss": 2.151,
+      "step": 1156
+    },
+    {
+      "epoch": 1.5660121868652674,
+      "grad_norm": 1.2355148875524264,
+      "learning_rate": 6.788146167246594e-06,
+      "loss": 2.2848,
+      "step": 1157
+    },
+    {
+      "epoch": 1.5673662830060935,
+      "grad_norm": 1.134385408399196,
+      "learning_rate": 6.783107663311566e-06,
+      "loss": 1.9886,
+      "step": 1158
+    },
+    {
+      "epoch": 1.5687203791469195,
+      "grad_norm": 1.1445134576197042,
+      "learning_rate": 6.7780670840113434e-06,
+      "loss": 2.2995,
+      "step": 1159
+    },
+    {
+      "epoch": 1.5700744752877456,
+      "grad_norm": 1.1766480093443303,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 2.2219,
+      "step": 1160
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 1.5615268999323522,
+      "learning_rate": 6.767979722784725e-06,
+      "loss": 2.0472,
+      "step": 1161
+    },
+    {
+      "epoch": 1.5727826675693974,
+      "grad_norm": 1.1706173436180667,
+      "learning_rate": 6.762932952599043e-06,
+      "loss": 2.4683,
+      "step": 1162
+    },
+    {
+      "epoch": 1.5741367637102233,
+      "grad_norm": 1.1480375058434151,
+      "learning_rate": 6.757884130529584e-06,
+      "loss": 1.9804,
+      "step": 1163
+    },
+    {
+      "epoch": 1.5754908598510493,
+      "grad_norm": 1.1079469797078965,
+      "learning_rate": 6.75283326245269e-06,
+      "loss": 2.3322,
+      "step": 1164
+    },
+    {
+      "epoch": 1.5768449559918754,
+      "grad_norm": 1.1362279207408492,
+      "learning_rate": 6.7477803542470834e-06,
+      "loss": 1.998,
+      "step": 1165
+    },
+    {
+      "epoch": 1.5781990521327014,
+      "grad_norm": 0.982450102859919,
+      "learning_rate": 6.742725411793862e-06,
+      "loss": 2.3529,
+      "step": 1166
+    },
+    {
+      "epoch": 1.5795531482735274,
+      "grad_norm": 1.1119736244356198,
+      "learning_rate": 6.737668440976494e-06,
+      "loss": 2.0098,
+      "step": 1167
+    },
+    {
+      "epoch": 1.5809072444143535,
+      "grad_norm": 1.1735566109179396,
+      "learning_rate": 6.7326094476808e-06,
+      "loss": 2.4483,
+      "step": 1168
+    },
+    {
+      "epoch": 1.5822613405551795,
+      "grad_norm": 0.9600432883498903,
+      "learning_rate": 6.727548437794963e-06,
+      "loss": 2.4039,
+      "step": 1169
+    },
+    {
+      "epoch": 1.5836154366960056,
+      "grad_norm": 0.9713832736002586,
+      "learning_rate": 6.722485417209509e-06,
+      "loss": 2.129,
+      "step": 1170
+    },
+    {
+      "epoch": 1.5849695328368314,
+      "grad_norm": 1.536541973608301,
+      "learning_rate": 6.717420391817306e-06,
+      "loss": 2.0486,
+      "step": 1171
+    },
+    {
+      "epoch": 1.5863236289776574,
+      "grad_norm": 1.1215408259768265,
+      "learning_rate": 6.712353367513555e-06,
+      "loss": 1.8101,
+      "step": 1172
+    },
+    {
+      "epoch": 1.5876777251184833,
+      "grad_norm": 0.9400933392391618,
+      "learning_rate": 6.707284350195779e-06,
+      "loss": 2.129,
+      "step": 1173
+    },
+    {
+      "epoch": 1.5890318212593093,
+      "grad_norm": 1.0955863112973159,
+      "learning_rate": 6.70221334576383e-06,
+      "loss": 1.8608,
+      "step": 1174
+    },
+    {
+      "epoch": 1.5903859174001354,
+      "grad_norm": 1.290340133465167,
+      "learning_rate": 6.697140360119867e-06,
+      "loss": 2.6089,
+      "step": 1175
+    },
+    {
+      "epoch": 1.5917400135409614,
+      "grad_norm": 1.0935001366917838,
+      "learning_rate": 6.692065399168352e-06,
+      "loss": 2.3835,
+      "step": 1176
+    },
+    {
+      "epoch": 1.5930941096817874,
+      "grad_norm": 1.1310710971297633,
+      "learning_rate": 6.686988468816055e-06,
+      "loss": 2.3186,
+      "step": 1177
+    },
+    {
+      "epoch": 1.5944482058226135,
+      "grad_norm": 1.1106261511672113,
+      "learning_rate": 6.681909574972028e-06,
+      "loss": 2.278,
+      "step": 1178
+    },
+    {
+      "epoch": 1.5958023019634395,
+      "grad_norm": 1.1099983154470332,
+      "learning_rate": 6.676828723547614e-06,
+      "loss": 1.8412,
+      "step": 1179
+    },
+    {
+      "epoch": 1.5971563981042654,
+      "grad_norm": 1.115352426872196,
+      "learning_rate": 6.671745920456434e-06,
+      "loss": 2.6848,
+      "step": 1180
+    },
+    {
+      "epoch": 1.5985104942450914,
+      "grad_norm": 1.9617683740846232,
+      "learning_rate": 6.666661171614382e-06,
+      "loss": 2.4141,
+      "step": 1181
+    },
+    {
+      "epoch": 1.5998645903859174,
+      "grad_norm": 1.0340361767560753,
+      "learning_rate": 6.661574482939615e-06,
+      "loss": 2.029,
+      "step": 1182
+    },
+    {
+      "epoch": 1.6012186865267433,
+      "grad_norm": 1.1104672220605092,
+      "learning_rate": 6.656485860352544e-06,
+      "loss": 2.3655,
+      "step": 1183
+    },
+    {
+      "epoch": 1.6025727826675693,
+      "grad_norm": 1.2030765915104757,
+      "learning_rate": 6.651395309775837e-06,
+      "loss": 2.1954,
+      "step": 1184
+    },
+    {
+      "epoch": 1.6039268788083954,
+      "grad_norm": 1.2573381048562495,
+      "learning_rate": 6.6463028371344015e-06,
+      "loss": 2.5574,
+      "step": 1185
+    },
+    {
+      "epoch": 1.6052809749492214,
+      "grad_norm": 1.1722502953655352,
+      "learning_rate": 6.641208448355388e-06,
+      "loss": 1.9912,
+      "step": 1186
+    },
+    {
+      "epoch": 1.6066350710900474,
+      "grad_norm": 0.9237524987059714,
+      "learning_rate": 6.63611214936817e-06,
+      "loss": 1.7042,
+      "step": 1187
+    },
+    {
+      "epoch": 1.6079891672308735,
+      "grad_norm": 1.0114862987523483,
+      "learning_rate": 6.631013946104348e-06,
+      "loss": 2.1457,
+      "step": 1188
+    },
+    {
+      "epoch": 1.6093432633716995,
+      "grad_norm": 1.2504620286352166,
+      "learning_rate": 6.625913844497739e-06,
+      "loss": 2.9242,
+      "step": 1189
+    },
+    {
+      "epoch": 1.6106973595125254,
+      "grad_norm": 1.0716280700456458,
+      "learning_rate": 6.620811850484368e-06,
+      "loss": 2.1747,
+      "step": 1190
+    },
+    {
+      "epoch": 1.6120514556533514,
+      "grad_norm": 1.0218159941143157,
+      "learning_rate": 6.6157079700024665e-06,
+      "loss": 2.1052,
+      "step": 1191
+    },
+    {
+      "epoch": 1.6134055517941774,
+      "grad_norm": 0.9964735852419978,
+      "learning_rate": 6.6106022089924535e-06,
+      "loss": 2.1085,
+      "step": 1192
+    },
+    {
+      "epoch": 1.6147596479350033,
+      "grad_norm": 1.1164408166829398,
+      "learning_rate": 6.605494573396947e-06,
+      "loss": 2.146,
+      "step": 1193
+    },
+    {
+      "epoch": 1.6161137440758293,
+      "grad_norm": 1.1424753368875373,
+      "learning_rate": 6.600385069160739e-06,
+      "loss": 2.8387,
+      "step": 1194
+    },
+    {
+      "epoch": 1.6174678402166554,
+      "grad_norm": 1.3178474061043657,
+      "learning_rate": 6.5952737022308e-06,
+      "loss": 1.902,
+      "step": 1195
+    },
+    {
+      "epoch": 1.6188219363574814,
+      "grad_norm": 1.10966169321405,
+      "learning_rate": 6.590160478556269e-06,
+      "loss": 2.4412,
+      "step": 1196
+    },
+    {
