diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9254 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1316,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007598784194528875,
+      "grad_norm": 5.392373980748419,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.1593,
+      "step": 1
+    },
+    {
+      "epoch": 0.001519756838905775,
+      "grad_norm": 5.615475781504103,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.1588,
+      "step": 2
+    },
+    {
+      "epoch": 0.0022796352583586625,
+      "grad_norm": 4.914069653695411,
+      "learning_rate": 1.5e-06,
+      "loss": 1.1712,
+      "step": 3
+    },
+    {
+      "epoch": 0.00303951367781155,
+      "grad_norm": 4.069148923342371,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.1347,
+      "step": 4
+    },
+    {
+      "epoch": 0.003799392097264438,
+      "grad_norm": 4.589236750514864,
+      "learning_rate": 2.5e-06,
+      "loss": 1.1706,
+      "step": 5
+    },
+    {
+      "epoch": 0.004559270516717325,
+      "grad_norm": 4.9434297504777005,
+      "learning_rate": 3e-06,
+      "loss": 1.1772,
+      "step": 6
+    },
+    {
+      "epoch": 0.005319148936170213,
+      "grad_norm": 3.4416977119500243,
+      "learning_rate": 3.5e-06,
+      "loss": 1.0584,
+      "step": 7
+    },
+    {
+      "epoch": 0.0060790273556231,
+      "grad_norm": 4.070486113809039,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.1407,
+      "step": 8
+    },
+    {
+      "epoch": 0.006838905775075988,
+      "grad_norm": 2.7096606367740073,
+      "learning_rate": 4.5e-06,
+      "loss": 1.088,
+      "step": 9
+    },
+    {
+      "epoch": 0.007598784194528876,
+      "grad_norm": 2.889884611574226,
+      "learning_rate": 5e-06,
+      "loss": 1.0615,
+      "step": 10
+    },
+    {
+      "epoch": 0.008358662613981762,
+      "grad_norm": 3.0319033761268064,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.0594,
+      "step": 11
+    },
+    {
+      "epoch": 0.00911854103343465,
+      "grad_norm": 3.1212228696636144,
+      "learning_rate": 6e-06,
+      "loss": 0.954,
+      "step": 12
+    },
+    {
+      "epoch": 0.009878419452887538,
+      "grad_norm": 3.711991792308233,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.142,
+      "step": 13
+    },
+    {
+      "epoch": 0.010638297872340425,
+      "grad_norm": 2.439421741044695,
+      "learning_rate": 7e-06,
+      "loss": 0.9602,
+      "step": 14
+    },
+    {
+      "epoch": 0.011398176291793313,
+      "grad_norm": 1.9628371660327635,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.9027,
+      "step": 15
+    },
+    {
+      "epoch": 0.0121580547112462,
+      "grad_norm": 3.257417507004413,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.039,
+      "step": 16
+    },
+    {
+      "epoch": 0.012917933130699088,
+      "grad_norm": 2.4146003969998846,
+      "learning_rate": 8.5e-06,
+      "loss": 1.0201,
+      "step": 17
+    },
+    {
+      "epoch": 0.013677811550151976,
+      "grad_norm": 3.09862075736547,
+      "learning_rate": 9e-06,
+      "loss": 1.057,
+      "step": 18
+    },
+    {
+      "epoch": 0.014437689969604863,
+      "grad_norm": 2.7183441077347377,
+      "learning_rate": 9.5e-06,
+      "loss": 0.9938,
+      "step": 19
+    },
+    {
+      "epoch": 0.015197568389057751,
+      "grad_norm": 1.9806218870322116,
+      "learning_rate": 1e-05,
+      "loss": 0.9274,
+      "step": 20
+    },
+    {
+      "epoch": 0.015957446808510637,
+      "grad_norm": 2.213785716942658,
+      "learning_rate": 1.0500000000000001e-05,
+      "loss": 0.9291,
+      "step": 21
+    },
+    {
+      "epoch": 0.016717325227963525,
+      "grad_norm": 2.2031307140835095,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.9623,
+      "step": 22
+    },
+    {
+      "epoch": 0.017477203647416412,
+      "grad_norm": 2.143276503422077,
+      "learning_rate": 1.15e-05,
+      "loss": 0.9017,
+      "step": 23
+    },
+    {
+      "epoch": 0.0182370820668693,
+      "grad_norm": 2.5934685165822136,
+      "learning_rate": 1.2e-05,
+      "loss": 1.0336,
+      "step": 24
+    },
+    {
+      "epoch": 0.018996960486322188,
+      "grad_norm": 1.7265298518991723,
+      "learning_rate": 1.25e-05,
+      "loss": 0.826,
+      "step": 25
+    },
+    {
+      "epoch": 0.019756838905775075,
+      "grad_norm": 1.9637629397975633,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.0266,
+      "step": 26
+    },
+    {
+      "epoch": 0.020516717325227963,
+      "grad_norm": 1.6548259502003555,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 0.8499,
+      "step": 27
+    },
+    {
+      "epoch": 0.02127659574468085,
+      "grad_norm": 1.496950294252413,
+      "learning_rate": 1.4e-05,
+      "loss": 0.9429,
+      "step": 28
+    },
+    {
+      "epoch": 0.022036474164133738,
+      "grad_norm": 1.9111993504389688,
+      "learning_rate": 1.45e-05,
+      "loss": 1.0238,
+      "step": 29
+    },
+    {
+      "epoch": 0.022796352583586626,
+      "grad_norm": 2.118572184185085,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.8923,
+      "step": 30
+    },
+    {
+      "epoch": 0.023556231003039513,
+      "grad_norm": 1.7079943693934316,
+      "learning_rate": 1.55e-05,
+      "loss": 0.986,
+      "step": 31
+    },
+    {
+      "epoch": 0.0243161094224924,
+      "grad_norm": 1.4156759204322658,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.9132,
+      "step": 32
+    },
+    {
+      "epoch": 0.02507598784194529,
+      "grad_norm": 1.8467608530833703,
+      "learning_rate": 1.65e-05,
+      "loss": 0.9626,
+      "step": 33
+    },
+    {
+      "epoch": 0.025835866261398176,
+      "grad_norm": 2.292348819555964,
+      "learning_rate": 1.7e-05,
+      "loss": 0.9339,
+      "step": 34
+    },
+    {
+      "epoch": 0.026595744680851064,
+      "grad_norm": 2.025279283460441,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 0.9731,
+      "step": 35
+    },
+    {
+      "epoch": 0.02735562310030395,
+      "grad_norm": 1.7197081377527985,
+      "learning_rate": 1.8e-05,
+      "loss": 0.8812,
+      "step": 36
+    },
+    {
+      "epoch": 0.02811550151975684,
+      "grad_norm": 1.4784991973389618,
+      "learning_rate": 1.8500000000000002e-05,
+      "loss": 0.8295,
+      "step": 37
+    },
+    {
+      "epoch": 0.028875379939209727,
+      "grad_norm": 1.7999946297063045,
+      "learning_rate": 1.9e-05,
+      "loss": 0.942,
+      "step": 38
+    },
+    {
+      "epoch": 0.029635258358662615,
+      "grad_norm": 1.57035411083628,
+      "learning_rate": 1.95e-05,
+      "loss": 0.9143,
+      "step": 39
+    },
+    {
+      "epoch": 0.030395136778115502,
+      "grad_norm": 1.4742382818502906,
+      "learning_rate": 2e-05,
+      "loss": 0.8261,
+      "step": 40
+    },
+    {
+      "epoch": 0.03115501519756839,
+      "grad_norm": 1.685470546563188,
+      "learning_rate": 1.9999969691239106e-05,
+      "loss": 0.9266,
+      "step": 41
+    },
+    {
+      "epoch": 0.031914893617021274,
+      "grad_norm": 1.819891227503006,
+      "learning_rate": 1.999987876514015e-05,
+      "loss": 0.7966,
+      "step": 42
+    },
+    {
+      "epoch": 0.03267477203647416,
+      "grad_norm": 1.6403215187550444,
+      "learning_rate": 1.9999727222254298e-05,
+      "loss": 0.8766,
+      "step": 43
+    },
+    {
+      "epoch": 0.03343465045592705,
+      "grad_norm": 1.661111203316323,
+      "learning_rate": 1.999951506350017e-05,
+      "loss": 0.8838,
+      "step": 44
+    },
+    {
+      "epoch": 0.03419452887537994,
+      "grad_norm": 1.8665661460800034,
+      "learning_rate": 1.999924229016382e-05,
+      "loss": 0.9154,
+      "step": 45
+    },
+    {
+      "epoch": 0.034954407294832825,
+      "grad_norm": 1.2156637252590854,
+      "learning_rate": 1.999890890389873e-05,
+      "loss": 0.8792,
+      "step": 46
+    },
+    {
+      "epoch": 0.03571428571428571,
+      "grad_norm": 1.5973421574876991,
+      "learning_rate": 1.9998514906725805e-05,
+      "loss": 1.0,
+      "step": 47
+    },
+    {
+      "epoch": 0.0364741641337386,
+      "grad_norm": 1.8290400076895916,
+      "learning_rate": 1.9998060301033363e-05,
+      "loss": 0.8365,
+      "step": 48
+    },
+    {
+      "epoch": 0.03723404255319149,
+      "grad_norm": 1.5044749699898898,
+      "learning_rate": 1.9997545089577105e-05,
+      "loss": 0.8659,
+      "step": 49
+    },
+    {
+      "epoch": 0.037993920972644375,
+      "grad_norm": 1.5176053017436755,
+      "learning_rate": 1.9996969275480116e-05,
+      "loss": 0.828,
+      "step": 50
+    },
+    {
+      "epoch": 0.03875379939209726,
+      "grad_norm": 1.6475722779572208,
+      "learning_rate": 1.9996332862232843e-05,
+      "loss": 0.7738,
+      "step": 51
+    },
+    {
+      "epoch": 0.03951367781155015,
+      "grad_norm": 1.3692033978549916,
+      "learning_rate": 1.9995635853693057e-05,
+      "loss": 0.7984,
+      "step": 52
+    },
+    {
+      "epoch": 0.04027355623100304,
+      "grad_norm": 1.5086570356724653,
+      "learning_rate": 1.999487825408586e-05,
+      "loss": 0.8643,
+      "step": 53
+    },
+    {
+      "epoch": 0.041033434650455926,
+      "grad_norm": 1.7307732470201098,
+      "learning_rate": 1.999406006800363e-05,
+      "loss": 0.9023,
+      "step": 54
+    },
+    {
+      "epoch": 0.04179331306990881,
+      "grad_norm": 1.4814422822505877,
+      "learning_rate": 1.9993181300406006e-05,
+      "loss": 0.9016,
+      "step": 55
+    },
+    {
+      "epoch": 0.0425531914893617,
+      "grad_norm": 1.649565494890097,
+      "learning_rate": 1.999224195661986e-05,
+      "loss": 0.9152,
+      "step": 56
+    },
+    {
+      "epoch": 0.04331306990881459,
+      "grad_norm": 1.428811652134738,
+      "learning_rate": 1.9991242042339265e-05,
+      "loss": 0.8827,
+      "step": 57
+    },
+    {
+      "epoch": 0.044072948328267476,
+      "grad_norm": 1.6088153906886453,
+      "learning_rate": 1.999018156362545e-05,
+      "loss": 0.8855,
+      "step": 58
+    },
+    {
+      "epoch": 0.044832826747720364,
+      "grad_norm": 1.21757428725773,
+      "learning_rate": 1.998906052690677e-05,
+      "loss": 0.8445,
+      "step": 59
+    },
+    {
+      "epoch": 0.04559270516717325,
+      "grad_norm": 1.4622543628469789,
+      "learning_rate": 1.9987878938978684e-05,
+      "loss": 0.8945,
+      "step": 60
+    },
+    {
+      "epoch": 0.04635258358662614,
+      "grad_norm": 1.3459144100217668,
+      "learning_rate": 1.9986636807003676e-05,
+      "loss": 0.7925,
+      "step": 61
+    },
+    {
+      "epoch": 0.04711246200607903,
+      "grad_norm": 1.8084721790905864,
+      "learning_rate": 1.998533413851124e-05,
+      "loss": 0.9144,
+      "step": 62
+    },
+    {
+      "epoch": 0.047872340425531915,
+      "grad_norm": 1.563168210948487,
+      "learning_rate": 1.9983970941397837e-05,
+      "loss": 0.9009,
+      "step": 63
+    },
+    {
+      "epoch": 0.0486322188449848,
+      "grad_norm": 1.2514060627145838,
+      "learning_rate": 1.9982547223926826e-05,
+      "loss": 0.8268,
+      "step": 64
+    },
+    {
+      "epoch": 0.04939209726443769,
+      "grad_norm": 1.5896045682936804,
+      "learning_rate": 1.998106299472843e-05,
+      "loss": 0.8506,
+      "step": 65
+    },
+    {
+      "epoch": 0.05015197568389058,
+      "grad_norm": 1.4741152211547537,
+      "learning_rate": 1.997951826279968e-05,
+      "loss": 0.86,
+      "step": 66
+    },
+    {
+      "epoch": 0.050911854103343465,
+      "grad_norm": 1.70832273064182,
+      "learning_rate": 1.9977913037504355e-05,
+      "loss": 0.9067,
+      "step": 67
+    },
+    {
+      "epoch": 0.05167173252279635,
+      "grad_norm": 1.3751246290228405,
+      "learning_rate": 1.997624732857294e-05,
+      "loss": 0.8035,
+      "step": 68
+    },
+    {
+      "epoch": 0.05243161094224924,
+      "grad_norm": 1.34252014451877,
+      "learning_rate": 1.9974521146102535e-05,
+      "loss": 0.8675,
+      "step": 69
+    },
+    {
+      "epoch": 0.05319148936170213,
+      "grad_norm": 1.4143421357527768,
+      "learning_rate": 1.9972734500556847e-05,
+      "loss": 0.8564,
+      "step": 70
+    },
+    {
+      "epoch": 0.053951367781155016,
+      "grad_norm": 1.6403313489416589,
+      "learning_rate": 1.997088740276607e-05,
+      "loss": 0.7951,
+      "step": 71
+    },
+    {
+      "epoch": 0.0547112462006079,
+      "grad_norm": 1.219845844216479,
+      "learning_rate": 1.9968979863926857e-05,
+      "loss": 0.8174,
+      "step": 72
+    },
+    {
+      "epoch": 0.05547112462006079,
+      "grad_norm": 1.59139585827731,
+      "learning_rate": 1.996701189560223e-05,
+      "loss": 0.8493,
+      "step": 73
+    },
+    {
+      "epoch": 0.05623100303951368,
+      "grad_norm": 1.4699277770542065,
+      "learning_rate": 1.9964983509721527e-05,
+      "loss": 0.9078,
+      "step": 74
+    },
+    {
+      "epoch": 0.056990881458966566,
+      "grad_norm": 1.5574253270254241,
+      "learning_rate": 1.9962894718580325e-05,
+      "loss": 0.7197,
+      "step": 75
+    },
+    {
+      "epoch": 0.057750759878419454,
+      "grad_norm": 1.5509163193218476,
+      "learning_rate": 1.9960745534840357e-05,
+      "loss": 0.8237,
+      "step": 76
+    },
+    {
+      "epoch": 0.05851063829787234,
+      "grad_norm": 1.4408780150628306,
+      "learning_rate": 1.9958535971529434e-05,
+      "loss": 1.0039,
+      "step": 77
+    },
+    {
+      "epoch": 0.05927051671732523,
+      "grad_norm": 1.453487595280164,
+      "learning_rate": 1.9956266042041394e-05,
+      "loss": 0.8981,
+      "step": 78
+    },
+    {
+      "epoch": 0.06003039513677812,
+      "grad_norm": 1.4539780740287995,
+      "learning_rate": 1.995393576013598e-05,
+      "loss": 0.849,
+      "step": 79
+    },
+    {
+      "epoch": 0.060790273556231005,
+      "grad_norm": 1.4456479361644199,
+      "learning_rate": 1.995154513993878e-05,
+      "loss": 0.818,
+      "step": 80
+    },
+    {
+      "epoch": 0.06155015197568389,
+      "grad_norm": 1.5015246396319477,
+      "learning_rate": 1.9949094195941152e-05,
+      "loss": 0.8152,
+      "step": 81
+    },
+    {
+      "epoch": 0.06231003039513678,
+      "grad_norm": 2.184036159077668,
+      "learning_rate": 1.99465829430001e-05,
+      "loss": 0.7642,
+      "step": 82
+    },
+    {
+      "epoch": 0.06306990881458967,
+      "grad_norm": 1.4495582363572672,
+      "learning_rate": 1.9944011396338223e-05,
+      "loss": 0.8664,
+      "step": 83
+    },
+    {
+      "epoch": 0.06382978723404255,
+      "grad_norm": 1.5529791956246375,
+      "learning_rate": 1.9941379571543597e-05,
+      "loss": 0.815,
+      "step": 84
+    },
+    {
+      "epoch": 0.06458966565349544,
+      "grad_norm": 1.543607432288304,
+      "learning_rate": 1.9938687484569694e-05,
+      "loss": 0.7809,
+      "step": 85
+    },
+    {
+      "epoch": 0.06534954407294832,
+      "grad_norm": 1.4084163251743675,
+      "learning_rate": 1.993593515173528e-05,
+      "loss": 0.8651,
+      "step": 86
+    },
+    {
+      "epoch": 0.06610942249240122,
+      "grad_norm": 1.5092848953424813,
+      "learning_rate": 1.99331225897243e-05,
+      "loss": 0.8116,
+      "step": 87
+    },
+    {
+      "epoch": 0.0668693009118541,
+      "grad_norm": 1.198520037496096,
+      "learning_rate": 1.993024981558583e-05,
+      "loss": 0.7682,
+      "step": 88
+    },
+    {
+      "epoch": 0.067629179331307,
+      "grad_norm": 1.338492147841421,
+      "learning_rate": 1.9927316846733902e-05,
+      "loss": 0.8698,
+      "step": 89
+    },
+    {
+      "epoch": 0.06838905775075987,
+      "grad_norm": 1.21422692756782,
+      "learning_rate": 1.9924323700947446e-05,
+      "loss": 0.8362,
+      "step": 90
+    },
+    {
+      "epoch": 0.06914893617021277,
+      "grad_norm": 1.4137231652492608,
+      "learning_rate": 1.9921270396370175e-05,
+      "loss": 0.9347,
+      "step": 91
+    },
+    {
+      "epoch": 0.06990881458966565,
+      "grad_norm": 1.4021530103978004,
+      "learning_rate": 1.991815695151046e-05,
+      "loss": 0.8722,
+      "step": 92
+    },
+    {
+      "epoch": 0.07066869300911854,
+      "grad_norm": 1.595914265986467,
+      "learning_rate": 1.9914983385241235e-05,
+      "loss": 0.8869,
+      "step": 93
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 1.343571323940674,
+      "learning_rate": 1.991174971679987e-05,
+      "loss": 0.7631,
+      "step": 94
+    },
+    {
+      "epoch": 0.07218844984802432,
+      "grad_norm": 1.4708678977572762,
+      "learning_rate": 1.990845596578807e-05,
+      "loss": 0.8463,
+      "step": 95
+    },
+    {
+      "epoch": 0.0729483282674772,
+      "grad_norm": 1.332039881729776,
+      "learning_rate": 1.9905102152171728e-05,
+      "loss": 0.9065,
+      "step": 96
+    },
+    {
+      "epoch": 0.0737082066869301,
+      "grad_norm": 1.2277713334137743,
+      "learning_rate": 1.990168829628083e-05,
+      "loss": 0.752,
+      "step": 97
+    },
+    {
+      "epoch": 0.07446808510638298,
+      "grad_norm": 1.6099144553470797,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.9118,
+      "step": 98
+    },
+    {
+      "epoch": 0.07522796352583587,
+      "grad_norm": 1.5664820017750818,
+      "learning_rate": 1.989468054081501e-05,
+      "loss": 0.8291,
+      "step": 99
+    },
+    {
+      "epoch": 0.07598784194528875,
+      "grad_norm": 1.4140769704353136,
+      "learning_rate": 1.9891086683719362e-05,
+      "loss": 0.9577,
+      "step": 100
+    },
+    {
+      "epoch": 0.07674772036474165,
+      "grad_norm": 1.3864767137337772,
+      "learning_rate": 1.988743286930746e-05,
+      "loss": 0.8285,
+      "step": 101
+    },
+    {
+      "epoch": 0.07750759878419453,
+      "grad_norm": 1.5455729545188828,
+      "learning_rate": 1.988371911972782e-05,
+      "loss": 0.8088,
+      "step": 102
+    },
+    {
+      "epoch": 0.07826747720364742,
+      "grad_norm": 1.3171862271945194,
+      "learning_rate": 1.987994545749227e-05,
+      "loss": 0.7693,
+      "step": 103
+    },
+    {
+      "epoch": 0.0790273556231003,
+      "grad_norm": 1.3893324250233694,
+      "learning_rate": 1.9876111905475816e-05,
+      "loss": 0.7947,
+      "step": 104
+    },
+    {
+      "epoch": 0.0797872340425532,
+      "grad_norm": 1.4024118979965372,
+      "learning_rate": 1.98722184869165e-05,
+      "loss": 0.8308,
+      "step": 105
+    },
+    {
+      "epoch": 0.08054711246200608,
+      "grad_norm": 2.012851667754775,
+      "learning_rate": 1.9868265225415263e-05,
+      "loss": 0.8298,
+      "step": 106
+    },
+    {
+      "epoch": 0.08130699088145897,
+      "grad_norm": 1.3482414109433463,
+      "learning_rate": 1.9864252144935795e-05,
+      "loss": 0.7672,
+      "step": 107
+    },
+    {
+      "epoch": 0.08206686930091185,
+      "grad_norm": 1.2317674879997527,
+      "learning_rate": 1.9860179269804394e-05,
+      "loss": 0.7817,
+      "step": 108
+    },
+    {
+      "epoch": 0.08282674772036475,
+      "grad_norm": 1.3981203786062606,
+      "learning_rate": 1.985604662470982e-05,
+      "loss": 0.6889,
+      "step": 109
+    },
+    {
+      "epoch": 0.08358662613981763,
+      "grad_norm": 1.303454367915867,
+      "learning_rate": 1.9851854234703146e-05,
+      "loss": 0.8761,
+      "step": 110
+    },
+    {
+      "epoch": 0.08434650455927052,
+      "grad_norm": 1.3103465623740185,
+      "learning_rate": 1.9847602125197597e-05,
+      "loss": 0.882,
+      "step": 111
+    },
+    {
+      "epoch": 0.0851063829787234,
+      "grad_norm": 1.2873199820635277,
+      "learning_rate": 1.984329032196841e-05,
+      "loss": 0.8828,
+      "step": 112
+    },
+    {
+      "epoch": 0.0858662613981763,
+      "grad_norm": 1.3713946170441742,
+      "learning_rate": 1.9838918851152668e-05,
+      "loss": 0.811,
+      "step": 113
+    },
+    {
+      "epoch": 0.08662613981762918,
+      "grad_norm": 1.2812852657951437,
+      "learning_rate": 1.9834487739249146e-05,
+      "loss": 0.7949,
+      "step": 114
+    },
+    {
+      "epoch": 0.08738601823708207,
+      "grad_norm": 1.5623819005523776,
+      "learning_rate": 1.982999701311814e-05,
+      "loss": 0.8683,
+      "step": 115
+    },
+    {
+      "epoch": 0.08814589665653495,
+      "grad_norm": 1.6172726857032396,
+      "learning_rate": 1.982544669998132e-05,
+      "loss": 0.847,
+      "step": 116
+    },
+    {
+      "epoch": 0.08890577507598785,
+      "grad_norm": 1.3213604777791341,
+      "learning_rate": 1.982083682742156e-05,
+      "loss": 0.798,
+      "step": 117
+    },
+    {
+      "epoch": 0.08966565349544073,
+      "grad_norm": 1.3084947029055982,
+      "learning_rate": 1.9816167423382766e-05,
+      "loss": 0.8813,
+      "step": 118
+    },
+    {
+      "epoch": 0.09042553191489362,
+      "grad_norm": 1.523391465498402,
+      "learning_rate": 1.9811438516169703e-05,
+      "loss": 0.9239,
+      "step": 119
+    },
+    {
+      "epoch": 0.0911854103343465,
+      "grad_norm": 1.238259855411423,
+      "learning_rate": 1.9806650134447837e-05,
+      "loss": 0.8683,
+      "step": 120
+    },
+    {
+      "epoch": 0.0919452887537994,
+      "grad_norm": 1.3192009377288854,
+      "learning_rate": 1.9801802307243153e-05,
+      "loss": 0.7316,
+      "step": 121
+    },
+    {
+      "epoch": 0.09270516717325228,
+      "grad_norm": 2.0096495892585815,
+      "learning_rate": 1.9796895063941978e-05,
+      "loss": 0.7898,
+      "step": 122
+    },
+    {
+      "epoch": 0.09346504559270517,
+      "grad_norm": 1.3969255987787645,
+      "learning_rate": 1.97919284342908e-05,
+      "loss": 0.8277,
+      "step": 123
+    },
+    {
+      "epoch": 0.09422492401215805,
+      "grad_norm": 1.397460771205679,
+      "learning_rate": 1.9786902448396102e-05,
+      "loss": 0.9012,
+      "step": 124
+    },
+    {
+      "epoch": 0.09498480243161095,
+      "grad_norm": 1.4257091875606565,
+      "learning_rate": 1.9781817136724166e-05,
+      "loss": 0.8008,
+      "step": 125
+    },
+    {
+      "epoch": 0.09574468085106383,
+      "grad_norm": 1.244569600044323,
+      "learning_rate": 1.9776672530100886e-05,
+      "loss": 0.8717,
+      "step": 126
+    },
+    {
+      "epoch": 0.09650455927051672,
+      "grad_norm": 1.188946081032126,
+      "learning_rate": 1.9771468659711595e-05,
+      "loss": 0.7327,
+      "step": 127
+    },
+    {
+      "epoch": 0.0972644376899696,
+      "grad_norm": 1.554362846503631,
+      "learning_rate": 1.976620555710087e-05,
+      "loss": 0.8527,
+      "step": 128
+    },
+    {
+      "epoch": 0.0980243161094225,
+      "grad_norm": 1.555728719202439,
+      "learning_rate": 1.9760883254172327e-05,
+      "loss": 0.7737,
+      "step": 129
+    },
+    {
+      "epoch": 0.09878419452887538,
+      "grad_norm": 1.503153607995042,
+      "learning_rate": 1.975550178318845e-05,
+      "loss": 0.8309,
+      "step": 130
+    },
+    {
+      "epoch": 0.09954407294832827,
+      "grad_norm": 1.3172064729258408,
+      "learning_rate": 1.9750061176770385e-05,
+      "loss": 0.7211,
+      "step": 131
+    },
+    {
+      "epoch": 0.10030395136778116,
+      "grad_norm": 1.3853806174052035,
+      "learning_rate": 1.9744561467897735e-05,
+      "loss": 0.7606,
+      "step": 132
+    },
+    {
+      "epoch": 0.10106382978723404,
+      "grad_norm": 1.2585010964278525,
+      "learning_rate": 1.9739002689908377e-05,
+      "loss": 0.8745,
+      "step": 133
+    },
+    {
+      "epoch": 0.10182370820668693,
+      "grad_norm": 1.3911881806666273,
+      "learning_rate": 1.9733384876498248e-05,
+      "loss": 0.7501,
+      "step": 134
+    },
+    {
+      "epoch": 0.10258358662613981,
+      "grad_norm": 1.251211848009401,
+      "learning_rate": 1.9727708061721132e-05,
+      "loss": 0.8233,
+      "step": 135
+    },
+    {
+      "epoch": 0.1033434650455927,
+      "grad_norm": 1.2337054183810516,
+      "learning_rate": 1.972197227998848e-05,
+      "loss": 0.826,
+      "step": 136
+    },
+    {
+      "epoch": 0.10410334346504559,
+      "grad_norm": 1.5041090979198204,
+      "learning_rate": 1.9716177566069174e-05,
+      "loss": 0.8783,
+      "step": 137
+    },
+    {
+      "epoch": 0.10486322188449848,
+      "grad_norm": 1.5101046814770216,
+      "learning_rate": 1.9710323955089343e-05,
+      "loss": 0.7804,
+      "step": 138
+    },
+    {
+      "epoch": 0.10562310030395136,
+      "grad_norm": 1.5072660387561263,
+      "learning_rate": 1.9704411482532116e-05,
+      "loss": 0.734,
+      "step": 139
+    },
+    {
+      "epoch": 0.10638297872340426,
+      "grad_norm": 1.2921081964890746,
+      "learning_rate": 1.9698440184237442e-05,
+      "loss": 0.849,
+      "step": 140
+    },
+    {
+      "epoch": 0.10714285714285714,
+      "grad_norm": 1.541853671980931,
+      "learning_rate": 1.9692410096401852e-05,
+      "loss": 0.8224,
+      "step": 141
+    },
+    {
+      "epoch": 0.10790273556231003,
+      "grad_norm": 1.3071587615374576,
+      "learning_rate": 1.968632125557824e-05,
+      "loss": 0.7587,
+      "step": 142
+    },
+    {
+      "epoch": 0.10866261398176291,
+      "grad_norm": 1.1838862172669808,
+      "learning_rate": 1.968017369867565e-05,
+      "loss": 0.7568,
+      "step": 143
+    },
+    {
+      "epoch": 0.1094224924012158,
+      "grad_norm": 1.4366564156184205,
+      "learning_rate": 1.9673967462959052e-05,
+      "loss": 0.7987,
+      "step": 144
+    },
+    {
+      "epoch": 0.11018237082066869,
+      "grad_norm": 1.3671489418655876,
+      "learning_rate": 1.966770258604911e-05,
+      "loss": 0.7297,
+      "step": 145
+    },
+    {
+      "epoch": 0.11094224924012158,
+      "grad_norm": 1.4206705119179035,
+      "learning_rate": 1.9661379105921948e-05,
+      "loss": 0.8031,
+      "step": 146
+    },
+    {
+      "epoch": 0.11170212765957446,
+      "grad_norm": 1.2580241072105247,
+      "learning_rate": 1.9654997060908946e-05,
+      "loss": 0.8033,
+      "step": 147
+    },
+    {
+      "epoch": 0.11246200607902736,
+      "grad_norm": 1.4646612072754077,
+      "learning_rate": 1.9648556489696472e-05,
+      "loss": 0.7896,
+      "step": 148
+    },
+    {
+      "epoch": 0.11322188449848024,
+      "grad_norm": 1.6275151468853166,
+      "learning_rate": 1.9642057431325675e-05,
+      "loss": 0.8652,
+      "step": 149
+    },
+    {
+      "epoch": 0.11398176291793313,
+      "grad_norm": 1.3685803210010863,
+      "learning_rate": 1.963549992519223e-05,
+      "loss": 0.7809,
+      "step": 150
+    },
+    {
+      "epoch": 0.11474164133738601,
+      "grad_norm": 1.5354586417497063,
+      "learning_rate": 1.9628884011046123e-05,
+      "loss": 0.7323,
+      "step": 151
+    },
+    {
+      "epoch": 0.11550151975683891,
+      "grad_norm": 1.3889822903157472,
+      "learning_rate": 1.9622209728991382e-05,
+      "loss": 0.8355,
+      "step": 152
+    },
+    {
+      "epoch": 0.11626139817629179,
+      "grad_norm": 1.2631071084505394,
+      "learning_rate": 1.9615477119485855e-05,
+      "loss": 0.7515,
+      "step": 153
+    },
+    {
+      "epoch": 0.11702127659574468,
+      "grad_norm": 1.4221865919993402,
+      "learning_rate": 1.9608686223340944e-05,
+      "loss": 0.8088,
+      "step": 154
+    },
+    {
+      "epoch": 0.11778115501519756,
+      "grad_norm": 1.273295713297955,
+      "learning_rate": 1.9601837081721387e-05,
+      "loss": 0.766,
+      "step": 155
+    },
+    {
+      "epoch": 0.11854103343465046,
+      "grad_norm": 1.2893316371044268,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.8533,
+      "step": 156
+    },
+    {
+      "epoch": 0.11930091185410334,
+      "grad_norm": 1.3268146271180419,
+      "learning_rate": 1.958796422848233e-05,
+      "loss": 0.6719,
+      "step": 157
+    },
+    {
+      "epoch": 0.12006079027355623,
+      "grad_norm": 1.1960561447896023,
+      "learning_rate": 1.9580940600956636e-05,
+      "loss": 0.7584,
+      "step": 158
+    },
+    {
+      "epoch": 0.12082066869300911,
+      "grad_norm": 1.4411873524033865,
+      "learning_rate": 1.9573858896143376e-05,
+      "loss": 0.8278,
+      "step": 159
+    },
+    {
+      "epoch": 0.12158054711246201,
+      "grad_norm": 1.3416181531983082,
+      "learning_rate": 1.9566719156970095e-05,
+      "loss": 0.8561,
+      "step": 160
+    },
+    {
+      "epoch": 0.12234042553191489,
+      "grad_norm": 1.2036183203443838,
+      "learning_rate": 1.955952142671612e-05,
+      "loss": 0.7858,
+      "step": 161
+    },
+    {
+      "epoch": 0.12310030395136778,
+      "grad_norm": 1.1870046808780073,
+      "learning_rate": 1.9552265749012306e-05,
+      "loss": 0.7806,
+      "step": 162
+    },
+    {
+      "epoch": 0.12386018237082067,
+      "grad_norm": 1.222302205896466,
+      "learning_rate": 1.9544952167840777e-05,
+      "loss": 0.8627,
+      "step": 163
+    },
+    {
+      "epoch": 0.12462006079027356,
+      "grad_norm": 1.5806660739206777,
+      "learning_rate": 1.9537580727534643e-05,
+      "loss": 0.783,
+      "step": 164
+    },
+    {
+      "epoch": 0.12537993920972645,
+      "grad_norm": 1.3781176998478113,
+      "learning_rate": 1.953015147277776e-05,
+      "loss": 0.8269,
+      "step": 165
+    },
+    {
+      "epoch": 0.12613981762917933,
+      "grad_norm": 1.4195510675367895,
+      "learning_rate": 1.9522664448604417e-05,
+      "loss": 0.8748,
+      "step": 166
+    },
+    {
+      "epoch": 0.12689969604863222,
