diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16429 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998717955292536,
+  "eval_steps": 400,
+  "global_step": 6093,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004923051676658069,
+      "grad_norm": 52.76170349121094,
+      "learning_rate": 2.9508196721311474e-06,
+      "loss": 10.8937,
+      "num_input_tokens_seen": 393216,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009846103353316137,
+      "grad_norm": 28.751510620117188,
+      "learning_rate": 5.901639344262295e-06,
+      "loss": 10.5303,
+      "num_input_tokens_seen": 786432,
+      "step": 6
+    },
+    {
+      "epoch": 0.0014769155029974205,
+      "grad_norm": 11.043315887451172,
+      "learning_rate": 8.852459016393442e-06,
+      "loss": 9.5849,
+      "num_input_tokens_seen": 1179648,
+      "step": 9
+    },
+    {
+      "epoch": 0.0019692206706632275,
+      "grad_norm": 6.976105213165283,
+      "learning_rate": 1.180327868852459e-05,
+      "loss": 9.0558,
+      "num_input_tokens_seen": 1572864,
+      "step": 12
+    },
+    {
+      "epoch": 0.0024615258383290342,
+      "grad_norm": 4.415246486663818,
+      "learning_rate": 1.4754098360655736e-05,
+      "loss": 8.6827,
+      "num_input_tokens_seen": 1966080,
+      "step": 15
+    },
+    {
+      "epoch": 0.002953831005994841,
+      "grad_norm": 2.8027164936065674,
+      "learning_rate": 1.7704918032786883e-05,
+      "loss": 8.3359,
+      "num_input_tokens_seen": 2359296,
+      "step": 18
+    },
+    {
+      "epoch": 0.0034461361736606477,
+      "grad_norm": 1.793583631515503,
+      "learning_rate": 2.065573770491803e-05,
+      "loss": 8.0831,
+      "num_input_tokens_seen": 2752512,
+      "step": 21
+    },
+    {
+      "epoch": 0.003938441341326455,
+      "grad_norm": 1.6349351406097412,
+      "learning_rate": 2.360655737704918e-05,
+      "loss": 7.8729,
+      "num_input_tokens_seen": 3145728,
+      "step": 24
+    },
+    {
+      "epoch": 0.004430746508992262,
+      "grad_norm": 1.0413860082626343,
+      "learning_rate": 2.6557377049180323e-05,
+      "loss": 7.7278,
+      "num_input_tokens_seen": 3538944,
+      "step": 27
+    },
+    {
+      "epoch": 0.0049230516766580685,
+      "grad_norm": 0.8472433090209961,
+      "learning_rate": 2.950819672131147e-05,
+      "loss": 7.5833,
+      "num_input_tokens_seen": 3932160,
+      "step": 30
+    },
+    {
+      "epoch": 0.005415356844323875,
+      "grad_norm": 0.8556374907493591,
+      "learning_rate": 3.245901639344262e-05,
+      "loss": 7.4282,
+      "num_input_tokens_seen": 4325376,
+      "step": 33
+    },
+    {
+      "epoch": 0.005907662011989682,
+      "grad_norm": 0.7362769842147827,
+      "learning_rate": 3.540983606557377e-05,
+      "loss": 7.3597,
+      "num_input_tokens_seen": 4718592,
+      "step": 36
+    },
+    {
+      "epoch": 0.006399967179655489,
+      "grad_norm": 0.5901055335998535,
+      "learning_rate": 3.836065573770491e-05,
+      "loss": 7.2626,
+      "num_input_tokens_seen": 5111808,
+      "step": 39
+    },
+    {
+      "epoch": 0.0068922723473212955,
+      "grad_norm": 0.6458311676979065,
+      "learning_rate": 4.131147540983606e-05,
+      "loss": 7.2543,
+      "num_input_tokens_seen": 5505024,
+      "step": 42
+    },
+    {
+      "epoch": 0.007384577514987102,
+      "grad_norm": 0.5546324253082275,
+      "learning_rate": 4.4262295081967207e-05,
+      "loss": 7.2457,
+      "num_input_tokens_seen": 5898240,
+      "step": 45
+    },
+    {
+      "epoch": 0.00787688268265291,
+      "grad_norm": 0.5260677337646484,
+      "learning_rate": 4.721311475409836e-05,
+      "loss": 7.1727,
+      "num_input_tokens_seen": 6291456,
+      "step": 48
+    },
+    {
+      "epoch": 0.008369187850318717,
+      "grad_norm": 0.55791175365448,
+      "learning_rate": 5.01639344262295e-05,
+      "loss": 7.1299,
+      "num_input_tokens_seen": 6684672,