+      "epoch": 1.6201760324983074,
+      "grad_norm": 1.1587333892126894,
+      "learning_rate": 6.585045404088442e-06,
+      "loss": 2.2305,
+      "step": 1197
+    },
+    {
+      "epoch": 1.6215301286391335,
+      "grad_norm": 1.0092191905987942,
+      "learning_rate": 6.579928484780773e-06,
+      "loss": 2.1454,
+      "step": 1198
+    },
+    {
+      "epoch": 1.6228842247799595,
+      "grad_norm": 1.0732773418214976,
+      "learning_rate": 6.5748097265888624e-06,
+      "loss": 2.3336,
+      "step": 1199
+    },
+    {
+      "epoch": 1.6242383209207854,
+      "grad_norm": 1.174781988020797,
+      "learning_rate": 6.569689135470451e-06,
+      "loss": 2.6344,
+      "step": 1200
+    },
+    {
+      "epoch": 1.6255924170616114,
+      "grad_norm": 1.1510963746609166,
+      "learning_rate": 6.564566717385412e-06,
+      "loss": 2.2199,
+      "step": 1201
+    },
+    {
+      "epoch": 1.6269465132024372,
+      "grad_norm": 0.8890344864420157,
+      "learning_rate": 6.559442478295745e-06,
+      "loss": 1.9674,
+      "step": 1202
+    },
+    {
+      "epoch": 1.6283006093432633,
+      "grad_norm": 1.0562702319805766,
+      "learning_rate": 6.5543164241655724e-06,
+      "loss": 2.0231,
+      "step": 1203
+    },
+    {
+      "epoch": 1.6296547054840893,
+      "grad_norm": 1.1686330096691793,
+      "learning_rate": 6.549188560961124e-06,
+      "loss": 2.401,
+      "step": 1204
+    },
+    {
+      "epoch": 1.6310088016249153,
+      "grad_norm": 0.9835702255574785,
+      "learning_rate": 6.544058894650736e-06,
+      "loss": 1.9566,
+      "step": 1205
+    },
+    {
+      "epoch": 1.6323628977657414,
+      "grad_norm": 1.1799470524015878,
+      "learning_rate": 6.5389274312048476e-06,
+      "loss": 2.3821,
+      "step": 1206
+    },
+    {
+      "epoch": 1.6337169939065674,
+      "grad_norm": 1.0305646424878727,
+      "learning_rate": 6.533794176595987e-06,
+      "loss": 1.7532,
+      "step": 1207
+    },
+    {
+      "epoch": 1.6350710900473935,
+      "grad_norm": 0.9411916034993967,
+      "learning_rate": 6.5286591367987655e-06,
+      "loss": 2.1668,
+      "step": 1208
+    },
+    {
+      "epoch": 1.6364251861882195,
+      "grad_norm": 1.1619947231820393,
+      "learning_rate": 6.523522317789874e-06,
+      "loss": 2.0694,
+      "step": 1209
+    },
+    {
+      "epoch": 1.6377792823290453,
+      "grad_norm": 1.1054945884631686,
+      "learning_rate": 6.518383725548074e-06,
+      "loss": 2.1855,
+      "step": 1210
+    },
+    {
+      "epoch": 1.6391333784698714,
+      "grad_norm": 1.0751426497234493,
+      "learning_rate": 6.513243366054191e-06,
+      "loss": 2.2038,
+      "step": 1211
+    },
+    {
+      "epoch": 1.6404874746106972,
+      "grad_norm": 1.0718361928319553,
+      "learning_rate": 6.508101245291109e-06,
+      "loss": 2.2788,
+      "step": 1212
+    },
+    {
+      "epoch": 1.6418415707515233,
+      "grad_norm": 1.1446897139146628,
+      "learning_rate": 6.502957369243757e-06,
+      "loss": 1.8886,
+      "step": 1213
+    },
+    {
+      "epoch": 1.6431956668923493,
+      "grad_norm": 1.0780227955508757,
+      "learning_rate": 6.497811743899112e-06,
+      "loss": 2.9019,
+      "step": 1214
+    },
+    {
+      "epoch": 1.6445497630331753,
+      "grad_norm": 1.1093588985005516,
+      "learning_rate": 6.492664375246185e-06,
+      "loss": 2.3378,
+      "step": 1215
+    },
+    {
+      "epoch": 1.6459038591740014,
+      "grad_norm": 1.2680934914767075,
+      "learning_rate": 6.487515269276015e-06,
+      "loss": 2.2777,
+      "step": 1216
+    },
+    {
+      "epoch": 1.6472579553148274,
+      "grad_norm": 1.0496344613229611,
+      "learning_rate": 6.482364431981667e-06,
+      "loss": 2.1846,
+      "step": 1217
+    },
+    {
+      "epoch": 1.6486120514556535,
+      "grad_norm": 0.8830408833414931,
+      "learning_rate": 6.4772118693582155e-06,
+      "loss": 2.4639,
+      "step": 1218
+    },
+    {
+      "epoch": 1.6499661475964793,
+      "grad_norm": 1.3617840820836493,
+      "learning_rate": 6.472057587402748e-06,
+      "loss": 2.1004,
+      "step": 1219
+    },
+    {
+      "epoch": 1.6513202437373053,
+      "grad_norm": 1.0634770023039701,
+      "learning_rate": 6.46690159211435e-06,
+      "loss": 2.3582,
+      "step": 1220
+    },
+    {
+      "epoch": 1.6526743398781314,
+      "grad_norm": 1.0462993743834461,
+      "learning_rate": 6.461743889494103e-06,
+      "loss": 2.1793,
+      "step": 1221
+    },
+    {
+      "epoch": 1.6540284360189572,
+      "grad_norm": 1.12270603128975,
+      "learning_rate": 6.456584485545075e-06,
+      "loss": 2.0312,
+      "step": 1222
+    },
+    {
+      "epoch": 1.6553825321597833,
+      "grad_norm": 1.0474924934857333,
+      "learning_rate": 6.451423386272312e-06,
+      "loss": 2.1907,
+      "step": 1223
+    },
+    {
+      "epoch": 1.6567366283006093,
+      "grad_norm": 1.2139372873126406,
+      "learning_rate": 6.4462605976828395e-06,
+      "loss": 2.1359,
+      "step": 1224
+    },
+    {
+      "epoch": 1.6580907244414353,
+      "grad_norm": 1.2375074680264349,
+      "learning_rate": 6.441096125785641e-06,
+      "loss": 2.7886,
+      "step": 1225
+    },
+    {
+      "epoch": 1.6594448205822614,
+      "grad_norm": 1.2726063282105258,
+      "learning_rate": 6.435929976591665e-06,
+      "loss": 1.8611,
+      "step": 1226
+    },
+    {
+      "epoch": 1.6607989167230874,
+      "grad_norm": 1.0325742896045942,
+      "learning_rate": 6.4307621561138114e-06,
+      "loss": 2.1808,
+      "step": 1227
+    },
+    {
+      "epoch": 1.6621530128639135,
+      "grad_norm": 1.1451127258616212,
+      "learning_rate": 6.425592670366923e-06,
+      "loss": 2.3795,
+      "step": 1228
+    },
+    {
+      "epoch": 1.6635071090047393,
+      "grad_norm": 1.0588080148815049,
+      "learning_rate": 6.420421525367783e-06,
+      "loss": 2.3691,
+      "step": 1229
+    },
+    {
+      "epoch": 1.6648612051455653,
+      "grad_norm": 1.1248420963082115,
+      "learning_rate": 6.415248727135103e-06,
+      "loss": 2.2255,
+      "step": 1230
+    },
+    {
+      "epoch": 1.6662153012863912,
+      "grad_norm": 1.12193324651476,
+      "learning_rate": 6.410074281689522e-06,
+      "loss": 2.2692,
+      "step": 1231
+    },
+    {
+      "epoch": 1.6675693974272172,
+      "grad_norm": 1.1007368907096031,
+      "learning_rate": 6.4048981950535975e-06,
+      "loss": 2.0518,
+      "step": 1232
+    },
+    {
+      "epoch": 1.6689234935680433,
+      "grad_norm": 1.06269170578331,
+      "learning_rate": 6.3997204732517915e-06,
+      "loss": 2.3531,
+      "step": 1233
+    },
+    {
+      "epoch": 1.6702775897088693,
+      "grad_norm": 1.1163412698869442,
+      "learning_rate": 6.394541122310473e-06,
+      "loss": 2.2524,
+      "step": 1234
+    },
+    {
+      "epoch": 1.6716316858496953,
+      "grad_norm": 1.0554457718352965,
+      "learning_rate": 6.389360148257908e-06,
+      "loss": 2.4751,
+      "step": 1235
+    },
+    {
+      "epoch": 1.6729857819905214,
+      "grad_norm": 1.1997008894725945,
+      "learning_rate": 6.384177557124247e-06,
+      "loss": 1.9252,
+      "step": 1236
+    },
+    {