+      "grad_norm": 1.5174887968385815,
+      "learning_rate": 1.9515119700399107e-05,
+      "loss": 0.8833,
+      "step": 167
+    },
+    {
+      "epoch": 0.1276595744680851,
+      "grad_norm": 1.3438248097416066,
+      "learning_rate": 1.9507517273896224e-05,
+      "loss": 0.8357,
+      "step": 168
+    },
+    {
+      "epoch": 0.128419452887538,
+      "grad_norm": 1.3500264760925325,
+      "learning_rate": 1.9499857215179788e-05,
+      "loss": 0.7527,
+      "step": 169
+    },
+    {
+      "epoch": 0.12917933130699089,
+      "grad_norm": 1.2983392832133898,
+      "learning_rate": 1.949213957068318e-05,
+      "loss": 0.8242,
+      "step": 170
+    },
+    {
+      "epoch": 0.12993920972644377,
+      "grad_norm": 1.6869508051753024,
+      "learning_rate": 1.9484364387188848e-05,
+      "loss": 0.8743,
+      "step": 171
+    },
+    {
+      "epoch": 0.13069908814589665,
+      "grad_norm": 1.4958875336688886,
+      "learning_rate": 1.9476531711828027e-05,
+      "loss": 0.8884,
+      "step": 172
+    },
+    {
+      "epoch": 0.13145896656534956,
+      "grad_norm": 1.339278362169919,
+      "learning_rate": 1.9468641592080452e-05,
+      "loss": 0.7361,
+      "step": 173
+    },
+    {
+      "epoch": 0.13221884498480244,
+      "grad_norm": 1.2137954227618424,
+      "learning_rate": 1.9460694075774082e-05,
+      "loss": 0.8389,
+      "step": 174
+    },
+    {
+      "epoch": 0.13297872340425532,
+      "grad_norm": 1.251947457987828,
+      "learning_rate": 1.9452689211084775e-05,
+      "loss": 0.7373,
+      "step": 175
+    },
+    {
+      "epoch": 0.1337386018237082,
+      "grad_norm": 1.4159733752838533,
+      "learning_rate": 1.9444627046536055e-05,
+      "loss": 0.8107,
+      "step": 176
+    },
+    {
+      "epoch": 0.1344984802431611,
+      "grad_norm": 1.308294027207144,
+      "learning_rate": 1.9436507630998758e-05,
+      "loss": 0.7178,
+      "step": 177
+    },
+    {
+      "epoch": 0.135258358662614,
+      "grad_norm": 1.502145995616604,
+      "learning_rate": 1.9428331013690763e-05,
+      "loss": 0.8856,
+      "step": 178
+    },
+    {
+      "epoch": 0.13601823708206687,
+      "grad_norm": 1.2598852554569113,
+      "learning_rate": 1.9420097244176708e-05,
+      "loss": 0.8292,
+      "step": 179
+    },
+    {
+      "epoch": 0.13677811550151975,
+      "grad_norm": 1.6076315002668338,
+      "learning_rate": 1.9411806372367656e-05,
+      "loss": 0.8087,
+      "step": 180
+    },
+    {
+      "epoch": 0.13753799392097266,
+      "grad_norm": 1.124290885756185,
+      "learning_rate": 1.940345844852082e-05,
+      "loss": 0.7432,
+      "step": 181
+    },
+    {
+      "epoch": 0.13829787234042554,
+      "grad_norm": 1.0686912126196735,
+      "learning_rate": 1.9395053523239243e-05,
+      "loss": 0.7699,
+      "step": 182
+    },
+    {
+      "epoch": 0.13905775075987842,
+      "grad_norm": 1.1573699691152508,
+      "learning_rate": 1.9386591647471508e-05,
+      "loss": 0.7677,
+      "step": 183
+    },
+    {
+      "epoch": 0.1398176291793313,
+      "grad_norm": 1.3648977894265473,
+      "learning_rate": 1.9378072872511397e-05,
+      "loss": 0.8221,
+      "step": 184
+    },
+    {
+      "epoch": 0.1405775075987842,
+      "grad_norm": 1.1364664694163902,
+      "learning_rate": 1.936949724999762e-05,
+      "loss": 0.8384,
+      "step": 185
+    },
+    {
+      "epoch": 0.1413373860182371,
+      "grad_norm": 1.3416024694427788,
+      "learning_rate": 1.936086483191347e-05,
+      "loss": 0.7966,
+      "step": 186
+    },
+    {
+      "epoch": 0.14209726443768997,
+      "grad_norm": 1.4957047699486865,
+      "learning_rate": 1.9352175670586534e-05,
+      "loss": 0.7893,
+      "step": 187
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 1.244185236914744,
+      "learning_rate": 1.934342981868835e-05,
+      "loss": 0.8186,
+      "step": 188
+    },
+    {
+      "epoch": 0.14361702127659576,
+      "grad_norm": 1.0898922714772983,
+      "learning_rate": 1.93346273292341e-05,
+      "loss": 0.7498,
+      "step": 189
+    },
+    {
+      "epoch": 0.14437689969604864,
+      "grad_norm": 1.5550980224719488,
+      "learning_rate": 1.93257682555823e-05,
+      "loss": 0.8653,
+      "step": 190
+    },
+    {
+      "epoch": 0.14513677811550152,
+      "grad_norm": 1.0239577576099583,
+      "learning_rate": 1.9316852651434463e-05,
+      "loss": 0.7954,
+      "step": 191
+    },
+    {
+      "epoch": 0.1458966565349544,
+      "grad_norm": 1.249219097874476,
+      "learning_rate": 1.9307880570834762e-05,
+      "loss": 0.8113,
+      "step": 192
+    },
+    {
+      "epoch": 0.1466565349544073,
+      "grad_norm": 1.2536737063563324,
+      "learning_rate": 1.929885206816973e-05,
+      "loss": 0.8924,
+      "step": 193
+    },
+    {
+      "epoch": 0.1474164133738602,
+      "grad_norm": 1.248144446624885,
+      "learning_rate": 1.9289767198167918e-05,
+      "loss": 0.8558,
+      "step": 194
+    },
+    {
+      "epoch": 0.14817629179331307,
+      "grad_norm": 1.3218071038847665,
+      "learning_rate": 1.9280626015899548e-05,
+      "loss": 0.6789,
+      "step": 195
+    },
+    {
+      "epoch": 0.14893617021276595,
+      "grad_norm": 1.3457663209794053,
+      "learning_rate": 1.9271428576776206e-05,
+      "loss": 0.909,
+      "step": 196
+    },
+    {
+      "epoch": 0.14969604863221886,
+      "grad_norm": 1.2355616152753617,
+      "learning_rate": 1.9262174936550485e-05,
+      "loss": 0.8809,
+      "step": 197
+    },
+    {
+      "epoch": 0.15045592705167174,
+      "grad_norm": 1.2568018009001072,
+      "learning_rate": 1.9252865151315667e-05,
+      "loss": 0.8291,
+      "step": 198
+    },
+    {
+      "epoch": 0.15121580547112462,
+      "grad_norm": 1.3065339012403299,
+      "learning_rate": 1.9243499277505355e-05,
+      "loss": 0.7787,
+      "step": 199
+    },
+    {
+      "epoch": 0.1519756838905775,
+      "grad_norm": 1.1262100235788826,
+      "learning_rate": 1.9234077371893156e-05,
+      "loss": 0.8359,
+      "step": 200
+    },
+    {
+      "epoch": 0.15273556231003038,
+      "grad_norm": 1.1213547156508892,
+      "learning_rate": 1.922459949159233e-05,
+      "loss": 0.7428,
+      "step": 201
+    },
+    {
+      "epoch": 0.1534954407294833,
+      "grad_norm": 1.1864132070158617,
+      "learning_rate": 1.921506569405544e-05,
+      "loss": 0.8355,
+      "step": 202
+    },
+    {
+      "epoch": 0.15425531914893617,
+      "grad_norm": 1.292709508551478,
+      "learning_rate": 1.9205476037073997e-05,
+      "loss": 0.8431,
+      "step": 203
+    },
+    {
+      "epoch": 0.15501519756838905,
+      "grad_norm": 1.1868138734025648,
+      "learning_rate": 1.9195830578778133e-05,
+      "loss": 0.745,
+      "step": 204
+    },
+    {
+      "epoch": 0.15577507598784193,
+      "grad_norm": 1.243939822562013,
+      "learning_rate": 1.918612937763622e-05,
+      "loss": 0.8194,
+      "step": 205
+    },
+    {
+      "epoch": 0.15653495440729484,
+      "grad_norm": 1.149116048090664,
+      "learning_rate": 1.917637249245454e-05,
+      "loss": 0.8315,
+      "step": 206
+    },
+    {
+      "epoch": 0.15729483282674772,
+      "grad_norm": 1.2601717567550066,
+      "learning_rate": 1.9166559982376905e-05,
+      "loss": 0.8913,
+      "step": 207
+    },
+    {
+      "epoch": 0.1580547112462006,
+      "grad_norm": 1.328584177588409,
+      "learning_rate": 1.9156691906884327e-05,
+      "loss": 0.8223,
+      "step": 208
+    },
+    {
+      "epoch": 0.15881458966565348,
+      "grad_norm": 1.5547321827587175,
+      "learning_rate": 1.914676832579463e-05,
+      "loss": 0.791,
+      "step": 209
+    },
+    {
+      "epoch": 0.1595744680851064,
+      "grad_norm": 1.2058407361433814,
+      "learning_rate": 1.913678929926211e-05,
+      "loss": 0.779,
+      "step": 210
+    },
+    {
+      "epoch": 0.16033434650455927,
+      "grad_norm": 1.3396307771537412,
+      "learning_rate": 1.912675488777714e-05,
+      "loss": 0.7232,
+      "step": 211
+    },
+    {
+      "epoch": 0.16109422492401215,
+      "grad_norm": 1.2998190586100276,
+      "learning_rate": 1.911666515216585e-05,
+      "loss": 0.7749,
+      "step": 212
+    },
+    {
+      "epoch": 0.16185410334346503,
+      "grad_norm": 1.1849324726680373,
+      "learning_rate": 1.9106520153589708e-05,
+      "loss": 0.8955,
+      "step": 213
+    },
+    {
+      "epoch": 0.16261398176291794,
+      "grad_norm": 1.3790790669413162,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.9048,
+      "step": 214
+    },
+    {
+      "epoch": 0.16337386018237082,
+      "grad_norm": 1.1937403185789235,
+      "learning_rate": 1.9086064613863366e-05,
+      "loss": 0.7869,
+      "step": 215
+    },
+    {
+      "epoch": 0.1641337386018237,
+      "grad_norm": 1.3427073302617243,
+      "learning_rate": 1.9075754196709574e-05,
+      "loss": 0.8516,
+      "step": 216
+    },
+    {
+      "epoch": 0.16489361702127658,
+      "grad_norm": 1.3609249588495251,
+      "learning_rate": 1.9065388764583003e-05,
+      "loss": 0.8163,
+      "step": 217
+    },
+    {
+      "epoch": 0.1656534954407295,
+      "grad_norm": 1.082447919451176,
+      "learning_rate": 1.9054968380316342e-05,
+      "loss": 0.7296,
+      "step": 218
+    },
+    {
+      "epoch": 0.16641337386018237,
+      "grad_norm": 1.1720051941365763,
+      "learning_rate": 1.9044493107075367e-05,
+      "loss": 0.7865,
+      "step": 219
+    },
+    {
+      "epoch": 0.16717325227963525,
+      "grad_norm": 1.1536113908201957,
+      "learning_rate": 1.90339630083586e-05,
+      "loss": 0.8147,
+      "step": 220
+    },
+    {
+      "epoch": 0.16793313069908813,
+      "grad_norm": 1.30860579752053,
+      "learning_rate": 1.902337814799688e-05,
+      "loss": 0.9001,
+      "step": 221
+    },
+    {
+      "epoch": 0.16869300911854104,
+      "grad_norm": 1.2957687830149693,
+      "learning_rate": 1.901273859015301e-05,
+      "loss": 0.9007,
+      "step": 222
+    },
+    {
+      "epoch": 0.16945288753799392,
+      "grad_norm": 2.6842584865115526,
+      "learning_rate": 1.900204439932136e-05,
+      "loss": 0.8743,
+      "step": 223
+    },
+    {
+      "epoch": 0.1702127659574468,
+      "grad_norm": 1.4396478669672228,
+      "learning_rate": 1.899129564032745e-05,
+      "loss": 0.8249,
+      "step": 224
+    },
+    {
+      "epoch": 0.17097264437689969,
+      "grad_norm": 1.4422431849699353,
+      "learning_rate": 1.898049237832761e-05,
+      "loss": 0.7921,
+      "step": 225
+    },
+    {
+      "epoch": 0.1717325227963526,
+      "grad_norm": 1.2168879309113474,
+      "learning_rate": 1.8969634678808523e-05,
+      "loss": 0.7845,
+      "step": 226
+    },
+    {
+      "epoch": 0.17249240121580547,
+      "grad_norm": 1.2456034586319875,
+      "learning_rate": 1.8958722607586883e-05,
+      "loss": 0.7799,
+      "step": 227
+    },
+    {
+      "epoch": 0.17325227963525835,
+      "grad_norm": 1.1600154305212664,
+      "learning_rate": 1.8947756230808955e-05,
+      "loss": 0.8335,
+      "step": 228
+    },
+    {
+      "epoch": 0.17401215805471124,
+      "grad_norm": 1.494016771059299,
+      "learning_rate": 1.8936735614950196e-05,
+      "loss": 0.8819,
+      "step": 229
+    },
+    {
+      "epoch": 0.17477203647416414,
+      "grad_norm": 1.2756174524563888,
+      "learning_rate": 1.8925660826814855e-05,
+      "loss": 0.8888,
+      "step": 230
+    },
+    {
+      "epoch": 0.17553191489361702,
+      "grad_norm": 1.1619746990383337,
+      "learning_rate": 1.8914531933535548e-05,
+      "loss": 0.8657,
+      "step": 231
+    },
+    {
+      "epoch": 0.1762917933130699,
+      "grad_norm": 1.1424858446652142,
+      "learning_rate": 1.8903349002572873e-05,
+      "loss": 0.8787,
+      "step": 232
+    },
+    {
+      "epoch": 0.1770516717325228,
+      "grad_norm": 1.189939887590901,
+      "learning_rate": 1.889211210171498e-05,
+      "loss": 0.8428,
+      "step": 233
+    },
+    {
+      "epoch": 0.1778115501519757,
+      "grad_norm": 1.1936796343318596,
+      "learning_rate": 1.8880821299077184e-05,
+      "loss": 0.8588,
+      "step": 234
+    },
+    {
+      "epoch": 0.17857142857142858,
+      "grad_norm": 1.4065647340327965,
+      "learning_rate": 1.8869476663101525e-05,
+      "loss": 0.8189,
+      "step": 235
+    },
+    {
+      "epoch": 0.17933130699088146,
+      "grad_norm": 1.0408203774913998,
+      "learning_rate": 1.885807826255638e-05,
+      "loss": 0.8276,
+      "step": 236
+    },
+    {
+      "epoch": 0.18009118541033434,
+      "grad_norm": 1.081409053735091,
+      "learning_rate": 1.8846626166536027e-05,
+      "loss": 0.8443,
+      "step": 237
+    },
+    {
+      "epoch": 0.18085106382978725,
+      "grad_norm": 1.1432310944964015,
+      "learning_rate": 1.883512044446023e-05,
+      "loss": 0.7497,
+      "step": 238
+    },
+    {
+      "epoch": 0.18161094224924013,
+      "grad_norm": 1.3470961995619208,
+      "learning_rate": 1.882356116607383e-05,
+      "loss": 0.8103,
+      "step": 239
+    },
+    {
+      "epoch": 0.182370820668693,
+      "grad_norm": 1.411957400701631,
+      "learning_rate": 1.8811948401446312e-05,
+      "loss": 0.8736,
+      "step": 240
+    },
+    {
+      "epoch": 0.1831306990881459,
+      "grad_norm": 1.2352325761124088,
+      "learning_rate": 1.8800282220971368e-05,
+      "loss": 0.8187,
+      "step": 241
+    },
+    {
+      "epoch": 0.1838905775075988,
+      "grad_norm": 1.0889469436250492,
+      "learning_rate": 1.8788562695366495e-05,
+      "loss": 0.8589,
+      "step": 242
+    },
+    {
+      "epoch": 0.18465045592705168,
+      "grad_norm": 1.3133401090492782,
+      "learning_rate": 1.8776789895672557e-05,
+      "loss": 0.8244,
+      "step": 243
+    },
+    {
+      "epoch": 0.18541033434650456,
+      "grad_norm": 1.5074821242863536,
+      "learning_rate": 1.8764963893253346e-05,
+      "loss": 0.8579,
+      "step": 244
+    },
+    {
+      "epoch": 0.18617021276595744,
+      "grad_norm": 1.00865864329584,
+      "learning_rate": 1.875308475979516e-05,
+      "loss": 0.7791,
+      "step": 245
+    },
+    {
+      "epoch": 0.18693009118541035,
+      "grad_norm": 1.253557021475991,
+      "learning_rate": 1.8741152567306356e-05,
+      "loss": 0.7365,
+      "step": 246
+    },
+    {
+      "epoch": 0.18768996960486323,
+      "grad_norm": 1.3369745721969943,
+      "learning_rate": 1.8729167388116934e-05,
+      "loss": 0.8689,
+      "step": 247
+    },
+    {
+      "epoch": 0.1884498480243161,
+      "grad_norm": 1.3071305924038121,
+      "learning_rate": 1.8717129294878075e-05,
+      "loss": 0.8417,
+      "step": 248
+    },
+    {
+      "epoch": 0.189209726443769,
+      "grad_norm": 1.1501306153544906,
+      "learning_rate": 1.8705038360561724e-05,
+      "loss": 0.8011,
+      "step": 249
+    },
+    {
+      "epoch": 0.1899696048632219,
+      "grad_norm": 1.2045201301924366,
+      "learning_rate": 1.869289465846012e-05,
+      "loss": 0.9313,
+      "step": 250
+    },
+    {
+      "epoch": 0.19072948328267478,
+      "grad_norm": 1.2535595856961743,
+      "learning_rate": 1.868069826218538e-05,
+      "loss": 0.8452,
+      "step": 251
+    },
+    {
+      "epoch": 0.19148936170212766,
+      "grad_norm": 1.3220176709231053,
+      "learning_rate": 1.866844924566904e-05,
+      "loss": 0.8362,
+      "step": 252
+    },
+    {
+      "epoch": 0.19224924012158054,
+      "grad_norm": 1.1019657196787032,
+      "learning_rate": 1.8656147683161594e-05,
+      "loss": 0.8708,
+      "step": 253
+    },
+    {
+      "epoch": 0.19300911854103345,
+      "grad_norm": 1.424630197873639,
+      "learning_rate": 1.8643793649232072e-05,
+      "loss": 0.806,
+      "step": 254
+    },
+    {
+      "epoch": 0.19376899696048633,
+      "grad_norm": 1.378063597833453,
+      "learning_rate": 1.8631387218767564e-05,
+      "loss": 0.7881,
+      "step": 255
+    },
+    {
+      "epoch": 0.1945288753799392,
+      "grad_norm": 1.282933125284444,
+      "learning_rate": 1.8618928466972773e-05,
+      "loss": 0.8304,
+      "step": 256
+    },
+    {
+      "epoch": 0.1952887537993921,
+      "grad_norm": 2.4772765829403123,
+      "learning_rate": 1.860641746936957e-05,
+      "loss": 0.8523,
+      "step": 257
+    },
+    {
+      "epoch": 0.196048632218845,
+      "grad_norm": 1.418144991348534,
+      "learning_rate": 1.859385430179652e-05,
+      "loss": 0.8348,
+      "step": 258
+    },
+    {
+      "epoch": 0.19680851063829788,
+      "grad_norm": 1.213603855886597,
+      "learning_rate": 1.8581239040408433e-05,
+      "loss": 0.758,
+      "step": 259
+    },
+    {
+      "epoch": 0.19756838905775076,
+      "grad_norm": 1.153413472883319,
+      "learning_rate": 1.8568571761675893e-05,
+      "loss": 0.7345,
+      "step": 260
+    },
+    {
+      "epoch": 0.19832826747720364,
+      "grad_norm": 1.344343857832059,
+      "learning_rate": 1.855585254238481e-05,
+      "loss": 0.685,
+      "step": 261
+    },
+    {
+      "epoch": 0.19908814589665655,
+      "grad_norm": 1.4308863366658235,
+      "learning_rate": 1.8543081459635937e-05,
+      "loss": 0.735,
+      "step": 262
+    },
+    {
+      "epoch": 0.19984802431610943,
+      "grad_norm": 1.2987140195370532,
+      "learning_rate": 1.853025859084441e-05,
+      "loss": 0.7916,
+      "step": 263
+    },
+    {
+      "epoch": 0.2006079027355623,
+      "grad_norm": 2.4292975031700292,
+      "learning_rate": 1.8517384013739287e-05,
+      "loss": 0.7567,
+      "step": 264
+    },
+    {
+      "epoch": 0.2013677811550152,
+      "grad_norm": 1.346335767671634,
+      "learning_rate": 1.8504457806363058e-05,
+      "loss": 0.7877,
+      "step": 265
+    },
+    {
+      "epoch": 0.20212765957446807,
+      "grad_norm": 1.3311699322400439,
+      "learning_rate": 1.8491480047071192e-05,
+      "loss": 0.8149,
+      "step": 266
+    },
+    {
+      "epoch": 0.20288753799392098,
+      "grad_norm": 1.5753723269120146,
+      "learning_rate": 1.847845081453165e-05,
+      "loss": 0.8317,
+      "step": 267
+    },
+    {
+      "epoch": 0.20364741641337386,
+      "grad_norm": 1.3240965527913418,
+      "learning_rate": 1.846537018772441e-05,
+      "loss": 0.8982,
+      "step": 268
+    },
+    {
+      "epoch": 0.20440729483282674,
+      "grad_norm": 1.5986433343548423,
+      "learning_rate": 1.845223824594099e-05,
+      "loss": 0.9481,
+      "step": 269
+    },
+    {
+      "epoch": 0.20516717325227962,
+      "grad_norm": 1.289128583925758,
+      "learning_rate": 1.8439055068783966e-05,
+      "loss": 0.7662,
+      "step": 270
+    },
+    {
+      "epoch": 0.20592705167173253,
+      "grad_norm": 1.3158592270793201,
+      "learning_rate": 1.8425820736166492e-05,
+      "loss": 0.8643,
+      "step": 271
+    },
+    {
+      "epoch": 0.2066869300911854,
+      "grad_norm": 1.1872514587697873,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.8401,
+      "step": 272
+    },
+    {
+      "epoch": 0.2074468085106383,
+      "grad_norm": 1.8169105863005799,
+      "learning_rate": 1.839919892575278e-05,
+      "loss": 0.817,
+      "step": 273
+    },
+    {
+      "epoch": 0.20820668693009117,
+      "grad_norm": 1.2769137841393565,
+      "learning_rate": 1.8385811609331355e-05,
+      "loss": 0.8177,
+      "step": 274
+    },
+    {
+      "epoch": 0.20896656534954408,
+      "grad_norm": 1.3546661685046748,
+      "learning_rate": 1.837237346019814e-05,
+      "loss": 0.7926,
+      "step": 275
+    },
+    {
+      "epoch": 0.20972644376899696,
+      "grad_norm": 1.4388025793800308,
+      "learning_rate": 1.8358884559811855e-05,
+      "loss": 0.7365,
+      "step": 276
+    },
+    {
+      "epoch": 0.21048632218844984,
+      "grad_norm": 1.7306062464308192,
+      "learning_rate": 1.834534498993888e-05,
+      "loss": 0.8317,
+      "step": 277
+    },
+    {
+      "epoch": 0.21124620060790272,
+      "grad_norm": 1.442442585820585,
+      "learning_rate": 1.833175483265273e-05,
+      "loss": 0.8254,
+      "step": 278
+    },
+    {
+      "epoch": 0.21200607902735563,
+      "grad_norm": 1.4238653530865713,
+      "learning_rate": 1.831811417033357e-05,
+      "loss": 0.8327,
+      "step": 279
+    },
+    {
+      "epoch": 0.2127659574468085,
+      "grad_norm": 1.1764715502052638,
+      "learning_rate": 1.8304423085667713e-05,
+      "loss": 0.753,
+      "step": 280
+    },
+    {
+      "epoch": 0.2135258358662614,
+      "grad_norm": 1.1855529270479457,
+      "learning_rate": 1.8290681661647124e-05,
+      "loss": 0.7281,
+      "step": 281
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 1.4055428341031602,
+      "learning_rate": 1.827688998156891e-05,
+      "loss": 0.8775,
+      "step": 282
+    },
+    {
+      "epoch": 0.21504559270516718,
+      "grad_norm": 1.1278610066021444,
+      "learning_rate": 1.826304812903481e-05,
+      "loss": 0.7605,
+      "step": 283
+    },
+    {
+      "epoch": 0.21580547112462006,
+      "grad_norm": 1.230292541864071,
+      "learning_rate": 1.8249156187950717e-05,
+      "loss": 0.8114,
+      "step": 284
+    },
+    {
+      "epoch": 0.21656534954407294,
+      "grad_norm": 1.0624171978120978,
+      "learning_rate": 1.8235214242526125e-05,
+      "loss": 0.7305,
+      "step": 285
+    },
+    {
+      "epoch": 0.21732522796352582,
+      "grad_norm": 1.248817652679081,
+      "learning_rate": 1.8221222377273656e-05,
+      "loss": 0.703,
+      "step": 286
+    },
+    {
+      "epoch": 0.21808510638297873,
+      "grad_norm": 1.339926916421626,
+      "learning_rate": 1.8207180677008528e-05,
+      "loss": 0.8791,
+      "step": 287
+    },
+    {
+      "epoch": 0.2188449848024316,
+      "grad_norm": 1.3120865498034557,
+      "learning_rate": 1.819308922684805e-05,
+      "loss": 0.8117,
+      "step": 288
+    },
+    {
+      "epoch": 0.2196048632218845,
+      "grad_norm": 1.586444358599728,
+      "learning_rate": 1.8178948112211104e-05,
+      "loss": 0.7686,
+      "step": 289
+    },
+    {
+      "epoch": 0.22036474164133737,
+      "grad_norm": 1.1782894592032886,
+      "learning_rate": 1.816475741881761e-05,
+      "loss": 0.7765,
+      "step": 290
+    },
+    {
+      "epoch": 0.22112462006079028,
+      "grad_norm": 1.024174827969512,
+      "learning_rate": 1.815051723268805e-05,
+      "loss": 0.7618,
+      "step": 291
+    },
+    {
+      "epoch": 0.22188449848024316,
+      "grad_norm": 1.2118229336816007,
+      "learning_rate": 1.8136227640142895e-05,
+      "loss": 0.7632,
+      "step": 292
+    },
+    {
+      "epoch": 0.22264437689969604,
+      "grad_norm": 2.831901255089281,
+      "learning_rate": 1.8121888727802113e-05,
+      "loss": 0.8469,
+      "step": 293
+    },
+    {
+      "epoch": 0.22340425531914893,
+      "grad_norm": 1.0974630686236104,
+      "learning_rate": 1.8107500582584642e-05,
+      "loss": 0.7237,
+      "step": 294
+    },
+    {
+      "epoch": 0.22416413373860183,
+      "grad_norm": 1.1089723309653179,
+      "learning_rate": 1.8093063291707847e-05,
+      "loss": 0.8126,
+      "step": 295
+    },
+    {
+      "epoch": 0.22492401215805471,
+      "grad_norm": 1.239233754298523,
+      "learning_rate": 1.807857694268701e-05,
+      "loss": 0.6578,
+      "step": 296
+    },
+    {
+      "epoch": 0.2256838905775076,
+      "grad_norm": 1.218648492045814,
+      "learning_rate": 1.806404162333479e-05,
+      "loss": 0.7293,
+      "step": 297
+    },
+    {
+      "epoch": 0.22644376899696048,
+      "grad_norm": 1.4388090752273834,
+      "learning_rate": 1.804945742176069e-05,
+      "loss": 0.8582,
+      "step": 298
+    },
+    {
+      "epoch": 0.22720364741641338,
+      "grad_norm": 1.1338612730074744,
+      "learning_rate": 1.8034824426370522e-05,
+      "loss": 0.6526,
+      "step": 299
+    },
+    {
+      "epoch": 0.22796352583586627,
+      "grad_norm": 1.1095651278513057,
+      "learning_rate": 1.802014272586589e-05,
+      "loss": 0.7955,
+      "step": 300
+    },
+    {
+      "epoch": 0.22872340425531915,
+      "grad_norm": 1.0179556574023763,
+      "learning_rate": 1.8005412409243604e-05,
+      "loss": 0.8085,
+      "step": 301
+    },
+    {
+      "epoch": 0.22948328267477203,
+      "grad_norm": 1.1342699502336406,
+      "learning_rate": 1.799063356579521e-05,
+      "loss": 0.8013,
+      "step": 302
+    },
+    {
+      "epoch": 0.23024316109422494,
+      "grad_norm": 1.3480287218090232,
+      "learning_rate": 1.797580628510639e-05,
+      "loss": 0.7289,
+      "step": 303
+    },
+    {
+      "epoch": 0.23100303951367782,
+      "grad_norm": 1.3481284749528273,
+      "learning_rate": 1.796093065705644e-05,
+      "loss": 0.789,
+      "step": 304
+    },
+    {
+      "epoch": 0.2317629179331307,
+      "grad_norm": 1.134281183785409,
+      "learning_rate": 1.7946006771817733e-05,
+      "loss": 0.7622,
+      "step": 305
+    },
+    {
+      "epoch": 0.23252279635258358,
+      "grad_norm": 1.3034552078744333,
+      "learning_rate": 1.7931034719855166e-05,
+      "loss": 0.7145,
+      "step": 306
+    },
+    {
+      "epoch": 0.23328267477203649,
+      "grad_norm": 4.164441591688541,
+      "learning_rate": 1.7916014591925605e-05,
+      "loss": 0.7498,
+      "step": 307
+    },
+    {
+      "epoch": 0.23404255319148937,
+      "grad_norm": 1.274502113758588,
+      "learning_rate": 1.7900946479077345e-05,
+      "loss": 0.813,
+      "step": 308
+    },
+    {
+      "epoch": 0.23480243161094225,
+      "grad_norm": 1.2561873096870315,
+      "learning_rate": 1.788583047264955e-05,
+      "loss": 0.7454,
+      "step": 309
+    },
+    {
+      "epoch": 0.23556231003039513,
+      "grad_norm": 1.3636582841755223,
+      "learning_rate": 1.7870666664271706e-05,
+      "loss": 0.8052,
+      "step": 310
+    },
+    {
+      "epoch": 0.23632218844984804,
+      "grad_norm": 1.266779632537306,
+      "learning_rate": 1.7855455145863064e-05,
+      "loss": 0.8781,
+      "step": 311
+    },
+    {
+      "epoch": 0.23708206686930092,
+      "grad_norm": 1.2674372090536117,
+      "learning_rate": 1.784019600963207e-05,
+      "loss": 0.671,
+      "step": 312
+    },
+    {
+      "epoch": 0.2378419452887538,
+      "grad_norm": 1.2644363201775775,
+      "learning_rate": 1.782488934807584e-05,
+      "loss": 0.708,
+      "step": 313
+    },
+    {
+      "epoch": 0.23860182370820668,
+      "grad_norm": 1.3595379806788828,
+      "learning_rate": 1.7809535253979548e-05,
+      "loss": 0.7788,
+      "step": 314
+    },
+    {
+      "epoch": 0.2393617021276596,
+      "grad_norm": 1.235124119841333,
+      "learning_rate": 1.7794133820415916e-05,
+      "loss": 0.7672,
+      "step": 315
+    },
+    {
+      "epoch": 0.24012158054711247,
+      "grad_norm": 1.2547608529516823,
+      "learning_rate": 1.777868514074462e-05,
+      "loss": 0.6577,
+      "step": 316
+    },
+    {
+      "epoch": 0.24088145896656535,
+      "grad_norm": 1.2464188679916115,
+      "learning_rate": 1.776318930861172e-05,
+      "loss": 0.8099,
+      "step": 317
+    },
+    {
+      "epoch": 0.24164133738601823,
+      "grad_norm": 1.253977613923761,
+      "learning_rate": 1.7747646417949114e-05,
+      "loss": 0.8013,
+      "step": 318
+    },
+    {
+      "epoch": 0.24240121580547114,
+      "grad_norm": 1.2580058097893763,
+      "learning_rate": 1.7732056562973956e-05,
+      "loss": 0.7319,
+      "step": 319
+    },
+    {
+      "epoch": 0.24316109422492402,
+      "grad_norm": 1.312382310386678,
+      "learning_rate": 1.771641983818808e-05,
+      "loss": 0.7896,
+      "step": 320
+    },
+    {
+      "epoch": 0.2439209726443769,
+      "grad_norm": 1.2391223661498425,
+      "learning_rate": 1.7700736338377435e-05,
+      "loss": 0.7844,
+      "step": 321
+    },
+    {
+      "epoch": 0.24468085106382978,
+      "grad_norm": 1.2098548099697453,
+      "learning_rate": 1.7685006158611514e-05,
+      "loss": 0.8045,
+      "step": 322
+    },
+    {
+      "epoch": 0.2454407294832827,
+      "grad_norm": 1.3471453672810556,
+      "learning_rate": 1.7669229394242767e-05,
+      "loss": 0.7868,
+      "step": 323
+    },
+    {
+      "epoch": 0.24620060790273557,
+      "grad_norm": 1.2136618191577149,
+      "learning_rate": 1.7653406140906027e-05,
+      "loss": 0.7945,
+      "step": 324
+    },
+    {
+      "epoch": 0.24696048632218845,
+      "grad_norm": 1.1649325831768262,
+      "learning_rate": 1.763753649451794e-05,
+      "loss": 0.776,
+      "step": 325
+    },
+    {
+      "epoch": 0.24772036474164133,
+      "grad_norm": 1.195030183514188,
+      "learning_rate": 1.7621620551276366e-05,
+      "loss": 0.7754,
+      "step": 326
+    },
+    {
+      "epoch": 0.24848024316109424,
+      "grad_norm": 1.159257066391459,
+      "learning_rate": 1.760565840765981e-05,
+      "loss": 0.7639,
+      "step": 327
+    },
+    {
+      "epoch": 0.24924012158054712,
+      "grad_norm": 1.231598878991431,
+      "learning_rate": 1.758965016042683e-05,
+      "loss": 0.7727,
+      "step": 328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1844601831849817,
+      "learning_rate": 1.757359590661545e-05,
+      "loss": 0.7067,
+      "step": 329
+    },
+    {
+      "epoch": 0.2507598784194529,
+      "grad_norm": 1.5924474170665193,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.8584,
+      "step": 330
+    },
+    {
+      "epoch": 0.25151975683890576,
+      "grad_norm": 1.1493715177062336,