+      "step": 51
+    },
+    {
+      "epoch": 0.008861493017984523,
+      "grad_norm": 0.5637151002883911,
+      "learning_rate": 5.3114754098360647e-05,
+      "loss": 7.1604,
+      "num_input_tokens_seen": 7077888,
+      "step": 54
+    },
+    {
+      "epoch": 0.00935379818565033,
+      "grad_norm": 0.54989093542099,
+      "learning_rate": 5.60655737704918e-05,
+      "loss": 7.0051,
+      "num_input_tokens_seen": 7471104,
+      "step": 57
+    },
+    {
+      "epoch": 0.009846103353316137,
+      "grad_norm": 0.46848881244659424,
+      "learning_rate": 5.901639344262294e-05,
+      "loss": 7.02,
+      "num_input_tokens_seen": 7864320,
+      "step": 60
+    },
+    {
+      "epoch": 0.010338408520981944,
+      "grad_norm": 0.4989602565765381,
+      "learning_rate": 6.19672131147541e-05,
+      "loss": 6.9589,
+      "num_input_tokens_seen": 8257536,
+      "step": 63
+    },
+    {
+      "epoch": 0.01083071368864775,
+      "grad_norm": 0.5755507349967957,
+      "learning_rate": 6.491803278688524e-05,
+      "loss": 7.008,
+      "num_input_tokens_seen": 8650752,
+      "step": 66
+    },
+    {
+      "epoch": 0.011323018856313557,
+      "grad_norm": 0.543915867805481,
+      "learning_rate": 6.786885245901639e-05,
+      "loss": 6.9092,
+      "num_input_tokens_seen": 9043968,
+      "step": 69
+    },
+    {
+      "epoch": 0.011815324023979364,
+      "grad_norm": 0.5782525539398193,
+      "learning_rate": 7.081967213114753e-05,
+      "loss": 6.8488,
+      "num_input_tokens_seen": 9437184,
+      "step": 72
+    },
+    {
+      "epoch": 0.01230762919164517,
+      "grad_norm": 0.5549654364585876,
+      "learning_rate": 7.377049180327868e-05,
+      "loss": 6.8617,
+      "num_input_tokens_seen": 9830400,
+      "step": 75
+    },
+    {
+      "epoch": 0.012799934359310977,
+      "grad_norm": 0.4978010058403015,
+      "learning_rate": 7.672131147540982e-05,
+      "loss": 6.8245,
+      "num_input_tokens_seen": 10223616,
+      "step": 78
+    },
+    {
+      "epoch": 0.013292239526976784,
+      "grad_norm": 0.5749168395996094,
+      "learning_rate": 7.967213114754097e-05,
+      "loss": 6.8257,
+      "num_input_tokens_seen": 10616832,
+      "step": 81
+    },
+    {
+      "epoch": 0.013784544694642591,
+      "grad_norm": 0.4636499285697937,
+      "learning_rate": 8.262295081967212e-05,
+      "loss": 6.829,
+      "num_input_tokens_seen": 11010048,
+      "step": 84
+    },
+    {
+      "epoch": 0.014276849862308398,
+      "grad_norm": 0.5625278949737549,
+      "learning_rate": 8.557377049180327e-05,
+      "loss": 6.8155,
+      "num_input_tokens_seen": 11403264,
+      "step": 87
+    },
+    {
+      "epoch": 0.014769155029974204,
+      "grad_norm": 0.5715211629867554,
+      "learning_rate": 8.852459016393441e-05,
+      "loss": 6.8136,
+      "num_input_tokens_seen": 11796480,
+      "step": 90
+    },
+    {
+      "epoch": 0.015261460197640011,
+      "grad_norm": 0.6819092035293579,
+      "learning_rate": 9.147540983606556e-05,
+      "loss": 6.7438,
+      "num_input_tokens_seen": 12189696,
+      "step": 93
+    },
+    {
+      "epoch": 0.01575376536530582,
+      "grad_norm": 0.5896216034889221,
+      "learning_rate": 9.442622950819672e-05,
+      "loss": 6.7681,
+      "num_input_tokens_seen": 12582912,
+      "step": 96
+    },
+    {
+      "epoch": 0.016246070532971627,
+      "grad_norm": 0.6056619882583618,
+      "learning_rate": 9.737704918032786e-05,
+      "loss": 6.7136,
+      "num_input_tokens_seen": 12976128,
+      "step": 99
+    },
+    {
+      "epoch": 0.016738375700637433,
+      "grad_norm": 0.6384982466697693,
+      "learning_rate": 0.000100327868852459,
+      "loss": 6.7073,
+      "num_input_tokens_seen": 13369344,
+      "step": 102
+    },
+    {
+      "epoch": 0.01723068086830324,
+      "grad_norm": 0.5903695821762085,
+      "learning_rate": 0.00010327868852459015,
+      "loss": 6.7011,