+      "epoch": 1.6743398781313474,
+      "grad_norm": 1.079055031664645,
+      "learning_rate": 6.378993354941529e-06,
+      "loss": 2.3641,
+      "step": 1237
+    },
+    {
+      "epoch": 1.6756939742721735,
+      "grad_norm": 1.1896072037787495,
+      "learning_rate": 6.3738075477436655e-06,
+      "loss": 1.9261,
+      "step": 1238
+    },
+    {
+      "epoch": 1.6770480704129993,
+      "grad_norm": 0.9256492202916144,
+      "learning_rate": 6.368620141566432e-06,
+      "loss": 2.3076,
+      "step": 1239
+    },
+    {
+      "epoch": 1.6784021665538253,
+      "grad_norm": 1.1122465104394996,
+      "learning_rate": 6.363431142447469e-06,
+      "loss": 2.0237,
+      "step": 1240
+    },
+    {
+      "epoch": 1.6797562626946512,
+      "grad_norm": 0.9678398341378385,
+      "learning_rate": 6.358240556426274e-06,
+      "loss": 2.4974,
+      "step": 1241
+    },
+    {
+      "epoch": 1.6811103588354772,
+      "grad_norm": 0.9925297077996342,
+      "learning_rate": 6.353048389544185e-06,
+      "loss": 2.244,
+      "step": 1242
+    },
+    {
+      "epoch": 1.6824644549763033,
+      "grad_norm": 1.0616822439065718,
+      "learning_rate": 6.3478546478443844e-06,
+      "loss": 2.3992,
+      "step": 1243
+    },
+    {
+      "epoch": 1.6838185511171293,
+      "grad_norm": 1.1733624673460672,
+      "learning_rate": 6.342659337371884e-06,
+      "loss": 1.8391,
+      "step": 1244
+    },
+    {
+      "epoch": 1.6851726472579553,
+      "grad_norm": 1.2040137843680947,
+      "learning_rate": 6.3374624641735275e-06,
+      "loss": 1.8877,
+      "step": 1245
+    },
+    {
+      "epoch": 1.6865267433987814,
+      "grad_norm": 1.0453714976813977,
+      "learning_rate": 6.33226403429797e-06,
+      "loss": 2.0552,
+      "step": 1246
+    },
+    {
+      "epoch": 1.6878808395396074,
+      "grad_norm": 1.1426720858647532,
+      "learning_rate": 6.327064053795684e-06,
+      "loss": 2.2107,
+      "step": 1247
+    },
+    {
+      "epoch": 1.6892349356804333,
+      "grad_norm": 1.199062064451871,
+      "learning_rate": 6.321862528718945e-06,
+      "loss": 2.0693,
+      "step": 1248
+    },
+    {
+      "epoch": 1.6905890318212593,
+      "grad_norm": 1.091725236727943,
+      "learning_rate": 6.3166594651218235e-06,
+      "loss": 2.1291,
+      "step": 1249
+    },
+    {
+      "epoch": 1.6919431279620853,
+      "grad_norm": 1.1275607912830679,
+      "learning_rate": 6.311454869060188e-06,
+      "loss": 2.2288,
+      "step": 1250
+    },
+    {
+      "epoch": 1.6932972241029112,
+      "grad_norm": 1.236386694185598,
+      "learning_rate": 6.3062487465916825e-06,
+      "loss": 2.4416,
+      "step": 1251
+    },
+    {
+      "epoch": 1.6946513202437372,
+      "grad_norm": 0.9827244912709826,
+      "learning_rate": 6.301041103775731e-06,
+      "loss": 2.3147,
+      "step": 1252
+    },
+    {
+      "epoch": 1.6960054163845633,
+      "grad_norm": 1.2144425166077264,
+      "learning_rate": 6.295831946673527e-06,
+      "loss": 2.4805,
+      "step": 1253
+    },
+    {
+      "epoch": 1.6973595125253893,
+      "grad_norm": 0.9983619215236832,
+      "learning_rate": 6.290621281348028e-06,
+      "loss": 2.3773,
+      "step": 1254
+    },
+    {
+      "epoch": 1.6987136086662153,
+      "grad_norm": 1.465782148598409,
+      "learning_rate": 6.285409113863944e-06,
+      "loss": 2.6457,
+      "step": 1255
+    },
+    {
+      "epoch": 1.7000677048070414,
+      "grad_norm": 0.9944582045815571,
+      "learning_rate": 6.280195450287736e-06,
+      "loss": 2.3837,
+      "step": 1256
+    },
+    {
+      "epoch": 1.7014218009478674,
+      "grad_norm": 1.1585939861872792,
+      "learning_rate": 6.274980296687605e-06,
+      "loss": 2.2119,
+      "step": 1257
+    },
+    {
+      "epoch": 1.7027758970886933,
+      "grad_norm": 1.1437189700305217,
+      "learning_rate": 6.269763659133486e-06,
+      "loss": 1.8079,
+      "step": 1258
+    },
+    {
+      "epoch": 1.7041299932295193,
+      "grad_norm": 1.2582747800473197,
+      "learning_rate": 6.264545543697039e-06,
+      "loss": 2.5012,
+      "step": 1259
+    },
+    {
+      "epoch": 1.7054840893703453,
+      "grad_norm": 1.2978378068690752,
+      "learning_rate": 6.259325956451651e-06,
+      "loss": 2.9485,
+      "step": 1260
+    },
+    {
+      "epoch": 1.7068381855111712,
+      "grad_norm": 1.0424842976922624,
+      "learning_rate": 6.254104903472417e-06,
+      "loss": 2.2624,
+      "step": 1261
+    },
+    {
+      "epoch": 1.7081922816519972,
+      "grad_norm": 1.0868588376583306,
+      "learning_rate": 6.248882390836135e-06,
+      "loss": 2.3471,
+      "step": 1262
+    },
+    {
+      "epoch": 1.7095463777928233,
+      "grad_norm": 1.0151211458884728,
+      "learning_rate": 6.243658424621308e-06,
+      "loss": 1.9188,
+      "step": 1263
+    },
+    {
+      "epoch": 1.7109004739336493,
+      "grad_norm": 1.1267719574629707,
+      "learning_rate": 6.238433010908131e-06,
+      "loss": 2.0653,
+      "step": 1264
+    },
+    {
+      "epoch": 1.7122545700744753,
+      "grad_norm": 1.1710384477346152,
+      "learning_rate": 6.233206155778476e-06,
+      "loss": 2.8079,
+      "step": 1265
+    },
+    {
+      "epoch": 1.7136086662153014,
+      "grad_norm": 1.2399841383381511,
+      "learning_rate": 6.227977865315902e-06,
+      "loss": 2.3833,
+      "step": 1266
+    },
+    {
+      "epoch": 1.7149627623561274,
+      "grad_norm": 1.0116588623292786,
+      "learning_rate": 6.222748145605631e-06,
+      "loss": 2.4164,
+      "step": 1267
+    },
+    {
+      "epoch": 1.7163168584969533,
+      "grad_norm": 0.8527451840083117,
+      "learning_rate": 6.217517002734554e-06,
+      "loss": 2.1877,
+      "step": 1268
+    },
+    {
+      "epoch": 1.7176709546377793,
+      "grad_norm": 1.079554004708601,
+      "learning_rate": 6.212284442791217e-06,
+      "loss": 1.7948,
+      "step": 1269
+    },
+    {
+      "epoch": 1.7190250507786051,
+      "grad_norm": 0.9137877079458239,
+      "learning_rate": 6.207050471865814e-06,
+      "loss": 1.841,
+      "step": 1270
+    },
+    {
+      "epoch": 1.7203791469194312,
+      "grad_norm": 1.1473597440593684,
+      "learning_rate": 6.201815096050181e-06,
+      "loss": 1.9544,
+      "step": 1271
+    },
+    {
+      "epoch": 1.7217332430602572,
+      "grad_norm": 1.189027252854009,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 2.2543,
+      "step": 1272
+    },
+    {
+      "epoch": 1.7230873392010833,
+      "grad_norm": 0.8466476995754849,
+      "learning_rate": 6.1913401541237414e-06,
+      "loss": 1.8128,
+      "step": 1273
+    },
+    {
+      "epoch": 1.7244414353419093,
+      "grad_norm": 1.320497220055551,
+      "learning_rate": 6.186100600204757e-06,
+      "loss": 1.9606,
+      "step": 1274
+    },
+    {
+      "epoch": 1.7257955314827353,
+      "grad_norm": 1.1503966715251264,
+      "learning_rate": 6.180859665779173e-06,
+      "loss": 2.0067,
+      "step": 1275
+    },
+    {
+      "epoch": 1.7271496276235614,
+      "grad_norm": 1.197795743633595,
+      "learning_rate": 6.175617356946928e-06,
+      "loss": 2.4943,
+      "step": 1276
+    },
+    {
+      "epoch": 1.7285037237643874,