+      "learning_rate": 1.754134976880343e-05,
+      "loss": 0.7477,
+      "step": 331
+    },
+    {
+      "epoch": 0.25227963525835867,
+      "grad_norm": 1.2275942317850876,
+      "learning_rate": 1.752515808027088e-05,
+      "loss": 0.7986,
+      "step": 332
+    },
+    {
+      "epoch": 0.2530395136778115,
+      "grad_norm": 1.2516077599858326,
+      "learning_rate": 1.7508920776094943e-05,
+      "loss": 0.7934,
+      "step": 333
+    },
+    {
+      "epoch": 0.25379939209726443,
+      "grad_norm": 1.1819855497843523,
+      "learning_rate": 1.749263795470213e-05,
+      "loss": 0.744,
+      "step": 334
+    },
+    {
+      "epoch": 0.25455927051671734,
+      "grad_norm": 1.2143224653512694,
+      "learning_rate": 1.7476309714794874e-05,
+      "loss": 0.7713,
+      "step": 335
+    },
+    {
+      "epoch": 0.2553191489361702,
+      "grad_norm": 1.2321509393176535,
+      "learning_rate": 1.7459936155350908e-05,
+      "loss": 0.8308,
+      "step": 336
+    },
+    {
+      "epoch": 0.2560790273556231,
+      "grad_norm": 1.2988231987462813,
+      "learning_rate": 1.7443517375622706e-05,
+      "loss": 0.7681,
+      "step": 337
+    },
+    {
+      "epoch": 0.256838905775076,
+      "grad_norm": 1.1910359567755102,
+      "learning_rate": 1.742705347513683e-05,
+      "loss": 0.7908,
+      "step": 338
+    },
+    {
+      "epoch": 0.25759878419452886,
+      "grad_norm": 1.2710111666197443,
+      "learning_rate": 1.7410544553693368e-05,
+      "loss": 0.8822,
+      "step": 339
+    },
+    {
+      "epoch": 0.25835866261398177,
+      "grad_norm": 1.1028923587629835,
+      "learning_rate": 1.7393990711365312e-05,
+      "loss": 0.8332,
+      "step": 340
+    },
+    {
+      "epoch": 0.2591185410334346,
+      "grad_norm": 1.0426416469688176,
+      "learning_rate": 1.7377392048497954e-05,
+      "loss": 0.7304,
+      "step": 341
+    },
+    {
+      "epoch": 0.25987841945288753,
+      "grad_norm": 1.2949586899018366,
+      "learning_rate": 1.7360748665708268e-05,
+      "loss": 0.7674,
+      "step": 342
+    },
+    {
+      "epoch": 0.26063829787234044,
+      "grad_norm": 1.1027637940420714,
+      "learning_rate": 1.7344060663884325e-05,
+      "loss": 0.7404,
+      "step": 343
+    },
+    {
+      "epoch": 0.2613981762917933,
+      "grad_norm": 1.3927957909651685,
+      "learning_rate": 1.7327328144184648e-05,
+      "loss": 0.7895,
+      "step": 344
+    },
+    {
+      "epoch": 0.2621580547112462,
+      "grad_norm": 1.1621493894342037,
+      "learning_rate": 1.7310551208037627e-05,
+      "loss": 0.8042,
+      "step": 345
+    },
+    {
+      "epoch": 0.2629179331306991,
+      "grad_norm": 1.2081606306645722,
+      "learning_rate": 1.729372995714089e-05,
+      "loss": 0.6491,
+      "step": 346
+    },
+    {
+      "epoch": 0.26367781155015196,
+      "grad_norm": 1.219598890320066,
+      "learning_rate": 1.7276864493460702e-05,
+      "loss": 0.8626,
+      "step": 347
+    },
+    {
+      "epoch": 0.26443768996960487,
+      "grad_norm": 1.2033463802362787,
+      "learning_rate": 1.725995491923131e-05,
+      "loss": 0.7272,
+      "step": 348
+    },
+    {
+      "epoch": 0.2651975683890577,
+      "grad_norm": 1.346822838996504,
+      "learning_rate": 1.724300133695437e-05,
+      "loss": 0.7944,
+      "step": 349
+    },
+    {
+      "epoch": 0.26595744680851063,
+      "grad_norm": 1.4462444891707886,
+      "learning_rate": 1.7226003849398292e-05,
+      "loss": 0.8156,
+      "step": 350
+    },
+    {
+      "epoch": 0.26671732522796354,
+      "grad_norm": 1.2005302941250175,
+      "learning_rate": 1.720896255959764e-05,
+      "loss": 0.8301,
+      "step": 351
+    },
+    {
+      "epoch": 0.2674772036474164,
+      "grad_norm": 1.0894697360010115,
+      "learning_rate": 1.7191877570852482e-05,
+      "loss": 0.7737,
+      "step": 352
+    },
+    {
+      "epoch": 0.2682370820668693,
+      "grad_norm": 1.0806869356296902,
+      "learning_rate": 1.717474898672779e-05,
+      "loss": 0.72,
+      "step": 353
+    },
+    {
+      "epoch": 0.2689969604863222,
+      "grad_norm": 1.3169214061141539,
+      "learning_rate": 1.7157576911052796e-05,
+      "loss": 0.8269,
+      "step": 354
+    },
+    {
+      "epoch": 0.26975683890577506,
+      "grad_norm": 1.2420921150436637,
+      "learning_rate": 1.7140361447920363e-05,
+      "loss": 0.805,
+      "step": 355
+    },
+    {
+      "epoch": 0.270516717325228,
+      "grad_norm": 1.0641952583078091,
+      "learning_rate": 1.712310270168637e-05,
+      "loss": 0.6925,
+      "step": 356
+    },
+    {
+      "epoch": 0.2712765957446808,
+      "grad_norm": 1.08072912866875,
+      "learning_rate": 1.7105800776969053e-05,
+      "loss": 0.7469,
+      "step": 357
+    },
+    {
+      "epoch": 0.27203647416413373,
+      "grad_norm": 1.077651076838975,
+      "learning_rate": 1.7088455778648397e-05,
+      "loss": 0.8513,
+      "step": 358
+    },
+    {
+      "epoch": 0.27279635258358664,
+      "grad_norm": 1.115403034371285,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.7293,
+      "step": 359
+    },
+    {
+      "epoch": 0.2735562310030395,
+      "grad_norm": 1.0448107198849983,
+      "learning_rate": 1.7053636982021844e-05,
+      "loss": 0.7339,
+      "step": 360
+    },
+    {
+      "epoch": 0.2743161094224924,
+      "grad_norm": 1.1040024265383643,
+      "learning_rate": 1.7036163394778865e-05,
+      "loss": 0.7801,
+      "step": 361
+    },
+    {
+      "epoch": 0.2750759878419453,
+      "grad_norm": 1.1830956591962893,
+      "learning_rate": 1.7018647156057095e-05,
+      "loss": 0.7069,
+      "step": 362
+    },
+    {
+      "epoch": 0.27583586626139817,
+      "grad_norm": 1.221649805199723,
+      "learning_rate": 1.7001088372035637e-05,
+      "loss": 0.7232,
+      "step": 363
+    },
+    {
+      "epoch": 0.2765957446808511,
+      "grad_norm": 1.2331709557416881,
+      "learning_rate": 1.6983487149151486e-05,
+      "loss": 0.7239,
+      "step": 364
+    },
+    {
+      "epoch": 0.2773556231003039,
+      "grad_norm": 1.2190524746498974,
+      "learning_rate": 1.696584359409889e-05,
+      "loss": 0.6422,
+      "step": 365
+    },
+    {
+      "epoch": 0.27811550151975684,
+      "grad_norm": 1.3212175401909012,
+      "learning_rate": 1.6948157813828718e-05,
+      "loss": 0.7876,
+      "step": 366
+    },
+    {
+      "epoch": 0.27887537993920974,
+      "grad_norm": 1.340964092097885,
+      "learning_rate": 1.693042991554777e-05,
+      "loss": 0.8319,
+      "step": 367
+    },
+    {
+      "epoch": 0.2796352583586626,
+      "grad_norm": 1.4400012740896617,
+      "learning_rate": 1.6912660006718186e-05,
+      "loss": 0.8326,
+      "step": 368
+    },
+    {
+      "epoch": 0.2803951367781155,
+      "grad_norm": 1.3071365110935433,
+      "learning_rate": 1.6894848195056747e-05,
+      "loss": 0.802,
+      "step": 369
+    },
+    {
+      "epoch": 0.2811550151975684,
+      "grad_norm": 1.143992366688308,
+      "learning_rate": 1.6876994588534234e-05,
+      "loss": 0.8713,
+      "step": 370
+    },
+    {
+      "epoch": 0.28191489361702127,
+      "grad_norm": 1.421416338312035,
+      "learning_rate": 1.685909929537479e-05,
+      "loss": 0.6818,
+      "step": 371
+    },
+    {
+      "epoch": 0.2826747720364742,
+      "grad_norm": 1.1835459613627775,
+      "learning_rate": 1.684116242405525e-05,
+      "loss": 0.7498,
+      "step": 372
+    },
+    {
+      "epoch": 0.28343465045592703,
+      "grad_norm": 1.322714460411963,
+      "learning_rate": 1.6823184083304482e-05,
+      "loss": 0.7272,
+      "step": 373
+    },
+    {
+      "epoch": 0.28419452887537994,
+      "grad_norm": 1.21680776076326,
+      "learning_rate": 1.680516438210273e-05,
+      "loss": 0.8131,
+      "step": 374
+    },
+    {
+      "epoch": 0.28495440729483285,
+      "grad_norm": 1.0478828789518506,
+      "learning_rate": 1.6787103429680955e-05,
+      "loss": 0.7681,
+      "step": 375
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.1967472445752216,
+      "learning_rate": 1.676900133552018e-05,
+      "loss": 0.7526,
+      "step": 376
+    },
+    {
+      "epoch": 0.2864741641337386,
+      "grad_norm": 1.0345058357987296,
+      "learning_rate": 1.6750858209350808e-05,
+      "loss": 0.8153,
+      "step": 377
+    },
+    {
+      "epoch": 0.2872340425531915,
+      "grad_norm": 1.208809639529449,
+      "learning_rate": 1.673267416115198e-05,
+      "loss": 0.8176,
+      "step": 378
+    },
+    {
+      "epoch": 0.28799392097264437,
+      "grad_norm": 1.1712691389380345,
+      "learning_rate": 1.6714449301150883e-05,
+      "loss": 0.7528,
+      "step": 379
+    },
+    {
+      "epoch": 0.2887537993920973,
+      "grad_norm": 1.1839785201033064,
+      "learning_rate": 1.6696183739822108e-05,
+      "loss": 0.8387,
+      "step": 380
+    },
+    {
+      "epoch": 0.28951367781155013,
+      "grad_norm": 1.2516534759009132,
+      "learning_rate": 1.6677877587886956e-05,
+      "loss": 0.7668,
+      "step": 381
+    },
+    {
+      "epoch": 0.29027355623100304,
+      "grad_norm": 1.410017906394451,
+      "learning_rate": 1.665953095631279e-05,
+      "loss": 0.7434,
+      "step": 382
+    },
+    {
+      "epoch": 0.29103343465045595,
+      "grad_norm": 1.2800424937579746,
+      "learning_rate": 1.6641143956312337e-05,
+      "loss": 0.8265,
+      "step": 383
+    },
+    {
+      "epoch": 0.2917933130699088,
+      "grad_norm": 1.5058975032467392,
+      "learning_rate": 1.6622716699343032e-05,
+      "loss": 0.8508,
+      "step": 384
+    },
+    {
+      "epoch": 0.2925531914893617,
+      "grad_norm": 1.303773721114821,
+      "learning_rate": 1.660424929710635e-05,
+      "loss": 0.7675,
+      "step": 385
+    },
+    {
+      "epoch": 0.2933130699088146,
+      "grad_norm": 1.4036528322023598,
+      "learning_rate": 1.6585741861547102e-05,
+      "loss": 0.7774,
+      "step": 386
+    },
+    {
+      "epoch": 0.29407294832826747,
+      "grad_norm": 1.081350837248616,
+      "learning_rate": 1.6567194504852778e-05,
+      "loss": 0.7436,
+      "step": 387
+    },
+    {
+      "epoch": 0.2948328267477204,
+      "grad_norm": 1.3699282034155558,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.7488,
+      "step": 388
+    },
+    {
+      "epoch": 0.29559270516717323,
+      "grad_norm": 1.3175813961312064,
+      "learning_rate": 1.652998047801812e-05,
+      "loss": 0.8234,
+      "step": 389
+    },
+    {
+      "epoch": 0.29635258358662614,
+      "grad_norm": 1.4991870376005667,
+      "learning_rate": 1.6511314033459994e-05,
+      "loss": 0.8478,
+      "step": 390
+    },
+    {
+      "epoch": 0.29711246200607905,
+      "grad_norm": 1.274575767033035,
+      "learning_rate": 1.649260811892984e-05,
+      "loss": 0.9165,
+      "step": 391
+    },
+    {
+      "epoch": 0.2978723404255319,
+      "grad_norm": 1.095438448087222,
+      "learning_rate": 1.647386284781828e-05,
+      "loss": 0.7289,
+      "step": 392
+    },
+    {
+      "epoch": 0.2986322188449848,
+      "grad_norm": 1.2565901625292208,
+      "learning_rate": 1.645507833375449e-05,
+      "loss": 0.7155,
+      "step": 393
+    },
+    {
+      "epoch": 0.2993920972644377,
+      "grad_norm": 1.1346280159229207,
+      "learning_rate": 1.643625469060555e-05,
+      "loss": 0.8248,
+      "step": 394
+    },
+    {
+      "epoch": 0.30015197568389057,
+      "grad_norm": 1.1990888274207598,
+      "learning_rate": 1.6417392032475715e-05,
+      "loss": 0.8198,
+      "step": 395
+    },
+    {
+      "epoch": 0.3009118541033435,
+      "grad_norm": 1.0215794971897059,
+      "learning_rate": 1.6398490473705742e-05,
+      "loss": 0.8245,
+      "step": 396
+    },
+    {
+      "epoch": 0.30167173252279633,
+      "grad_norm": 1.1115585693979149,
+      "learning_rate": 1.6379550128872202e-05,
+      "loss": 0.7206,
+      "step": 397
+    },
+    {
+      "epoch": 0.30243161094224924,
+      "grad_norm": 1.004768858656318,
+      "learning_rate": 1.6360571112786768e-05,
+      "loss": 0.7819,
+      "step": 398
+    },
+    {
+      "epoch": 0.30319148936170215,
+      "grad_norm": 1.1508323939293101,
+      "learning_rate": 1.6341553540495533e-05,
+      "loss": 0.7656,
+      "step": 399
+    },
+    {
+      "epoch": 0.303951367781155,
+      "grad_norm": 1.3202901349614393,
+      "learning_rate": 1.6322497527278308e-05,
+      "loss": 0.7249,
+      "step": 400
+    },
+    {
+      "epoch": 0.3047112462006079,
+      "grad_norm": 1.085179473978901,
+      "learning_rate": 1.6303403188647914e-05,
+      "loss": 0.7614,
+      "step": 401
+    },
+    {
+      "epoch": 0.30547112462006076,
+      "grad_norm": 1.3733912801023669,
+      "learning_rate": 1.6284270640349516e-05,
+      "loss": 0.7673,
+      "step": 402
+    },
+    {
+      "epoch": 0.30623100303951367,
+      "grad_norm": 1.005601823506343,
+      "learning_rate": 1.6265099998359868e-05,
+      "loss": 0.7739,
+      "step": 403
+    },
+    {
+      "epoch": 0.3069908814589666,
+      "grad_norm": 1.2084167262300816,
+      "learning_rate": 1.6245891378886655e-05,
+      "loss": 0.7411,
+      "step": 404
+    },
+    {
+      "epoch": 0.30775075987841943,
+      "grad_norm": 1.026803819956731,
+      "learning_rate": 1.6226644898367767e-05,
+      "loss": 0.8073,
+      "step": 405
+    },
+    {
+      "epoch": 0.30851063829787234,
+      "grad_norm": 1.174796304680086,
+      "learning_rate": 1.62073606734706e-05,
+      "loss": 0.749,
+      "step": 406
+    },
+    {
+      "epoch": 0.30927051671732525,
+      "grad_norm": 1.0463661252192882,
+      "learning_rate": 1.6188038821091346e-05,
+      "loss": 0.7496,
+      "step": 407
+    },
+    {
+      "epoch": 0.3100303951367781,
+      "grad_norm": 1.171960934171062,
+      "learning_rate": 1.6168679458354284e-05,
+      "loss": 0.747,
+      "step": 408
+    },
+    {
+      "epoch": 0.310790273556231,
+      "grad_norm": 1.3147049412489755,
+      "learning_rate": 1.6149282702611077e-05,
+      "loss": 0.7915,
+      "step": 409
+    },
+    {
+      "epoch": 0.31155015197568386,
+      "grad_norm": 1.3350489898674418,
+      "learning_rate": 1.6129848671440047e-05,
+      "loss": 0.6741,
+      "step": 410
+    },
+    {
+      "epoch": 0.3123100303951368,
+      "grad_norm": 1.103528249440529,
+      "learning_rate": 1.611037748264548e-05,
+      "loss": 0.8204,
+      "step": 411
+    },
+    {
+      "epoch": 0.3130699088145897,
+      "grad_norm": 1.3771958067762746,
+      "learning_rate": 1.6090869254256892e-05,
+      "loss": 0.6846,
+      "step": 412
+    },
+    {
+      "epoch": 0.31382978723404253,
+      "grad_norm": 1.3058389924401281,
+      "learning_rate": 1.6071324104528333e-05,
+      "loss": 0.8816,
+      "step": 413
+    },
+    {
+      "epoch": 0.31458966565349544,
+      "grad_norm": 1.1591394916763322,
+      "learning_rate": 1.6051742151937655e-05,
+      "loss": 0.7428,
+      "step": 414
+    },
+    {
+      "epoch": 0.31534954407294835,
+      "grad_norm": 1.1638239136948474,
+      "learning_rate": 1.60321235151858e-05,
+      "loss": 0.8085,
+      "step": 415
+    },
+    {
+      "epoch": 0.3161094224924012,
+      "grad_norm": 1.1336620717950094,
+      "learning_rate": 1.6012468313196086e-05,
+      "loss": 0.8214,
+      "step": 416
+    },
+    {
+      "epoch": 0.3168693009118541,
+      "grad_norm": 1.1940294495807557,
+      "learning_rate": 1.599277666511347e-05,
+      "loss": 0.8215,
+      "step": 417
+    },
+    {
+      "epoch": 0.31762917933130697,
+      "grad_norm": 1.1802830931942,
+      "learning_rate": 1.5973048690303848e-05,
+      "loss": 0.7415,
+      "step": 418
+    },
+    {
+      "epoch": 0.3183890577507599,
+      "grad_norm": 1.1182604763564765,
+      "learning_rate": 1.5953284508353316e-05,
+      "loss": 0.7352,
+      "step": 419
+    },
+    {
+      "epoch": 0.3191489361702128,
+      "grad_norm": 1.2492278687757044,
+      "learning_rate": 1.593348423906745e-05,
+      "loss": 0.776,
+      "step": 420
+    },
+    {
+      "epoch": 0.31990881458966564,
+      "grad_norm": 1.1954141273737178,
+      "learning_rate": 1.5913648002470562e-05,
+      "loss": 0.7464,
+      "step": 421
+    },
+    {
+      "epoch": 0.32066869300911854,
+      "grad_norm": 1.3015246134897502,
+      "learning_rate": 1.589377591880501e-05,
+      "loss": 0.7622,
+      "step": 422
+    },
+    {
+      "epoch": 0.32142857142857145,
+      "grad_norm": 1.1224692664097269,
+      "learning_rate": 1.5873868108530443e-05,
+      "loss": 0.6181,
+      "step": 423
+    },
+    {
+      "epoch": 0.3221884498480243,
+      "grad_norm": 1.2533483748063843,
+      "learning_rate": 1.585392469232307e-05,
+      "loss": 0.858,
+      "step": 424
+    },
+    {
+      "epoch": 0.3229483282674772,
+      "grad_norm": 1.122961911873335,
+      "learning_rate": 1.5833945791074943e-05,
+      "loss": 0.7809,
+      "step": 425
+    },
+    {
+      "epoch": 0.32370820668693007,
+      "grad_norm": 1.3135799173833802,
+      "learning_rate": 1.58139315258932e-05,
+      "loss": 0.7554,
+      "step": 426
+    },
+    {
+      "epoch": 0.324468085106383,
+      "grad_norm": 1.273584627487933,
+      "learning_rate": 1.5793882018099365e-05,
+      "loss": 0.8533,
+      "step": 427
+    },
+    {
+      "epoch": 0.3252279635258359,
+      "grad_norm": 1.2874066666555242,
+      "learning_rate": 1.5773797389228583e-05,
+      "loss": 0.8293,
+      "step": 428
+    },
+    {
+      "epoch": 0.32598784194528874,
+      "grad_norm": 1.1526970687848226,
+      "learning_rate": 1.5753677761028896e-05,
+      "loss": 0.7106,
+      "step": 429
+    },
+    {
+      "epoch": 0.32674772036474165,
+      "grad_norm": 1.194532089923422,
+      "learning_rate": 1.5733523255460506e-05,
+      "loss": 0.6916,
+      "step": 430
+    },
+    {
+      "epoch": 0.32750759878419455,
+      "grad_norm": 1.1191253614825432,
+      "learning_rate": 1.571333399469503e-05,
+      "loss": 0.6864,
+      "step": 431
+    },
+    {
+      "epoch": 0.3282674772036474,
+      "grad_norm": 1.002599107035822,
+      "learning_rate": 1.5693110101114763e-05,
+      "loss": 0.6757,
+      "step": 432
+    },
+    {
+      "epoch": 0.3290273556231003,
+      "grad_norm": 1.1606653068590027,
+      "learning_rate": 1.5672851697311935e-05,
+      "loss": 0.8724,
+      "step": 433
+    },
+    {
+      "epoch": 0.32978723404255317,
+      "grad_norm": 1.0461255227947164,
+      "learning_rate": 1.565255890608797e-05,
+      "loss": 0.7232,
+      "step": 434
+    },
+    {
+      "epoch": 0.3305471124620061,
+      "grad_norm": 1.1337841433196596,
+      "learning_rate": 1.5632231850452745e-05,
+      "loss": 0.7884,
+      "step": 435
+    },
+    {
+      "epoch": 0.331306990881459,
+      "grad_norm": 1.2481479032019933,
+      "learning_rate": 1.5611870653623826e-05,
+      "loss": 0.7777,
+      "step": 436
+    },
+    {
+      "epoch": 0.33206686930091184,
+      "grad_norm": 1.1538715959835062,
+      "learning_rate": 1.5591475439025745e-05,
+      "loss": 0.8474,
+      "step": 437
+    },
+    {
+      "epoch": 0.33282674772036475,
+      "grad_norm": 1.1278764402856465,
+      "learning_rate": 1.557104633028924e-05,
+      "loss": 0.7918,
+      "step": 438
+    },
+    {
+      "epoch": 0.33358662613981765,
+      "grad_norm": 1.2177653458117224,
+      "learning_rate": 1.5550583451250504e-05,
+      "loss": 0.7433,
+      "step": 439
+    },
+    {
+      "epoch": 0.3343465045592705,
+      "grad_norm": 1.2816295962098154,
+      "learning_rate": 1.5530086925950435e-05,
+      "loss": 0.7341,
+      "step": 440
+    },
+    {
+      "epoch": 0.3351063829787234,
+      "grad_norm": 1.2855058438606246,
+      "learning_rate": 1.5509556878633894e-05,
+      "loss": 0.7964,
+      "step": 441
+    },
+    {
+      "epoch": 0.33586626139817627,
+      "grad_norm": 1.210793739638994,
+      "learning_rate": 1.5488993433748944e-05,
+      "loss": 0.7667,
+      "step": 442
+    },
+    {
+      "epoch": 0.3366261398176292,
+      "grad_norm": 1.2354508105613038,
+      "learning_rate": 1.5468396715946082e-05,
+      "loss": 0.8007,
+      "step": 443
+    },
+    {
+      "epoch": 0.3373860182370821,
+      "grad_norm": 1.2157655449115916,
+      "learning_rate": 1.5447766850077517e-05,
+      "loss": 0.7962,
+      "step": 444
+    },
+    {
+      "epoch": 0.33814589665653494,
+      "grad_norm": 1.1814788790840558,
+      "learning_rate": 1.5427103961196378e-05,
+      "loss": 0.6983,
+      "step": 445
+    },
+    {
+      "epoch": 0.33890577507598785,
+      "grad_norm": 1.0656501038586574,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.7435,
+      "step": 446
+    },
+    {
+      "epoch": 0.33966565349544076,
+      "grad_norm": 1.05447829161781,
+      "learning_rate": 1.5385679615609045e-05,
+      "loss": 0.8181,
+      "step": 447
+    },
+    {
+      "epoch": 0.3404255319148936,
+      "grad_norm": 1.3259139309112116,
+      "learning_rate": 1.5364918410006967e-05,
+      "loss": 0.8196,
+      "step": 448
+    },
+    {
+      "epoch": 0.3411854103343465,
+      "grad_norm": 1.3179288092906976,
+      "learning_rate": 1.534412468359903e-05,
+      "loss": 0.7695,
+      "step": 449
+    },
+    {
+      "epoch": 0.34194528875379937,
+      "grad_norm": 1.294894096237413,
+      "learning_rate": 1.5323298562431646e-05,
+      "loss": 0.8215,
+      "step": 450
+    },
+    {
+      "epoch": 0.3427051671732523,
+      "grad_norm": 1.0755513853730156,
+      "learning_rate": 1.5302440172747606e-05,
+      "loss": 0.7272,
+      "step": 451
+    },
+    {
+      "epoch": 0.3434650455927052,
+      "grad_norm": 1.282859085633421,
+      "learning_rate": 1.5281549640985295e-05,
+      "loss": 0.6895,
+      "step": 452
+    },
+    {
+      "epoch": 0.34422492401215804,
+      "grad_norm": 1.1930661243488856,
+      "learning_rate": 1.5260627093777936e-05,
+      "loss": 0.8268,
+      "step": 453
+    },
+    {
+      "epoch": 0.34498480243161095,
+      "grad_norm": 1.0866451187615458,
+      "learning_rate": 1.5239672657952833e-05,
+      "loss": 0.7809,
+      "step": 454
+    },
+    {
+      "epoch": 0.34574468085106386,
+      "grad_norm": 1.220607386293379,
+      "learning_rate": 1.5218686460530579e-05,
+      "loss": 0.7719,
+      "step": 455
+    },
+    {
+      "epoch": 0.3465045592705167,
+      "grad_norm": 1.3393480631399062,
+      "learning_rate": 1.5197668628724302e-05,
+      "loss": 0.7731,
+      "step": 456
+    },
+    {
+      "epoch": 0.3472644376899696,
+      "grad_norm": 1.290488709701274,
+      "learning_rate": 1.517661928993889e-05,
+      "loss": 0.8544,
+      "step": 457
+    },
+    {
+      "epoch": 0.34802431610942247,
+      "grad_norm": 1.2186437415720965,
+      "learning_rate": 1.515553857177022e-05,
+      "loss": 0.7633,
+      "step": 458
+    },
+    {
+      "epoch": 0.3487841945288754,
+      "grad_norm": 1.0632730996984814,
+      "learning_rate": 1.5134426602004378e-05,
+      "loss": 0.7913,
+      "step": 459
+    },
+    {
+      "epoch": 0.3495440729483283,
+      "grad_norm": 1.3885119597502555,
+      "learning_rate": 1.5113283508616895e-05,
+      "loss": 0.728,
+      "step": 460
+    },
+    {
+      "epoch": 0.35030395136778114,
+      "grad_norm": 1.1049399011437813,
+      "learning_rate": 1.5092109419771962e-05,
+      "loss": 0.688,
+      "step": 461
+    },
+    {
+      "epoch": 0.35106382978723405,
+      "grad_norm": 1.1804058487766822,
+      "learning_rate": 1.5070904463821658e-05,
+      "loss": 0.8329,
+      "step": 462
+    },
+    {
+      "epoch": 0.3518237082066869,
+      "grad_norm": 1.1751007297581517,
+      "learning_rate": 1.5049668769305172e-05,
+      "loss": 0.7934,
+      "step": 463
+    },
+    {
+      "epoch": 0.3525835866261398,
+      "grad_norm": 1.0871412331260804,
+      "learning_rate": 1.5028402464948023e-05,
+      "loss": 0.7124,
+      "step": 464
+    },
+    {
+      "epoch": 0.3533434650455927,
+      "grad_norm": 1.1814869569600126,
+      "learning_rate": 1.5007105679661276e-05,
+      "loss": 0.7608,
+      "step": 465
+    },
+    {
+      "epoch": 0.3541033434650456,
+      "grad_norm": 1.2768595272009262,
+      "learning_rate": 1.4985778542540764e-05,
+      "loss": 0.7963,
+      "step": 466
+    },
+    {
+      "epoch": 0.3548632218844985,
+      "grad_norm": 1.2637062865213613,
+      "learning_rate": 1.4964421182866312e-05,
+      "loss": 0.8799,
+      "step": 467
+    },
+    {
+      "epoch": 0.3556231003039514,
+      "grad_norm": 1.095234475660764,
+      "learning_rate": 1.4943033730100936e-05,
+      "loss": 0.6945,
+      "step": 468
+    },
+    {
+      "epoch": 0.35638297872340424,
+      "grad_norm": 1.2023691776289789,
+      "learning_rate": 1.4921616313890073e-05,
+      "loss": 0.6206,
+      "step": 469
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 0.9692133018570475,
+      "learning_rate": 1.4900169064060804e-05,
+      "loss": 0.7085,
+      "step": 470
+    },
+    {
+      "epoch": 0.35790273556231,
+      "grad_norm": 1.2970972735015418,
+      "learning_rate": 1.4878692110621028e-05,
+      "loss": 0.7751,
+      "step": 471
+    },
+    {
+      "epoch": 0.3586626139817629,
+      "grad_norm": 1.192748655045959,
+      "learning_rate": 1.4857185583758722e-05,
+      "loss": 0.7444,
+      "step": 472
+    },
+    {
+      "epoch": 0.3594224924012158,
+      "grad_norm": 1.3135419687207899,
+      "learning_rate": 1.483564961384112e-05,
+      "loss": 0.7526,
+      "step": 473
+    },
+    {
+      "epoch": 0.3601823708206687,
+      "grad_norm": 1.2715561806372109,
+      "learning_rate": 1.4814084331413938e-05,
+      "loss": 0.801,
+      "step": 474
+    },
+    {
+      "epoch": 0.3609422492401216,
+      "grad_norm": 1.1055960137648684,
+      "learning_rate": 1.479248986720057e-05,
+      "loss": 0.7801,
+      "step": 475
+    },
+    {
+      "epoch": 0.3617021276595745,
+      "grad_norm": 1.2252469713673722,
+      "learning_rate": 1.4770866352101308e-05,
+      "loss": 0.7273,
+      "step": 476
+    },
+    {
+      "epoch": 0.36246200607902734,
+      "grad_norm": 1.1487823254499623,
+      "learning_rate": 1.474921391719254e-05,
+      "loss": 0.8097,
+      "step": 477
+    },
+    {
+      "epoch": 0.36322188449848025,
+      "grad_norm": 1.1386182161632215,
+      "learning_rate": 1.472753269372596e-05,
+      "loss": 0.7327,
+      "step": 478
+    },
+    {
+      "epoch": 0.3639817629179331,
+      "grad_norm": 1.1552140594901108,
+      "learning_rate": 1.4705822813127776e-05,
+      "loss": 0.818,
+      "step": 479
+    },
+    {
+      "epoch": 0.364741641337386,
+      "grad_norm": 0.9898996357312346,
+      "learning_rate": 1.4684084406997903e-05,
+      "loss": 0.7945,
+      "step": 480
+    },
+    {
+      "epoch": 0.3655015197568389,
+      "grad_norm": 0.9887730384363455,
+      "learning_rate": 1.466231760710917e-05,
+      "loss": 0.7552,
+      "step": 481
+    },
+    {
+      "epoch": 0.3662613981762918,
+      "grad_norm": 1.3676239745133492,
+      "learning_rate": 1.4640522545406519e-05,
+      "loss": 0.7643,
+      "step": 482
+    },
+    {
+      "epoch": 0.3670212765957447,
+      "grad_norm": 1.256157177775327,
+      "learning_rate": 1.4618699354006223e-05,
+      "loss": 0.8437,
+      "step": 483
+    },
+    {
+      "epoch": 0.3677811550151976,
+      "grad_norm": 1.1539469195117926,
+      "learning_rate": 1.4596848165195052e-05,
+      "loss": 0.6623,
+      "step": 484
+    },
+    {
+      "epoch": 0.36854103343465044,
+      "grad_norm": 1.2507320110146378,
+      "learning_rate": 1.45749691114295e-05,
+      "loss": 0.7909,
+      "step": 485
+    },
+    {
+      "epoch": 0.36930091185410335,
+      "grad_norm": 1.1367719149570354,
+      "learning_rate": 1.4553062325334968e-05,
+      "loss": 0.8016,
+      "step": 486
+    },
+    {
+      "epoch": 0.3700607902735562,
+      "grad_norm": 1.0817239050227132,
+      "learning_rate": 1.4531127939704965e-05,
+      "loss": 0.7795,
+      "step": 487
+    },
+    {
+      "epoch": 0.3708206686930091,
+      "grad_norm": 1.1319940927957448,
+      "learning_rate": 1.4509166087500305e-05,
+      "loss": 0.8259,
+      "step": 488
+    },
+    {
+      "epoch": 0.371580547112462,
+      "grad_norm": 1.111174747008594,
+      "learning_rate": 1.4487176901848285e-05,
+      "loss": 0.7436,
+      "step": 489
+    },
+    {
+      "epoch": 0.3723404255319149,
+      "grad_norm": 1.2088816855900237,
+      "learning_rate": 1.4465160516041905e-05,
+      "loss": 0.7708,
+      "step": 490
+    },
+    {
+      "epoch": 0.3731003039513678,
+      "grad_norm": 1.225764934403788,
+      "learning_rate": 1.4443117063539039e-05,
+      "loss": 0.7472,
+      "step": 491
+    },
+    {
+      "epoch": 0.3738601823708207,
+      "grad_norm": 1.1162502285286449,
+      "learning_rate": 1.4421046677961627e-05,
+      "loss": 0.8919,
+      "step": 492
+    },
+    {
+      "epoch": 0.37462006079027355,
+      "grad_norm": 1.1158602272833609,
+      "learning_rate": 1.439894949309489e-05,
+      "loss": 0.6754,
+      "step": 493
+    },
+    {
+      "epoch": 0.37537993920972645,
+      "grad_norm": 1.0891222481558076,
+      "learning_rate": 1.4376825642886473e-05,
+      "loss": 0.835,
+      "step": 494
+    },
+    {
+      "epoch": 0.3761398176291793,
+      "grad_norm": 1.0774289009642644,
+      "learning_rate": 1.435467526144568e-05,
+      "loss": 0.7633,