+      "num_input_tokens_seen": 13762560,
+      "step": 105
+    },
+    {
+      "epoch": 0.017722986035969047,
+      "grad_norm": 0.5709877014160156,
+      "learning_rate": 0.00010622950819672129,
+      "loss": 6.7258,
+      "num_input_tokens_seen": 14155776,
+      "step": 108
+    },
+    {
+      "epoch": 0.018215291203634854,
+      "grad_norm": 0.621760904788971,
+      "learning_rate": 0.00010918032786885245,
+      "loss": 6.6358,
+      "num_input_tokens_seen": 14548992,
+      "step": 111
+    },
+    {
+      "epoch": 0.01870759637130066,
+      "grad_norm": 0.5689477920532227,
+      "learning_rate": 0.0001121311475409836,
+      "loss": 6.608,
+      "num_input_tokens_seen": 14942208,
+      "step": 114
+    },
+    {
+      "epoch": 0.019199901538966467,
+      "grad_norm": 0.5650547742843628,
+      "learning_rate": 0.00011508196721311474,
+      "loss": 6.6641,
+      "num_input_tokens_seen": 15335424,
+      "step": 117
+    },
+    {
+      "epoch": 0.019692206706632274,
+      "grad_norm": 0.6167349219322205,
+      "learning_rate": 0.00011803278688524588,
+      "loss": 6.5976,
+      "num_input_tokens_seen": 15728640,
+      "step": 120
+    },
+    {
+      "epoch": 0.02018451187429808,
+      "grad_norm": 0.5514015555381775,
+      "learning_rate": 0.00012098360655737703,
+      "loss": 6.5955,
+      "num_input_tokens_seen": 16121856,
+      "step": 123
+    },
+    {
+      "epoch": 0.020676817041963887,
+      "grad_norm": 0.8216041326522827,
+      "learning_rate": 0.0001239344262295082,
+      "loss": 6.6051,
+      "num_input_tokens_seen": 16515072,
+      "step": 126
+    },
+    {
+      "epoch": 0.021169122209629694,
+      "grad_norm": 0.6293138265609741,
+      "learning_rate": 0.00012688524590163933,
+      "loss": 6.6251,
+      "num_input_tokens_seen": 16908288,
+      "step": 129
+    },
+    {
+      "epoch": 0.0216614273772955,
+      "grad_norm": 0.6225654482841492,
+      "learning_rate": 0.0001298360655737705,
+      "loss": 6.5213,
+      "num_input_tokens_seen": 17301504,
+      "step": 132
+    },
+    {
+      "epoch": 0.022153732544961308,
+      "grad_norm": 0.7211737632751465,
+      "learning_rate": 0.00013278688524590162,
+      "loss": 6.5802,
+      "num_input_tokens_seen": 17694720,
+      "step": 135
+    },
+    {
+      "epoch": 0.022646037712627114,
+      "grad_norm": 0.9310851097106934,
+      "learning_rate": 0.00013573770491803278,
+      "loss": 6.5739,
+      "num_input_tokens_seen": 18087936,
+      "step": 138
+    },
+    {
+      "epoch": 0.02313834288029292,
+      "grad_norm": 0.9796333909034729,
+      "learning_rate": 0.00013868852459016394,
+      "loss": 6.6066,
+      "num_input_tokens_seen": 18481152,
+      "step": 141
+    },
+    {
+      "epoch": 0.023630648047958728,
+      "grad_norm": 0.9219833016395569,
+      "learning_rate": 0.00014163934426229507,
+      "loss": 6.5536,
+      "num_input_tokens_seen": 18874368,
+      "step": 144
+    },
+    {
+      "epoch": 0.024122953215624535,
+      "grad_norm": 0.9721167087554932,
+      "learning_rate": 0.00014459016393442622,
+      "loss": 6.5368,
+      "num_input_tokens_seen": 19267584,
+      "step": 147
+    },
+    {
+      "epoch": 0.02461525838329034,
+      "grad_norm": 0.9795536994934082,
+      "learning_rate": 0.00014754098360655736,
+      "loss": 6.5266,
+      "num_input_tokens_seen": 19660800,
+      "step": 150
+    },
+    {
+      "epoch": 0.025107563550956148,
+      "grad_norm": 0.7936388850212097,
+      "learning_rate": 0.0001504918032786885,
+      "loss": 6.5445,
+      "num_input_tokens_seen": 20054016,
+      "step": 153
+    },
+    {
+      "epoch": 0.025599868718621955,
+      "grad_norm": 0.6322927474975586,
+      "learning_rate": 0.00015344262295081964,
+      "loss": 6.5434,
+      "num_input_tokens_seen": 20447232,
+      "step": 156
+    },
+    {
+      "epoch": 0.02609217388628776,
+      "grad_norm": 0.6971911191940308,