+      "grad_norm": 1.153983741344922,
+      "learning_rate": 6.170373679809565e-06,
+      "loss": 2.6468,
+      "step": 1277
+    },
+    {
+      "epoch": 1.7298578199052133,
+      "grad_norm": 1.1521006556182283,
+      "learning_rate": 6.165128640470217e-06,
+      "loss": 2.3347,
+      "step": 1278
+    },
+    {
+      "epoch": 1.7312119160460393,
+      "grad_norm": 0.9816413381160185,
+      "learning_rate": 6.159882245033606e-06,
+      "loss": 2.4498,
+      "step": 1279
+    },
+    {
+      "epoch": 1.7325660121868651,
+      "grad_norm": 1.07736534022899,
+      "learning_rate": 6.1546344996060294e-06,
+      "loss": 2.1578,
+      "step": 1280
+    },
+    {
+      "epoch": 1.7339201083276912,
+      "grad_norm": 1.1324233134707422,
+      "learning_rate": 6.149385410295353e-06,
+      "loss": 2.1432,
+      "step": 1281
+    },
+    {
+      "epoch": 1.7352742044685172,
+      "grad_norm": 0.9905290656285861,
+      "learning_rate": 6.144134983211012e-06,
+      "loss": 1.872,
+      "step": 1282
+    },
+    {
+      "epoch": 1.7366283006093433,
+      "grad_norm": 1.121201508813207,
+      "learning_rate": 6.138883224463999e-06,
+      "loss": 1.9648,
+      "step": 1283
+    },
+    {
+      "epoch": 1.7379823967501693,
+      "grad_norm": 1.1862628290117698,
+      "learning_rate": 6.13363014016685e-06,
+      "loss": 2.1744,
+      "step": 1284
+    },
+    {
+      "epoch": 1.7393364928909953,
+      "grad_norm": 0.9830015960619232,
+      "learning_rate": 6.128375736433652e-06,
+      "loss": 2.5532,
+      "step": 1285
+    },
+    {
+      "epoch": 1.7406905890318214,
+      "grad_norm": 1.3321125394117272,
+      "learning_rate": 6.123120019380021e-06,
+      "loss": 1.8057,
+      "step": 1286
+    },
+    {
+      "epoch": 1.7420446851726472,
+      "grad_norm": 0.9980976929176649,
+      "learning_rate": 6.117862995123105e-06,
+      "loss": 2.337,
+      "step": 1287
+    },
+    {
+      "epoch": 1.7433987813134733,
+      "grad_norm": 1.3173652394689053,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 2.0408,
+      "step": 1288
+    },
+    {
+      "epoch": 1.7447528774542993,
+      "grad_norm": 1.126084316928292,
+      "learning_rate": 6.107345049475609e-06,
+      "loss": 2.079,
+      "step": 1289
+    },
+    {
+      "epoch": 1.7461069735951251,
+      "grad_norm": 1.0413819596877105,
+      "learning_rate": 6.102084140326901e-06,
+      "loss": 2.0356,
+      "step": 1290
+    },
+    {
+      "epoch": 1.7474610697359512,
+      "grad_norm": 1.189218565313232,
+      "learning_rate": 6.0968219484586415e-06,
+      "loss": 2.1497,
+      "step": 1291
+    },
+    {
+      "epoch": 1.7488151658767772,
+      "grad_norm": 1.2345681545643294,
+      "learning_rate": 6.091558479995512e-06,
+      "loss": 2.0314,
+      "step": 1292
+    },
+    {
+      "epoch": 1.7501692620176033,
+      "grad_norm": 1.2541579573695316,
+      "learning_rate": 6.086293741063685e-06,
+      "loss": 2.2602,
+      "step": 1293
+    },
+    {
+      "epoch": 1.7515233581584293,
+      "grad_norm": 1.4163932516550155,
+      "learning_rate": 6.081027737790804e-06,
+      "loss": 2.649,
+      "step": 1294
+    },
+    {
+      "epoch": 1.7528774542992553,
+      "grad_norm": 1.0677445213609345,
+      "learning_rate": 6.075760476305992e-06,
+      "loss": 1.6769,
+      "step": 1295
+    },
+    {
+      "epoch": 1.7542315504400814,
+      "grad_norm": 1.1305409453258632,
+      "learning_rate": 6.070491962739831e-06,
+      "loss": 2.1413,
+      "step": 1296
+    },
+    {
+      "epoch": 1.7555856465809072,
+      "grad_norm": 0.8589085134430466,
+      "learning_rate": 6.065222203224363e-06,
+      "loss": 2.2007,
+      "step": 1297
+    },
+    {
+      "epoch": 1.7569397427217333,
+      "grad_norm": 1.198177616657121,
+      "learning_rate": 6.05995120389308e-06,
+      "loss": 2.463,
+      "step": 1298
+    },
+    {
+      "epoch": 1.758293838862559,
+      "grad_norm": 1.0517431847975949,
+      "learning_rate": 6.054678970880915e-06,
+      "loss": 1.8492,
+      "step": 1299
+    },
+    {
+      "epoch": 1.7596479350033851,
+      "grad_norm": 1.0877464246501793,
+      "learning_rate": 6.049405510324237e-06,
+      "loss": 2.2445,
+      "step": 1300
+    },
+    {
+      "epoch": 1.7610020311442112,
+      "grad_norm": 0.9182819543184888,
+      "learning_rate": 6.04413082836085e-06,
+      "loss": 1.923,
+      "step": 1301
+    },
+    {
+      "epoch": 1.7623561272850372,
+      "grad_norm": 0.9479159720049781,
+      "learning_rate": 6.0388549311299696e-06,
+      "loss": 1.8528,
+      "step": 1302
+    },
+    {
+      "epoch": 1.7637102234258633,
+      "grad_norm": 1.1330012872888824,
+      "learning_rate": 6.033577824772234e-06,
+      "loss": 2.3405,
+      "step": 1303
+    },
+    {
+      "epoch": 1.7650643195666893,
+      "grad_norm": 1.156479517032631,
+      "learning_rate": 6.028299515429683e-06,
+      "loss": 2.4133,
+      "step": 1304
+    },
+    {
+      "epoch": 1.7664184157075153,
+      "grad_norm": 1.2849276017784446,
+      "learning_rate": 6.023020009245761e-06,
+      "loss": 2.0552,
+      "step": 1305
+    },
+    {
+      "epoch": 1.7677725118483414,
+      "grad_norm": 1.1081263537914658,
+      "learning_rate": 6.017739312365304e-06,
+      "loss": 1.9643,
+      "step": 1306
+    },
+    {
+      "epoch": 1.7691266079891672,
+      "grad_norm": 1.192468338343465,
+      "learning_rate": 6.012457430934532e-06,
+      "loss": 2.2097,
+      "step": 1307
+    },
+    {
+      "epoch": 1.7704807041299933,
+      "grad_norm": 1.3083267149971682,
+      "learning_rate": 6.007174371101045e-06,
+      "loss": 2.0671,
+      "step": 1308
+    },
+    {
+      "epoch": 1.771834800270819,
+      "grad_norm": 1.0164272321056016,
+      "learning_rate": 6.001890139013816e-06,
+      "loss": 2.1438,
+      "step": 1309
+    },
+    {
+      "epoch": 1.7731888964116451,
+      "grad_norm": 1.0169184208499216,
+      "learning_rate": 5.99660474082318e-06,
+      "loss": 1.6461,
+      "step": 1310
+    },
+    {
+      "epoch": 1.7745429925524712,
+      "grad_norm": 1.1208962249464,
+      "learning_rate": 5.99131818268083e-06,
+      "loss": 2.0891,
+      "step": 1311
+    },
+    {
+      "epoch": 1.7758970886932972,
+      "grad_norm": 0.9880243308990688,
+      "learning_rate": 5.986030470739811e-06,
+      "loss": 2.2771,
+      "step": 1312
+    },
+    {
+      "epoch": 1.7772511848341233,
+      "grad_norm": 1.0164372276741276,
+      "learning_rate": 5.980741611154508e-06,
+      "loss": 1.9089,
+      "step": 1313
+    },
+    {
+      "epoch": 1.7786052809749493,
+      "grad_norm": 1.1780962664030357,
+      "learning_rate": 5.975451610080643e-06,
+      "loss": 2.2283,
+      "step": 1314
+    },
+    {
+      "epoch": 1.7799593771157753,
+      "grad_norm": 1.2337816535300354,
+      "learning_rate": 5.970160473675266e-06,
+      "loss": 2.0219,
+      "step": 1315
+    },
+    {
+      "epoch": 1.7813134732566012,
+      "grad_norm": 1.231311259477204,
+      "learning_rate": 5.96486820809675e-06,
+      "loss": 2.263,
+      "step": 1316
+    },
+    {
+      "epoch": 1.7826675693974272,
+      "grad_norm": 1.1820145455048345,