+      "step": 495
+    },
+    {
+      "epoch": 0.3768996960486322,
+      "grad_norm": 1.1010053587410074,
+      "learning_rate": 1.4332498483042639e-05,
+      "loss": 0.7572,
+      "step": 496
+    },
+    {
+      "epoch": 0.3776595744680851,
+      "grad_norm": 1.3042486518362346,
+      "learning_rate": 1.4310295442107472e-05,
+      "loss": 0.7659,
+      "step": 497
+    },
+    {
+      "epoch": 0.378419452887538,
+      "grad_norm": 1.1632619767172343,
+      "learning_rate": 1.428806627322952e-05,
+      "loss": 0.8763,
+      "step": 498
+    },
+    {
+      "epoch": 0.3791793313069909,
+      "grad_norm": 1.2766467060051423,
+      "learning_rate": 1.4265811111156491e-05,
+      "loss": 0.7718,
+      "step": 499
+    },
+    {
+      "epoch": 0.3799392097264438,
+      "grad_norm": 1.1013403801657626,
+      "learning_rate": 1.4243530090793667e-05,
+      "loss": 0.7466,
+      "step": 500
+    },
+    {
+      "epoch": 0.38069908814589665,
+      "grad_norm": 1.1081380726596666,
+      "learning_rate": 1.4221223347203067e-05,
+      "loss": 0.7945,
+      "step": 501
+    },
+    {
+      "epoch": 0.38145896656534956,
+      "grad_norm": 1.1600362541283282,
+      "learning_rate": 1.4198891015602648e-05,
+      "loss": 0.8559,
+      "step": 502
+    },
+    {
+      "epoch": 0.3822188449848024,
+      "grad_norm": 1.1679070483079377,
+      "learning_rate": 1.4176533231365463e-05,
+      "loss": 0.7827,
+      "step": 503
+    },
+    {
+      "epoch": 0.3829787234042553,
+      "grad_norm": 1.087576877411926,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.6864,
+      "step": 504
+    },
+    {
+      "epoch": 0.3837386018237082,
+      "grad_norm": 1.1756331296030478,
+      "learning_rate": 1.4131741847243665e-05,
+      "loss": 0.8091,
+      "step": 505
+    },
+    {
+      "epoch": 0.3844984802431611,
+      "grad_norm": 1.2645845492058057,
+      "learning_rate": 1.4109308518873321e-05,
+      "loss": 0.8335,
+      "step": 506
+    },
+    {
+      "epoch": 0.385258358662614,
+      "grad_norm": 1.163925232622094,
+      "learning_rate": 1.4086850280893107e-05,
+      "loss": 0.6869,
+      "step": 507
+    },
+    {
+      "epoch": 0.3860182370820669,
+      "grad_norm": 1.3704424537243407,
+      "learning_rate": 1.40643672694393e-05,
+      "loss": 0.7459,
+      "step": 508
+    },
+    {
+      "epoch": 0.38677811550151975,
+      "grad_norm": 1.14432417961867,
+      "learning_rate": 1.4041859620798341e-05,
+      "loss": 0.7432,
+      "step": 509
+    },
+    {
+      "epoch": 0.38753799392097266,
+      "grad_norm": 1.223881973008391,
+      "learning_rate": 1.4019327471406021e-05,
+      "loss": 0.7779,
+      "step": 510
+    },
+    {
+      "epoch": 0.3882978723404255,
+      "grad_norm": 1.1692544458251006,
+      "learning_rate": 1.3996770957846643e-05,
+      "loss": 0.7751,
+      "step": 511
+    },
+    {
+      "epoch": 0.3890577507598784,
+      "grad_norm": 1.166975919514324,
+      "learning_rate": 1.3974190216852203e-05,
+      "loss": 0.7696,
+      "step": 512
+    },
+    {
+      "epoch": 0.3898176291793313,
+      "grad_norm": 1.1562713426590367,
+      "learning_rate": 1.3951585385301557e-05,
+      "loss": 0.7419,
+      "step": 513
+    },
+    {
+      "epoch": 0.3905775075987842,
+      "grad_norm": 1.187327520978288,
+      "learning_rate": 1.3928956600219593e-05,
+      "loss": 0.7498,
+      "step": 514
+    },
+    {
+      "epoch": 0.3913373860182371,
+      "grad_norm": 1.2230763170723078,
+      "learning_rate": 1.3906303998776392e-05,
+      "loss": 0.6194,
+      "step": 515
+    },
+    {
+      "epoch": 0.39209726443769,
+      "grad_norm": 1.3399677206766483,
+      "learning_rate": 1.388362771828642e-05,
+      "loss": 0.826,
+      "step": 516
+    },
+    {
+      "epoch": 0.39285714285714285,
+      "grad_norm": 1.2045222889734593,
+      "learning_rate": 1.3860927896207665e-05,
+      "loss": 0.6885,
+      "step": 517
+    },
+    {
+      "epoch": 0.39361702127659576,
+      "grad_norm": 1.249508900751444,
+      "learning_rate": 1.383820467014082e-05,
+      "loss": 0.826,
+      "step": 518
+    },
+    {
+      "epoch": 0.3943768996960486,
+      "grad_norm": 1.463197283717251,
+      "learning_rate": 1.3815458177828455e-05,
+      "loss": 0.7469,
+      "step": 519
+    },
+    {
+      "epoch": 0.3951367781155015,
+      "grad_norm": 1.4240981666509096,
+      "learning_rate": 1.3792688557154166e-05,
+      "loss": 0.7172,
+      "step": 520
+    },
+    {
+      "epoch": 0.3958966565349544,
+      "grad_norm": 1.037596684194348,
+      "learning_rate": 1.3769895946141753e-05,
+      "loss": 0.7861,
+      "step": 521
+    },
+    {
+      "epoch": 0.3966565349544073,
+      "grad_norm": 1.2627813756704651,
+      "learning_rate": 1.3747080482954378e-05,
+      "loss": 0.8146,
+      "step": 522
+    },
+    {
+      "epoch": 0.3974164133738602,
+      "grad_norm": 1.196949297922996,
+      "learning_rate": 1.3724242305893716e-05,
+      "loss": 0.8127,
+      "step": 523
+    },
+    {
+      "epoch": 0.3981762917933131,
+      "grad_norm": 1.1349164147949475,
+      "learning_rate": 1.3701381553399147e-05,
+      "loss": 0.699,
+      "step": 524
+    },
+    {
+      "epoch": 0.39893617021276595,
+      "grad_norm": 1.049413976516623,
+      "learning_rate": 1.3678498364046877e-05,
+      "loss": 0.7486,
+      "step": 525
+    },
+    {
+      "epoch": 0.39969604863221886,
+      "grad_norm": 1.0959358568221518,
+      "learning_rate": 1.3655592876549135e-05,
+      "loss": 0.7678,
+      "step": 526
+    },
+    {
+      "epoch": 0.4004559270516717,
+      "grad_norm": 1.0203803122275903,
+      "learning_rate": 1.363266522975331e-05,
+      "loss": 0.725,
+      "step": 527
+    },
+    {
+      "epoch": 0.4012158054711246,
+      "grad_norm": 1.2044356250068233,
+      "learning_rate": 1.3609715562641116e-05,
+      "loss": 0.811,
+      "step": 528
+    },
+    {
+      "epoch": 0.40197568389057753,
+      "grad_norm": 1.3222295856000086,
+      "learning_rate": 1.358674401432774e-05,
+      "loss": 0.5911,
+      "step": 529
+    },
+    {
+      "epoch": 0.4027355623100304,
+      "grad_norm": 1.0913691366241578,
+      "learning_rate": 1.3563750724061025e-05,
+      "loss": 0.7614,
+      "step": 530
+    },
+    {
+      "epoch": 0.4034954407294833,
+      "grad_norm": 1.1746171382577624,
+      "learning_rate": 1.354073583122059e-05,
+      "loss": 0.7411,
+      "step": 531
+    },
+    {
+      "epoch": 0.40425531914893614,
+      "grad_norm": 1.4158235294830364,
+      "learning_rate": 1.3517699475317016e-05,
+      "loss": 0.8055,
+      "step": 532
+    },
+    {
+      "epoch": 0.40501519756838905,
+      "grad_norm": 1.1926144941953758,
+      "learning_rate": 1.3494641795990986e-05,
+      "loss": 0.7008,
+      "step": 533
+    },
+    {
+      "epoch": 0.40577507598784196,
+      "grad_norm": 1.1977355387652078,
+      "learning_rate": 1.3471562933012432e-05,
+      "loss": 0.7389,
+      "step": 534
+    },
+    {
+      "epoch": 0.4065349544072948,
+      "grad_norm": 1.1074293443823684,
+      "learning_rate": 1.3448463026279706e-05,
+      "loss": 0.6826,
+      "step": 535
+    },
+    {
+      "epoch": 0.4072948328267477,
+      "grad_norm": 1.1579000196463276,
+      "learning_rate": 1.3425342215818718e-05,
+      "loss": 0.8204,
+      "step": 536
+    },
+    {
+      "epoch": 0.40805471124620063,
+      "grad_norm": 1.1320408734153626,
+      "learning_rate": 1.3402200641782089e-05,
+      "loss": 0.666,
+      "step": 537
+    },
+    {
+      "epoch": 0.4088145896656535,
+      "grad_norm": 1.102357738478694,
+      "learning_rate": 1.3379038444448307e-05,
+      "loss": 0.7923,
+      "step": 538
+    },
+    {
+      "epoch": 0.4095744680851064,
+      "grad_norm": 1.115969398834189,
+      "learning_rate": 1.335585576422087e-05,
+      "loss": 0.795,
+      "step": 539
+    },
+    {
+      "epoch": 0.41033434650455924,
+      "grad_norm": 1.1177016135967561,
+      "learning_rate": 1.3332652741627445e-05,
+      "loss": 0.8207,
+      "step": 540
+    },
+    {
+      "epoch": 0.41109422492401215,
+      "grad_norm": 1.0183563617545865,
+      "learning_rate": 1.3309429517318999e-05,
+      "loss": 0.6787,
+      "step": 541
+    },
+    {
+      "epoch": 0.41185410334346506,
+      "grad_norm": 1.1849766766962344,
+      "learning_rate": 1.3286186232068972e-05,
+      "loss": 0.7712,
+      "step": 542
+    },
+    {
+      "epoch": 0.4126139817629179,
+      "grad_norm": 1.0649504403154333,
+      "learning_rate": 1.326292302677239e-05,
+      "loss": 0.776,
+      "step": 543
+    },
+    {
+      "epoch": 0.4133738601823708,
+      "grad_norm": 1.1103778255621153,
+      "learning_rate": 1.3239640042445037e-05,
+      "loss": 0.7236,
+      "step": 544
+    },
+    {
+      "epoch": 0.41413373860182373,
+      "grad_norm": 1.0781850790804375,
+      "learning_rate": 1.3216337420222602e-05,
+      "loss": 0.6777,
+      "step": 545
+    },
+    {
+      "epoch": 0.4148936170212766,
+      "grad_norm": 1.1779486641454955,
+      "learning_rate": 1.31930153013598e-05,
+      "loss": 0.8779,
+      "step": 546
+    },
+    {
+      "epoch": 0.4156534954407295,
+      "grad_norm": 1.1121183142586448,
+      "learning_rate": 1.3169673827229539e-05,
+      "loss": 0.7979,
+      "step": 547
+    },
+    {
+      "epoch": 0.41641337386018235,
+      "grad_norm": 1.0452578616803079,
+      "learning_rate": 1.3146313139322051e-05,
+      "loss": 0.8233,
+      "step": 548
+    },
+    {
+      "epoch": 0.41717325227963525,
+      "grad_norm": 1.1305161368686159,
+      "learning_rate": 1.3122933379244036e-05,
+      "loss": 0.7697,
+      "step": 549
+    },
+    {
+      "epoch": 0.41793313069908816,
+      "grad_norm": 1.1332904441347997,
+      "learning_rate": 1.3099534688717806e-05,
+      "loss": 0.7204,
+      "step": 550
+    },
+    {
+      "epoch": 0.418693009118541,
+      "grad_norm": 1.0544281101793018,
+      "learning_rate": 1.3076117209580418e-05,
+      "loss": 0.7957,
+      "step": 551
+    },
+    {
+      "epoch": 0.4194528875379939,
+      "grad_norm": 1.2097653291564578,
+      "learning_rate": 1.3052681083782837e-05,
+      "loss": 0.6848,
+      "step": 552
+    },
+    {
+      "epoch": 0.42021276595744683,
+      "grad_norm": 1.0942027920905422,
+      "learning_rate": 1.3029226453389044e-05,
+      "loss": 0.9204,
+      "step": 553
+    },
+    {
+      "epoch": 0.4209726443768997,
+      "grad_norm": 1.1629671487499098,
+      "learning_rate": 1.3005753460575195e-05,
+      "loss": 0.7929,
+      "step": 554
+    },
+    {
+      "epoch": 0.4217325227963526,
+      "grad_norm": 1.1615017757832076,
+      "learning_rate": 1.2982262247628758e-05,
+      "loss": 0.6554,
+      "step": 555
+    },
+    {
+      "epoch": 0.42249240121580545,
+      "grad_norm": 1.2350464342641283,
+      "learning_rate": 1.2958752956947645e-05,
+      "loss": 0.6981,
+      "step": 556
+    },
+    {
+      "epoch": 0.42325227963525835,
+      "grad_norm": 1.074491759854292,
+      "learning_rate": 1.2935225731039349e-05,
+      "loss": 0.8622,
+      "step": 557
+    },
+    {
+      "epoch": 0.42401215805471126,
+      "grad_norm": 1.1647847863807008,
+      "learning_rate": 1.2911680712520082e-05,
+      "loss": 0.8068,
+      "step": 558
+    },
+    {
+      "epoch": 0.4247720364741641,
+      "grad_norm": 1.1447247309035253,
+      "learning_rate": 1.2888118044113913e-05,
+      "loss": 0.7515,
+      "step": 559
+    },
+    {
+      "epoch": 0.425531914893617,
+      "grad_norm": 1.0904101769686227,
+      "learning_rate": 1.2864537868651894e-05,
+      "loss": 0.7788,
+      "step": 560
+    },
+    {
+      "epoch": 0.42629179331306993,
+      "grad_norm": 1.1684433478464358,
+      "learning_rate": 1.2840940329071213e-05,
+      "loss": 0.671,
+      "step": 561
+    },
+    {
+      "epoch": 0.4270516717325228,
+      "grad_norm": 1.2017550465747733,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.7427,
+      "step": 562
+    },
+    {
+      "epoch": 0.4278115501519757,
+      "grad_norm": 1.0420135459940278,
+      "learning_rate": 1.2793693729827984e-05,
+      "loss": 0.7451,
+      "step": 563
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 1.0449468298970457,
+      "learning_rate": 1.2770044956562613e-05,
+      "loss": 0.8562,
+      "step": 564
+    },
+    {
+      "epoch": 0.42933130699088146,
+      "grad_norm": 1.2136363762203919,
+      "learning_rate": 1.2746379391971191e-05,
+      "loss": 0.6584,
+      "step": 565
+    },
+    {
+      "epoch": 0.43009118541033436,
+      "grad_norm": 1.2362341757280944,
+      "learning_rate": 1.272269717950851e-05,
+      "loss": 0.6923,
+      "step": 566
+    },
+    {
+      "epoch": 0.4308510638297872,
+      "grad_norm": 1.2038813073156636,
+      "learning_rate": 1.2698998462730265e-05,
+      "loss": 0.7222,
+      "step": 567
+    },
+    {
+      "epoch": 0.4316109422492401,
+      "grad_norm": 1.1402926150117914,
+      "learning_rate": 1.2675283385292212e-05,
+      "loss": 0.7143,
+      "step": 568
+    },
+    {
+      "epoch": 0.43237082066869303,
+      "grad_norm": 1.277234664355934,
+      "learning_rate": 1.2651552090949265e-05,
+      "loss": 0.7828,
+      "step": 569
+    },
+    {
+      "epoch": 0.4331306990881459,
+      "grad_norm": 1.083121761869766,
+      "learning_rate": 1.2627804723554653e-05,
+      "loss": 0.8248,
+      "step": 570
+    },
+    {
+      "epoch": 0.4338905775075988,
+      "grad_norm": 1.1046276920654474,
+      "learning_rate": 1.2604041427059037e-05,
+      "loss": 0.8345,
+      "step": 571
+    },
+    {
+      "epoch": 0.43465045592705165,
+      "grad_norm": 1.1001442593016568,
+      "learning_rate": 1.2580262345509621e-05,
+      "loss": 0.7206,
+      "step": 572
+    },
+    {
+      "epoch": 0.43541033434650456,
+      "grad_norm": 1.14236207055973,
+      "learning_rate": 1.2556467623049313e-05,
+      "loss": 0.7629,
+      "step": 573
+    },
+    {
+      "epoch": 0.43617021276595747,
+      "grad_norm": 1.1605887190397732,
+      "learning_rate": 1.2532657403915821e-05,
+      "loss": 0.747,
+      "step": 574
+    },
+    {
+      "epoch": 0.4369300911854103,
+      "grad_norm": 1.117710187766665,
+      "learning_rate": 1.2508831832440795e-05,
+      "loss": 0.78,
+      "step": 575
+    },
+    {
+      "epoch": 0.4376899696048632,
+      "grad_norm": 1.2482054789186818,
+      "learning_rate": 1.248499105304894e-05,
+      "loss": 0.7959,
+      "step": 576
+    },
+    {
+      "epoch": 0.43844984802431614,
+      "grad_norm": 1.2256788050979497,
+      "learning_rate": 1.2461135210257156e-05,
+      "loss": 0.5836,
+      "step": 577
+    },
+    {
+      "epoch": 0.439209726443769,
+      "grad_norm": 1.4775730013813828,
+      "learning_rate": 1.2437264448673647e-05,
+      "loss": 0.7441,
+      "step": 578
+    },
+    {
+      "epoch": 0.4399696048632219,
+      "grad_norm": 1.0844347032086228,
+      "learning_rate": 1.2413378912997058e-05,
+      "loss": 0.7404,
+      "step": 579
+    },
+    {
+      "epoch": 0.44072948328267475,
+      "grad_norm": 1.220072154238479,
+      "learning_rate": 1.2389478748015584e-05,
+      "loss": 0.7549,
+      "step": 580
+    },
+    {
+      "epoch": 0.44148936170212766,
+      "grad_norm": 1.3797855512370152,
+      "learning_rate": 1.2365564098606103e-05,
+      "loss": 0.702,
+      "step": 581
+    },
+    {
+      "epoch": 0.44224924012158057,
+      "grad_norm": 1.1048282390434343,
+      "learning_rate": 1.2341635109733292e-05,
+      "loss": 0.681,
+      "step": 582
+    },
+    {
+      "epoch": 0.4430091185410334,
+      "grad_norm": 1.2410407226194968,
+      "learning_rate": 1.2317691926448753e-05,
+      "loss": 0.6772,
+      "step": 583
+    },
+    {
+      "epoch": 0.44376899696048633,
+      "grad_norm": 1.1242367894543606,
+      "learning_rate": 1.2293734693890131e-05,
+      "loss": 0.6518,
+      "step": 584
+    },
+    {
+      "epoch": 0.44452887537993924,
+      "grad_norm": 0.9905778948996463,
+      "learning_rate": 1.226976355728023e-05,
+      "loss": 0.6832,
+      "step": 585
+    },
+    {
+      "epoch": 0.4452887537993921,
+      "grad_norm": 1.1201973075655094,
+      "learning_rate": 1.2245778661926138e-05,
+      "loss": 0.7694,
+      "step": 586
+    },
+    {
+      "epoch": 0.446048632218845,
+      "grad_norm": 1.2210274246473196,
+      "learning_rate": 1.222178015321835e-05,
+      "loss": 0.7827,
+      "step": 587
+    },
+    {
+      "epoch": 0.44680851063829785,
+      "grad_norm": 1.0076097698137845,
+      "learning_rate": 1.2197768176629876e-05,
+      "loss": 0.762,
+      "step": 588
+    },
+    {
+      "epoch": 0.44756838905775076,
+      "grad_norm": 1.0624166505363433,
+      "learning_rate": 1.2173742877715374e-05,
+      "loss": 0.7607,
+      "step": 589
+    },
+    {
+      "epoch": 0.44832826747720367,
+      "grad_norm": 1.2321772853254775,
+      "learning_rate": 1.2149704402110243e-05,
+      "loss": 0.7371,
+      "step": 590
+    },
+    {
+      "epoch": 0.4490881458966565,
+      "grad_norm": 1.2595966240381076,
+      "learning_rate": 1.2125652895529766e-05,
+      "loss": 0.7215,
+      "step": 591
+    },
+    {
+      "epoch": 0.44984802431610943,
+      "grad_norm": 1.1213688750304858,
+      "learning_rate": 1.2101588503768224e-05,
+      "loss": 0.7963,
+      "step": 592
+    },
+    {
+      "epoch": 0.4506079027355623,
+      "grad_norm": 0.9769789823292284,
+      "learning_rate": 1.2077511372697986e-05,
+      "loss": 0.6596,
+      "step": 593
+    },
+    {
+      "epoch": 0.4513677811550152,
+      "grad_norm": 1.3705621757358848,
+      "learning_rate": 1.2053421648268661e-05,
+      "loss": 0.7226,
+      "step": 594
+    },
+    {
+      "epoch": 0.4521276595744681,
+      "grad_norm": 1.0877206767535184,
+      "learning_rate": 1.2029319476506183e-05,
+      "loss": 0.8156,
+      "step": 595
+    },
+    {
+      "epoch": 0.45288753799392095,
+      "grad_norm": 1.1559961003190877,
+      "learning_rate": 1.2005205003511948e-05,
+      "loss": 0.6944,
+      "step": 596
+    },
+    {
+      "epoch": 0.45364741641337386,
+      "grad_norm": 1.055306325871312,
+      "learning_rate": 1.1981078375461917e-05,
+      "loss": 0.7877,
+      "step": 597
+    },
+    {
+      "epoch": 0.45440729483282677,
+      "grad_norm": 1.2138338408638794,
+      "learning_rate": 1.1956939738605722e-05,
+      "loss": 0.8352,
+      "step": 598
+    },
+    {
+      "epoch": 0.4551671732522796,
+      "grad_norm": 1.2164380288185532,
+      "learning_rate": 1.1932789239265803e-05,
+      "loss": 0.7387,
+      "step": 599
+    },
+    {
+      "epoch": 0.45592705167173253,
+      "grad_norm": 1.0622016134428793,
+      "learning_rate": 1.1908627023836504e-05,
+      "loss": 0.7813,
+      "step": 600
+    },
+    {
+      "epoch": 0.4566869300911854,
+      "grad_norm": 1.0354380350319183,
+      "learning_rate": 1.1884453238783185e-05,
+      "loss": 0.6817,
+      "step": 601
+    },
+    {
+      "epoch": 0.4574468085106383,
+      "grad_norm": 1.1146324749938001,
+      "learning_rate": 1.1860268030641337e-05,
+      "loss": 0.6585,
+      "step": 602
+    },
+    {
+      "epoch": 0.4582066869300912,
+      "grad_norm": 1.1733953270814297,
+      "learning_rate": 1.1836071546015704e-05,
+      "loss": 0.7152,
+      "step": 603
+    },
+    {
+      "epoch": 0.45896656534954405,
+      "grad_norm": 1.1063408683598057,
+      "learning_rate": 1.1811863931579377e-05,
+      "loss": 0.7577,
+      "step": 604
+    },
+    {
+      "epoch": 0.45972644376899696,
+      "grad_norm": 1.2502742143674057,
+      "learning_rate": 1.1787645334072912e-05,
+      "loss": 0.6801,
+      "step": 605
+    },
+    {
+      "epoch": 0.46048632218844987,
+      "grad_norm": 1.0856442473471655,
+      "learning_rate": 1.176341590030345e-05,
+      "loss": 0.7897,
+      "step": 606
+    },
+    {
+      "epoch": 0.4612462006079027,
+      "grad_norm": 1.269493236892582,
+      "learning_rate": 1.1739175777143813e-05,
+      "loss": 0.7531,
+      "step": 607
+    },
+    {
+      "epoch": 0.46200607902735563,
+      "grad_norm": 1.1160454691982469,
+      "learning_rate": 1.1714925111531619e-05,
+      "loss": 0.653,
+      "step": 608
+    },
+    {
+      "epoch": 0.4627659574468085,
+      "grad_norm": 1.1847312086049953,
+      "learning_rate": 1.169066405046839e-05,
+      "loss": 0.6859,
+      "step": 609
+    },
+    {
+      "epoch": 0.4635258358662614,
+      "grad_norm": 1.230192197807551,
+      "learning_rate": 1.1666392741018675e-05,
+      "loss": 0.8168,
+      "step": 610
+    },
+    {
+      "epoch": 0.4642857142857143,
+      "grad_norm": 1.1096867804048502,
+      "learning_rate": 1.1642111330309129e-05,
+      "loss": 0.7538,
+      "step": 611
+    },
+    {
+      "epoch": 0.46504559270516715,
+      "grad_norm": 1.0792534800316143,
+      "learning_rate": 1.161781996552765e-05,
+      "loss": 0.7887,
+      "step": 612
+    },
+    {
+      "epoch": 0.46580547112462006,
+      "grad_norm": 1.4139217480179833,
+      "learning_rate": 1.159351879392247e-05,
+      "loss": 0.7883,
+      "step": 613
+    },
+    {
+      "epoch": 0.46656534954407297,
+      "grad_norm": 1.2339602383114001,
+      "learning_rate": 1.1569207962801264e-05,
+      "loss": 0.7791,
+      "step": 614
+    },
+    {
+      "epoch": 0.4673252279635258,
+      "grad_norm": 1.0168116006915144,
+      "learning_rate": 1.1544887619530277e-05,
+      "loss": 0.7971,
+      "step": 615
+    },
+    {
+      "epoch": 0.46808510638297873,
+      "grad_norm": 0.9641437311687417,
+      "learning_rate": 1.152055791153339e-05,
+      "loss": 0.7587,
+      "step": 616
+    },
+    {
+      "epoch": 0.4688449848024316,
+      "grad_norm": 1.2384311319102208,
+      "learning_rate": 1.1496218986291274e-05,
+      "loss": 0.6966,
+      "step": 617
+    },
+    {
+      "epoch": 0.4696048632218845,
+      "grad_norm": 1.123483046561224,
+      "learning_rate": 1.1471870991340459e-05,
+      "loss": 0.7111,
+      "step": 618
+    },
+    {
+      "epoch": 0.4703647416413374,
+      "grad_norm": 1.3323715236633211,
+      "learning_rate": 1.1447514074272452e-05,
+      "loss": 0.9058,
+      "step": 619
+    },
+    {
+      "epoch": 0.47112462006079026,
+      "grad_norm": 1.0458444755972274,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.6511,
+      "step": 620
+    },
+    {
+      "epoch": 0.47188449848024316,
+      "grad_norm": 1.1212928070334884,
+      "learning_rate": 1.1398774064420444e-05,
+      "loss": 0.7476,
+      "step": 621
+    },
+    {
+      "epoch": 0.4726443768996961,
+      "grad_norm": 1.0909705721586769,
+      "learning_rate": 1.1374391267086301e-05,
+      "loss": 0.7604,
+      "step": 622
+    },
+    {
+      "epoch": 0.4734042553191489,
+      "grad_norm": 1.1751938977778684,
+      "learning_rate": 1.1350000138532902e-05,
+      "loss": 0.8169,
+      "step": 623
+    },
+    {
+      "epoch": 0.47416413373860183,
+      "grad_norm": 1.2461151672781425,
+      "learning_rate": 1.1325600826613221e-05,
+      "loss": 0.7262,
+      "step": 624
+    },
+    {
+      "epoch": 0.4749240121580547,
+      "grad_norm": 0.9967543669312325,
+      "learning_rate": 1.1301193479229842e-05,
+      "loss": 0.7393,
+      "step": 625
+    },
+    {
+      "epoch": 0.4756838905775076,
+      "grad_norm": 1.1946765867292415,
+      "learning_rate": 1.1276778244334055e-05,
+      "loss": 0.8378,
+      "step": 626
+    },
+    {
+      "epoch": 0.4764437689969605,
+      "grad_norm": 1.1800912079687036,
+      "learning_rate": 1.1252355269924965e-05,
+      "loss": 0.7834,
+      "step": 627
+    },
+    {
+      "epoch": 0.47720364741641336,
+      "grad_norm": 1.0682660612153005,
+      "learning_rate": 1.1227924704048585e-05,
+      "loss": 0.6977,
+      "step": 628
+    },
+    {
+      "epoch": 0.47796352583586627,
+      "grad_norm": 1.2082657620713166,
+      "learning_rate": 1.1203486694796956e-05,
+      "loss": 0.7893,
+      "step": 629
+    },
+    {
+      "epoch": 0.4787234042553192,
+      "grad_norm": 1.5186537845384005,
+      "learning_rate": 1.1179041390307235e-05,
+      "loss": 0.8235,
+      "step": 630
+    },
+    {
+      "epoch": 0.479483282674772,
+      "grad_norm": 1.1171118006243392,
+      "learning_rate": 1.1154588938760795e-05,
+      "loss": 0.8361,
+      "step": 631
+    },
+    {
+      "epoch": 0.48024316109422494,
+      "grad_norm": 1.0720592098179131,
+      "learning_rate": 1.1130129488382341e-05,
+      "loss": 0.8329,
+      "step": 632
+    },
+    {
+      "epoch": 0.4810030395136778,
+      "grad_norm": 1.149089667943674,
+      "learning_rate": 1.1105663187438997e-05,
+      "loss": 0.7355,
+      "step": 633
+    },
+    {
+      "epoch": 0.4817629179331307,
+      "grad_norm": 1.0824070043425529,
+      "learning_rate": 1.1081190184239418e-05,
+      "loss": 0.6931,
+      "step": 634
+    },
+    {
+      "epoch": 0.4825227963525836,
+      "grad_norm": 1.1997002322685555,
+      "learning_rate": 1.1056710627132885e-05,
+      "loss": 0.6972,
+      "step": 635
+    },
+    {
+      "epoch": 0.48328267477203646,
+      "grad_norm": 1.1943832275144262,
+      "learning_rate": 1.1032224664508406e-05,
+      "loss": 0.6955,
+      "step": 636
+    },
+    {
+      "epoch": 0.48404255319148937,
+      "grad_norm": 1.478006634154453,
+      "learning_rate": 1.1007732444793815e-05,
+      "loss": 0.7545,
+      "step": 637
+    },
+    {
+      "epoch": 0.4848024316109423,
+      "grad_norm": 1.3061954095680135,
+      "learning_rate": 1.0983234116454885e-05,
+      "loss": 0.7204,
+      "step": 638
+    },
+    {
+      "epoch": 0.48556231003039513,
+      "grad_norm": 1.2531293584528425,
+      "learning_rate": 1.0958729827994406e-05,
+      "loss": 0.844,
+      "step": 639
+    },
+    {
+      "epoch": 0.48632218844984804,
+      "grad_norm": 1.0687070140958945,
+      "learning_rate": 1.09342197279513e-05,
+      "loss": 0.7528,
+      "step": 640
+    },
+    {
+      "epoch": 0.4870820668693009,
+      "grad_norm": 1.2505257621824397,
+      "learning_rate": 1.090970396489973e-05,
+      "loss": 0.7184,
+      "step": 641
+    },
+    {
+      "epoch": 0.4878419452887538,
+      "grad_norm": 1.2507111589246462,
+      "learning_rate": 1.0885182687448162e-05,
+      "loss": 0.7329,
+      "step": 642
+    },
+    {
+      "epoch": 0.4886018237082067,
+      "grad_norm": 1.080529498499154,
+      "learning_rate": 1.086065604423851e-05,
+      "loss": 0.7596,
+      "step": 643
+    },
+    {
+      "epoch": 0.48936170212765956,
+      "grad_norm": 1.0020667948655044,
+      "learning_rate": 1.0836124183945209e-05,
+      "loss": 0.7774,
+      "step": 644
+    },
+    {
+      "epoch": 0.49012158054711247,
+      "grad_norm": 1.2598415490818535,
+      "learning_rate": 1.0811587255274313e-05,
+      "loss": 0.7406,
+      "step": 645
+    },
+    {
+      "epoch": 0.4908814589665654,
+      "grad_norm": 1.2288655437787905,
+      "learning_rate": 1.0787045406962605e-05,
+      "loss": 0.71,
+      "step": 646
+    },
+    {
+      "epoch": 0.49164133738601823,
+      "grad_norm": 1.190316068046688,
+      "learning_rate": 1.0762498787776688e-05,
+      "loss": 0.7821,
+      "step": 647
+    },
+    {
+      "epoch": 0.49240121580547114,
+      "grad_norm": 1.1505791158862277,
+      "learning_rate": 1.073794754651208e-05,
+      "loss": 0.7304,
+      "step": 648
+    },
+    {
+      "epoch": 0.493161094224924,
+      "grad_norm": 1.0576484726630266,
+      "learning_rate": 1.0713391831992324e-05,
+      "loss": 0.7889,
+      "step": 649
+    },
+    {
+      "epoch": 0.4939209726443769,
+      "grad_norm": 1.377761448021965,
+      "learning_rate": 1.0688831793068078e-05,
+      "loss": 0.7526,
+      "step": 650
+    },
+    {
+      "epoch": 0.4946808510638298,
+      "grad_norm": 1.248091367107788,
+      "learning_rate": 1.0664267578616208e-05,
+      "loss": 0.6932,
+      "step": 651
+    },
+    {
+      "epoch": 0.49544072948328266,
+      "grad_norm": 1.1413076054917706,
+      "learning_rate": 1.0639699337538897e-05,
+      "loss": 0.6946,
+      "step": 652
+    },
+    {
+      "epoch": 0.49620060790273557,
+      "grad_norm": 1.1391480731885173,
+      "learning_rate": 1.0615127218762733e-05,
+      "loss": 0.6469,
+      "step": 653
+    },
+    {
+      "epoch": 0.4969604863221885,
+      "grad_norm": 1.2298577878062031,
+      "learning_rate": 1.059055137123781e-05,
+      "loss": 0.6232,
+      "step": 654
+    },
+    {
+      "epoch": 0.49772036474164133,
+      "grad_norm": 1.1016633016463047,
+      "learning_rate": 1.0565971943936826e-05,
+      "loss": 0.6501,
+      "step": 655
+    },
+    {
+      "epoch": 0.49848024316109424,
+      "grad_norm": 1.13210944329585,
+      "learning_rate": 1.0541389085854177e-05,
+      "loss": 0.7832,
+      "step": 656
+    },
+    {
+      "epoch": 0.4992401215805471,
+      "grad_norm": 1.06302593615317,
+      "learning_rate": 1.0516802946005059e-05,
+      "loss": 0.7029,
+      "step": 657
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.120242254263207,
+      "learning_rate": 1.0492213673424554e-05,
+      "loss": 0.7361,
+      "step": 658
+    },
+    {
+      "epoch": 0.5007598784194529,
+      "grad_norm": 1.1330827914748247,
+      "learning_rate": 1.0467621417166744e-05,
+      "loss": 0.8563,
+      "step": 659