+      "learning_rate": 0.0001563934426229508,
+      "loss": 6.5021,
+      "num_input_tokens_seen": 20840448,
+      "step": 159
+    },
+    {
+      "epoch": 0.02658447905395357,
+      "grad_norm": 0.6323833465576172,
+      "learning_rate": 0.00015934426229508193,
+      "loss": 6.4679,
+      "num_input_tokens_seen": 21233664,
+      "step": 162
+    },
+    {
+      "epoch": 0.027076784221619375,
+      "grad_norm": 0.590775728225708,
+      "learning_rate": 0.00016229508196721312,
+      "loss": 6.4743,
+      "num_input_tokens_seen": 21626880,
+      "step": 165
+    },
+    {
+      "epoch": 0.027569089389285182,
+      "grad_norm": 0.8365766406059265,
+      "learning_rate": 0.00016524590163934425,
+      "loss": 6.4743,
+      "num_input_tokens_seen": 22020096,
+      "step": 168
+    },
+    {
+      "epoch": 0.02806139455695099,
+      "grad_norm": 0.9162667393684387,
+      "learning_rate": 0.0001681967213114754,
+      "loss": 6.4554,
+      "num_input_tokens_seen": 22413312,
+      "step": 171
+    },
+    {
+      "epoch": 0.028553699724616795,
+      "grad_norm": 0.8129913210868835,
+      "learning_rate": 0.00017114754098360654,
+      "loss": 6.4733,
+      "num_input_tokens_seen": 22806528,
+      "step": 174
+    },
+    {
+      "epoch": 0.029046004892282602,
+      "grad_norm": 1.118039608001709,
+      "learning_rate": 0.0001740983606557377,
+      "loss": 6.4847,
+      "num_input_tokens_seen": 23199744,
+      "step": 177
+    },
+    {
+      "epoch": 0.02953831005994841,
+      "grad_norm": 1.1731743812561035,
+      "learning_rate": 0.00017704918032786883,
+      "loss": 6.469,
+      "num_input_tokens_seen": 23592960,
+      "step": 180
+    },
+    {
+      "epoch": 0.030030615227614216,
+      "grad_norm": 0.8692623972892761,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 6.4902,
+      "num_input_tokens_seen": 23986176,
+      "step": 183
+    },
+    {
+      "epoch": 0.030522920395280023,
+      "grad_norm": 0.924551784992218,
+      "learning_rate": 0.00018295081967213112,
+      "loss": 6.4257,
+      "num_input_tokens_seen": 24379392,
+      "step": 186
+    },
+    {
+      "epoch": 0.031015225562945833,
+      "grad_norm": 0.7523306012153625,
+      "learning_rate": 0.00018590163934426227,
+      "loss": 6.4386,
+      "num_input_tokens_seen": 24772608,
+      "step": 189
+    },
+    {
+      "epoch": 0.03150753073061164,
+      "grad_norm": 0.7796934843063354,
+      "learning_rate": 0.00018885245901639343,
+      "loss": 6.4095,
+      "num_input_tokens_seen": 25165824,
+      "step": 192
+    },
+    {
+      "epoch": 0.03199983589827744,
+      "grad_norm": 1.026790738105774,
+      "learning_rate": 0.0001918032786885246,
+      "loss": 6.3698,
+      "num_input_tokens_seen": 25559040,
+      "step": 195
+    },
+    {
+      "epoch": 0.03249214106594325,
+      "grad_norm": 0.9334837794303894,
+      "learning_rate": 0.00019475409836065572,
+      "loss": 6.3773,
+      "num_input_tokens_seen": 25952256,
+      "step": 198
+    },
+    {
+      "epoch": 0.032984446233609056,
+      "grad_norm": 1.4499056339263916,
+      "learning_rate": 0.00019770491803278688,
+      "loss": 6.4132,
+      "num_input_tokens_seen": 26345472,
+      "step": 201
+    },
+    {
+      "epoch": 0.03347675140127487,
+      "grad_norm": 1.2045769691467285,
+      "learning_rate": 0.000200655737704918,
+      "loss": 6.4163,
+      "num_input_tokens_seen": 26738688,
+      "step": 204
+    },
+    {
+      "epoch": 0.03396905656894067,
+      "grad_norm": 1.3456003665924072,
+      "learning_rate": 0.00020360655737704917,
+      "loss": 6.4426,
+      "num_input_tokens_seen": 27131904,
+      "step": 207
+    },
+    {
+      "epoch": 0.03446136173660648,
+      "grad_norm": 1.1916530132293701,
+      "learning_rate": 0.0002065573770491803,
+      "loss": 6.4234,
+      "num_input_tokens_seen": 27525120,
+      "step": 210
+    },
+    {