+      "learning_rate": 5.959574819504782e-06,
+      "loss": 2.1722,
+      "step": 1317
+    },
+    {
+      "epoch": 1.7840216655382533,
+      "grad_norm": 1.0602621258357094,
+      "learning_rate": 5.954280314060353e-06,
+      "loss": 2.0836,
+      "step": 1318
+    },
+    {
+      "epoch": 1.785375761679079,
+      "grad_norm": 1.0769310348498187,
+      "learning_rate": 5.9489846979257596e-06,
+      "loss": 2.1107,
+      "step": 1319
+    },
+    {
+      "epoch": 1.7867298578199051,
+      "grad_norm": 1.1685570897175144,
+      "learning_rate": 5.943687977264584e-06,
+      "loss": 2.3701,
+      "step": 1320
+    },
+    {
+      "epoch": 1.7880839539607312,
+      "grad_norm": 1.0639504620411393,
+      "learning_rate": 5.938390158241701e-06,
+      "loss": 1.8519,
+      "step": 1321
+    },
+    {
+      "epoch": 1.7894380501015572,
+      "grad_norm": 1.0471961016733244,
+      "learning_rate": 5.933091247023258e-06,
+      "loss": 2.3008,
+      "step": 1322
+    },
+    {
+      "epoch": 1.7907921462423833,
+      "grad_norm": 1.079334192472969,
+      "learning_rate": 5.927791249776678e-06,
+      "loss": 1.9762,
+      "step": 1323
+    },
+    {
+      "epoch": 1.7921462423832093,
+      "grad_norm": 1.0698384472116922,
+      "learning_rate": 5.922490172670645e-06,
+      "loss": 2.3354,
+      "step": 1324
+    },
+    {
+      "epoch": 1.7935003385240353,
+      "grad_norm": 0.9830724595088538,
+      "learning_rate": 5.917188021875101e-06,
+      "loss": 2.0081,
+      "step": 1325
+    },
+    {
+      "epoch": 1.7948544346648612,
+      "grad_norm": 1.340171291705083,
+      "learning_rate": 5.91188480356124e-06,
+      "loss": 2.1949,
+      "step": 1326
+    },
+    {
+      "epoch": 1.7962085308056872,
+      "grad_norm": 1.0913921699387112,
+      "learning_rate": 5.906580523901493e-06,
+      "loss": 2.0597,
+      "step": 1327
+    },
+    {
+      "epoch": 1.7975626269465133,
+      "grad_norm": 1.0873779097080578,
+      "learning_rate": 5.90127518906953e-06,
+      "loss": 2.8079,
+      "step": 1328
+    },
+    {
+      "epoch": 1.798916723087339,
+      "grad_norm": 1.143892818041929,
+      "learning_rate": 5.89596880524025e-06,
+      "loss": 2.1394,
+      "step": 1329
+    },
+    {
+      "epoch": 1.8002708192281651,
+      "grad_norm": 0.9347818074040029,
+      "learning_rate": 5.89066137858977e-06,
+      "loss": 2.0658,
+      "step": 1330
+    },
+    {
+      "epoch": 1.8016249153689912,
+      "grad_norm": 1.0104844185810784,
+      "learning_rate": 5.885352915295426e-06,
+      "loss": 2.4573,
+      "step": 1331
+    },
+    {
+      "epoch": 1.8029790115098172,
+      "grad_norm": 1.2621189889416795,
+      "learning_rate": 5.880043421535751e-06,
+      "loss": 2.2771,
+      "step": 1332
+    },
+    {
+      "epoch": 1.8043331076506433,
+      "grad_norm": 0.9776230416310355,
+      "learning_rate": 5.874732903490489e-06,
+      "loss": 2.2122,
+      "step": 1333
+    },
+    {
+      "epoch": 1.8056872037914693,
+      "grad_norm": 1.1935436794161547,
+      "learning_rate": 5.869421367340565e-06,
+      "loss": 2.6218,
+      "step": 1334
+    },
+    {
+      "epoch": 1.8070412999322953,
+      "grad_norm": 1.166409774943222,
+      "learning_rate": 5.864108819268098e-06,
+      "loss": 2.1516,
+      "step": 1335
+    },
+    {
+      "epoch": 1.8083953960731212,
+      "grad_norm": 0.9563193799285069,
+      "learning_rate": 5.858795265456382e-06,
+      "loss": 2.1492,
+      "step": 1336
+    },
+    {
+      "epoch": 1.8097494922139472,
+      "grad_norm": 1.0173612877579,
+      "learning_rate": 5.853480712089875e-06,
+      "loss": 2.5045,
+      "step": 1337
+    },
+    {
+      "epoch": 1.811103588354773,
+      "grad_norm": 1.075936558534605,
+      "learning_rate": 5.8481651653542105e-06,
+      "loss": 1.8214,
+      "step": 1338
+    },
+    {
+      "epoch": 1.812457684495599,
+      "grad_norm": 1.2662166496893965,
+      "learning_rate": 5.842848631436169e-06,
+      "loss": 2.3341,
+      "step": 1339
+    },
+    {
+      "epoch": 1.8138117806364251,
+      "grad_norm": 1.114110647827857,
+      "learning_rate": 5.837531116523683e-06,
+      "loss": 1.9848,
+      "step": 1340
+    },
+    {
+      "epoch": 1.8151658767772512,
+      "grad_norm": 1.133467071972051,
+      "learning_rate": 5.832212626805825e-06,
+      "loss": 1.7728,
+      "step": 1341
+    },
+    {
+      "epoch": 1.8165199729180772,
+      "grad_norm": 1.1904363405626495,
+      "learning_rate": 5.826893168472807e-06,
+      "loss": 2.2031,
+      "step": 1342
+    },
+    {
+      "epoch": 1.8178740690589033,
+      "grad_norm": 1.16982781000476,
+      "learning_rate": 5.821572747715961e-06,
+      "loss": 2.2635,
+      "step": 1343
+    },
+    {
+      "epoch": 1.8192281651997293,
+      "grad_norm": 2.3913374538342995,
+      "learning_rate": 5.816251370727748e-06,
+      "loss": 2.7674,
+      "step": 1344
+    },
+    {
+      "epoch": 1.8205822613405553,
+      "grad_norm": 1.155071954814068,
+      "learning_rate": 5.810929043701733e-06,
+      "loss": 1.6222,
+      "step": 1345
+    },
+    {
+      "epoch": 1.8219363574813812,
+      "grad_norm": 1.1195068129462349,
+      "learning_rate": 5.805605772832592e-06,
+      "loss": 2.2405,
+      "step": 1346
+    },
+    {
+      "epoch": 1.8232904536222072,
+      "grad_norm": 1.0607883154488784,
+      "learning_rate": 5.800281564316098e-06,
+      "loss": 2.3167,
+      "step": 1347
+    },
+    {
+      "epoch": 1.824644549763033,
+      "grad_norm": 0.9198180067594192,
+      "learning_rate": 5.794956424349117e-06,
+      "loss": 2.2837,
+      "step": 1348
+    },
+    {
+      "epoch": 1.825998645903859,
+      "grad_norm": 1.1141503225834095,
+      "learning_rate": 5.789630359129599e-06,
+      "loss": 1.9332,
+      "step": 1349
+    },
+    {
+      "epoch": 1.8273527420446851,
+      "grad_norm": 1.1494560282733117,
+      "learning_rate": 5.784303374856566e-06,
+      "loss": 2.0177,
+      "step": 1350
+    },
+    {
+      "epoch": 1.8287068381855112,
+      "grad_norm": 0.9838628228119614,
+      "learning_rate": 5.778975477730117e-06,
+      "loss": 1.9256,
+      "step": 1351
+    },
+    {
+      "epoch": 1.8300609343263372,
+      "grad_norm": 1.0379092917122728,
+      "learning_rate": 5.773646673951406e-06,
+      "loss": 2.0192,
+      "step": 1352
+    },
+    {
+      "epoch": 1.8314150304671633,
+      "grad_norm": 1.0735535500692883,
+      "learning_rate": 5.768316969722651e-06,
+      "loss": 1.7708,
+      "step": 1353
+    },
+    {
+      "epoch": 1.8327691266079893,
+      "grad_norm": 1.1525689914627582,
+      "learning_rate": 5.762986371247111e-06,
+      "loss": 2.0803,
+      "step": 1354
+    },
+    {
+      "epoch": 1.8341232227488151,
+      "grad_norm": 1.2969233616374964,
+      "learning_rate": 5.757654884729087e-06,
+      "loss": 1.7907,
+      "step": 1355
+    },
+    {
+      "epoch": 1.8354773188896412,
+      "grad_norm": 1.2729837968386013,
+      "learning_rate": 5.752322516373916e-06,
+      "loss": 1.9638,
+      "step": 1356
+    },
+    {
+      "epoch": 1.8368314150304672,
+      "grad_norm": 1.123620270200479,