+    },
+    {
+      "epoch": 0.5015197568389058,
+      "grad_norm": 1.3511630849170222,
+      "learning_rate": 1.0443026326303789e-05,
+      "loss": 0.7762,
+      "step": 660
+    },
+    {
+      "epoch": 0.5022796352583586,
+      "grad_norm": 1.1291585501546435,
+      "learning_rate": 1.0418428549925033e-05,
+      "loss": 0.6979,
+      "step": 661
+    },
+    {
+      "epoch": 0.5030395136778115,
+      "grad_norm": 1.0884622705560356,
+      "learning_rate": 1.0393828237136108e-05,
+      "loss": 0.7682,
+      "step": 662
+    },
+    {
+      "epoch": 0.5037993920972644,
+      "grad_norm": 1.2429830639549164,
+      "learning_rate": 1.0369225537058004e-05,
+      "loss": 0.8714,
+      "step": 663
+    },
+    {
+      "epoch": 0.5045592705167173,
+      "grad_norm": 1.0576093612851172,
+      "learning_rate": 1.0344620598826199e-05,
+      "loss": 0.6366,
+      "step": 664
+    },
+    {
+      "epoch": 0.5053191489361702,
+      "grad_norm": 1.236175622011617,
+      "learning_rate": 1.0320013571589727e-05,
+      "loss": 0.7086,
+      "step": 665
+    },
+    {
+      "epoch": 0.506079027355623,
+      "grad_norm": 1.5006644344885889,
+      "learning_rate": 1.0295404604510286e-05,
+      "loss": 0.7595,
+      "step": 666
+    },
+    {
+      "epoch": 0.506838905775076,
+      "grad_norm": 1.2767719535087934,
+      "learning_rate": 1.0270793846761347e-05,
+      "loss": 0.7759,
+      "step": 667
+    },
+    {
+      "epoch": 0.5075987841945289,
+      "grad_norm": 1.1155010522074496,
+      "learning_rate": 1.0246181447527213e-05,
+      "loss": 0.7257,
+      "step": 668
+    },
+    {
+      "epoch": 0.5083586626139818,
+      "grad_norm": 1.1552502485248597,
+      "learning_rate": 1.0221567556002154e-05,
+      "loss": 0.8051,
+      "step": 669
+    },
+    {
+      "epoch": 0.5091185410334347,
+      "grad_norm": 1.0214004297340544,
+      "learning_rate": 1.0196952321389482e-05,
+      "loss": 0.7516,
+      "step": 670
+    },
+    {
+      "epoch": 0.5098784194528876,
+      "grad_norm": 1.0781221550155105,
+      "learning_rate": 1.0172335892900645e-05,
+      "loss": 0.7074,
+      "step": 671
+    },
+    {
+      "epoch": 0.5106382978723404,
+      "grad_norm": 1.2210785973855254,
+      "learning_rate": 1.0147718419754335e-05,
+      "loss": 0.7237,
+      "step": 672
+    },
+    {
+      "epoch": 0.5113981762917933,
+      "grad_norm": 1.171152830006125,
+      "learning_rate": 1.0123100051175567e-05,
+      "loss": 0.7524,
+      "step": 673
+    },
+    {
+      "epoch": 0.5121580547112462,
+      "grad_norm": 1.2062323077371615,
+      "learning_rate": 1.0098480936394801e-05,
+      "loss": 0.6973,
+      "step": 674
+    },
+    {
+      "epoch": 0.5129179331306991,
+      "grad_norm": 1.4246802896434376,
+      "learning_rate": 1.0073861224647002e-05,
+      "loss": 0.6907,
+      "step": 675
+    },
+    {
+      "epoch": 0.513677811550152,
+      "grad_norm": 1.180418488238661,
+      "learning_rate": 1.0049241065170765e-05,
+      "loss": 0.81,
+      "step": 676
+    },
+    {
+      "epoch": 0.5144376899696048,
+      "grad_norm": 1.1737335268386981,
+      "learning_rate": 1.0024620607207393e-05,
+      "loss": 0.701,
+      "step": 677
+    },
+    {
+      "epoch": 0.5151975683890577,
+      "grad_norm": 0.9601725612238251,
+      "learning_rate": 1e-05,
+      "loss": 0.6626,
+      "step": 678
+    },
+    {
+      "epoch": 0.5159574468085106,
+      "grad_norm": 1.2102241908113591,
+      "learning_rate": 9.97537939279261e-06,
+      "loss": 0.7688,
+      "step": 679
+    },
+    {
+      "epoch": 0.5167173252279635,
+      "grad_norm": 1.3845812329157814,
+      "learning_rate": 9.950758934829242e-06,
+      "loss": 0.6194,
+      "step": 680
+    },
+    {
+      "epoch": 0.5174772036474165,
+      "grad_norm": 1.302664216243635,
+      "learning_rate": 9.926138775352998e-06,
+      "loss": 0.6957,
+      "step": 681
+    },
+    {
+      "epoch": 0.5182370820668692,
+      "grad_norm": 1.274488027701452,
+      "learning_rate": 9.901519063605204e-06,
+      "loss": 0.7888,
+      "step": 682
+    },
+    {
+      "epoch": 0.5189969604863222,
+      "grad_norm": 1.0335838940477384,
+      "learning_rate": 9.876899948824435e-06,
+      "loss": 0.7131,
+      "step": 683
+    },
+    {
+      "epoch": 0.5197568389057751,
+      "grad_norm": 1.0212845146712737,
+      "learning_rate": 9.85228158024567e-06,
+      "loss": 0.7249,
+      "step": 684
+    },
+    {
+      "epoch": 0.520516717325228,
+      "grad_norm": 1.122222342354429,
+      "learning_rate": 9.82766410709936e-06,
+      "loss": 0.7623,
+      "step": 685
+    },
+    {
+      "epoch": 0.5212765957446809,
+      "grad_norm": 1.0687245736709179,
+      "learning_rate": 9.80304767861052e-06,
+      "loss": 0.6708,
+      "step": 686
+    },
+    {
+      "epoch": 0.5220364741641338,
+      "grad_norm": 1.204374697486561,
+      "learning_rate": 9.77843244399785e-06,
+      "loss": 0.7664,
+      "step": 687
+    },
+    {
+      "epoch": 0.5227963525835866,
+      "grad_norm": 1.2736332945104485,
+      "learning_rate": 9.75381855247279e-06,
+      "loss": 0.8362,
+      "step": 688
+    },
+    {
+      "epoch": 0.5235562310030395,
+      "grad_norm": 0.9624540616080698,
+      "learning_rate": 9.729206153238658e-06,
+      "loss": 0.7137,
+      "step": 689
+    },
+    {
+      "epoch": 0.5243161094224924,
+      "grad_norm": 1.0421247849343214,
+      "learning_rate": 9.704595395489714e-06,
+      "loss": 0.7408,
+      "step": 690
+    },
+    {
+      "epoch": 0.5250759878419453,
+      "grad_norm": 1.013035300998222,
+      "learning_rate": 9.679986428410276e-06,
+      "loss": 0.5776,
+      "step": 691
+    },
+    {
+      "epoch": 0.5258358662613982,
+      "grad_norm": 1.0871447199817086,
+      "learning_rate": 9.655379401173804e-06,
+      "loss": 0.7964,
+      "step": 692
+    },
+    {
+      "epoch": 0.526595744680851,
+      "grad_norm": 1.4237651645969185,
+      "learning_rate": 9.630774462942002e-06,
+      "loss": 0.6829,
+      "step": 693
+    },
+    {
+      "epoch": 0.5273556231003039,
+      "grad_norm": 1.0855572690595994,
+      "learning_rate": 9.606171762863899e-06,
+      "loss": 0.7261,
+      "step": 694
+    },
+    {
+      "epoch": 0.5281155015197568,
+      "grad_norm": 1.2445924867674263,
+      "learning_rate": 9.581571450074969e-06,
+      "loss": 0.7748,
+      "step": 695
+    },
+    {
+      "epoch": 0.5288753799392097,
+      "grad_norm": 1.1859548928831545,
+      "learning_rate": 9.556973673696214e-06,
+      "loss": 0.6548,
+      "step": 696
+    },
+    {
+      "epoch": 0.5296352583586627,
+      "grad_norm": 1.0687423053782519,
+      "learning_rate": 9.53237858283326e-06,
+      "loss": 0.8082,
+      "step": 697
+    },
+    {
+      "epoch": 0.5303951367781155,
+      "grad_norm": 1.2799853643131716,
+      "learning_rate": 9.507786326575451e-06,
+      "loss": 0.7998,
+      "step": 698
+    },
+    {
+      "epoch": 0.5311550151975684,
+      "grad_norm": 1.0776337137073257,
+      "learning_rate": 9.483197053994948e-06,
+      "loss": 0.7741,
+      "step": 699
+    },
+    {
+      "epoch": 0.5319148936170213,
+      "grad_norm": 0.9874137067298976,
+      "learning_rate": 9.458610914145826e-06,
+      "loss": 0.7094,
+      "step": 700
+    },
+    {
+      "epoch": 0.5326747720364742,
+      "grad_norm": 1.1138458268182485,
+      "learning_rate": 9.434028056063178e-06,
+      "loss": 0.7488,
+      "step": 701
+    },
+    {
+      "epoch": 0.5334346504559271,
+      "grad_norm": 1.196985562955274,
+      "learning_rate": 9.409448628762194e-06,
+      "loss": 0.7638,
+      "step": 702
+    },
+    {
+      "epoch": 0.53419452887538,
+      "grad_norm": 1.2136898659918227,
+      "learning_rate": 9.38487278123727e-06,
+      "loss": 0.6823,
+      "step": 703
+    },
+    {
+      "epoch": 0.5349544072948328,
+      "grad_norm": 1.3892406474102839,
+      "learning_rate": 9.360300662461104e-06,
+      "loss": 0.6746,
+      "step": 704
+    },
+    {
+      "epoch": 0.5357142857142857,
+      "grad_norm": 1.1017160819969196,
+      "learning_rate": 9.335732421383794e-06,
+      "loss": 0.7706,
+      "step": 705
+    },
+    {
+      "epoch": 0.5364741641337386,
+      "grad_norm": 1.327308285762868,
+      "learning_rate": 9.311168206931925e-06,
+      "loss": 0.8612,
+      "step": 706
+    },
+    {
+      "epoch": 0.5372340425531915,
+      "grad_norm": 1.1205800863120123,
+      "learning_rate": 9.286608168007678e-06,
+      "loss": 0.6615,
+      "step": 707
+    },
+    {
+      "epoch": 0.5379939209726444,
+      "grad_norm": 1.1868064806482874,
+      "learning_rate": 9.262052453487924e-06,
+      "loss": 0.6602,
+      "step": 708
+    },
+    {
+      "epoch": 0.5387537993920972,
+      "grad_norm": 1.2255973213368228,
+      "learning_rate": 9.237501212223314e-06,
+      "loss": 0.7448,
+      "step": 709
+    },
+    {
+      "epoch": 0.5395136778115501,
+      "grad_norm": 1.0637993903719365,
+      "learning_rate": 9.212954593037396e-06,
+      "loss": 0.6991,
+      "step": 710
+    },
+    {
+      "epoch": 0.540273556231003,
+      "grad_norm": 1.3108472120873003,
+      "learning_rate": 9.18841274472569e-06,
+      "loss": 0.722,
+      "step": 711
+    },
+    {
+      "epoch": 0.541033434650456,
+      "grad_norm": 1.0000410345632875,
+      "learning_rate": 9.163875816054795e-06,
+      "loss": 0.7951,
+      "step": 712
+    },
+    {
+      "epoch": 0.5417933130699089,
+      "grad_norm": 1.0898157550563874,
+      "learning_rate": 9.139343955761493e-06,
+      "loss": 0.7302,
+      "step": 713
+    },
+    {
+      "epoch": 0.5425531914893617,
+      "grad_norm": 1.2462590031779057,
+      "learning_rate": 9.11481731255184e-06,
+      "loss": 0.8054,
+      "step": 714
+    },
+    {
+      "epoch": 0.5433130699088146,
+      "grad_norm": 1.0842623103782374,
+      "learning_rate": 9.090296035100275e-06,
+      "loss": 0.7488,
+      "step": 715
+    },
+    {
+      "epoch": 0.5440729483282675,
+      "grad_norm": 1.126828508481787,
+      "learning_rate": 9.065780272048703e-06,
+      "loss": 0.7089,
+      "step": 716
+    },
+    {
+      "epoch": 0.5448328267477204,
+      "grad_norm": 1.1161194449285288,
+      "learning_rate": 9.041270172005599e-06,
+      "loss": 0.8428,
+      "step": 717
+    },
+    {
+      "epoch": 0.5455927051671733,
+      "grad_norm": 1.117177427711117,
+      "learning_rate": 9.016765883545116e-06,
+      "loss": 0.7703,
+      "step": 718
+    },
+    {
+      "epoch": 0.5463525835866262,
+      "grad_norm": 1.0526580123128249,
+      "learning_rate": 8.992267555206185e-06,
+      "loss": 0.7697,
+      "step": 719
+    },
+    {
+      "epoch": 0.547112462006079,
+      "grad_norm": 1.144212356521607,
+      "learning_rate": 8.967775335491596e-06,
+      "loss": 0.7864,
+      "step": 720
+    },
+    {
+      "epoch": 0.5478723404255319,
+      "grad_norm": 1.080723967095971,
+      "learning_rate": 8.943289372867118e-06,
+      "loss": 0.6892,
+      "step": 721
+    },
+    {
+      "epoch": 0.5486322188449848,
+      "grad_norm": 1.0718604544768817,
+      "learning_rate": 8.918809815760585e-06,
+      "loss": 0.7212,
+      "step": 722
+    },
+    {
+      "epoch": 0.5493920972644377,
+      "grad_norm": 1.1174684303999995,
+      "learning_rate": 8.894336812561006e-06,
+      "loss": 0.7087,
+      "step": 723
+    },
+    {
+      "epoch": 0.5501519756838906,
+      "grad_norm": 1.0963600814875032,
+      "learning_rate": 8.869870511617662e-06,
+      "loss": 0.7443,
+      "step": 724
+    },
+    {
+      "epoch": 0.5509118541033434,
+      "grad_norm": 1.228940173574746,
+      "learning_rate": 8.845411061239208e-06,
+      "loss": 0.7431,
+      "step": 725
+    },
+    {
+      "epoch": 0.5516717325227963,
+      "grad_norm": 1.021217582634966,
+      "learning_rate": 8.82095860969277e-06,
+      "loss": 0.7284,
+      "step": 726
+    },
+    {
+      "epoch": 0.5524316109422492,
+      "grad_norm": 1.0842966533534397,
+      "learning_rate": 8.796513305203049e-06,
+      "loss": 0.7526,
+      "step": 727
+    },
+    {
+      "epoch": 0.5531914893617021,
+      "grad_norm": 1.3455009430552207,
+      "learning_rate": 8.772075295951416e-06,
+      "loss": 0.6505,
+      "step": 728
+    },
+    {
+      "epoch": 0.5539513677811551,
+      "grad_norm": 0.9568903741747474,
+      "learning_rate": 8.74764473007504e-06,
+      "loss": 0.6225,
+      "step": 729
+    },
+    {
+      "epoch": 0.5547112462006079,
+      "grad_norm": 1.1662620746932606,
+      "learning_rate": 8.723221755665948e-06,
+      "loss": 0.7685,
+      "step": 730
+    },
+    {
+      "epoch": 0.5554711246200608,
+      "grad_norm": 1.0179421849609032,
+      "learning_rate": 8.698806520770161e-06,
+      "loss": 0.6808,
+      "step": 731
+    },
+    {
+      "epoch": 0.5562310030395137,
+      "grad_norm": 1.0580889299423355,
+      "learning_rate": 8.674399173386779e-06,
+      "loss": 0.7488,
+      "step": 732
+    },
+    {
+      "epoch": 0.5569908814589666,
+      "grad_norm": 0.9153959677609983,
+      "learning_rate": 8.6499998614671e-06,
+      "loss": 0.6928,
+      "step": 733
+    },
+    {
+      "epoch": 0.5577507598784195,
+      "grad_norm": 1.2461673348388922,
+      "learning_rate": 8.625608732913702e-06,
+      "loss": 0.7417,
+      "step": 734
+    },
+    {
+      "epoch": 0.5585106382978723,
+      "grad_norm": 1.1843093976279093,
+      "learning_rate": 8.60122593557956e-06,
+      "loss": 0.7276,
+      "step": 735
+    },
+    {
+      "epoch": 0.5592705167173252,
+      "grad_norm": 1.1612172390993862,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.6551,
+      "step": 736
+    },
+    {
+      "epoch": 0.5600303951367781,
+      "grad_norm": 1.0345523205995302,
+      "learning_rate": 8.55248592572755e-06,
+      "loss": 0.7154,
+      "step": 737
+    },
+    {
+      "epoch": 0.560790273556231,
+      "grad_norm": 1.155661385060302,
+      "learning_rate": 8.528129008659543e-06,
+      "loss": 0.8645,
+      "step": 738
+    },
+    {
+      "epoch": 0.5615501519756839,
+      "grad_norm": 1.1374514153416948,
+      "learning_rate": 8.50378101370873e-06,
+      "loss": 0.7155,
+      "step": 739
+    },
+    {
+      "epoch": 0.5623100303951368,
+      "grad_norm": 1.1213383547786449,
+      "learning_rate": 8.479442088466612e-06,
+      "loss": 0.6853,
+      "step": 740
+    },
+    {
+      "epoch": 0.5630699088145896,
+      "grad_norm": 1.435739682577007,
+      "learning_rate": 8.455112380469728e-06,
+      "loss": 0.6802,
+      "step": 741
+    },
+    {
+      "epoch": 0.5638297872340425,
+      "grad_norm": 1.2143939462650815,
+      "learning_rate": 8.430792037198738e-06,
+      "loss": 0.6548,
+      "step": 742
+    },
+    {
+      "epoch": 0.5645896656534954,
+      "grad_norm": 0.9600860052959144,
+      "learning_rate": 8.406481206077535e-06,
+      "loss": 0.6814,
+      "step": 743
+    },
+    {
+      "epoch": 0.5653495440729484,
+      "grad_norm": 1.302452799435894,
+      "learning_rate": 8.382180034472353e-06,
+      "loss": 0.7308,
+      "step": 744
+    },
+    {
+      "epoch": 0.5661094224924013,
+      "grad_norm": 1.0653760433270474,
+      "learning_rate": 8.357888669690876e-06,
+      "loss": 0.7746,
+      "step": 745
+    },
+    {
+      "epoch": 0.5668693009118541,
+      "grad_norm": 1.3315367773834987,
+      "learning_rate": 8.333607258981328e-06,
+      "loss": 0.6738,
+      "step": 746
+    },
+    {
+      "epoch": 0.567629179331307,
+      "grad_norm": 1.2657872209400918,
+      "learning_rate": 8.309335949531609e-06,
+      "loss": 0.7687,
+      "step": 747
+    },
+    {
+      "epoch": 0.5683890577507599,
+      "grad_norm": 1.035527551805823,
+      "learning_rate": 8.285074888468385e-06,
+      "loss": 0.6656,
+      "step": 748
+    },
+    {
+      "epoch": 0.5691489361702128,
+      "grad_norm": 1.1689385914495711,
+      "learning_rate": 8.260824222856189e-06,
+      "loss": 0.7617,
+      "step": 749
+    },
+    {
+      "epoch": 0.5699088145896657,
+      "grad_norm": 1.182533572120935,
+      "learning_rate": 8.236584099696553e-06,
+      "loss": 0.6259,
+      "step": 750
+    },
+    {
+      "epoch": 0.5706686930091185,
+      "grad_norm": 1.2539832334039402,
+      "learning_rate": 8.212354665927088e-06,
+      "loss": 0.8626,
+      "step": 751
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 1.16793885360512,
+      "learning_rate": 8.188136068420627e-06,
+      "loss": 0.7743,
+      "step": 752
+    },
+    {
+      "epoch": 0.5721884498480243,
+      "grad_norm": 1.3045884813334427,
+      "learning_rate": 8.163928453984298e-06,
+      "loss": 0.6861,
+      "step": 753
+    },
+    {
+      "epoch": 0.5729483282674772,
+      "grad_norm": 1.1578334058918511,
+      "learning_rate": 8.139731969358665e-06,
+      "loss": 0.6578,
+      "step": 754
+    },
+    {
+      "epoch": 0.5737082066869301,
+      "grad_norm": 1.085006394436604,
+      "learning_rate": 8.115546761216822e-06,
+      "loss": 0.7372,
+      "step": 755
+    },
+    {
+      "epoch": 0.574468085106383,
+      "grad_norm": 1.2728905675998503,
+      "learning_rate": 8.091372976163496e-06,
+      "loss": 0.7905,
+      "step": 756
+    },
+    {
+      "epoch": 0.5752279635258358,
+      "grad_norm": 1.2577270128595894,
+      "learning_rate": 8.067210760734199e-06,
+      "loss": 0.6852,
+      "step": 757
+    },
+    {
+      "epoch": 0.5759878419452887,
+      "grad_norm": 1.091320885267308,
+      "learning_rate": 8.043060261394283e-06,
+      "loss": 0.6856,
+      "step": 758
+    },
+    {
+      "epoch": 0.5767477203647416,
+      "grad_norm": 1.0634220427111116,
+      "learning_rate": 8.01892162453809e-06,
+      "loss": 0.7451,
+      "step": 759
+    },
+    {
+      "epoch": 0.5775075987841946,
+      "grad_norm": 0.965968319145033,
+      "learning_rate": 7.994794996488055e-06,
+      "loss": 0.7295,
+      "step": 760
+    },
+    {
+      "epoch": 0.5782674772036475,
+      "grad_norm": 1.4557035005068617,
+      "learning_rate": 7.970680523493817e-06,
+      "loss": 0.8156,
+      "step": 761
+    },
+    {
+      "epoch": 0.5790273556231003,
+      "grad_norm": 1.228414568854264,
+      "learning_rate": 7.94657835173134e-06,
+      "loss": 0.7676,
+      "step": 762
+    },
+    {
+      "epoch": 0.5797872340425532,
+      "grad_norm": 1.0662627348682954,
+      "learning_rate": 7.922488627302016e-06,
+      "loss": 0.8173,
+      "step": 763
+    },
+    {
+      "epoch": 0.5805471124620061,
+      "grad_norm": 1.0943911331242442,
+      "learning_rate": 7.898411496231781e-06,
+      "loss": 0.6445,
+      "step": 764
+    },
+    {
+      "epoch": 0.581306990881459,
+      "grad_norm": 1.1845902784067897,
+      "learning_rate": 7.874347104470234e-06,
+      "loss": 0.735,
+      "step": 765
+    },
+    {
+      "epoch": 0.5820668693009119,
+      "grad_norm": 1.0037170474828196,
+      "learning_rate": 7.85029559788976e-06,
+      "loss": 0.7256,
+      "step": 766
+    },
+    {
+      "epoch": 0.5828267477203647,
+      "grad_norm": 1.1300711418023441,
+      "learning_rate": 7.82625712228463e-06,
+      "loss": 0.6534,
+      "step": 767
+    },
+    {
+      "epoch": 0.5835866261398176,
+      "grad_norm": 1.2021840303759719,
+      "learning_rate": 7.802231823370126e-06,
+      "loss": 0.7048,
+      "step": 768
+    },
+    {
+      "epoch": 0.5843465045592705,
+      "grad_norm": 1.2229196377630545,
+      "learning_rate": 7.778219846781654e-06,
+      "loss": 0.6571,
+      "step": 769
+    },
+    {
+      "epoch": 0.5851063829787234,
+      "grad_norm": 1.22428433043158,
+      "learning_rate": 7.754221338073863e-06,
+      "loss": 0.7309,
+      "step": 770
+    },
+    {
+      "epoch": 0.5858662613981763,
+      "grad_norm": 1.1636052982261818,
+      "learning_rate": 7.730236442719774e-06,
+      "loss": 0.8828,
+      "step": 771
+    },
+    {
+      "epoch": 0.5866261398176292,
+      "grad_norm": 1.0557514909654202,
+      "learning_rate": 7.706265306109872e-06,
+      "loss": 0.7323,
+      "step": 772
+    },
+    {
+      "epoch": 0.587386018237082,
+      "grad_norm": 1.128705574309213,
+      "learning_rate": 7.682308073551252e-06,
+      "loss": 0.801,
+      "step": 773
+    },
+    {
+      "epoch": 0.5881458966565349,
+      "grad_norm": 1.085263535196822,
+      "learning_rate": 7.658364890266713e-06,
+      "loss": 0.6932,
+      "step": 774
+    },
+    {
+      "epoch": 0.5889057750759878,
+      "grad_norm": 1.1571362771119302,
+      "learning_rate": 7.634435901393899e-06,
+      "loss": 0.6665,
+      "step": 775
+    },
+    {
+      "epoch": 0.5896656534954408,
+      "grad_norm": 1.1363472000581079,
+      "learning_rate": 7.6105212519844195e-06,
+      "loss": 0.8424,
+      "step": 776
+    },
+    {
+      "epoch": 0.5904255319148937,
+      "grad_norm": 1.1893971887933428,
+      "learning_rate": 7.586621087002945e-06,
+      "loss": 0.7843,
+      "step": 777
+    },
+    {
+      "epoch": 0.5911854103343465,
+      "grad_norm": 1.3545350834474912,
+      "learning_rate": 7.5627355513263545e-06,
+      "loss": 0.6648,
+      "step": 778
+    },
+    {
+      "epoch": 0.5919452887537994,
+      "grad_norm": 1.2821343466577149,
+      "learning_rate": 7.5388647897428445e-06,
+      "loss": 0.7347,
+      "step": 779
+    },
+    {
+      "epoch": 0.5927051671732523,
+      "grad_norm": 1.0665277209889865,
+      "learning_rate": 7.51500894695106e-06,
+      "loss": 0.7554,
+      "step": 780
+    },
+    {
+      "epoch": 0.5934650455927052,
+      "grad_norm": 0.9871422560187552,
+      "learning_rate": 7.491168167559208e-06,
+      "loss": 0.8435,
+      "step": 781
+    },
+    {
+      "epoch": 0.5942249240121581,
+      "grad_norm": 1.2007422072233171,
+      "learning_rate": 7.467342596084179e-06,
+      "loss": 0.6699,
+      "step": 782
+    },
+    {
+      "epoch": 0.5949848024316109,
+      "grad_norm": 1.2017986976755273,
+      "learning_rate": 7.443532376950688e-06,
+      "loss": 0.6769,
+      "step": 783
+    },
+    {
+      "epoch": 0.5957446808510638,
+      "grad_norm": 1.2069515723864184,
+      "learning_rate": 7.419737654490379e-06,
+      "loss": 0.6098,
+      "step": 784
+    },
+    {
+      "epoch": 0.5965045592705167,
+      "grad_norm": 1.213967528264772,
+      "learning_rate": 7.3959585729409664e-06,
+      "loss": 0.7866,
+      "step": 785
+    },
+    {
+      "epoch": 0.5972644376899696,
+      "grad_norm": 1.190147133494981,
+      "learning_rate": 7.372195276445349e-06,
+      "loss": 0.7867,
+      "step": 786
+    },
+    {
+      "epoch": 0.5980243161094225,
+      "grad_norm": 1.1198316602618827,
+      "learning_rate": 7.34844790905074e-06,
+      "loss": 0.6284,
+      "step": 787
+    },
+    {
+      "epoch": 0.5987841945288754,
+      "grad_norm": 1.054934350969657,
+      "learning_rate": 7.324716614707794e-06,
+      "loss": 0.7569,
+      "step": 788
+    },
+    {
+      "epoch": 0.5995440729483282,
+      "grad_norm": 1.184969480600631,
+      "learning_rate": 7.301001537269736e-06,
+      "loss": 0.6951,
+      "step": 789
+    },
+    {
+      "epoch": 0.6003039513677811,
+      "grad_norm": 1.207437517303436,
+      "learning_rate": 7.277302820491492e-06,
+      "loss": 0.6511,
+      "step": 790
+    },
+    {
+      "epoch": 0.601063829787234,
+      "grad_norm": 1.0028977572232165,
+      "learning_rate": 7.253620608028811e-06,
+      "loss": 0.6651,
+      "step": 791
+    },
+    {
+      "epoch": 0.601823708206687,
+      "grad_norm": 1.364380210991326,
+      "learning_rate": 7.229955043437391e-06,
+      "loss": 0.7429,
+      "step": 792
+    },
+    {
+      "epoch": 0.6025835866261399,
+      "grad_norm": 1.214893916268479,
+      "learning_rate": 7.206306270172019e-06,
+      "loss": 0.6863,
+      "step": 793
+    },
+    {
+      "epoch": 0.6033434650455927,
+      "grad_norm": 1.1620810745177512,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.7676,
+      "step": 794
+    },
+    {
+      "epoch": 0.6041033434650456,
+      "grad_norm": 1.3414636551310861,
+      "learning_rate": 7.1590596709287905e-06,
+      "loss": 0.8243,
+      "step": 795
+    },
+    {
+      "epoch": 0.6048632218844985,
+      "grad_norm": 1.0261374718405134,
+      "learning_rate": 7.135462131348107e-06,
+      "loss": 0.6541,
+      "step": 796
+    },
+    {
+      "epoch": 0.6056231003039514,
+      "grad_norm": 1.050953400295176,
+      "learning_rate": 7.111881955886093e-06,
+      "loss": 0.7374,
+      "step": 797
+    },
+    {
+      "epoch": 0.6063829787234043,
+      "grad_norm": 1.186612811549936,
+      "learning_rate": 7.088319287479919e-06,
+      "loss": 0.7259,
+      "step": 798
+    },
+    {
+      "epoch": 0.6071428571428571,
+      "grad_norm": 1.1878827605247304,
+      "learning_rate": 7.064774268960654e-06,
+      "loss": 0.6904,
+      "step": 799
+    },
+    {
+      "epoch": 0.60790273556231,
+      "grad_norm": 1.198071994450376,
+      "learning_rate": 7.041247043052357e-06,
+      "loss": 0.771,
+      "step": 800
+    },
+    {
+      "epoch": 0.6086626139817629,
+      "grad_norm": 1.1138795709931302,
+      "learning_rate": 7.017737752371243e-06,
+      "loss": 0.7284,
+      "step": 801
+    },
+    {
+      "epoch": 0.6094224924012158,
+      "grad_norm": 1.1239188974925398,
+      "learning_rate": 6.994246539424809e-06,
+      "loss": 0.7718,
+      "step": 802
+    },
+    {
+      "epoch": 0.6101823708206687,
+      "grad_norm": 1.0269669817827287,
+      "learning_rate": 6.970773546610958e-06,
+      "loss": 0.6545,
+      "step": 803
+    },
+    {
+      "epoch": 0.6109422492401215,
+      "grad_norm": 0.9849457179037013,
+      "learning_rate": 6.947318916217167e-06,
+      "loss": 0.7391,
+      "step": 804
+    },
+    {
+      "epoch": 0.6117021276595744,
+      "grad_norm": 1.2560459615613944,
+      "learning_rate": 6.923882790419585e-06,
+      "loss": 0.7587,
+      "step": 805
+    },
+    {
+      "epoch": 0.6124620060790273,
+      "grad_norm": 1.1244463345154598,
+      "learning_rate": 6.9004653112822e-06,
+      "loss": 0.7012,
+      "step": 806
+    },
+    {
+      "epoch": 0.6132218844984803,
+      "grad_norm": 1.2910529956660979,
+      "learning_rate": 6.87706662075597e-06,
+      "loss": 0.7811,
+      "step": 807
+    },
+    {
+      "epoch": 0.6139817629179332,
+      "grad_norm": 1.1279784294118858,
+      "learning_rate": 6.853686860677949e-06,
+      "loss": 0.7147,
+      "step": 808
+    },
+    {
+      "epoch": 0.6147416413373861,
+      "grad_norm": 1.023180421150499,
+      "learning_rate": 6.830326172770463e-06,
+      "loss": 0.7056,
+      "step": 809
+    },
+    {
+      "epoch": 0.6155015197568389,
+      "grad_norm": 1.1226610159276051,
+      "learning_rate": 6.806984698640202e-06,
+      "loss": 0.8144,
+      "step": 810
+    },
+    {
+      "epoch": 0.6162613981762918,
+      "grad_norm": 1.2161411828987057,
+      "learning_rate": 6.783662579777402e-06,
+      "loss": 0.6917,
+      "step": 811
+    },
+    {
+      "epoch": 0.6170212765957447,
+      "grad_norm": 1.1613987005065696,
+      "learning_rate": 6.760359957554964e-06,
+      "loss": 0.6466,
+      "step": 812
+    },
+    {
+      "epoch": 0.6177811550151976,
+      "grad_norm": 1.018688143415983,
+      "learning_rate": 6.737076973227614e-06,
+      "loss": 0.7506,
+      "step": 813
+    },
+    {
+      "epoch": 0.6185410334346505,
+      "grad_norm": 1.1689520435806657,
+      "learning_rate": 6.713813767931032e-06,
+      "loss": 0.6895,
+      "step": 814
+    },
+    {
+      "epoch": 0.6193009118541033,
+      "grad_norm": 1.1856195924843178,
+      "learning_rate": 6.690570482681003e-06,
+      "loss": 0.6359,
+      "step": 815
+    },
+    {
+      "epoch": 0.6200607902735562,
+      "grad_norm": 1.0650650175975844,
+      "learning_rate": 6.66734725837256e-06,
+      "loss": 0.7376,
+      "step": 816
+    },
+    {
+      "epoch": 0.6208206686930091,
+      "grad_norm": 1.1105670335666409,
+      "learning_rate": 6.6441442357791315e-06,
+      "loss": 0.6092,
+      "step": 817
+    },
+    {
+      "epoch": 0.621580547112462,
+      "grad_norm": 1.125965907106106,
+      "learning_rate": 6.620961555551697e-06,
+      "loss": 0.6722,
+      "step": 818
+    },
+    {
+      "epoch": 0.6223404255319149,
+      "grad_norm": 1.0691928517490077,
+      "learning_rate": 6.597799358217915e-06,
+      "loss": 0.5996,
+      "step": 819
+    },
+    {
+      "epoch": 0.6231003039513677,
+      "grad_norm": 1.0869838294675218,
+      "learning_rate": 6.574657784181287e-06,
+      "loss": 0.666,
+      "step": 820
+    },
+    {
+      "epoch": 0.6238601823708206,
+      "grad_norm": 1.0560312851091436,
+      "learning_rate": 6.551536973720298e-06,
+      "loss": 0.7692,
+      "step": 821
+    },
+    {
+      "epoch": 0.6246200607902735,
+      "grad_norm": 1.1435912397842571,
+      "learning_rate": 6.5284370669875706e-06,
+      "loss": 0.6902,
+      "step": 822
+    },
+    {
+      "epoch": 0.6253799392097265,
+      "grad_norm": 1.1566242840694625,
+      "learning_rate": 6.505358204009018e-06,
+      "loss": 0.6189,
+      "step": 823
+    },
+    {
+      "epoch": 0.6261398176291794,
+      "grad_norm": 1.1794303233552845,
+      "learning_rate": 6.482300524682985e-06,
+      "loss": 0.7805,
+      "step": 824
+    },
+    {
+      "epoch": 0.6268996960486323,