+      "epoch": 0.03495366690427228,
+      "grad_norm": 1.1489585638046265,
+      "learning_rate": 0.00020950819672131146,
+      "loss": 6.4218,
+      "num_input_tokens_seen": 27918336,
+      "step": 213
+    },
+    {
+      "epoch": 0.035445972071938094,
+      "grad_norm": 0.6667603850364685,
+      "learning_rate": 0.00021245901639344259,
+      "loss": 6.3939,
+      "num_input_tokens_seen": 28311552,
+      "step": 216
+    },
+    {
+      "epoch": 0.0359382772396039,
+      "grad_norm": 0.8607211112976074,
+      "learning_rate": 0.00021540983606557374,
+      "loss": 6.3799,
+      "num_input_tokens_seen": 28704768,
+      "step": 219
+    },
+    {
+      "epoch": 0.03643058240726971,
+      "grad_norm": 0.837247908115387,
+      "learning_rate": 0.0002183606557377049,
+      "loss": 6.3738,
+      "num_input_tokens_seen": 29097984,
+      "step": 222
+    },
+    {
+      "epoch": 0.03692288757493551,
+      "grad_norm": 0.7893315553665161,
+      "learning_rate": 0.00022131147540983606,
+      "loss": 6.3407,
+      "num_input_tokens_seen": 29491200,
+      "step": 225
+    },
+    {
+      "epoch": 0.03741519274260132,
+      "grad_norm": 0.8632616400718689,
+      "learning_rate": 0.0002242622950819672,
+      "loss": 6.3659,
+      "num_input_tokens_seen": 29884416,
+      "step": 228
+    },
+    {
+      "epoch": 0.037907497910267124,
+      "grad_norm": 0.8238740563392639,
+      "learning_rate": 0.00022721311475409835,
+      "loss": 6.3464,
+      "num_input_tokens_seen": 30277632,
+      "step": 231
+    },
+    {
+      "epoch": 0.038399803077932934,
+      "grad_norm": 0.9048452377319336,
+      "learning_rate": 0.00023016393442622948,
+      "loss": 6.3345,
+      "num_input_tokens_seen": 30670848,
+      "step": 234
+    },
+    {
+      "epoch": 0.03889210824559874,
+      "grad_norm": 0.8803463578224182,
+      "learning_rate": 0.00023311475409836064,
+      "loss": 6.3739,
+      "num_input_tokens_seen": 31064064,
+      "step": 237
+    },
+    {
+      "epoch": 0.03938441341326455,
+      "grad_norm": 0.9749881029129028,
+      "learning_rate": 0.00023606557377049177,
+      "loss": 6.3295,
+      "num_input_tokens_seen": 31457280,
+      "step": 240
+    },
+    {
+      "epoch": 0.03987671858093035,
+      "grad_norm": 0.903581976890564,
+      "learning_rate": 0.00023901639344262293,
+      "loss": 6.3686,
+      "num_input_tokens_seen": 31850496,
+      "step": 243
+    },
+    {
+      "epoch": 0.04036902374859616,
+      "grad_norm": 0.7235903143882751,
+      "learning_rate": 0.00024196721311475406,
+      "loss": 6.3202,
+      "num_input_tokens_seen": 32243712,
+      "step": 246
+    },
+    {
+      "epoch": 0.040861328916261964,
+      "grad_norm": 0.9854725003242493,
+      "learning_rate": 0.0002449180327868852,
+      "loss": 6.35,
+      "num_input_tokens_seen": 32636928,
+      "step": 249
+    },
+    {
+      "epoch": 0.041353634083927775,
+      "grad_norm": 1.2127262353897095,
+      "learning_rate": 0.0002478688524590164,
+      "loss": 6.3013,
+      "num_input_tokens_seen": 33030144,
+      "step": 252
+    },
+    {
+      "epoch": 0.041845939251593585,
+      "grad_norm": 1.090453028678894,
+      "learning_rate": 0.00025081967213114756,
+      "loss": 6.3094,
+      "num_input_tokens_seen": 33423360,
+      "step": 255
+    },
+    {
+      "epoch": 0.04233824441925939,
+      "grad_norm": 1.2343584299087524,
+      "learning_rate": 0.00025377049180327866,
+      "loss": 6.3217,
+      "num_input_tokens_seen": 33816576,
+      "step": 258
+    },
+    {
+      "epoch": 0.0428305495869252,
+      "grad_norm": 1.1911264657974243,
+      "learning_rate": 0.0002567213114754098,
+      "loss": 6.2728,
+      "num_input_tokens_seen": 34209792,
+      "step": 261
+    },
+    {
+      "epoch": 0.043322854754591,
+      "grad_norm": 1.042060375213623,
+      "learning_rate": 0.000259672131147541,
+      "loss": 6.3082,
+      "num_input_tokens_seen": 34603008,