+      "learning_rate": 5.746989272387959e-06,
+      "loss": 2.4644,
+      "step": 1357
+    },
+    {
+      "epoch": 1.838185511171293,
+      "grad_norm": 1.1478076870568643,
+      "learning_rate": 5.741655158978598e-06,
+      "loss": 2.3496,
+      "step": 1358
+    },
+    {
+      "epoch": 1.839539607312119,
+      "grad_norm": 0.9606164841242715,
+      "learning_rate": 5.736320182354227e-06,
+      "loss": 2.3232,
+      "step": 1359
+    },
+    {
+      "epoch": 1.8408937034529451,
+      "grad_norm": 0.9434545645826506,
+      "learning_rate": 5.730984348724242e-06,
+      "loss": 2.1405,
+      "step": 1360
+    },
+    {
+      "epoch": 1.8422477995937712,
+      "grad_norm": 1.1057923006106751,
+      "learning_rate": 5.725647664299039e-06,
+      "loss": 2.9284,
+      "step": 1361
+    },
+    {
+      "epoch": 1.8436018957345972,
+      "grad_norm": 1.0188797974653283,
+      "learning_rate": 5.720310135290002e-06,
+      "loss": 2.1123,
+      "step": 1362
+    },
+    {
+      "epoch": 1.8449559918754233,
+      "grad_norm": 1.4800000718890725,
+      "learning_rate": 5.7149717679095026e-06,
+      "loss": 2.5262,
+      "step": 1363
+    },
+    {
+      "epoch": 1.8463100880162493,
+      "grad_norm": 1.4201072710479334,
+      "learning_rate": 5.709632568370884e-06,
+      "loss": 1.9675,
+      "step": 1364
+    },
+    {
+      "epoch": 1.8476641841570751,
+      "grad_norm": 1.0271104317135864,
+      "learning_rate": 5.704292542888458e-06,
+      "loss": 2.3992,
+      "step": 1365
+    },
+    {
+      "epoch": 1.8490182802979012,
+      "grad_norm": 1.1894861709189837,
+      "learning_rate": 5.698951697677498e-06,
+      "loss": 2.3255,
+      "step": 1366
+    },
+    {
+      "epoch": 1.850372376438727,
+      "grad_norm": 0.9818606413266471,
+      "learning_rate": 5.6936100389542345e-06,
+      "loss": 2.068,
+      "step": 1367
+    },
+    {
+      "epoch": 1.851726472579553,
+      "grad_norm": 1.1613550136148025,
+      "learning_rate": 5.688267572935843e-06,
+      "loss": 2.0782,
+      "step": 1368
+    },
+    {
+      "epoch": 1.853080568720379,
+      "grad_norm": 1.2604918956006983,
+      "learning_rate": 5.682924305840435e-06,
+      "loss": 2.3411,
+      "step": 1369
+    },
+    {
+      "epoch": 1.8544346648612051,
+      "grad_norm": 1.0429238519002082,
+      "learning_rate": 5.6775802438870596e-06,
+      "loss": 2.3507,
+      "step": 1370
+    },
+    {
+      "epoch": 1.8557887610020312,
+      "grad_norm": 1.1476866829926053,
+      "learning_rate": 5.6722353932956895e-06,
+      "loss": 2.1169,
+      "step": 1371
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 1.2071680463300853,
+      "learning_rate": 5.6668897602872145e-06,
+      "loss": 2.3788,
+      "step": 1372
+    },
+    {
+      "epoch": 1.8584969532836833,
+      "grad_norm": 1.330454708741957,
+      "learning_rate": 5.661543351083435e-06,
+      "loss": 2.3059,
+      "step": 1373
+    },
+    {
+      "epoch": 1.8598510494245093,
+      "grad_norm": 1.0666387576576266,
+      "learning_rate": 5.656196171907055e-06,
+      "loss": 1.988,
+      "step": 1374
+    },
+    {
+      "epoch": 1.8612051455653351,
+      "grad_norm": 1.1222282028435293,
+      "learning_rate": 5.650848228981676e-06,
+      "loss": 2.3535,
+      "step": 1375
+    },
+    {
+      "epoch": 1.8625592417061612,
+      "grad_norm": 1.0691636063361554,
+      "learning_rate": 5.645499528531785e-06,
+      "loss": 2.3795,
+      "step": 1376
+    },
+    {
+      "epoch": 1.863913337846987,
+      "grad_norm": 1.0416420607905428,
+      "learning_rate": 5.640150076782755e-06,
+      "loss": 2.2143,
+      "step": 1377
+    },
+    {
+      "epoch": 1.865267433987813,
+      "grad_norm": 1.1126322928122026,
+      "learning_rate": 5.634799879960833e-06,
+      "loss": 1.8687,
+      "step": 1378
+    },
+    {
+      "epoch": 1.866621530128639,
+      "grad_norm": 1.0283378507513847,
+      "learning_rate": 5.629448944293128e-06,
+      "loss": 2.3348,
+      "step": 1379
+    },
+    {
+      "epoch": 1.8679756262694651,
+      "grad_norm": 1.1086979474304333,
+      "learning_rate": 5.624097276007614e-06,
+      "loss": 2.4503,
+      "step": 1380
+    },
+    {
+      "epoch": 1.8693297224102912,
+      "grad_norm": 1.0114717774494804,
+      "learning_rate": 5.618744881333117e-06,
+      "loss": 2.2857,
+      "step": 1381
+    },
+    {
+      "epoch": 1.8706838185511172,
+      "grad_norm": 1.271521217949957,
+      "learning_rate": 5.613391766499308e-06,
+      "loss": 2.1678,
+      "step": 1382
+    },
+    {
+      "epoch": 1.8720379146919433,
+      "grad_norm": 1.0690493430782453,
+      "learning_rate": 5.6080379377366936e-06,
+      "loss": 2.0153,
+      "step": 1383
+    },
+    {
+      "epoch": 1.873392010832769,
+      "grad_norm": 1.139241558729131,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 2.6149,
+      "step": 1384
+    },
+    {
+      "epoch": 1.8747461069735951,
+      "grad_norm": 1.0202843691038828,
+      "learning_rate": 5.597328163351237e-06,
+      "loss": 2.3195,
+      "step": 1385
+    },
+    {
+      "epoch": 1.8761002031144212,
+      "grad_norm": 1.0587205093081162,
+      "learning_rate": 5.59197223019354e-06,
+      "loss": 2.369,
+      "step": 1386
+    },
+    {
+      "epoch": 1.877454299255247,
+      "grad_norm": 1.0246205891822728,
+      "learning_rate": 5.586615608037309e-06,
+      "loss": 1.7189,
+      "step": 1387
+    },
+    {
+      "epoch": 1.878808395396073,
+      "grad_norm": 1.2556625533663983,
+      "learning_rate": 5.581258303117138e-06,
+      "loss": 2.6877,
+      "step": 1388
+    },
+    {
+      "epoch": 1.880162491536899,
+      "grad_norm": 1.0753392702185522,
+      "learning_rate": 5.575900321668414e-06,
+      "loss": 2.2089,
+      "step": 1389
+    },
+    {
+      "epoch": 1.8815165876777251,
+      "grad_norm": 1.0364442011380954,
+      "learning_rate": 5.5705416699273085e-06,
+      "loss": 2.2061,
+      "step": 1390
+    },
+    {
+      "epoch": 1.8828706838185512,
+      "grad_norm": 0.9721557814407294,
+      "learning_rate": 5.565182354130776e-06,
+      "loss": 2.0579,
+      "step": 1391
+    },
+    {
+      "epoch": 1.8842247799593772,
+      "grad_norm": 1.0946729786176126,
+      "learning_rate": 5.559822380516539e-06,
+      "loss": 1.8052,
+      "step": 1392
+    },
+    {
+      "epoch": 1.8855788761002032,
+      "grad_norm": 1.2816555939803111,
+      "learning_rate": 5.554461755323095e-06,
+      "loss": 2.5817,
+      "step": 1393
+    },
+    {
+      "epoch": 1.886932972241029,
+      "grad_norm": 1.132710357655246,
+      "learning_rate": 5.54910048478969e-06,
+      "loss": 2.0912,
+      "step": 1394
+    },
+    {
+      "epoch": 1.8882870683818551,
+      "grad_norm": 1.4036561544628778,
+      "learning_rate": 5.5437385751563265e-06,
+      "loss": 1.9642,
+      "step": 1395
+    },
+    {
+      "epoch": 1.8896411645226812,
+      "grad_norm": 0.9338991067427596,
+      "learning_rate": 5.5383760326637525e-06,
+      "loss": 2.107,
+      "step": 1396
+    },
+    {
+      "epoch": 1.890995260663507,
+      "grad_norm": 1.0328732328422332,
+      "learning_rate": 5.533012863553445e-06,
+      "loss": 2.0279,