+      "grad_norm": 1.10223527961512,
+      "learning_rate": 6.459264168779416e-06,
+      "loss": 0.7555,
+      "step": 825
+    },
+    {
+      "epoch": 0.6276595744680851,
+      "grad_norm": 1.2528347286286639,
+      "learning_rate": 6.436249275938977e-06,
+      "loss": 0.7455,
+      "step": 826
+    },
+    {
+      "epoch": 0.628419452887538,
+      "grad_norm": 1.2248288878026101,
+      "learning_rate": 6.413255985672262e-06,
+      "loss": 0.6467,
+      "step": 827
+    },
+    {
+      "epoch": 0.6291793313069909,
+      "grad_norm": 1.1104651445554183,
+      "learning_rate": 6.390284437358889e-06,
+      "loss": 0.7305,
+      "step": 828
+    },
+    {
+      "epoch": 0.6299392097264438,
+      "grad_norm": 1.1907494972095103,
+      "learning_rate": 6.367334770246692e-06,
+      "loss": 0.6441,
+      "step": 829
+    },
+    {
+      "epoch": 0.6306990881458967,
+      "grad_norm": 1.3876439867214416,
+      "learning_rate": 6.344407123450867e-06,
+      "loss": 0.7022,
+      "step": 830
+    },
+    {
+      "epoch": 0.6314589665653495,
+      "grad_norm": 1.287098647943779,
+      "learning_rate": 6.321501635953124e-06,
+      "loss": 0.7059,
+      "step": 831
+    },
+    {
+      "epoch": 0.6322188449848024,
+      "grad_norm": 1.1581410475012297,
+      "learning_rate": 6.298618446600856e-06,
+      "loss": 0.6567,
+      "step": 832
+    },
+    {
+      "epoch": 0.6329787234042553,
+      "grad_norm": 1.4502182146441993,
+      "learning_rate": 6.275757694106286e-06,
+      "loss": 0.7023,
+      "step": 833
+    },
+    {
+      "epoch": 0.6337386018237082,
+      "grad_norm": 1.1216577138105488,
+      "learning_rate": 6.252919517045626e-06,
+      "loss": 0.7353,
+      "step": 834
+    },
+    {
+      "epoch": 0.6344984802431611,
+      "grad_norm": 1.1944810866958329,
+      "learning_rate": 6.230104053858248e-06,
+      "loss": 0.7686,
+      "step": 835
+    },
+    {
+      "epoch": 0.6352583586626139,
+      "grad_norm": 1.1299914280579517,
+      "learning_rate": 6.207311442845834e-06,
+      "loss": 0.6831,
+      "step": 836
+    },
+    {
+      "epoch": 0.6360182370820668,
+      "grad_norm": 1.1382941614712434,
+      "learning_rate": 6.1845418221715455e-06,
+      "loss": 0.79,
+      "step": 837
+    },
+    {
+      "epoch": 0.6367781155015197,
+      "grad_norm": 1.195754553541583,
+      "learning_rate": 6.161795329859184e-06,
+      "loss": 0.6401,
+      "step": 838
+    },
+    {
+      "epoch": 0.6375379939209727,
+      "grad_norm": 1.0571798668620749,
+      "learning_rate": 6.13907210379234e-06,
+      "loss": 0.6749,
+      "step": 839
+    },
+    {
+      "epoch": 0.6382978723404256,
+      "grad_norm": 1.2248375773498381,
+      "learning_rate": 6.116372281713581e-06,
+      "loss": 0.7163,
+      "step": 840
+    },
+    {
+      "epoch": 0.6390577507598785,
+      "grad_norm": 1.185301403413712,
+      "learning_rate": 6.093696001223609e-06,
+      "loss": 0.76,
+      "step": 841
+    },
+    {
+      "epoch": 0.6398176291793313,
+      "grad_norm": 1.2269450132243764,
+      "learning_rate": 6.071043399780412e-06,
+      "loss": 0.6985,
+      "step": 842
+    },
+    {
+      "epoch": 0.6405775075987842,
+      "grad_norm": 1.1859661222475697,
+      "learning_rate": 6.0484146146984475e-06,
+      "loss": 0.7265,
+      "step": 843
+    },
+    {
+      "epoch": 0.6413373860182371,
+      "grad_norm": 0.9394291150111883,
+      "learning_rate": 6.025809783147803e-06,
+      "loss": 0.6968,
+      "step": 844
+    },
+    {
+      "epoch": 0.64209726443769,
+      "grad_norm": 1.292071531499517,
+      "learning_rate": 6.00322904215336e-06,
+      "loss": 0.7807,
+      "step": 845
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 1.1305562407779628,
+      "learning_rate": 5.980672528593981e-06,
+      "loss": 0.6842,
+      "step": 846
+    },
+    {
+      "epoch": 0.6436170212765957,
+      "grad_norm": 1.3063299823402525,
+      "learning_rate": 5.95814037920166e-06,
+      "loss": 0.7281,
+      "step": 847
+    },
+    {
+      "epoch": 0.6443768996960486,
+      "grad_norm": 0.9750204069424666,
+      "learning_rate": 5.935632730560702e-06,
+      "loss": 0.7723,
+      "step": 848
+    },
+    {
+      "epoch": 0.6451367781155015,
+      "grad_norm": 1.0243836626345881,
+      "learning_rate": 5.913149719106896e-06,
+      "loss": 0.6473,
+      "step": 849
+    },
+    {
+      "epoch": 0.6458966565349544,
+      "grad_norm": 1.071468090018402,
+      "learning_rate": 5.89069148112668e-06,
+      "loss": 0.7328,
+      "step": 850
+    },
+    {
+      "epoch": 0.6466565349544073,
+      "grad_norm": 1.252162612887728,
+      "learning_rate": 5.868258152756336e-06,
+      "loss": 0.7987,
+      "step": 851
+    },
+    {
+      "epoch": 0.6474164133738601,
+      "grad_norm": 1.1027125679904999,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.6656,
+      "step": 852
+    },
+    {
+      "epoch": 0.648176291793313,
+      "grad_norm": 1.0378684860406113,
+      "learning_rate": 5.823466768634538e-06,
+      "loss": 0.7004,
+      "step": 853
+    },
+    {
+      "epoch": 0.648936170212766,
+      "grad_norm": 1.3431434778834976,
+      "learning_rate": 5.801108984397355e-06,
+      "loss": 0.6317,
+      "step": 854
+    },
+    {
+      "epoch": 0.6496960486322189,
+      "grad_norm": 0.9368845010269132,
+      "learning_rate": 5.778776652796936e-06,
+      "loss": 0.6568,
+      "step": 855
+    },
+    {
+      "epoch": 0.6504559270516718,
+      "grad_norm": 1.0269610836317398,
+      "learning_rate": 5.7564699092063345e-06,
+      "loss": 0.7142,
+      "step": 856
+    },
+    {
+      "epoch": 0.6512158054711246,
+      "grad_norm": 1.1480218899479488,
+      "learning_rate": 5.734188888843512e-06,
+      "loss": 0.6641,
+      "step": 857
+    },
+    {
+      "epoch": 0.6519756838905775,
+      "grad_norm": 1.0163519592455907,
+      "learning_rate": 5.711933726770487e-06,
+      "loss": 0.6621,
+      "step": 858
+    },
+    {
+      "epoch": 0.6527355623100304,
+      "grad_norm": 1.1105102755810814,
+      "learning_rate": 5.689704557892528e-06,
+      "loss": 0.7129,
+      "step": 859
+    },
+    {
+      "epoch": 0.6534954407294833,
+      "grad_norm": 1.0773301612733883,
+      "learning_rate": 5.667501516957365e-06,
+      "loss": 0.7075,
+      "step": 860
+    },
+    {
+      "epoch": 0.6542553191489362,
+      "grad_norm": 1.021179632469384,
+      "learning_rate": 5.645324738554321e-06,
+      "loss": 0.7549,
+      "step": 861
+    },
+    {
+      "epoch": 0.6550151975683891,
+      "grad_norm": 1.1504842543031002,
+      "learning_rate": 5.623174357113528e-06,
+      "loss": 0.7804,
+      "step": 862
+    },
+    {
+      "epoch": 0.6557750759878419,
+      "grad_norm": 1.2178677190514033,
+      "learning_rate": 5.601050506905114e-06,
+      "loss": 0.784,
+      "step": 863
+    },
+    {
+      "epoch": 0.6565349544072948,
+      "grad_norm": 1.1197155608727563,
+      "learning_rate": 5.578953322038372e-06,
+      "loss": 0.7202,
+      "step": 864
+    },
+    {
+      "epoch": 0.6572948328267477,
+      "grad_norm": 1.2098866589506239,
+      "learning_rate": 5.556882936460966e-06,
+      "loss": 0.702,
+      "step": 865
+    },
+    {
+      "epoch": 0.6580547112462006,
+      "grad_norm": 1.1204183667271879,
+      "learning_rate": 5.5348394839580986e-06,
+      "loss": 0.7055,
+      "step": 866
+    },
+    {
+      "epoch": 0.6588145896656535,
+      "grad_norm": 1.093622128522684,
+      "learning_rate": 5.51282309815172e-06,
+      "loss": 0.6023,
+      "step": 867
+    },
+    {
+      "epoch": 0.6595744680851063,
+      "grad_norm": 1.1103643197759976,
+      "learning_rate": 5.4908339124997005e-06,
+      "loss": 0.7477,
+      "step": 868
+    },
+    {
+      "epoch": 0.6603343465045592,
+      "grad_norm": 1.1612183387805102,
+      "learning_rate": 5.468872060295034e-06,
+      "loss": 0.7221,
+      "step": 869
+    },
+    {
+      "epoch": 0.6610942249240122,
+      "grad_norm": 1.1105119069016545,
+      "learning_rate": 5.446937674665034e-06,
+      "loss": 0.6597,
+      "step": 870
+    },
+    {
+      "epoch": 0.6618541033434651,
+      "grad_norm": 1.1494042117037129,
+      "learning_rate": 5.425030888570506e-06,
+      "loss": 0.701,
+      "step": 871
+    },
+    {
+      "epoch": 0.662613981762918,
+      "grad_norm": 1.2061798873302412,
+      "learning_rate": 5.403151834804951e-06,
+      "loss": 0.6631,
+      "step": 872
+    },
+    {
+      "epoch": 0.6633738601823708,
+      "grad_norm": 1.1556068599008187,
+      "learning_rate": 5.381300645993779e-06,
+      "loss": 0.7402,
+      "step": 873
+    },
+    {
+      "epoch": 0.6641337386018237,
+      "grad_norm": 1.138979966124941,
+      "learning_rate": 5.359477454593483e-06,
+      "loss": 0.7994,
+      "step": 874
+    },
+    {
+      "epoch": 0.6648936170212766,
+      "grad_norm": 1.0738647881906687,
+      "learning_rate": 5.337682392890832e-06,
+      "loss": 0.6859,
+      "step": 875
+    },
+    {
+      "epoch": 0.6656534954407295,
+      "grad_norm": 1.2662111227678947,
+      "learning_rate": 5.3159155930021e-06,
+      "loss": 0.7938,
+      "step": 876
+    },
+    {
+      "epoch": 0.6664133738601824,
+      "grad_norm": 1.3243150908230195,
+      "learning_rate": 5.294177186872227e-06,
+      "loss": 0.6169,
+      "step": 877
+    },
+    {
+      "epoch": 0.6671732522796353,
+      "grad_norm": 1.2313652148182201,
+      "learning_rate": 5.27246730627404e-06,
+      "loss": 0.7532,
+      "step": 878
+    },
+    {
+      "epoch": 0.6679331306990881,
+      "grad_norm": 1.199184492052818,
+      "learning_rate": 5.250786082807462e-06,
+      "loss": 0.8094,
+      "step": 879
+    },
+    {
+      "epoch": 0.668693009118541,
+      "grad_norm": 1.1766620461831854,
+      "learning_rate": 5.229133647898696e-06,
+      "loss": 0.8151,
+      "step": 880
+    },
+    {
+      "epoch": 0.6694528875379939,
+      "grad_norm": 1.1689636430398356,
+      "learning_rate": 5.207510132799436e-06,
+      "loss": 0.7648,
+      "step": 881
+    },
+    {
+      "epoch": 0.6702127659574468,
+      "grad_norm": 1.0004296601669387,
+      "learning_rate": 5.185915668586066e-06,
+      "loss": 0.6519,
+      "step": 882
+    },
+    {
+      "epoch": 0.6709726443768997,
+      "grad_norm": 1.0056620349400942,
+      "learning_rate": 5.164350386158881e-06,
+      "loss": 0.6794,
+      "step": 883
+    },
+    {
+      "epoch": 0.6717325227963525,
+      "grad_norm": 1.1405291924443834,
+      "learning_rate": 5.1428144162412815e-06,
+      "loss": 0.7398,
+      "step": 884
+    },
+    {
+      "epoch": 0.6724924012158054,
+      "grad_norm": 1.076546875498532,
+      "learning_rate": 5.121307889378975e-06,
+      "loss": 0.7544,
+      "step": 885
+    },
+    {
+      "epoch": 0.6732522796352584,
+      "grad_norm": 1.1985162180916387,
+      "learning_rate": 5.099830935939203e-06,
+      "loss": 0.6901,
+      "step": 886
+    },
+    {
+      "epoch": 0.6740121580547113,
+      "grad_norm": 1.1479988991791108,
+      "learning_rate": 5.078383686109927e-06,
+      "loss": 0.6769,
+      "step": 887
+    },
+    {
+      "epoch": 0.6747720364741642,
+      "grad_norm": 1.0843032147912806,
+      "learning_rate": 5.05696626989907e-06,
+      "loss": 0.7671,
+      "step": 888
+    },
+    {
+      "epoch": 0.675531914893617,
+      "grad_norm": 1.2193163798667448,
+      "learning_rate": 5.035578817133692e-06,
+      "loss": 0.7064,
+      "step": 889
+    },
+    {
+      "epoch": 0.6762917933130699,
+      "grad_norm": 1.139022165568261,
+      "learning_rate": 5.01422145745924e-06,
+      "loss": 0.6694,
+      "step": 890
+    },
+    {
+      "epoch": 0.6770516717325228,
+      "grad_norm": 1.1386427257420573,
+      "learning_rate": 4.992894320338727e-06,
+      "loss": 0.6988,
+      "step": 891
+    },
+    {
+      "epoch": 0.6778115501519757,
+      "grad_norm": 1.3423131800256485,
+      "learning_rate": 4.971597535051977e-06,
+      "loss": 0.7786,
+      "step": 892
+    },
+    {
+      "epoch": 0.6785714285714286,
+      "grad_norm": 1.3569449163170135,
+      "learning_rate": 4.9503312306948294e-06,
+      "loss": 0.6839,
+      "step": 893
+    },
+    {
+      "epoch": 0.6793313069908815,
+      "grad_norm": 1.258419000446593,
+      "learning_rate": 4.929095536178347e-06,
+      "loss": 0.7035,
+      "step": 894
+    },
+    {
+      "epoch": 0.6800911854103343,
+      "grad_norm": 1.264671663052783,
+      "learning_rate": 4.907890580228042e-06,
+      "loss": 0.72,
+      "step": 895
+    },
+    {
+      "epoch": 0.6808510638297872,
+      "grad_norm": 0.9729021624772011,
+      "learning_rate": 4.886716491383111e-06,
+      "loss": 0.7169,
+      "step": 896
+    },
+    {
+      "epoch": 0.6816109422492401,
+      "grad_norm": 1.1728477700061473,
+      "learning_rate": 4.865573397995626e-06,
+      "loss": 0.7557,
+      "step": 897
+    },
+    {
+      "epoch": 0.682370820668693,
+      "grad_norm": 1.0926411727337517,
+      "learning_rate": 4.844461428229782e-06,
+      "loss": 0.6697,
+      "step": 898
+    },
+    {
+      "epoch": 0.6831306990881459,
+      "grad_norm": 1.071162133552548,
+      "learning_rate": 4.823380710061112e-06,
+      "loss": 0.7911,
+      "step": 899
+    },
+    {
+      "epoch": 0.6838905775075987,
+      "grad_norm": 1.0062565374222423,
+      "learning_rate": 4.802331371275703e-06,
+      "loss": 0.7092,
+      "step": 900
+    },
+    {
+      "epoch": 0.6846504559270516,
+      "grad_norm": 1.1386791316794218,
+      "learning_rate": 4.7813135394694235e-06,
+      "loss": 0.6663,
+      "step": 901
+    },
+    {
+      "epoch": 0.6854103343465046,
+      "grad_norm": 0.9831080114325188,
+      "learning_rate": 4.760327342047167e-06,
+      "loss": 0.6426,
+      "step": 902
+    },
+    {
+      "epoch": 0.6861702127659575,
+      "grad_norm": 1.0602127458280242,
+      "learning_rate": 4.739372906222066e-06,
+      "loss": 0.6984,
+      "step": 903
+    },
+    {
+      "epoch": 0.6869300911854104,
+      "grad_norm": 1.0843705272627073,
+      "learning_rate": 4.718450359014713e-06,
+      "loss": 0.6831,
+      "step": 904
+    },
+    {
+      "epoch": 0.6876899696048632,
+      "grad_norm": 1.098679645048744,
+      "learning_rate": 4.697559827252398e-06,
+      "loss": 0.8123,
+      "step": 905
+    },
+    {
+      "epoch": 0.6884498480243161,
+      "grad_norm": 1.3142719602757702,
+      "learning_rate": 4.676701437568354e-06,
+      "loss": 0.7175,
+      "step": 906
+    },
+    {
+      "epoch": 0.689209726443769,
+      "grad_norm": 1.1643291238805322,
+      "learning_rate": 4.655875316400974e-06,
+      "loss": 0.6634,
+      "step": 907
+    },
+    {
+      "epoch": 0.6899696048632219,
+      "grad_norm": 1.1947809431912677,
+      "learning_rate": 4.635081589993033e-06,
+      "loss": 0.6458,
+      "step": 908
+    },
+    {
+      "epoch": 0.6907294832826748,
+      "grad_norm": 1.3230117296353991,
+      "learning_rate": 4.614320384390959e-06,
+      "loss": 0.7312,
+      "step": 909
+    },
+    {
+      "epoch": 0.6914893617021277,
+      "grad_norm": 1.1844063741159998,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.7545,
+      "step": 910
+    },
+    {
+      "epoch": 0.6922492401215805,
+      "grad_norm": 1.0646849778857395,
+      "learning_rate": 4.572896038803622e-06,
+      "loss": 0.6961,
+      "step": 911
+    },
+    {
+      "epoch": 0.6930091185410334,
+      "grad_norm": 1.0945617066849158,
+      "learning_rate": 4.552233149922483e-06,
+      "loss": 0.7279,
+      "step": 912
+    },
+    {
+      "epoch": 0.6937689969604863,
+      "grad_norm": 1.0442736101302776,
+      "learning_rate": 4.531603284053919e-06,
+      "loss": 0.6863,
+      "step": 913
+    },
+    {
+      "epoch": 0.6945288753799392,
+      "grad_norm": 1.264079733440778,
+      "learning_rate": 4.51100656625106e-06,
+      "loss": 0.7263,
+      "step": 914
+    },
+    {
+      "epoch": 0.6952887537993921,
+      "grad_norm": 1.1671458044603604,
+      "learning_rate": 4.490443121366105e-06,
+      "loss": 0.7512,
+      "step": 915
+    },
+    {
+      "epoch": 0.6960486322188449,
+      "grad_norm": 1.381619409291954,
+      "learning_rate": 4.4699130740495675e-06,
+      "loss": 0.7452,
+      "step": 916
+    },
+    {
+      "epoch": 0.6968085106382979,
+      "grad_norm": 1.1536720377711638,
+      "learning_rate": 4.4494165487495035e-06,
+      "loss": 0.6957,
+      "step": 917
+    },
+    {
+      "epoch": 0.6975683890577508,
+      "grad_norm": 1.0571111154420783,
+      "learning_rate": 4.428953669710764e-06,
+      "loss": 0.7022,
+      "step": 918
+    },
+    {
+      "epoch": 0.6983282674772037,
+      "grad_norm": 1.0389429368627123,
+      "learning_rate": 4.40852456097426e-06,
+      "loss": 0.707,
+      "step": 919
+    },
+    {
+      "epoch": 0.6990881458966566,
+      "grad_norm": 1.1875919850127112,
+      "learning_rate": 4.388129346376177e-06,
+      "loss": 0.7821,
+      "step": 920
+    },
+    {
+      "epoch": 0.6998480243161094,
+      "grad_norm": 1.2473748616867024,
+      "learning_rate": 4.367768149547256e-06,
+      "loss": 0.8524,
+      "step": 921
+    },
+    {
+      "epoch": 0.7006079027355623,
+      "grad_norm": 1.0994514001752107,
+      "learning_rate": 4.34744109391203e-06,
+      "loss": 0.69,
+      "step": 922
+    },
+    {
+      "epoch": 0.7013677811550152,
+      "grad_norm": 1.1799993750972086,
+      "learning_rate": 4.327148302688069e-06,
+      "loss": 0.6326,
+      "step": 923
+    },
+    {
+      "epoch": 0.7021276595744681,
+      "grad_norm": 1.0037350990689164,
+      "learning_rate": 4.306889898885241e-06,
+      "loss": 0.6607,
+      "step": 924
+    },
+    {
+      "epoch": 0.702887537993921,
+      "grad_norm": 1.0331513526916414,
+      "learning_rate": 4.286666005304971e-06,
+      "loss": 0.6456,
+      "step": 925
+    },
+    {
+      "epoch": 0.7036474164133738,
+      "grad_norm": 1.2114272294038748,
+      "learning_rate": 4.2664767445394965e-06,
+      "loss": 0.7048,
+      "step": 926
+    },
+    {
+      "epoch": 0.7044072948328267,
+      "grad_norm": 0.948969197372682,
+      "learning_rate": 4.2463222389711045e-06,
+      "loss": 0.7238,
+      "step": 927
+    },
+    {
+      "epoch": 0.7051671732522796,
+      "grad_norm": 1.259145840103064,
+      "learning_rate": 4.226202610771419e-06,
+      "loss": 0.6822,
+      "step": 928
+    },
+    {
+      "epoch": 0.7059270516717325,
+      "grad_norm": 1.147743823697958,
+      "learning_rate": 4.206117981900636e-06,
+      "loss": 0.6834,
+      "step": 929
+    },
+    {
+      "epoch": 0.7066869300911854,
+      "grad_norm": 1.3613688973426528,
+      "learning_rate": 4.186068474106803e-06,
+      "loss": 0.6058,
+      "step": 930
+    },
+    {
+      "epoch": 0.7074468085106383,
+      "grad_norm": 1.0641233144969966,
+      "learning_rate": 4.16605420892506e-06,
+      "loss": 0.6037,
+      "step": 931
+    },
+    {
+      "epoch": 0.7082066869300911,
+      "grad_norm": 1.4404076348677173,
+      "learning_rate": 4.146075307676932e-06,
+      "loss": 0.7231,
+      "step": 932
+    },
+    {
+      "epoch": 0.708966565349544,
+      "grad_norm": 1.1265784012494935,
+      "learning_rate": 4.126131891469561e-06,
+      "loss": 0.7518,
+      "step": 933
+    },
+    {
+      "epoch": 0.709726443768997,
+      "grad_norm": 1.1259639811618485,
+      "learning_rate": 4.10622408119499e-06,
+      "loss": 0.7479,
+      "step": 934
+    },
+    {
+      "epoch": 0.7104863221884499,
+      "grad_norm": 1.12906985920995,
+      "learning_rate": 4.086351997529441e-06,
+      "loss": 0.7117,
+      "step": 935
+    },
+    {
+      "epoch": 0.7112462006079028,
+      "grad_norm": 1.1141762682174512,
+      "learning_rate": 4.0665157609325565e-06,
+      "loss": 0.6223,
+      "step": 936
+    },
+    {
+      "epoch": 0.7120060790273556,
+      "grad_norm": 0.9843606839695441,
+      "learning_rate": 4.0467154916466835e-06,
+      "loss": 0.6804,
+      "step": 937
+    },
+    {
+      "epoch": 0.7127659574468085,
+      "grad_norm": 1.0791368453161987,
+      "learning_rate": 4.026951309696152e-06,
+      "loss": 0.8217,
+      "step": 938
+    },
+    {
+      "epoch": 0.7135258358662614,
+      "grad_norm": 1.17924828361289,
+      "learning_rate": 4.007223334886531e-06,
+      "loss": 0.6319,
+      "step": 939
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 1.086662042991453,
+      "learning_rate": 3.98753168680392e-06,
+      "loss": 0.7768,
+      "step": 940
+    },
+    {
+      "epoch": 0.7150455927051672,
+      "grad_norm": 1.1637104562064116,
+      "learning_rate": 3.967876484814202e-06,
+      "loss": 0.699,
+      "step": 941
+    },
+    {
+      "epoch": 0.71580547112462,
+      "grad_norm": 1.1821074776029854,
+      "learning_rate": 3.948257848062351e-06,
+      "loss": 0.7243,
+      "step": 942
+    },
+    {
+      "epoch": 0.7165653495440729,
+      "grad_norm": 1.2758482975392278,
+      "learning_rate": 3.92867589547167e-06,
+      "loss": 0.6933,
+      "step": 943
+    },
+    {
+      "epoch": 0.7173252279635258,
+      "grad_norm": 1.136673931415126,
+      "learning_rate": 3.909130745743108e-06,
+      "loss": 0.6787,
+      "step": 944
+    },
+    {
+      "epoch": 0.7180851063829787,
+      "grad_norm": 1.2505316932380202,
+      "learning_rate": 3.889622517354523e-06,
+      "loss": 0.7131,
+      "step": 945
+    },
+    {
+      "epoch": 0.7188449848024316,
+      "grad_norm": 1.1532069822589315,
+      "learning_rate": 3.870151328559956e-06,
+      "loss": 0.7163,
+      "step": 946
+    },
+    {
+      "epoch": 0.7196048632218845,
+      "grad_norm": 1.3010144592978872,
+      "learning_rate": 3.850717297388926e-06,
+      "loss": 0.6895,
+      "step": 947
+    },
+    {
+      "epoch": 0.7203647416413373,
+      "grad_norm": 1.2213309503213339,
+      "learning_rate": 3.831320541645717e-06,
+      "loss": 0.7386,
+      "step": 948
+    },
+    {
+      "epoch": 0.7211246200607903,
+      "grad_norm": 1.1249828865944345,
+      "learning_rate": 3.8119611789086576e-06,
+      "loss": 0.7562,
+      "step": 949
+    },
+    {
+      "epoch": 0.7218844984802432,
+      "grad_norm": 1.112085150276745,
+      "learning_rate": 3.7926393265294016e-06,
+      "loss": 0.7472,
+      "step": 950
+    },
+    {
+      "epoch": 0.7226443768996961,
+      "grad_norm": 1.1691517022721343,
+      "learning_rate": 3.773355101632236e-06,
+      "loss": 0.6366,
+      "step": 951
+    },
+    {
+      "epoch": 0.723404255319149,
+      "grad_norm": 1.065081114749969,
+      "learning_rate": 3.7541086211133504e-06,
+      "loss": 0.7091,
+      "step": 952
+    },
+    {
+      "epoch": 0.7241641337386018,
+      "grad_norm": 1.3061023932315732,
+      "learning_rate": 3.734900001640135e-06,
+      "loss": 0.7162,
+      "step": 953
+    },
+    {
+      "epoch": 0.7249240121580547,
+      "grad_norm": 1.309023206323555,
+      "learning_rate": 3.7157293596504863e-06,
+      "loss": 0.6626,
+      "step": 954
+    },
+    {
+      "epoch": 0.7256838905775076,
+      "grad_norm": 1.1235930459888246,
+      "learning_rate": 3.696596811352087e-06,
+      "loss": 0.6648,
+      "step": 955
+    },
+    {
+      "epoch": 0.7264437689969605,
+      "grad_norm": 1.1237782636304048,
+      "learning_rate": 3.677502472721699e-06,
+      "loss": 0.6832,
+      "step": 956
+    },
+    {
+      "epoch": 0.7272036474164134,
+      "grad_norm": 1.0942128565760725,
+      "learning_rate": 3.6584464595044713e-06,
+      "loss": 0.7049,
+      "step": 957
+    },
+    {
+      "epoch": 0.7279635258358662,
+      "grad_norm": 1.1946352330560444,
+      "learning_rate": 3.6394288872132335e-06,
+      "loss": 0.6031,
+      "step": 958
+    },
+    {
+      "epoch": 0.7287234042553191,
+      "grad_norm": 0.9925979700840246,
+      "learning_rate": 3.6204498711278014e-06,
+      "loss": 0.7806,
+      "step": 959
+    },
+    {
+      "epoch": 0.729483282674772,
+      "grad_norm": 1.1783905654043534,
+      "learning_rate": 3.6015095262942578e-06,
+      "loss": 0.7694,
+      "step": 960
+    },
+    {
+      "epoch": 0.7302431610942249,
+      "grad_norm": 1.1285297640686531,
+      "learning_rate": 3.5826079675242897e-06,
+      "loss": 0.7035,
+      "step": 961
+    },
+    {
+      "epoch": 0.7310030395136778,
+      "grad_norm": 1.2766209090145464,
+      "learning_rate": 3.563745309394452e-06,
+      "loss": 0.8306,
+      "step": 962
+    },
+    {
+      "epoch": 0.7317629179331308,
+      "grad_norm": 1.2802069631147242,
+      "learning_rate": 3.54492166624551e-06,
+      "loss": 0.6821,
+      "step": 963
+    },
+    {
+      "epoch": 0.7325227963525835,
+      "grad_norm": 1.4528178238422362,
+      "learning_rate": 3.5261371521817247e-06,
+      "loss": 0.8057,
+      "step": 964
+    },
+    {
+      "epoch": 0.7332826747720365,
+      "grad_norm": 1.1360673012907418,
+      "learning_rate": 3.507391881070161e-06,
+      "loss": 0.6928,
+      "step": 965
+    },
+    {
+      "epoch": 0.7340425531914894,
+      "grad_norm": 1.0857177186934868,
+      "learning_rate": 3.4886859665400075e-06,
+      "loss": 0.6418,
+      "step": 966
+    },
+    {
+      "epoch": 0.7348024316109423,
+      "grad_norm": 1.2429148105905763,
+      "learning_rate": 3.470019521981882e-06,
+      "loss": 0.7205,
+      "step": 967
+    },
+    {
+      "epoch": 0.7355623100303952,
+      "grad_norm": 1.1381324753940012,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.6926,
+      "step": 968
+    },
+    {
+      "epoch": 0.736322188449848,
+      "grad_norm": 1.2286178986824534,
+      "learning_rate": 3.4328054951472267e-06,
+      "loss": 0.6778,
+      "step": 969
+    },
+    {
+      "epoch": 0.7370820668693009,
+      "grad_norm": 1.249604798766476,
+      "learning_rate": 3.4142581384528984e-06,
+      "loss": 0.6991,
+      "step": 970
+    },
+    {
+      "epoch": 0.7378419452887538,
+      "grad_norm": 1.2152282939220294,
+      "learning_rate": 3.395750702893651e-06,
+      "loss": 0.8301,
+      "step": 971
+    },
+    {
+      "epoch": 0.7386018237082067,
+      "grad_norm": 1.1292510512641436,
+      "learning_rate": 3.377283300656967e-06,
+      "loss": 0.6164,
+      "step": 972
+    },
+    {
+      "epoch": 0.7393617021276596,
+      "grad_norm": 1.129932752453749,
+      "learning_rate": 3.358856043687666e-06,
+      "loss": 0.667,
+      "step": 973
+    },
+    {
+      "epoch": 0.7401215805471124,
+      "grad_norm": 1.2040509606498078,
+      "learning_rate": 3.340469043687213e-06,
+      "loss": 0.7359,
+      "step": 974
+    },
+    {
+      "epoch": 0.7408814589665653,
+      "grad_norm": 1.0966554016954166,
+      "learning_rate": 3.322122412113047e-06,
+      "loss": 0.6991,
+      "step": 975
+    },
+    {
+      "epoch": 0.7416413373860182,
+      "grad_norm": 1.2752114699698873,
+      "learning_rate": 3.3038162601778944e-06,
+      "loss": 0.7226,
+      "step": 976
+    },
+    {
+      "epoch": 0.7424012158054711,
+      "grad_norm": 1.1973005258559652,
+      "learning_rate": 3.285550698849117e-06,
+      "loss": 0.7434,
+      "step": 977
+    },
+    {
+      "epoch": 0.743161094224924,
+      "grad_norm": 1.2115166712763157,
+      "learning_rate": 3.2673258388480235e-06,
+      "loss": 0.6354,
+      "step": 978
+    },
+    {
+      "epoch": 0.743920972644377,
+      "grad_norm": 1.2342273069639336,
+      "learning_rate": 3.2491417906491916e-06,
+      "loss": 0.6306,
+      "step": 979
+    },
+    {
+      "epoch": 0.7446808510638298,
+      "grad_norm": 1.2359619179075527,
+      "learning_rate": 3.230998664479823e-06,
+      "loss": 0.6276,
+      "step": 980
+    },
+    {
+      "epoch": 0.7454407294832827,
+      "grad_norm": 1.0415850804754119,
+      "learning_rate": 3.212896570319045e-06,
+      "loss": 0.6845,
+      "step": 981
+    },
+    {
+      "epoch": 0.7462006079027356,
+      "grad_norm": 1.1802739171012435,
+      "learning_rate": 3.194835617897273e-06,
+      "loss": 0.622,
+      "step": 982
+    },
+    {
+      "epoch": 0.7469604863221885,
+      "grad_norm": 1.2439907155583805,
+      "learning_rate": 3.176815916695518e-06,
+      "loss": 0.7096,
+      "step": 983
+    },
+    {
+      "epoch": 0.7477203647416414,
+      "grad_norm": 1.147914771751723,
+      "learning_rate": 3.158837575944751e-06,
+      "loss": 0.7258,
+      "step": 984
+    },
+    {
+      "epoch": 0.7484802431610942,
+      "grad_norm": 1.2978049727772043,
+      "learning_rate": 3.1409007046252114e-06,
+      "loss": 0.6771,
+      "step": 985
+    },
+    {
+      "epoch": 0.7492401215805471,
+      "grad_norm": 1.1499766562757003,
+      "learning_rate": 3.123005411465766e-06,
+      "loss": 0.701,
+      "step": 986
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0693963917103544,
+      "learning_rate": 3.105151804943256e-06,
+      "loss": 0.7043,
+      "step": 987
+    },
+    {
+      "epoch": 0.7507598784194529,
+      "grad_norm": 1.2305140175608322,
+      "learning_rate": 3.087339993281816e-06,
+      "loss": 0.6566,
+      "step": 988
+    },
+    {
+      "epoch": 0.7515197568389058,
+      "grad_norm": 1.003867377853084,
+      "learning_rate": 3.06957008445223e-06,
+      "loss": 0.6741,
+      "step": 989
+    },
+    {
+      "epoch": 0.7522796352583586,
+      "grad_norm": 1.055359001129926,