+      "step": 264
+    },
+    {
+      "epoch": 0.04381515992225681,
+      "grad_norm": 0.9093752503395081,
+      "learning_rate": 0.0002626229508196721,
+      "loss": 6.266,
+      "num_input_tokens_seen": 34996224,
+      "step": 267
+    },
+    {
+      "epoch": 0.044307465089922615,
+      "grad_norm": 1.0649739503860474,
+      "learning_rate": 0.00026557377049180324,
+      "loss": 6.2528,
+      "num_input_tokens_seen": 35389440,
+      "step": 270
+    },
+    {
+      "epoch": 0.044799770257588425,
+      "grad_norm": 1.3432960510253906,
+      "learning_rate": 0.0002685245901639344,
+      "loss": 6.2903,
+      "num_input_tokens_seen": 35782656,
+      "step": 273
+    },
+    {
+      "epoch": 0.04529207542525423,
+      "grad_norm": 1.2440747022628784,
+      "learning_rate": 0.00027147540983606556,
+      "loss": 6.274,
+      "num_input_tokens_seen": 36175872,
+      "step": 276
+    },
+    {
+      "epoch": 0.04578438059292004,
+      "grad_norm": 1.3492511510849,
+      "learning_rate": 0.00027442622950819666,
+      "loss": 6.2834,
+      "num_input_tokens_seen": 36569088,
+      "step": 279
+    },
+    {
+      "epoch": 0.04627668576058584,
+      "grad_norm": 1.2173079252243042,
+      "learning_rate": 0.00027737704918032787,
+      "loss": 6.2886,
+      "num_input_tokens_seen": 36962304,
+      "step": 282
+    },
+    {
+      "epoch": 0.04676899092825165,
+      "grad_norm": 1.230082631111145,
+      "learning_rate": 0.00028032786885245903,
+      "loss": 6.2568,
+      "num_input_tokens_seen": 37355520,
+      "step": 285
+    },
+    {
+      "epoch": 0.047261296095917456,
+      "grad_norm": 1.3385438919067383,
+      "learning_rate": 0.00028327868852459013,
+      "loss": 6.2438,
+      "num_input_tokens_seen": 37748736,
+      "step": 288
+    },
+    {
+      "epoch": 0.047753601263583266,
+      "grad_norm": 1.4713610410690308,
+      "learning_rate": 0.0002862295081967213,
+      "loss": 6.307,
+      "num_input_tokens_seen": 38141952,
+      "step": 291
+    },
+    {
+      "epoch": 0.04824590643124907,
+      "grad_norm": 1.210629940032959,
+      "learning_rate": 0.00028918032786885245,
+      "loss": 6.2704,
+      "num_input_tokens_seen": 38535168,
+      "step": 294
+    },
+    {
+      "epoch": 0.04873821159891488,
+      "grad_norm": 1.1790496110916138,
+      "learning_rate": 0.00029213114754098355,
+      "loss": 6.1878,
+      "num_input_tokens_seen": 38928384,
+      "step": 297
+    },
+    {
+      "epoch": 0.04923051676658068,
+      "grad_norm": 1.0269746780395508,
+      "learning_rate": 0.0002950819672131147,
+      "loss": 6.279,
+      "num_input_tokens_seen": 39321600,
+      "step": 300
+    },
+    {
+      "epoch": 0.04972282193424649,
+      "grad_norm": 1.2352854013442993,
+      "learning_rate": 0.00029803278688524587,
+      "loss": 6.2379,
+      "num_input_tokens_seen": 39714816,
+      "step": 303
+    },
+    {
+      "epoch": 0.050215127101912296,
+      "grad_norm": 1.5501960515975952,
+      "learning_rate": 0.00029950940277884624,
+      "loss": 6.2097,
+      "num_input_tokens_seen": 40108032,
+      "step": 306
+    },
+    {
+      "epoch": 0.05070743226957811,
+      "grad_norm": 1.5629328489303589,
+      "learning_rate": 0.0002980519274494139,
+      "loss": 6.1982,
+      "num_input_tokens_seen": 40501248,
+      "step": 309
+    },
+    {
+      "epoch": 0.05119973743724391,
+      "grad_norm": 1.323115587234497,
+      "learning_rate": 0.0002966155242578669,
+      "loss": 6.2333,
+      "num_input_tokens_seen": 40894464,
+      "step": 312
+    },
+    {
+      "epoch": 0.05169204260490972,
+      "grad_norm": 2.0143845081329346,
+      "learning_rate": 0.00029519969028245457,
+      "loss": 6.2445,
+      "num_input_tokens_seen": 41287680,
+      "step": 315
+    },
+    {
+      "epoch": 0.05218434777257552,
+      "grad_norm": 1.3097857236862183,
+      "learning_rate": 0.0002938039392468745,