+      "step": 1397
+    },
+    {
+      "epoch": 1.892349356804333,
+      "grad_norm": 1.0913343945197256,
+      "learning_rate": 5.527649074067618e-06,
+      "loss": 2.037,
+      "step": 1398
+    },
+    {
+      "epoch": 1.893703452945159,
+      "grad_norm": 1.1464521194384492,
+      "learning_rate": 5.522284670449204e-06,
+      "loss": 1.9054,
+      "step": 1399
+    },
+    {
+      "epoch": 1.8950575490859851,
+      "grad_norm": 1.7305402471538387,
+      "learning_rate": 5.51691965894185e-06,
+      "loss": 2.1952,
+      "step": 1400
+    },
+    {
+      "epoch": 1.8964116452268112,
+      "grad_norm": 0.9382393928786104,
+      "learning_rate": 5.511554045789915e-06,
+      "loss": 1.937,
+      "step": 1401
+    },
+    {
+      "epoch": 1.8977657413676372,
+      "grad_norm": 1.042649655218192,
+      "learning_rate": 5.506187837238451e-06,
+      "loss": 1.9652,
+      "step": 1402
+    },
+    {
+      "epoch": 1.8991198375084632,
+      "grad_norm": 1.1266190255444637,
+      "learning_rate": 5.5008210395332095e-06,
+      "loss": 2.5897,
+      "step": 1403
+    },
+    {
+      "epoch": 1.900473933649289,
+      "grad_norm": 1.0568485706741606,
+      "learning_rate": 5.495453658920622e-06,
+      "loss": 2.1456,
+      "step": 1404
+    },
+    {
+      "epoch": 1.9018280297901151,
+      "grad_norm": 2.3042054688617517,
+      "learning_rate": 5.490085701647805e-06,
+      "loss": 2.4831,
+      "step": 1405
+    },
+    {
+      "epoch": 1.903182125930941,
+      "grad_norm": 1.0661084879580494,
+      "learning_rate": 5.484717173962538e-06,
+      "loss": 1.979,
+      "step": 1406
+    },
+    {
+      "epoch": 1.904536222071767,
+      "grad_norm": 1.8624940514969899,
+      "learning_rate": 5.479348082113273e-06,
+      "loss": 2.2721,
+      "step": 1407
+    },
+    {
+      "epoch": 1.905890318212593,
+      "grad_norm": 1.0349458998467256,
+      "learning_rate": 5.4739784323491115e-06,
+      "loss": 2.1855,
+      "step": 1408
+    },
+    {
+      "epoch": 1.907244414353419,
+      "grad_norm": 1.0420374247356121,
+      "learning_rate": 5.468608230919811e-06,
+      "loss": 1.9857,
+      "step": 1409
+    },
+    {
+      "epoch": 1.9085985104942451,
+      "grad_norm": 1.0571913625970193,
+      "learning_rate": 5.463237484075765e-06,
+      "loss": 1.9137,
+      "step": 1410
+    },
+    {
+      "epoch": 1.9099526066350712,
+      "grad_norm": 1.148192788565203,
+      "learning_rate": 5.457866198068006e-06,
+      "loss": 1.9036,
+      "step": 1411
+    },
+    {
+      "epoch": 1.9113067027758972,
+      "grad_norm": 0.9825967829135833,
+      "learning_rate": 5.45249437914819e-06,
+      "loss": 2.4171,
+      "step": 1412
+    },
+    {
+      "epoch": 1.9126607989167232,
+      "grad_norm": 1.070670848522326,
+      "learning_rate": 5.4471220335685985e-06,
+      "loss": 2.2573,
+      "step": 1413
+    },
+    {
+      "epoch": 1.914014895057549,
+      "grad_norm": 1.1613970345507367,
+      "learning_rate": 5.44174916758212e-06,
+      "loss": 2.1134,
+      "step": 1414
+    },
+    {
+      "epoch": 1.9153689911983751,
+      "grad_norm": 0.9179865288261003,
+      "learning_rate": 5.436375787442256e-06,
+      "loss": 2.2131,
+      "step": 1415
+    },
+    {
+      "epoch": 1.916723087339201,
+      "grad_norm": 1.276005671850693,
+      "learning_rate": 5.4310018994030974e-06,
+      "loss": 2.0346,
+      "step": 1416
+    },
+    {
+      "epoch": 1.918077183480027,
+      "grad_norm": 1.1781979348827227,
+      "learning_rate": 5.425627509719336e-06,
+      "loss": 2.0975,
+      "step": 1417
+    },
+    {
+      "epoch": 1.919431279620853,
+      "grad_norm": 1.1588736839015075,
+      "learning_rate": 5.420252624646238e-06,
+      "loss": 2.8176,
+      "step": 1418
+    },
+    {
+      "epoch": 1.920785375761679,
+      "grad_norm": 1.0396625706559195,
+      "learning_rate": 5.414877250439654e-06,
+      "loss": 2.2559,
+      "step": 1419
+    },
+    {
+      "epoch": 1.9221394719025051,
+      "grad_norm": 1.1691760754641374,
+      "learning_rate": 5.409501393356001e-06,
+      "loss": 2.134,
+      "step": 1420
+    },
+    {
+      "epoch": 1.9234935680433312,
+      "grad_norm": 1.0188156446108358,
+      "learning_rate": 5.404125059652255e-06,
+      "loss": 3.1594,
+      "step": 1421
+    },
+    {
+      "epoch": 1.9248476641841572,
+      "grad_norm": 1.1384064676805719,
+      "learning_rate": 5.398748255585952e-06,
+      "loss": 2.3359,
+      "step": 1422
+    },
+    {
+      "epoch": 1.926201760324983,
+      "grad_norm": 1.0703628157280833,
+      "learning_rate": 5.39337098741517e-06,
+      "loss": 2.322,
+      "step": 1423
+    },
+    {
+      "epoch": 1.927555856465809,
+      "grad_norm": 3.5972509087878834,
+      "learning_rate": 5.387993261398532e-06,
+      "loss": 2.0459,
+      "step": 1424
+    },
+    {
+      "epoch": 1.9289099526066351,
+      "grad_norm": 1.1278074196782333,
+      "learning_rate": 5.3826150837951925e-06,
+      "loss": 2.1835,
+      "step": 1425
+    },
+    {
+      "epoch": 1.930264048747461,
+      "grad_norm": 1.1684466435119685,
+      "learning_rate": 5.3772364608648304e-06,
+      "loss": 2.5017,
+      "step": 1426
+    },
+    {
+      "epoch": 1.931618144888287,
+      "grad_norm": 1.1988587658843286,
+      "learning_rate": 5.371857398867644e-06,
+      "loss": 2.4385,
+      "step": 1427
+    },
+    {
+      "epoch": 1.932972241029113,
+      "grad_norm": 0.943349610365291,
+      "learning_rate": 5.36647790406434e-06,
+      "loss": 2.1582,
+      "step": 1428
+    },
+    {
+      "epoch": 1.934326337169939,
+      "grad_norm": 1.0186861984009383,
+      "learning_rate": 5.361097982716133e-06,
+      "loss": 2.0697,
+      "step": 1429
+    },
+    {
+      "epoch": 1.9356804333107651,
+      "grad_norm": 1.0863559782982186,
+      "learning_rate": 5.355717641084733e-06,
+      "loss": 2.6047,
+      "step": 1430
+    },
+    {
+      "epoch": 1.9370345294515912,
+      "grad_norm": 1.2753139571991683,
+      "learning_rate": 5.350336885432337e-06,
+      "loss": 2.5542,
+      "step": 1431
+    },
+    {
+      "epoch": 1.9383886255924172,
+      "grad_norm": 1.2432570870357769,
+      "learning_rate": 5.3449557220216245e-06,
+      "loss": 2.465,
+      "step": 1432
+    },
+    {
+      "epoch": 1.939742721733243,
+      "grad_norm": 1.1214426550882808,
+      "learning_rate": 5.339574157115752e-06,
+      "loss": 1.8317,
+      "step": 1433
+    },
+    {
+      "epoch": 1.941096817874069,
+      "grad_norm": 1.1278145331998868,
+      "learning_rate": 5.334192196978341e-06,
+      "loss": 1.8582,
+      "step": 1434
+    },
+    {
+      "epoch": 1.942450914014895,
+      "grad_norm": 0.9703824311730281,
+      "learning_rate": 5.328809847873472e-06,
+      "loss": 2.0282,
+      "step": 1435
+    },
+    {
+      "epoch": 1.943805010155721,
+      "grad_norm": 1.1313991479024939,
+      "learning_rate": 5.32342711606568e-06,
+      "loss": 2.2019,
+      "step": 1436
+    },
+    {
+      "epoch": 1.945159106296547,
+      "grad_norm": 1.1285464646294614,
+      "learning_rate": 5.318044007819948e-06,
+      "loss": 2.2181,
+      "step": 1437
+    },