+      "learning_rate": 3.0518421861712845e-06,
+      "loss": 0.7424,
+      "step": 990
+    },
+    {
+      "epoch": 0.7530395136778115,
+      "grad_norm": 1.0489625087030288,
+      "learning_rate": 3.0341564059011086e-06,
+      "loss": 0.6396,
+      "step": 991
+    },
+    {
+      "epoch": 0.7537993920972644,
+      "grad_norm": 1.1893706178742733,
+      "learning_rate": 3.0165128508485166e-06,
+      "loss": 0.6847,
+      "step": 992
+    },
+    {
+      "epoch": 0.7545592705167173,
+      "grad_norm": 1.0135298428599318,
+      "learning_rate": 2.9989116279643637e-06,
+      "loss": 0.5938,
+      "step": 993
+    },
+    {
+      "epoch": 0.7553191489361702,
+      "grad_norm": 1.1610621297187727,
+      "learning_rate": 2.9813528439429074e-06,
+      "loss": 0.7356,
+      "step": 994
+    },
+    {
+      "epoch": 0.756079027355623,
+      "grad_norm": 1.306043230804643,
+      "learning_rate": 2.9638366052211387e-06,
+      "loss": 0.7352,
+      "step": 995
+    },
+    {
+      "epoch": 0.756838905775076,
+      "grad_norm": 1.3878230222334922,
+      "learning_rate": 2.946363017978159e-06,
+      "loss": 0.7463,
+      "step": 996
+    },
+    {
+      "epoch": 0.7575987841945289,
+      "grad_norm": 1.2866088794863302,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.7971,
+      "step": 997
+    },
+    {
+      "epoch": 0.7583586626139818,
+      "grad_norm": 1.2283459458688568,
+      "learning_rate": 2.911544221351608e-06,
+      "loss": 0.6631,
+      "step": 998
+    },
+    {
+      "epoch": 0.7591185410334347,
+      "grad_norm": 1.024583863335167,
+      "learning_rate": 2.894199223030948e-06,
+      "loss": 0.7676,
+      "step": 999
+    },
+    {
+      "epoch": 0.7598784194528876,
+      "grad_norm": 1.1118576213588252,
+      "learning_rate": 2.87689729831363e-06,
+      "loss": 0.6562,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7606382978723404,
+      "grad_norm": 0.9923751398641592,
+      "learning_rate": 2.8596385520796365e-06,
+      "loss": 0.6727,
+      "step": 1001
+    },
+    {
+      "epoch": 0.7613981762917933,
+      "grad_norm": 1.1727199316684949,
+      "learning_rate": 2.8424230889472047e-06,
+      "loss": 0.6297,
+      "step": 1002
+    },
+    {
+      "epoch": 0.7621580547112462,
+      "grad_norm": 1.2294185210188011,
+      "learning_rate": 2.825251013272212e-06,
+      "loss": 0.8385,
+      "step": 1003
+    },
+    {
+      "epoch": 0.7629179331306991,
+      "grad_norm": 1.2168106027942747,
+      "learning_rate": 2.8081224291475216e-06,
+      "loss": 0.762,
+      "step": 1004
+    },
+    {
+      "epoch": 0.763677811550152,
+      "grad_norm": 1.2253817236474687,
+      "learning_rate": 2.7910374404023634e-06,
+      "loss": 0.6375,
+      "step": 1005
+    },
+    {
+      "epoch": 0.7644376899696048,
+      "grad_norm": 1.0665312801720293,
+      "learning_rate": 2.7739961506017075e-06,
+      "loss": 0.7151,
+      "step": 1006
+    },
+    {
+      "epoch": 0.7651975683890577,
+      "grad_norm": 1.2265668347679475,
+      "learning_rate": 2.7569986630456334e-06,
+      "loss": 0.6618,
+      "step": 1007
+    },
+    {
+      "epoch": 0.7659574468085106,
+      "grad_norm": 0.9884700225329818,
+      "learning_rate": 2.740045080768694e-06,
+      "loss": 0.6926,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7667173252279635,
+      "grad_norm": 1.1992772744751734,
+      "learning_rate": 2.7231355065392996e-06,
+      "loss": 0.7414,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7674772036474165,
+      "grad_norm": 1.1104745481751945,
+      "learning_rate": 2.706270042859108e-06,
+      "loss": 0.7639,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7682370820668692,
+      "grad_norm": 0.9965536525277243,
+      "learning_rate": 2.689448791962377e-06,
+      "loss": 0.8284,
+      "step": 1011
+    },
+    {
+      "epoch": 0.7689969604863222,
+      "grad_norm": 1.1678112956685116,
+      "learning_rate": 2.672671855815355e-06,
+      "loss": 0.7828,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7697568389057751,
+      "grad_norm": 1.3156802308059738,
+      "learning_rate": 2.6559393361156804e-06,
+      "loss": 0.6398,
+      "step": 1013
+    },
+    {
+      "epoch": 0.770516717325228,
+      "grad_norm": 1.1401901016167089,
+      "learning_rate": 2.6392513342917327e-06,
+      "loss": 0.6525,
+      "step": 1014
+    },
+    {
+      "epoch": 0.7712765957446809,
+      "grad_norm": 1.0560539135353917,
+      "learning_rate": 2.6226079515020507e-06,
+      "loss": 0.7281,
+      "step": 1015
+    },
+    {
+      "epoch": 0.7720364741641338,
+      "grad_norm": 1.1909614312462784,
+      "learning_rate": 2.606009288634689e-06,
+      "loss": 0.6785,
+      "step": 1016
+    },
+    {
+      "epoch": 0.7727963525835866,
+      "grad_norm": 1.0645811009555441,
+      "learning_rate": 2.5894554463066356e-06,
+      "loss": 0.8196,
+      "step": 1017
+    },
+    {
+      "epoch": 0.7735562310030395,
+      "grad_norm": 1.0744237761868103,
+      "learning_rate": 2.5729465248631733e-06,
+      "loss": 0.7089,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7743161094224924,
+      "grad_norm": 1.1501549180149673,
+      "learning_rate": 2.5564826243772965e-06,
+      "loss": 0.7186,
+      "step": 1019
+    },
+    {
+      "epoch": 0.7750759878419453,
+      "grad_norm": 1.090718993467275,
+      "learning_rate": 2.540063844649092e-06,
+      "loss": 0.6967,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7758358662613982,
+      "grad_norm": 1.262645687632852,
+      "learning_rate": 2.5236902852051314e-06,
+      "loss": 0.6468,
+      "step": 1021
+    },
+    {
+      "epoch": 0.776595744680851,
+      "grad_norm": 0.9878664141286467,
+      "learning_rate": 2.5073620452978708e-06,
+      "loss": 0.7271,
+      "step": 1022
+    },
+    {
+      "epoch": 0.7773556231003039,
+      "grad_norm": 1.2611651073752586,
+      "learning_rate": 2.4910792239050575e-06,
+      "loss": 0.6891,
+      "step": 1023
+    },
+    {
+      "epoch": 0.7781155015197568,
+      "grad_norm": 1.3770427550654896,
+      "learning_rate": 2.474841919729122e-06,
+      "loss": 0.6701,
+      "step": 1024
+    },
+    {
+      "epoch": 0.7788753799392097,
+      "grad_norm": 1.1740807071337804,
+      "learning_rate": 2.458650231196572e-06,
+      "loss": 0.6969,
+      "step": 1025
+    },
+    {
+      "epoch": 0.7796352583586627,
+      "grad_norm": 1.15327703282196,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.7951,
+      "step": 1026
+    },
+    {
+      "epoch": 0.7803951367781155,
+      "grad_norm": 1.0976455733746449,
+      "learning_rate": 2.4264040933845535e-06,
+      "loss": 0.6239,
+      "step": 1027
+    },
+    {
+      "epoch": 0.7811550151975684,
+      "grad_norm": 1.1362177795153996,
+      "learning_rate": 2.410349839573175e-06,
+      "loss": 0.71,
+      "step": 1028
+    },
+    {
+      "epoch": 0.7819148936170213,
+      "grad_norm": 1.2026870438152502,
+      "learning_rate": 2.3943415923401923e-06,
+      "loss": 0.6649,
+      "step": 1029
+    },
+    {
+      "epoch": 0.7826747720364742,
+      "grad_norm": 1.2052658837448995,
+      "learning_rate": 2.3783794487236367e-06,
+      "loss": 0.7296,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7834346504559271,
+      "grad_norm": 1.1281897537962795,
+      "learning_rate": 2.3624635054820633e-06,
+      "loss": 0.7188,
+      "step": 1031
+    },
+    {
+      "epoch": 0.78419452887538,
+      "grad_norm": 1.0897801958842848,
+      "learning_rate": 2.346593859093974e-06,
+      "loss": 0.7076,
+      "step": 1032
+    },
+    {
+      "epoch": 0.7849544072948328,
+      "grad_norm": 1.0640259314164149,
+      "learning_rate": 2.3307706057572354e-06,
+      "loss": 0.7441,
+      "step": 1033
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 1.0344057270478118,
+      "learning_rate": 2.3149938413884886e-06,
+      "loss": 0.71,
+      "step": 1034
+    },
+    {
+      "epoch": 0.7864741641337386,
+      "grad_norm": 1.258022110174666,
+      "learning_rate": 2.299263661622566e-06,
+      "loss": 0.7375,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7872340425531915,
+      "grad_norm": 1.1067641335949387,
+      "learning_rate": 2.2835801618119247e-06,
+      "loss": 0.7371,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7879939209726444,
+      "grad_norm": 1.2002173951393722,
+      "learning_rate": 2.2679434370260457e-06,
+      "loss": 0.6505,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7887537993920972,
+      "grad_norm": 1.133912103250985,
+      "learning_rate": 2.2523535820508844e-06,
+      "loss": 0.7709,
+      "step": 1038
+    },
+    {
+      "epoch": 0.7895136778115501,
+      "grad_norm": 1.1460622281855655,
+      "learning_rate": 2.2368106913882814e-06,
+      "loss": 0.718,
+      "step": 1039
+    },
+    {
+      "epoch": 0.790273556231003,
+      "grad_norm": 1.3231479511702298,
+      "learning_rate": 2.2213148592553847e-06,
+      "loss": 0.6763,
+      "step": 1040
+    },
+    {
+      "epoch": 0.791033434650456,
+      "grad_norm": 1.0916282763634857,
+      "learning_rate": 2.205866179584084e-06,
+      "loss": 0.6576,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7917933130699089,
+      "grad_norm": 1.1640565182207296,
+      "learning_rate": 2.190464746020452e-06,
+      "loss": 0.7421,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7925531914893617,
+      "grad_norm": 1.0117814584644247,
+      "learning_rate": 2.175110651924165e-06,
+      "loss": 0.6422,
+      "step": 1043
+    },
+    {
+      "epoch": 0.7933130699088146,
+      "grad_norm": 1.0485542457299897,
+      "learning_rate": 2.159803990367931e-06,
+      "loss": 0.7094,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7940729483282675,
+      "grad_norm": 1.2116167735512808,
+      "learning_rate": 2.1445448541369396e-06,
+      "loss": 0.7358,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7948328267477204,
+      "grad_norm": 1.1462035059040563,
+      "learning_rate": 2.1293333357282954e-06,
+      "loss": 0.7469,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7955927051671733,
+      "grad_norm": 1.1941863771000016,
+      "learning_rate": 2.1141695273504503e-06,
+      "loss": 0.7137,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7963525835866262,
+      "grad_norm": 1.312889254090232,
+      "learning_rate": 2.0990535209226547e-06,
+      "loss": 0.8082,
+      "step": 1048
+    },
+    {
+      "epoch": 0.797112462006079,
+      "grad_norm": 1.072468781454342,
+      "learning_rate": 2.083985408074396e-06,
+      "loss": 0.7231,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7978723404255319,
+      "grad_norm": 1.1860345229851978,
+      "learning_rate": 2.068965280144837e-06,
+      "loss": 0.672,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7986322188449848,
+      "grad_norm": 1.2823965807740199,
+      "learning_rate": 2.0539932281822685e-06,
+      "loss": 0.7493,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7993920972644377,
+      "grad_norm": 1.1486470335368848,
+      "learning_rate": 2.0390693429435626e-06,
+      "loss": 0.6758,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8001519756838906,
+      "grad_norm": 1.2163530985688338,
+      "learning_rate": 2.024193714893614e-06,
+      "loss": 0.7542,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8009118541033434,
+      "grad_norm": 1.1418608494404552,
+      "learning_rate": 2.0093664342047903e-06,
+      "loss": 0.7042,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8016717325227963,
+      "grad_norm": 1.091740118029224,
+      "learning_rate": 1.994587590756397e-06,
+      "loss": 0.6547,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8024316109422492,
+      "grad_norm": 1.0548786543265325,
+      "learning_rate": 1.979857274134115e-06,
+      "loss": 0.6756,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8031914893617021,
+      "grad_norm": 1.1320023087914366,
+      "learning_rate": 1.9651755736294785e-06,
+      "loss": 0.6682,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8039513677811551,
+      "grad_norm": 1.2189572447533126,
+      "learning_rate": 1.9505425782393117e-06,
+      "loss": 0.7586,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8047112462006079,
+      "grad_norm": 1.0943551201130357,
+      "learning_rate": 1.9359583766652135e-06,
+      "loss": 0.714,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8054711246200608,
+      "grad_norm": 0.9997249186870598,
+      "learning_rate": 1.9214230573129944e-06,
+      "loss": 0.7033,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8062310030395137,
+      "grad_norm": 1.0062498233336101,
+      "learning_rate": 1.9069367082921542e-06,
+      "loss": 0.6967,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8069908814589666,
+      "grad_norm": 1.2074679469839953,
+      "learning_rate": 1.892499417415362e-06,
+      "loss": 0.6098,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8077507598784195,
+      "grad_norm": 0.974094819682152,
+      "learning_rate": 1.87811127219789e-06,
+      "loss": 0.7002,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8085106382978723,
+      "grad_norm": 0.999777172254122,
+      "learning_rate": 1.8637723598571078e-06,
+      "loss": 0.6874,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8092705167173252,
+      "grad_norm": 1.3233453645588935,
+      "learning_rate": 1.849482767311953e-06,
+      "loss": 0.733,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8100303951367781,
+      "grad_norm": 1.093917921272683,
+      "learning_rate": 1.8352425811823893e-06,
+      "loss": 0.5386,
+      "step": 1066
+    },
+    {
+      "epoch": 0.810790273556231,
+      "grad_norm": 1.203921911962039,
+      "learning_rate": 1.8210518877889016e-06,
+      "loss": 0.6852,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8115501519756839,
+      "grad_norm": 1.2387311631723403,
+      "learning_rate": 1.8069107731519507e-06,
+      "loss": 0.7721,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8123100303951368,
+      "grad_norm": 1.199552627375313,
+      "learning_rate": 1.7928193229914747e-06,
+      "loss": 0.6982,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8130699088145896,
+      "grad_norm": 0.9963886859919631,
+      "learning_rate": 1.7787776227263464e-06,
+      "loss": 0.5456,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8138297872340425,
+      "grad_norm": 1.3545208480064566,
+      "learning_rate": 1.7647857574738759e-06,
+      "loss": 0.7682,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8145896656534954,
+      "grad_norm": 1.2208367567982428,
+      "learning_rate": 1.7508438120492864e-06,
+      "loss": 0.5537,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8153495440729484,
+      "grad_norm": 1.1165843778790494,
+      "learning_rate": 1.7369518709651923e-06,
+      "loss": 0.7316,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8161094224924013,
+      "grad_norm": 1.1409271275287363,
+      "learning_rate": 1.7231100184310955e-06,
+      "loss": 0.6392,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8168693009118541,
+      "grad_norm": 1.0884083854734414,
+      "learning_rate": 1.7093183383528776e-06,
+      "loss": 0.6117,
+      "step": 1075
+    },
+    {
+      "epoch": 0.817629179331307,
+      "grad_norm": 1.068969102945993,
+      "learning_rate": 1.6955769143322898e-06,
+      "loss": 0.4683,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8183890577507599,
+      "grad_norm": 0.9831886853581654,
+      "learning_rate": 1.681885829666432e-06,
+      "loss": 0.7013,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8191489361702128,
+      "grad_norm": 1.051184380106491,
+      "learning_rate": 1.6682451673472734e-06,
+      "loss": 0.6305,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8199088145896657,
+      "grad_norm": 1.1223495353706434,
+      "learning_rate": 1.6546550100611237e-06,
+      "loss": 0.7346,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8206686930091185,
+      "grad_norm": 1.1441585867010564,
+      "learning_rate": 1.641115440188148e-06,
+      "loss": 0.6982,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8214285714285714,
+      "grad_norm": 1.2192432945723835,
+      "learning_rate": 1.6276265398018642e-06,
+      "loss": 0.6753,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8221884498480243,
+      "grad_norm": 1.415282448423663,
+      "learning_rate": 1.6141883906686484e-06,
+      "loss": 0.7307,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8229483282674772,
+      "grad_norm": 1.2236317595223443,
+      "learning_rate": 1.6008010742472257e-06,
+      "loss": 0.7669,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8237082066869301,
+      "grad_norm": 1.0133830637305623,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.7271,
+      "step": 1084
+    },
+    {
+      "epoch": 0.824468085106383,
+      "grad_norm": 1.1343068460941579,
+      "learning_rate": 1.5741792638335096e-06,
+      "loss": 0.6954,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8252279635258358,
+      "grad_norm": 1.073228514372066,
+      "learning_rate": 1.5609449312160364e-06,
+      "loss": 0.584,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8259878419452887,
+      "grad_norm": 1.2155988715843964,
+      "learning_rate": 1.5477617540590128e-06,
+      "loss": 0.6985,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8267477203647416,
+      "grad_norm": 1.0626067139609812,
+      "learning_rate": 1.5346298122755932e-06,
+      "loss": 0.6585,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8275075987841946,
+      "grad_norm": 1.147707100709591,
+      "learning_rate": 1.5215491854683529e-06,
+      "loss": 0.7551,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8282674772036475,
+      "grad_norm": 1.3468869928677656,
+      "learning_rate": 1.5085199529288097e-06,
+      "loss": 0.6943,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8290273556231003,
+      "grad_norm": 1.2096274012114665,
+      "learning_rate": 1.4955421936369452e-06,
+      "loss": 0.6694,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8297872340425532,
+      "grad_norm": 1.2410948107008724,
+      "learning_rate": 1.4826159862607182e-06,
+      "loss": 0.6617,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8305471124620061,
+      "grad_norm": 1.155591538053921,
+      "learning_rate": 1.4697414091555918e-06,
+      "loss": 0.7733,
+      "step": 1093
+    },
+    {
+      "epoch": 0.831306990881459,
+      "grad_norm": 1.0068055977995527,
+      "learning_rate": 1.456918540364065e-06,
+      "loss": 0.7311,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8320668693009119,
+      "grad_norm": 1.3257082207534072,
+      "learning_rate": 1.4441474576151915e-06,
+      "loss": 0.652,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8328267477203647,
+      "grad_norm": 1.2197567822706545,
+      "learning_rate": 1.4314282383241097e-06,
+      "loss": 0.6927,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8335866261398176,
+      "grad_norm": 1.2881915459202287,
+      "learning_rate": 1.4187609595915697e-06,
+      "loss": 0.7052,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8343465045592705,
+      "grad_norm": 1.2023867797487267,
+      "learning_rate": 1.4061456982034816e-06,
+      "loss": 0.7201,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8351063829787234,
+      "grad_norm": 1.1691805510845608,
+      "learning_rate": 1.3935825306304329e-06,
+      "loss": 0.7011,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8358662613981763,
+      "grad_norm": 1.5529787151876815,
+      "learning_rate": 1.3810715330272285e-06,
+      "loss": 0.7339,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8366261398176292,
+      "grad_norm": 1.1484356790580228,
+      "learning_rate": 1.36861278123244e-06,
+      "loss": 0.6866,
+      "step": 1101
+    },
+    {
+      "epoch": 0.837386018237082,
+      "grad_norm": 1.0559503515868425,
+      "learning_rate": 1.3562063507679323e-06,
+      "loss": 0.5985,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8381458966565349,
+      "grad_norm": 1.2989479834431623,
+      "learning_rate": 1.3438523168384076e-06,
+      "loss": 0.6846,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8389057750759878,
+      "grad_norm": 1.0806475644725628,
+      "learning_rate": 1.3315507543309624e-06,
+      "loss": 0.7625,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8396656534954408,
+      "grad_norm": 1.140920582486702,
+      "learning_rate": 1.319301737814621e-06,
+      "loss": 0.6834,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8404255319148937,
+      "grad_norm": 1.111450261346159,
+      "learning_rate": 1.3071053415398816e-06,
+      "loss": 0.7714,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8411854103343465,
+      "grad_norm": 1.075448474979407,
+      "learning_rate": 1.2949616394382802e-06,
+      "loss": 0.7337,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8419452887537994,
+      "grad_norm": 0.9601202189186884,
+      "learning_rate": 1.2828707051219257e-06,
+      "loss": 0.6875,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8427051671732523,
+      "grad_norm": 1.1368061357838368,
+      "learning_rate": 1.2708326118830706e-06,
+      "loss": 0.7314,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8434650455927052,
+      "grad_norm": 1.1618574536059776,
+      "learning_rate": 1.2588474326936461e-06,
+      "loss": 0.7322,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8442249240121581,
+      "grad_norm": 1.3830992981104948,
+      "learning_rate": 1.2469152402048446e-06,
+      "loss": 0.7353,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8449848024316109,
+      "grad_norm": 1.1182925134402968,
+      "learning_rate": 1.2350361067466554e-06,
+      "loss": 0.7403,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8457446808510638,
+      "grad_norm": 1.2115068644376168,
+      "learning_rate": 1.2232101043274437e-06,
+      "loss": 0.7798,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8465045592705167,
+      "grad_norm": 1.1090546276827984,
+      "learning_rate": 1.2114373046335059e-06,
+      "loss": 0.7117,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8472644376899696,
+      "grad_norm": 4.166507610410275,
+      "learning_rate": 1.1997177790286362e-06,
+      "loss": 0.6269,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8480243161094225,
+      "grad_norm": 1.0538132793554953,
+      "learning_rate": 1.1880515985536911e-06,
+      "loss": 0.7255,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8487841945288754,
+      "grad_norm": 1.0944056765505172,
+      "learning_rate": 1.176438833926169e-06,
+      "loss": 0.7143,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8495440729483282,
+      "grad_norm": 1.1702415211461719,
+      "learning_rate": 1.1648795555397719e-06,
+      "loss": 0.5761,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8503039513677811,
+      "grad_norm": 1.035813223646502,
+      "learning_rate": 1.1533738334639787e-06,
+      "loss": 0.667,
+      "step": 1119
+    },
+    {
+      "epoch": 0.851063829787234,
+      "grad_norm": 1.1226982400342818,
+      "learning_rate": 1.1419217374436231e-06,
+      "loss": 0.5816,
+      "step": 1120
+    },
+    {
+      "epoch": 0.851823708206687,
+      "grad_norm": 1.156939352942342,
+      "learning_rate": 1.1305233368984792e-06,
+      "loss": 0.6661,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8525835866261399,
+      "grad_norm": 1.0516669260442288,
+      "learning_rate": 1.1191787009228194e-06,
+      "loss": 0.7294,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8533434650455927,
+      "grad_norm": 1.2466026215142973,
+      "learning_rate": 1.1078878982850193e-06,
+      "loss": 0.6684,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8541033434650456,
+      "grad_norm": 1.1064008263905736,
+      "learning_rate": 1.0966509974271289e-06,
+      "loss": 0.7697,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8548632218844985,
+      "grad_norm": 1.2501255707066732,
+      "learning_rate": 1.0854680664644534e-06,
+      "loss": 0.7212,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8556231003039514,
+      "grad_norm": 1.236860045561205,
+      "learning_rate": 1.0743391731851471e-06,
+      "loss": 0.6379,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8563829787234043,
+      "grad_norm": 1.0468169815623685,
+      "learning_rate": 1.0632643850498048e-06,
+      "loss": 0.7328,
+      "step": 1127
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 1.347195466333952,
+      "learning_rate": 1.0522437691910493e-06,
+      "loss": 0.7241,
+      "step": 1128
+    },
+    {
+      "epoch": 0.85790273556231,
+      "grad_norm": 1.238148071226312,
+      "learning_rate": 1.0412773924131202e-06,
+      "loss": 0.7605,
+      "step": 1129
+    },
+    {
+      "epoch": 0.8586626139817629,
+      "grad_norm": 1.2506111985823192,
+      "learning_rate": 1.0303653211914788e-06,
+      "loss": 0.6754,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8594224924012158,
+      "grad_norm": 1.362081669421355,
+      "learning_rate": 1.0195076216723931e-06,
+      "loss": 0.7927,
+      "step": 1131
+    },
+    {
+      "epoch": 0.8601823708206687,
+      "grad_norm": 1.0632919738921134,
+      "learning_rate": 1.008704359672551e-06,
+      "loss": 0.6896,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8609422492401215,
+      "grad_norm": 1.2383897549954153,
+      "learning_rate": 9.97955600678644e-07,
+      "loss": 0.6869,
+      "step": 1133
+    },
+    {
+      "epoch": 0.8617021276595744,
+      "grad_norm": 1.1142541478646706,
+      "learning_rate": 9.872614098469912e-07,
+      "loss": 0.7687,
+      "step": 1134
+    },
+    {
+      "epoch": 0.8624620060790273,
+      "grad_norm": 1.19587245527059,
+      "learning_rate": 9.766218520031234e-07,
+      "loss": 0.7303,
+      "step": 1135
+    },
+    {
+      "epoch": 0.8632218844984803,
+      "grad_norm": 1.005657664076255,
+      "learning_rate": 9.660369916414015e-07,
+      "loss": 0.6556,
+      "step": 1136
+    },
+    {
+      "epoch": 0.8639817629179332,
+      "grad_norm": 1.2812819362765668,
+      "learning_rate": 9.555068929246324e-07,
+      "loss": 0.6985,
+      "step": 1137
+    },
+    {
+      "epoch": 0.8647416413373861,
+      "grad_norm": 1.2186831420915536,
+      "learning_rate": 9.450316196836618e-07,
+      "loss": 0.7435,
+      "step": 1138
+    },
+    {
+      "epoch": 0.8655015197568389,
+      "grad_norm": 1.1290186973602714,
+      "learning_rate": 9.346112354169978e-07,
+      "loss": 0.7659,
+      "step": 1139
+    },
+    {
+      "epoch": 0.8662613981762918,
+      "grad_norm": 1.0923307542579395,
+      "learning_rate": 9.242458032904311e-07,
+      "loss": 0.7297,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8670212765957447,
+      "grad_norm": 1.0764295889612814,
+      "learning_rate": 9.139353861366385e-07,
+      "loss": 0.612,
+      "step": 1141
+    },
+    {
+      "epoch": 0.8677811550151976,
+      "grad_norm": 1.1162467272674055,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.715,
+      "step": 1142
+    },
+    {
+      "epoch": 0.8685410334346505,
+      "grad_norm": 0.9378882497520132,
+      "learning_rate": 8.934798464102923e-07,
+      "loss": 0.7498,
+      "step": 1143
+    },
+    {
+      "epoch": 0.8693009118541033,
+      "grad_norm": 1.3018566485684548,
+      "learning_rate": 8.833348478341519e-07,
+      "loss": 0.6843,
+      "step": 1144
+    },
+    {
+      "epoch": 0.8700607902735562,
+      "grad_norm": 1.2038932549579577,
+      "learning_rate": 8.732451122228592e-07,
+      "loss": 0.7395,
+      "step": 1145
+    },
+    {
+      "epoch": 0.8708206686930091,
+      "grad_norm": 1.1821132797317102,
+      "learning_rate": 8.632107007378932e-07,
+      "loss": 0.6861,
+      "step": 1146
+    },
+    {
+      "epoch": 0.871580547112462,
+      "grad_norm": 1.103795186039256,
+      "learning_rate": 8.532316742053715e-07,
+      "loss": 0.7239,
+      "step": 1147
+    },
+    {
+      "epoch": 0.8723404255319149,
+      "grad_norm": 1.0363457541271124,
+      "learning_rate": 8.433080931156767e-07,
+      "loss": 0.7178,
+      "step": 1148
+    },
+    {
+      "epoch": 0.8731003039513677,
+      "grad_norm": 1.2387877877681717,
+      "learning_rate": 8.334400176230983e-07,
+      "loss": 0.6529,
+      "step": 1149
+    },
+    {
+      "epoch": 0.8738601823708206,
+      "grad_norm": 1.2619716618586272,
+      "learning_rate": 8.236275075454647e-07,
+      "loss": 0.6424,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8746200607902735,
+      "grad_norm": 1.2394430031206916,
+      "learning_rate": 8.138706223637827e-07,
+      "loss": 0.6702,
+      "step": 1151
+    },
+    {
+      "epoch": 0.8753799392097265,
+      "grad_norm": 1.3113427456180624,
+      "learning_rate": 8.041694212218698e-07,
+      "loss": 0.7173,
+      "step": 1152
+    },
+    {
+      "epoch": 0.8761398176291794,
+      "grad_norm": 1.3913267096519508,
+      "learning_rate": 7.945239629260037e-07,
+      "loss": 0.752,
+      "step": 1153
+    },
+    {
+      "epoch": 0.8768996960486323,
+      "grad_norm": 1.0704987317338162,
+      "learning_rate": 7.849343059445635e-07,
+      "loss": 0.718,
+      "step": 1154
+    },
+    {