+      "loss": 6.188,
+      "num_input_tokens_seen": 41680896,
+      "step": 318
+    },
+    {
+      "epoch": 0.052676652940241334,
+      "grad_norm": 1.5996229648590088,
+      "learning_rate": 0.000292427800818576,
+      "loss": 6.1792,
+      "num_input_tokens_seen": 42074112,
+      "step": 321
+    },
+    {
+      "epoch": 0.05316895810790714,
+      "grad_norm": 1.2864151000976562,
+      "learning_rate": 0.000291070819942883,
+      "loss": 6.1811,
+      "num_input_tokens_seen": 42467328,
+      "step": 324
+    },
+    {
+      "epoch": 0.05366126327557295,
+      "grad_norm": 1.683950662612915,
+      "learning_rate": 0.00028973255621079304,
+      "loss": 6.1795,
+      "num_input_tokens_seen": 42860544,
+      "step": 327
+    },
+    {
+      "epoch": 0.05415356844323875,
+      "grad_norm": 1.5458773374557495,
+      "learning_rate": 0.0002884125832584601,
+      "loss": 6.1799,
+      "num_input_tokens_seen": 43253760,
+      "step": 330
+    },
+    {
+      "epoch": 0.05464587361090456,
+      "grad_norm": 1.4698423147201538,
+      "learning_rate": 0.0002871104881964997,
+      "loss": 6.1903,
+      "num_input_tokens_seen": 43646976,
+      "step": 333
+    },
+    {
+      "epoch": 0.055138178778570364,
+      "grad_norm": 1.4713280200958252,
+      "learning_rate": 0.0002858258710673835,
+      "loss": 6.2024,
+      "num_input_tokens_seen": 44040192,
+      "step": 336
+    },
+    {
+      "epoch": 0.055630483946236174,
+      "grad_norm": 1.2734639644622803,
+      "learning_rate": 0.000284558344329302,
+      "loss": 6.1246,
+      "num_input_tokens_seen": 44433408,
+      "step": 339
+    },
+    {
+      "epoch": 0.05612278911390198,
+      "grad_norm": 1.0039596557617188,
+      "learning_rate": 0.00028330753236498467,
+      "loss": 6.1914,
+      "num_input_tokens_seen": 44826624,
+      "step": 342
+    },
+    {
+      "epoch": 0.05661509428156779,
+      "grad_norm": 1.0492082834243774,
+      "learning_rate": 0.0002820730710140625,
+      "loss": 6.1344,
+      "num_input_tokens_seen": 45219840,
+      "step": 345
+    },
+    {
+      "epoch": 0.05710739944923359,
+      "grad_norm": 0.9763434529304504,
+      "learning_rate": 0.0002808546071276517,
+      "loss": 6.159,
+      "num_input_tokens_seen": 45613056,
+      "step": 348
+    },
+    {
+      "epoch": 0.0575997046168994,
+      "grad_norm": 0.7468030452728271,
+      "learning_rate": 0.00027965179814392076,
+      "loss": 6.1567,
+      "num_input_tokens_seen": 46006272,
+      "step": 351
+    },
+    {
+      "epoch": 0.058092009784565204,
+      "grad_norm": 0.8997909426689148,
+      "learning_rate": 0.0002784643116834829,
+      "loss": 6.1655,
+      "num_input_tokens_seen": 46399488,
+      "step": 354
+    },
+    {
+      "epoch": 0.058584314952231015,
+      "grad_norm": 0.9638291001319885,
+      "learning_rate": 0.00027729182516352875,
+      "loss": 6.0947,
+      "num_input_tokens_seen": 46792704,
+      "step": 357
+    },
+    {
+      "epoch": 0.05907662011989682,
+      "grad_norm": 0.84872967004776,
+      "learning_rate": 0.0002761340254296815,
+      "loss": 6.1313,
+      "num_input_tokens_seen": 47185920,
+      "step": 360
+    },
+    {
+      "epoch": 0.05956892528756263,
+      "grad_norm": 2.2599406242370605,
+      "learning_rate": 0.0002749906084046213,
+      "loss": 6.1251,
+      "num_input_tokens_seen": 47579136,
+      "step": 363
+    },
+    {
+      "epoch": 0.06006123045522843,
+      "grad_norm": 2.0698065757751465,
+      "learning_rate": 0.00027386127875258305,
+      "loss": 6.1651,
+      "num_input_tokens_seen": 47972352,
+      "step": 366
+    },
+    {
+      "epoch": 0.06055353562289424,
+      "grad_norm": 0.8866053223609924,
+      "learning_rate": 0.0002727457495588868,
+      "loss": 6.1123,
+      "num_input_tokens_seen": 48365568,
+      "step": 369
+    },
+    {
+      "epoch": 0.061045840790560045,
+      "grad_norm": 1.2379621267318726,