+    {
+      "epoch": 1.946513202437373,
+      "grad_norm": 1.2286900340042946,
+      "learning_rate": 5.312660529401693e-06,
+      "loss": 1.8868,
+      "step": 1438
+    },
+    {
+      "epoch": 1.947867298578199,
+      "grad_norm": 1.0620287314695505,
+      "learning_rate": 5.307276687076762e-06,
+      "loss": 2.5188,
+      "step": 1439
+    },
+    {
+      "epoch": 1.9492213947190251,
+      "grad_norm": 1.0392077445928851,
+      "learning_rate": 5.301892487111431e-06,
+      "loss": 2.0289,
+      "step": 1440
+    },
+    {
+      "epoch": 1.9505754908598512,
+      "grad_norm": 1.3207900398189212,
+      "learning_rate": 5.296507935772386e-06,
+      "loss": 2.2526,
+      "step": 1441
+    },
+    {
+      "epoch": 1.9519295870006772,
+      "grad_norm": 1.407831485125338,
+      "learning_rate": 5.291123039326728e-06,
+      "loss": 2.3924,
+      "step": 1442
+    },
+    {
+      "epoch": 1.953283683141503,
+      "grad_norm": 1.092640851593957,
+      "learning_rate": 5.285737804041955e-06,
+      "loss": 2.2856,
+      "step": 1443
+    },
+    {
+      "epoch": 1.954637779282329,
+      "grad_norm": 1.0720656311911092,
+      "learning_rate": 5.2803522361859596e-06,
+      "loss": 2.3745,
+      "step": 1444
+    },
+    {
+      "epoch": 1.955991875423155,
+      "grad_norm": 1.0470209379216529,
+      "learning_rate": 5.274966342027025e-06,
+      "loss": 2.4369,
+      "step": 1445
+    },
+    {
+      "epoch": 1.957345971563981,
+      "grad_norm": 1.7795808290406938,
+      "learning_rate": 5.269580127833812e-06,
+      "loss": 2.2505,
+      "step": 1446
+    },
+    {
+      "epoch": 1.958700067704807,
+      "grad_norm": 0.949115783937015,
+      "learning_rate": 5.264193599875353e-06,
+      "loss": 2.1088,
+      "step": 1447
+    },
+    {
+      "epoch": 1.960054163845633,
+      "grad_norm": 1.0307094217484138,
+      "learning_rate": 5.258806764421048e-06,
+      "loss": 2.0897,
+      "step": 1448
+    },
+    {
+      "epoch": 1.961408259986459,
+      "grad_norm": 1.0944949048376986,
+      "learning_rate": 5.253419627740651e-06,
+      "loss": 2.2959,
+      "step": 1449
+    },
+    {
+      "epoch": 1.9627623561272851,
+      "grad_norm": 1.0529310025241865,
+      "learning_rate": 5.248032196104271e-06,
+      "loss": 2.5453,
+      "step": 1450
+    },
+    {
+      "epoch": 1.9641164522681112,
+      "grad_norm": 1.135872662601797,
+      "learning_rate": 5.2426444757823595e-06,
+      "loss": 2.1203,
+      "step": 1451
+    },
+    {
+      "epoch": 1.965470548408937,
+      "grad_norm": 1.0238104015543514,
+      "learning_rate": 5.2372564730457e-06,
+      "loss": 2.2604,
+      "step": 1452
+    },
+    {
+      "epoch": 1.966824644549763,
+      "grad_norm": 0.9503239116070696,
+      "learning_rate": 5.231868194165409e-06,
+      "loss": 2.4823,
+      "step": 1453
+    },
+    {
+      "epoch": 1.968178740690589,
+      "grad_norm": 1.1160937498521495,
+      "learning_rate": 5.226479645412923e-06,
+      "loss": 2.2582,
+      "step": 1454
+    },
+    {
+      "epoch": 1.969532836831415,
+      "grad_norm": 1.0681616120180237,
+      "learning_rate": 5.221090833059992e-06,
+      "loss": 2.1064,
+      "step": 1455
+    },
+    {
+      "epoch": 1.970886932972241,
+      "grad_norm": 1.2336888814525633,
+      "learning_rate": 5.215701763378673e-06,
+      "loss": 2.2147,
+      "step": 1456
+    },
+    {
+      "epoch": 1.972241029113067,
+      "grad_norm": 2.70975758573399,
+      "learning_rate": 5.210312442641327e-06,
+      "loss": 2.0919,
+      "step": 1457
+    },
+    {
+      "epoch": 1.973595125253893,
+      "grad_norm": 1.062018871284572,
+      "learning_rate": 5.204922877120597e-06,
+      "loss": 2.1447,
+      "step": 1458
+    },
+    {
+      "epoch": 1.974949221394719,
+      "grad_norm": 1.0661389981794334,
+      "learning_rate": 5.19953307308942e-06,
+      "loss": 1.9739,
+      "step": 1459
+    },
+    {
+      "epoch": 1.9763033175355451,
+      "grad_norm": 1.0931038252658167,
+      "learning_rate": 5.194143036821008e-06,
+      "loss": 1.9725,
+      "step": 1460
+    },
+    {
+      "epoch": 1.9776574136763712,
+      "grad_norm": 1.0840157871848553,
+      "learning_rate": 5.188752774588841e-06,
+      "loss": 2.2702,
+      "step": 1461
+    },
+    {
+      "epoch": 1.979011509817197,
+      "grad_norm": 1.0991100379851682,
+      "learning_rate": 5.183362292666667e-06,
+      "loss": 2.2273,
+      "step": 1462
+    },
+    {
+      "epoch": 1.980365605958023,
+      "grad_norm": 1.0930537215096972,
+      "learning_rate": 5.177971597328484e-06,
+      "loss": 2.2848,
+      "step": 1463
+    },
+    {
+      "epoch": 1.981719702098849,
+      "grad_norm": 1.1727805767142325,
+      "learning_rate": 5.172580694848541e-06,
+      "loss": 2.2295,
+      "step": 1464
+    },
+    {
+      "epoch": 1.983073798239675,
+      "grad_norm": 1.0331418445485365,
+      "learning_rate": 5.16718959150133e-06,
+      "loss": 2.1311,
+      "step": 1465
+    },
+    {
+      "epoch": 1.984427894380501,
+      "grad_norm": 1.1782815464908813,
+      "learning_rate": 5.161798293561573e-06,
+      "loss": 2.2843,
+      "step": 1466
+    },
+    {
+      "epoch": 1.985781990521327,
+      "grad_norm": 1.1287633720503536,
+      "learning_rate": 5.156406807304223e-06,
+      "loss": 2.2715,
+      "step": 1467
+    },
+    {
+      "epoch": 1.987136086662153,
+      "grad_norm": 1.190713586693506,
+      "learning_rate": 5.151015139004445e-06,
+      "loss": 1.8253,
+      "step": 1468
+    },
+    {
+      "epoch": 1.988490182802979,
+      "grad_norm": 1.070858654181211,
+      "learning_rate": 5.145623294937624e-06,
+      "loss": 2.1808,
+      "step": 1469
+    },
+    {
+      "epoch": 1.9898442789438051,
+      "grad_norm": 1.1615785503909837,
+      "learning_rate": 5.140231281379345e-06,
+      "loss": 1.8325,
+      "step": 1470
+    },
+    {
+      "epoch": 1.9911983750846312,
+      "grad_norm": 0.983503119817654,
+      "learning_rate": 5.134839104605394e-06,
+      "loss": 2.0685,
+      "step": 1471
+    },
+    {
+      "epoch": 1.992552471225457,
+      "grad_norm": 1.1464176386439313,
+      "learning_rate": 5.129446770891738e-06,
+      "loss": 2.325,
+      "step": 1472
+    },
+    {
+      "epoch": 1.993906567366283,
+      "grad_norm": 1.62246860837646,
+      "learning_rate": 5.124054286514538e-06,
+      "loss": 2.6138,
+      "step": 1473
+    },
+    {
+      "epoch": 1.9952606635071088,
+      "grad_norm": 1.0277305356811535,
+      "learning_rate": 5.118661657750122e-06,
+      "loss": 2.1659,
+      "step": 1474
+    },
+    {
+      "epoch": 1.996614759647935,
+      "grad_norm": 1.054556455145067,
+      "learning_rate": 5.113268890874994e-06,
+      "loss": 2.5919,
+      "step": 1475
+    },
+    {
+      "epoch": 2.001354096140826,
+      "grad_norm": 1.1099987816340893,
+      "learning_rate": 5.107875992165809e-06,
+      "loss": 1.6184,
+      "step": 1476
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2952,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 738,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.8997223198949376e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}