+      "epoch": 0.8776595744680851,
+      "grad_norm": 1.2345197139195363,
+      "learning_rate": 7.75400508407671e-07,
+      "loss": 0.7142,
+      "step": 1155
+    },
+    {
+      "epoch": 0.878419452887538,
+      "grad_norm": 1.0980818944427368,
+      "learning_rate": 7.659226281068444e-07,
+      "loss": 0.6966,
+      "step": 1156
+    },
+    {
+      "epoch": 0.8791793313069909,
+      "grad_norm": 1.0955001015674064,
+      "learning_rate": 7.565007224946486e-07,
+      "loss": 0.725,
+      "step": 1157
+    },
+    {
+      "epoch": 0.8799392097264438,
+      "grad_norm": 1.1977248745894193,
+      "learning_rate": 7.471348486843355e-07,
+      "loss": 0.774,
+      "step": 1158
+    },
+    {
+      "epoch": 0.8806990881458967,
+      "grad_norm": 1.3009467416912521,
+      "learning_rate": 7.378250634495144e-07,
+      "loss": 0.764,
+      "step": 1159
+    },
+    {
+      "epoch": 0.8814589665653495,
+      "grad_norm": 1.0304684855411061,
+      "learning_rate": 7.285714232237972e-07,
+      "loss": 0.6621,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8822188449848024,
+      "grad_norm": 1.293036714289914,
+      "learning_rate": 7.193739841004565e-07,
+      "loss": 0.6839,
+      "step": 1161
+    },
+    {
+      "epoch": 0.8829787234042553,
+      "grad_norm": 1.2337567888538918,
+      "learning_rate": 7.102328018320859e-07,
+      "loss": 0.7641,
+      "step": 1162
+    },
+    {
+      "epoch": 0.8837386018237082,
+      "grad_norm": 0.9603759876515855,
+      "learning_rate": 7.011479318302716e-07,
+      "loss": 0.7511,
+      "step": 1163
+    },
+    {
+      "epoch": 0.8844984802431611,
+      "grad_norm": 1.100495149114542,
+      "learning_rate": 6.9211942916524e-07,
+      "loss": 0.5739,
+      "step": 1164
+    },
+    {
+      "epoch": 0.8852583586626139,
+      "grad_norm": 1.2567058277652308,
+      "learning_rate": 6.831473485655393e-07,
+      "loss": 0.7024,
+      "step": 1165
+    },
+    {
+      "epoch": 0.8860182370820668,
+      "grad_norm": 1.1536219678957829,
+      "learning_rate": 6.742317444176994e-07,
+      "loss": 0.7553,
+      "step": 1166
+    },
+    {
+      "epoch": 0.8867781155015197,
+      "grad_norm": 1.1568842131425099,
+      "learning_rate": 6.653726707659014e-07,
+      "loss": 0.642,
+      "step": 1167
+    },
+    {
+      "epoch": 0.8875379939209727,
+      "grad_norm": 1.1748261916895442,
+      "learning_rate": 6.565701813116543e-07,
+      "loss": 0.736,
+      "step": 1168
+    },
+    {
+      "epoch": 0.8882978723404256,
+      "grad_norm": 1.2272240589598433,
+      "learning_rate": 6.478243294134678e-07,
+      "loss": 0.6785,
+      "step": 1169
+    },
+    {
+      "epoch": 0.8890577507598785,
+      "grad_norm": 1.0793925928940429,
+      "learning_rate": 6.39135168086531e-07,
+      "loss": 0.6815,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8898176291793313,
+      "grad_norm": 1.0113854296561349,
+      "learning_rate": 6.305027500023841e-07,
+      "loss": 0.6401,
+      "step": 1171
+    },
+    {
+      "epoch": 0.8905775075987842,
+      "grad_norm": 1.2571607435026524,
+      "learning_rate": 6.219271274886052e-07,
+      "loss": 0.4813,
+      "step": 1172
+    },
+    {
+      "epoch": 0.8913373860182371,
+      "grad_norm": 0.9180442946861914,
+      "learning_rate": 6.13408352528495e-07,
+      "loss": 0.7629,
+      "step": 1173
+    },
+    {
+      "epoch": 0.89209726443769,
+      "grad_norm": 1.120989173886215,
+      "learning_rate": 6.04946476760756e-07,
+      "loss": 0.5761,
+      "step": 1174
+    },
+    {
+      "epoch": 0.8928571428571429,
+      "grad_norm": 1.119368512790432,
+      "learning_rate": 5.965415514791817e-07,
+      "loss": 0.7439,
+      "step": 1175
+    },
+    {
+      "epoch": 0.8936170212765957,
+      "grad_norm": 1.143580106138461,
+      "learning_rate": 5.881936276323463e-07,
+      "loss": 0.7,
+      "step": 1176
+    },
+    {
+      "epoch": 0.8943768996960486,
+      "grad_norm": 0.9725000133492355,
+      "learning_rate": 5.79902755823295e-07,
+      "loss": 0.7308,
+      "step": 1177
+    },
+    {
+      "epoch": 0.8951367781155015,
+      "grad_norm": 0.994127857335495,
+      "learning_rate": 5.716689863092362e-07,
+      "loss": 0.6033,
+      "step": 1178
+    },
+    {
+      "epoch": 0.8958966565349544,
+      "grad_norm": 1.1075290736095693,
+      "learning_rate": 5.634923690012451e-07,
+      "loss": 0.7075,
+      "step": 1179
+    },
+    {
+      "epoch": 0.8966565349544073,
+      "grad_norm": 1.167844867686388,
+      "learning_rate": 5.553729534639463e-07,
+      "loss": 0.5536,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8974164133738601,
+      "grad_norm": 1.0591506870039167,
+      "learning_rate": 5.473107889152241e-07,
+      "loss": 0.6447,
+      "step": 1181
+    },
+    {
+      "epoch": 0.898176291793313,
+      "grad_norm": 1.0350933535569133,
+      "learning_rate": 5.393059242259236e-07,
+      "loss": 0.7002,
+      "step": 1182
+    },
+    {
+      "epoch": 0.898936170212766,
+      "grad_norm": 1.2822200143678832,
+      "learning_rate": 5.313584079195488e-07,
+      "loss": 0.674,
+      "step": 1183
+    },
+    {
+      "epoch": 0.8996960486322189,
+      "grad_norm": 1.0687168416496815,
+      "learning_rate": 5.234682881719766e-07,
+      "loss": 0.7696,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9004559270516718,
+      "grad_norm": 1.1775314415301947,
+      "learning_rate": 5.156356128111551e-07,
+      "loss": 0.7153,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9012158054711246,
+      "grad_norm": 1.0806460690940876,
+      "learning_rate": 5.078604293168232e-07,
+      "loss": 0.6388,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9019756838905775,
+      "grad_norm": 1.3658980832932874,
+      "learning_rate": 5.001427848202145e-07,
+      "loss": 0.6274,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9027355623100304,
+      "grad_norm": 1.0971465178481652,
+      "learning_rate": 4.924827261037779e-07,
+      "loss": 0.6832,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9034954407294833,
+      "grad_norm": 1.2530282400707315,
+      "learning_rate": 4.848802996008928e-07,
+      "loss": 0.606,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9042553191489362,
+      "grad_norm": 1.2286583520290673,
+      "learning_rate": 4.773355513955847e-07,
+      "loss": 0.721,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9050151975683891,
+      "grad_norm": 1.2106924050350691,
+      "learning_rate": 4.6984852722224307e-07,
+      "loss": 0.7536,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9057750759878419,
+      "grad_norm": 0.9008661784706631,
+      "learning_rate": 4.6241927246535645e-07,
+      "loss": 0.5507,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9065349544072948,
+      "grad_norm": 1.0875351613738515,
+      "learning_rate": 4.5504783215922775e-07,
+      "loss": 0.6835,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9072948328267477,
+      "grad_norm": 1.122110061636014,
+      "learning_rate": 4.4773425098769697e-07,
+      "loss": 0.7824,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9080547112462006,
+      "grad_norm": 1.0468917022544524,
+      "learning_rate": 4.4047857328388457e-07,
+      "loss": 0.7099,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9088145896656535,
+      "grad_norm": 1.1327395463199081,
+      "learning_rate": 4.332808430299085e-07,
+      "loss": 0.704,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9095744680851063,
+      "grad_norm": 1.200395501431566,
+      "learning_rate": 4.2614110385662544e-07,
+      "loss": 0.6801,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9103343465045592,
+      "grad_norm": 1.2966065561282096,
+      "learning_rate": 4.190593990433656e-07,
+      "loss": 0.6921,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9110942249240122,
+      "grad_norm": 1.072234796169011,
+      "learning_rate": 4.1203577151767037e-07,
+      "loss": 0.7167,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9118541033434651,
+      "grad_norm": 1.2277287300371735,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.6231,
+      "step": 1200
+    },
+    {
+      "epoch": 0.912613981762918,
+      "grad_norm": 1.2406804352583274,
+      "learning_rate": 3.981629182786162e-07,
+      "loss": 0.6251,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9133738601823708,
+      "grad_norm": 1.2565989611946244,
+      "learning_rate": 3.913137766590569e-07,
+      "loss": 0.6691,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9141337386018237,
+      "grad_norm": 1.0894521219118882,
+      "learning_rate": 3.8452288051414765e-07,
+      "loss": 0.6174,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9148936170212766,
+      "grad_norm": 1.0896043395915658,
+      "learning_rate": 3.7779027100861785e-07,
+      "loss": 0.7955,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9156534954407295,
+      "grad_norm": 1.1387911305845246,
+      "learning_rate": 3.711159889538774e-07,
+      "loss": 0.7165,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9164133738601824,
+      "grad_norm": 1.0581039732048783,
+      "learning_rate": 3.645000748077709e-07,
+      "loss": 0.6385,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9171732522796353,
+      "grad_norm": 1.142293054018674,
+      "learning_rate": 3.5794256867432985e-07,
+      "loss": 0.7349,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9179331306990881,
+      "grad_norm": 1.1115528973809246,
+      "learning_rate": 3.5144351030353077e-07,
+      "loss": 0.6063,
+      "step": 1208
+    },
+    {
+      "epoch": 0.918693009118541,
+      "grad_norm": 1.2711522307666283,
+      "learning_rate": 3.45002939091057e-07,
+      "loss": 0.6267,
+      "step": 1209
+    },
+    {
+      "epoch": 0.9194528875379939,
+      "grad_norm": 1.130354538612267,
+      "learning_rate": 3.386208940780522e-07,
+      "loss": 0.6871,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9202127659574468,
+      "grad_norm": 1.2610028948951497,
+      "learning_rate": 3.3229741395089276e-07,
+      "loss": 0.6161,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9209726443768997,
+      "grad_norm": 1.0519435539156983,
+      "learning_rate": 3.260325370409501e-07,
+      "loss": 0.7396,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9217325227963525,
+      "grad_norm": 1.173704186994253,
+      "learning_rate": 3.19826301324353e-07,
+      "loss": 0.667,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9224924012158054,
+      "grad_norm": 1.189649352239027,
+      "learning_rate": 3.1367874442176485e-07,
+      "loss": 0.7598,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9232522796352584,
+      "grad_norm": 1.0286871900514776,
+      "learning_rate": 3.075899035981533e-07,
+      "loss": 0.7767,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9240121580547113,
+      "grad_norm": 1.0572010661601794,
+      "learning_rate": 3.0155981576255986e-07,
+      "loss": 0.7177,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9247720364741642,
+      "grad_norm": 1.044342277658108,
+      "learning_rate": 2.955885174678852e-07,
+      "loss": 0.6784,
+      "step": 1217
+    },
+    {
+      "epoch": 0.925531914893617,
+      "grad_norm": 1.0189565217507277,
+      "learning_rate": 2.896760449106606e-07,
+      "loss": 0.7161,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9262917933130699,
+      "grad_norm": 1.2660088084368877,
+      "learning_rate": 2.83822433930826e-07,
+      "loss": 0.7671,
+      "step": 1219
+    },
+    {
+      "epoch": 0.9270516717325228,
+      "grad_norm": 1.2991870091587046,
+      "learning_rate": 2.7802772001152224e-07,
+      "loss": 0.6416,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9278115501519757,
+      "grad_norm": 1.1181116151401733,
+      "learning_rate": 2.7229193827886913e-07,
+      "loss": 0.7236,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 1.0163129451962587,
+      "learning_rate": 2.6661512350175556e-07,
+      "loss": 0.7145,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9293313069908815,
+      "grad_norm": 1.1673052192523565,
+      "learning_rate": 2.609973100916241e-07,
+      "loss": 0.6913,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9300911854103343,
+      "grad_norm": 1.023824191239343,
+      "learning_rate": 2.5543853210226566e-07,
+      "loss": 0.689,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9308510638297872,
+      "grad_norm": 1.0056260338184262,
+      "learning_rate": 2.499388232296174e-07,
+      "loss": 0.692,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9316109422492401,
+      "grad_norm": 1.112601374178031,
+      "learning_rate": 2.4449821681155284e-07,
+      "loss": 0.7191,
+      "step": 1226
+    },
+    {
+      "epoch": 0.932370820668693,
+      "grad_norm": 1.0663092173586473,
+      "learning_rate": 2.3911674582767553e-07,
+      "loss": 0.7152,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9331306990881459,
+      "grad_norm": 1.1835757827302227,
+      "learning_rate": 2.3379444289913344e-07,
+      "loss": 0.727,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9338905775075987,
+      "grad_norm": 1.3007405802649337,
+      "learning_rate": 2.2853134028840594e-07,
+      "loss": 0.6252,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9346504559270516,
+      "grad_norm": 1.1787216115316808,
+      "learning_rate": 2.233274698991139e-07,
+      "loss": 0.6884,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9354103343465046,
+      "grad_norm": 1.1226604599470809,
+      "learning_rate": 2.1818286327583537e-07,
+      "loss": 0.72,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9361702127659575,
+      "grad_norm": 1.1849786593331166,
+      "learning_rate": 2.13097551603898e-07,
+      "loss": 0.7118,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9369300911854104,
+      "grad_norm": 1.123307297377377,
+      "learning_rate": 2.0807156570920027e-07,
+      "loss": 0.689,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9376899696048632,
+      "grad_norm": 1.173356309084475,
+      "learning_rate": 2.0310493605802395e-07,
+      "loss": 0.6502,
+      "step": 1234
+    },
+    {
+      "epoch": 0.9384498480243161,
+      "grad_norm": 1.1913393551696687,
+      "learning_rate": 1.9819769275684742e-07,
+      "loss": 0.6642,
+      "step": 1235
+    },
+    {
+      "epoch": 0.939209726443769,
+      "grad_norm": 1.0249961412609725,
+      "learning_rate": 1.9334986555216374e-07,
+      "loss": 0.7158,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9399696048632219,
+      "grad_norm": 1.2126445450577445,
+      "learning_rate": 1.8856148383029848e-07,
+      "loss": 0.6607,
+      "step": 1237
+    },
+    {
+      "epoch": 0.9407294832826748,
+      "grad_norm": 1.4093796517345238,
+      "learning_rate": 1.8383257661723664e-07,
+      "loss": 0.6734,
+      "step": 1238
+    },
+    {
+      "epoch": 0.9414893617021277,
+      "grad_norm": 1.2786023133275601,
+      "learning_rate": 1.791631725784404e-07,
+      "loss": 0.6996,
+      "step": 1239
+    },
+    {
+      "epoch": 0.9422492401215805,
+      "grad_norm": 1.0960561462544673,
+      "learning_rate": 1.7455330001868054e-07,
+      "loss": 0.7487,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9430091185410334,
+      "grad_norm": 1.1055539525407532,
+      "learning_rate": 1.7000298688186312e-07,
+      "loss": 0.6456,
+      "step": 1241
+    },
+    {
+      "epoch": 0.9437689969604863,
+      "grad_norm": 0.9960735919562198,
+      "learning_rate": 1.6551226075085746e-07,
+      "loss": 0.7192,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9445288753799392,
+      "grad_norm": 1.275738801543123,
+      "learning_rate": 1.6108114884733184e-07,
+      "loss": 0.8265,
+      "step": 1243
+    },
+    {
+      "epoch": 0.9452887537993921,
+      "grad_norm": 1.1633162842566498,
+      "learning_rate": 1.567096780315891e-07,
+      "loss": 0.7043,
+      "step": 1244
+    },
+    {
+      "epoch": 0.9460486322188449,
+      "grad_norm": 1.2665944074906847,
+      "learning_rate": 1.5239787480240353e-07,
+      "loss": 0.6394,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9468085106382979,
+      "grad_norm": 1.1431525657028725,
+      "learning_rate": 1.4814576529685543e-07,
+      "loss": 0.6584,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9475683890577508,
+      "grad_norm": 1.0445102532211836,
+      "learning_rate": 1.4395337529018116e-07,
+      "loss": 0.627,
+      "step": 1247
+    },
+    {
+      "epoch": 0.9483282674772037,
+      "grad_norm": 1.1371799365903608,
+      "learning_rate": 1.3982073019560782e-07,
+      "loss": 0.6978,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9490881458966566,
+      "grad_norm": 1.0721131322888444,
+      "learning_rate": 1.3574785506420773e-07,
+      "loss": 0.7239,
+      "step": 1249
+    },
+    {
+      "epoch": 0.9498480243161094,
+      "grad_norm": 1.177780740815343,
+      "learning_rate": 1.317347745847386e-07,
+      "loss": 0.7071,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9506079027355623,
+      "grad_norm": 0.9764352141315131,
+      "learning_rate": 1.277815130835014e-07,
+      "loss": 0.5987,
+      "step": 1251
+    },
+    {
+      "epoch": 0.9513677811550152,
+      "grad_norm": 1.0597707069561533,
+      "learning_rate": 1.2388809452418716e-07,
+      "loss": 0.6819,
+      "step": 1252
+    },
+    {
+      "epoch": 0.9521276595744681,
+      "grad_norm": 1.0357219353126603,
+      "learning_rate": 1.2005454250773262e-07,
+      "loss": 0.7386,
+      "step": 1253
+    },
+    {
+      "epoch": 0.952887537993921,
+      "grad_norm": 1.3688991177073506,
+      "learning_rate": 1.1628088027218265e-07,
+      "loss": 0.8196,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9536474164133738,
+      "grad_norm": 1.0794978094116288,
+      "learning_rate": 1.1256713069254244e-07,
+      "loss": 0.6784,
+      "step": 1255
+    },
+    {
+      "epoch": 0.9544072948328267,
+      "grad_norm": 1.2978388589542718,
+      "learning_rate": 1.0891331628063884e-07,
+      "loss": 0.7526,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9551671732522796,
+      "grad_norm": 1.0991604780745179,
+      "learning_rate": 1.0531945918499265e-07,
+      "loss": 0.6504,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9559270516717325,
+      "grad_norm": 1.1316949589748224,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.6864,
+      "step": 1258
+    },
+    {
+      "epoch": 0.9566869300911854,
+      "grad_norm": 1.0535009629368148,
+      "learning_rate": 9.831170371917276e-08,
+      "loss": 0.7574,
+      "step": 1259
+    },
+    {
+      "epoch": 0.9574468085106383,
+      "grad_norm": 1.0988680557329953,
+      "learning_rate": 9.489784782827582e-08,
+      "loss": 0.6674,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9582066869300911,
+      "grad_norm": 1.2012576178042316,
+      "learning_rate": 9.154403421193226e-08,
+      "loss": 0.6475,
+      "step": 1261
+    },
+    {
+      "epoch": 0.958966565349544,
+      "grad_norm": 1.1647637296827538,
+      "learning_rate": 8.825028320012752e-08,
+      "loss": 0.6579,
+      "step": 1262
+    },
+    {
+      "epoch": 0.959726443768997,
+      "grad_norm": 1.1590888732361975,
+      "learning_rate": 8.5016614758765e-08,
+      "loss": 0.7138,
+      "step": 1263
+    },
+    {
+      "epoch": 0.9604863221884499,
+      "grad_norm": 1.2104828609323357,
+      "learning_rate": 8.18430484895405e-08,
+      "loss": 0.7079,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9612462006079028,
+      "grad_norm": 1.0676473427692756,
+      "learning_rate": 7.872960362982684e-08,
+      "loss": 0.6328,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9620060790273556,
+      "grad_norm": 1.1050638827297063,
+      "learning_rate": 7.567629905255503e-08,
+      "loss": 0.5908,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9627659574468085,
+      "grad_norm": 1.1570194134365799,
+      "learning_rate": 7.268315326609988e-08,
+      "loss": 0.7251,
+      "step": 1267
+    },
+    {
+      "epoch": 0.9635258358662614,
+      "grad_norm": 1.055972426540126,
+      "learning_rate": 6.975018441417126e-08,
+      "loss": 0.6951,
+      "step": 1268
+    },
+    {
+      "epoch": 0.9642857142857143,
+      "grad_norm": 1.1340255700749204,
+      "learning_rate": 6.68774102756975e-08,
+      "loss": 0.729,
+      "step": 1269
+    },
+    {
+      "epoch": 0.9650455927051672,
+      "grad_norm": 1.1116704872217635,
+      "learning_rate": 6.406484826472547e-08,
+      "loss": 0.7146,
+      "step": 1270
+    },
+    {
+      "epoch": 0.96580547112462,
+      "grad_norm": 1.2238885966555615,
+      "learning_rate": 6.131251543030848e-08,
+      "loss": 0.6778,
+      "step": 1271
+    },
+    {
+      "epoch": 0.9665653495440729,
+      "grad_norm": 1.1140657613414842,
+      "learning_rate": 5.862042845640403e-08,
+      "loss": 0.6196,
+      "step": 1272
+    },
+    {
+      "epoch": 0.9673252279635258,
+      "grad_norm": 1.4034971558332028,
+      "learning_rate": 5.59886036617785e-08,
+      "loss": 0.6726,
+      "step": 1273
+    },
+    {
+      "epoch": 0.9680851063829787,
+      "grad_norm": 1.247850365035091,
+      "learning_rate": 5.3417056999901515e-08,
+      "loss": 0.8024,
+      "step": 1274
+    },
+    {
+      "epoch": 0.9688449848024316,
+      "grad_norm": 1.1403707142366584,
+      "learning_rate": 5.0905804058850595e-08,
+      "loss": 0.6981,
+      "step": 1275
+    },
+    {
+      "epoch": 0.9696048632218845,
+      "grad_norm": 1.1512503282266044,
+      "learning_rate": 4.845486006121891e-08,
+      "loss": 0.7136,
+      "step": 1276
+    },
+    {
+      "epoch": 0.9703647416413373,
+      "grad_norm": 1.1205146371244328,
+      "learning_rate": 4.6064239864020975e-08,
+      "loss": 0.6572,
+      "step": 1277
+    },
+    {
+      "epoch": 0.9711246200607903,
+      "grad_norm": 1.0640740176630763,
+      "learning_rate": 4.3733957958607134e-08,
+      "loss": 0.6775,
+      "step": 1278
+    },
+    {
+      "epoch": 0.9718844984802432,
+      "grad_norm": 1.21170834732034,
+      "learning_rate": 4.146402847056474e-08,
+      "loss": 0.6142,
+      "step": 1279
+    },
+    {
+      "epoch": 0.9726443768996961,
+      "grad_norm": 1.2060891666752993,
+      "learning_rate": 3.925446515964604e-08,
+      "loss": 0.6963,
+      "step": 1280
+    },
+    {
+      "epoch": 0.973404255319149,
+      "grad_norm": 1.1413982945678454,
+      "learning_rate": 3.7105281419675954e-08,
+      "loss": 0.701,
+      "step": 1281
+    },
+    {
+      "epoch": 0.9741641337386018,
+      "grad_norm": 1.1353912953364804,
+      "learning_rate": 3.5016490278473316e-08,
+      "loss": 0.733,
+      "step": 1282
+    },
+    {
+      "epoch": 0.9749240121580547,
+      "grad_norm": 0.9731905027567818,
+      "learning_rate": 3.2988104397773115e-08,
+      "loss": 0.6818,
+      "step": 1283
+    },
+    {
+      "epoch": 0.9756838905775076,
+      "grad_norm": 1.1467760181112903,
+      "learning_rate": 3.1020136073146575e-08,
+      "loss": 0.755,
+      "step": 1284
+    },
+    {
+      "epoch": 0.9764437689969605,
+      "grad_norm": 1.2500496670477363,
+      "learning_rate": 2.9112597233931228e-08,
+      "loss": 0.7724,
+      "step": 1285
+    },
+    {
+      "epoch": 0.9772036474164134,
+      "grad_norm": 1.0884238091336034,
+      "learning_rate": 2.7265499443154264e-08,
+      "loss": 0.7486,
+      "step": 1286
+    },
+    {
+      "epoch": 0.9779635258358662,
+      "grad_norm": 1.0088684484023063,
+      "learning_rate": 2.547885389746485e-08,
+      "loss": 0.7156,
+      "step": 1287
+    },
+    {
+      "epoch": 0.9787234042553191,
+      "grad_norm": 1.0334713239157454,
+      "learning_rate": 2.3752671427065276e-08,
+      "loss": 0.6725,
+      "step": 1288
+    },
+    {
+      "epoch": 0.979483282674772,
+      "grad_norm": 1.1735061403799454,
+      "learning_rate": 2.208696249564657e-08,
+      "loss": 0.6288,
+      "step": 1289
+    },
+    {
+      "epoch": 0.9802431610942249,
+      "grad_norm": 1.0449694502580413,
+      "learning_rate": 2.048173720032298e-08,
+      "loss": 0.7597,
+      "step": 1290
+    },
+    {
+      "epoch": 0.9810030395136778,
+      "grad_norm": 1.1234942762561577,
+      "learning_rate": 1.8937005271572052e-08,
+      "loss": 0.5683,
+      "step": 1291
+    },
+    {
+      "epoch": 0.9817629179331308,
+      "grad_norm": 1.4071207404835981,
+      "learning_rate": 1.7452776073175748e-08,
+      "loss": 0.7275,
+      "step": 1292
+    },
+    {
+      "epoch": 0.9825227963525835,
+      "grad_norm": 1.4128938156204476,
+      "learning_rate": 1.602905860216497e-08,
+      "loss": 0.8493,
+      "step": 1293
+    },
+    {
+      "epoch": 0.9832826747720365,
+      "grad_norm": 1.1248277575727312,
+      "learning_rate": 1.4665861488761813e-08,
+      "loss": 0.7518,
+      "step": 1294
+    },
+    {
+      "epoch": 0.9840425531914894,
+      "grad_norm": 1.2244461973302783,
+      "learning_rate": 1.3363192996328488e-08,
+      "loss": 0.7051,
+      "step": 1295
+    },
+    {
+      "epoch": 0.9848024316109423,
+      "grad_norm": 1.1039705050782245,
+      "learning_rate": 1.2121061021318492e-08,
+      "loss": 0.7085,
+      "step": 1296
+    },
+    {
+      "epoch": 0.9855623100303952,
+      "grad_norm": 1.2017909438185823,
+      "learning_rate": 1.0939473093229958e-08,
+      "loss": 0.7568,
+      "step": 1297
+    },
+    {
+      "epoch": 0.986322188449848,
+      "grad_norm": 1.1888312357208546,
+      "learning_rate": 9.818436374553487e-09,
+      "loss": 0.6483,
+      "step": 1298
+    },
+    {
+      "epoch": 0.9870820668693009,
+      "grad_norm": 1.1075281495725198,
+      "learning_rate": 8.757957660737726e-09,
+      "loss": 0.6389,
+      "step": 1299
+    },
+    {
+      "epoch": 0.9878419452887538,
+      "grad_norm": 1.2252737686145327,
+      "learning_rate": 7.758043380140523e-09,
+      "loss": 0.6934,
+      "step": 1300
+    },
+    {
+      "epoch": 0.9886018237082067,
+      "grad_norm": 1.3500685962348613,
+      "learning_rate": 6.818699593996725e-09,
+      "loss": 0.743,
+      "step": 1301
+    },
+    {
+      "epoch": 0.9893617021276596,
+      "grad_norm": 1.1391941942219126,
+      "learning_rate": 5.939931996372661e-09,
+      "loss": 0.726,
+      "step": 1302
+    },
+    {
+      "epoch": 0.9901215805471124,
+      "grad_norm": 1.322232366894908,
+      "learning_rate": 5.1217459141406074e-09,
+      "loss": 0.6969,
+      "step": 1303
+    },
+    {
+      "epoch": 0.9908814589665653,
+      "grad_norm": 1.0905255226681096,
+      "learning_rate": 4.364146306943262e-09,
+      "loss": 0.7189,
+      "step": 1304
+    },
+    {
+      "epoch": 0.9916413373860182,
+      "grad_norm": 1.0414862403250007,
+      "learning_rate": 3.6671377671604337e-09,
+      "loss": 0.69,
+      "step": 1305
+    },
+    {
+      "epoch": 0.9924012158054711,
+      "grad_norm": 1.431716637357648,
+      "learning_rate": 3.0307245198857303e-09,
+      "loss": 0.746,
+      "step": 1306
+    },
+    {
+      "epoch": 0.993161094224924,
+      "grad_norm": 0.9315482774719349,
+      "learning_rate": 2.454910422897694e-09,
+      "loss": 0.6958,
+      "step": 1307
+    },
+    {
+      "epoch": 0.993920972644377,
+      "grad_norm": 0.9621526906912882,
+      "learning_rate": 1.9396989666398137e-09,
+      "loss": 0.568,
+      "step": 1308
+    },
+    {
+      "epoch": 0.9946808510638298,
+      "grad_norm": 1.1947376206574818,
+      "learning_rate": 1.485093274194993e-09,
+      "loss": 0.6485,
+      "step": 1309
+    },
+    {
+      "epoch": 0.9954407294832827,
+      "grad_norm": 1.193575775298368,
+      "learning_rate": 1.0910961012711163e-09,
+      "loss": 0.7499,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9962006079027356,
+      "grad_norm": 1.1202479836190646,
+      "learning_rate": 7.577098361810641e-10,
+      "loss": 0.6712,
+      "step": 1311
+    },
+    {
+      "epoch": 0.9969604863221885,
+      "grad_norm": 1.150541600033343,
+      "learning_rate": 4.849364998305017e-10,
+      "loss": 0.6162,
+      "step": 1312
+    },
+    {
+      "epoch": 0.9977203647416414,
+      "grad_norm": 1.1841504466478256,
+      "learning_rate": 2.7277774570233507e-10,
+      "loss": 0.6731,
+      "step": 1313
+    },
+    {
+      "epoch": 0.9984802431610942,
+      "grad_norm": 1.1634988910801727,
+      "learning_rate": 1.2123485985227058e-10,
+      "loss": 0.6473,
+      "step": 1314
+    },
+    {
+      "epoch": 0.9992401215805471,
+      "grad_norm": 1.264183108117609,
+      "learning_rate": 3.030876089438195e-11,
+      "loss": 0.5803,
+      "step": 1315
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.114121273386629,
+      "learning_rate": 0.0,
+      "loss": 0.6488,
+      "step": 1316
+    },
+    {
+      "epoch": 1.0,
+      "step": 1316,
+      "total_flos": 141830150160384.0,
+      "train_loss": 0.7534279622205485,
+      "train_runtime": 5282.496,
+      "train_samples_per_second": 15.938,
+      "train_steps_per_second": 0.249
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1316,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 395,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 141830150160384.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}