+      "learning_rate": 0.0002716437420237123,
+      "loss": 6.1493,
+      "num_input_tokens_seen": 48758784,
+      "step": 372
+    },
+    {
+      "epoch": 0.061538145958225855,
+      "grad_norm": 1.01674222946167,
+      "learning_rate": 0.00027055498516937365,
+      "loss": 6.1375,
+      "num_input_tokens_seen": 49152000,
+      "step": 375
+    },
+    {
+      "epoch": 0.062030451125891665,
+      "grad_norm": 0.9889538884162903,
+      "learning_rate": 0.0002694792155603983,
+      "loss": 6.0978,
+      "num_input_tokens_seen": 49545216,
+      "step": 378
+    },
+    {
+      "epoch": 0.06252275629355747,
+      "grad_norm": 0.6733151078224182,
+      "learning_rate": 0.00026841617703575205,
+      "loss": 6.134,
+      "num_input_tokens_seen": 49938432,
+      "step": 381
+    },
+    {
+      "epoch": 0.06301506146122328,
+      "grad_norm": 0.9297388792037964,
+      "learning_rate": 0.00026736562045259293,
+      "loss": 6.1135,
+      "num_input_tokens_seen": 50331648,
+      "step": 384
+    },
+    {
+      "epoch": 0.06350736662888909,
+      "grad_norm": 0.6994606256484985,
+      "learning_rate": 0.0002663273034409719,
+      "loss": 6.0961,
+      "num_input_tokens_seen": 50724864,
+      "step": 387
+    },
+    {
+      "epoch": 0.06399967179655489,
+      "grad_norm": 0.9161903262138367,
+      "learning_rate": 0.0002653009901689313,
+      "loss": 6.1339,
+      "num_input_tokens_seen": 51118080,
+      "step": 390
+    },
+    {
+      "epoch": 0.0644919769642207,
+      "grad_norm": 0.8527827262878418,
+      "learning_rate": 0.000264286451117485,
+      "loss": 6.0692,
+      "num_input_tokens_seen": 51511296,
+      "step": 393
+    },
+    {
+      "epoch": 0.0649842821318865,
+      "grad_norm": 0.7189248204231262,
+      "learning_rate": 0.0002632834628649923,
+      "loss": 6.0884,
+      "num_input_tokens_seen": 51904512,
+      "step": 396
+    },
+    {
+      "epoch": 0.06547658729955232,
+      "grad_norm": 0.9880132079124451,
+      "learning_rate": 0.00026229180788046543,
+      "loss": 6.0703,
+      "num_input_tokens_seen": 52297728,
+      "step": 399
+    },
+    {
+      "epoch": 0.06564068902210758,
+      "eval_accuracy": 0.1701481843347989,
+      "eval_loss": 6.233191013336182,
+      "eval_runtime": 110.6814,
+      "eval_samples_per_second": 2.71,
+      "eval_steps_per_second": 1.355,
+      "num_input_tokens_seen": 52428800,
+      "step": 400
+    },
+    {
+      "epoch": 0.06596889246721811,
+      "grad_norm": 1.2372077703475952,
+      "learning_rate": 0.0002613112743253766,
+      "loss": 6.0785,
+      "num_input_tokens_seen": 52690944,
+      "step": 402
+    },
+    {
+      "epoch": 0.06646119763488392,
+      "grad_norm": 1.0689289569854736,
+      "learning_rate": 0.0002603416558635551,
+      "loss": 6.0829,
+      "num_input_tokens_seen": 53084160,
+      "step": 405
+    },
+    {
+      "epoch": 0.06695350280254973,
+      "grad_norm": 1.0362907648086548,
+      "learning_rate": 0.0002593827514787864,
+      "loss": 6.0684,
+      "num_input_tokens_seen": 53477376,
+      "step": 408
+    },
+    {
+      "epoch": 0.06744580797021554,
+      "grad_norm": 0.9470035433769226,
+      "learning_rate": 0.00025843436529974725,
+      "loss": 6.0331,
+      "num_input_tokens_seen": 53870592,
+      "step": 411
+    },
+    {
+      "epoch": 0.06793811313788134,
+      "grad_norm": 0.9663071036338806,
+      "learning_rate": 0.00025749630643193106,
+      "loss": 6.0694,
+      "num_input_tokens_seen": 54263808,
+      "step": 414
+    },
+    {
+      "epoch": 0.06843041830554715,
+      "grad_norm": 1.1301859617233276,
+      "learning_rate": 0.0002565683887962357,
+      "loss": 6.0433,
+      "num_input_tokens_seen": 54657024,
+      "step": 417
+    },
+    {
+      "epoch": 0.06892272347321296,
+      "grad_norm": 1.1818194389343262,
+      "learning_rate": 0.000255650430973904,
+      "loss": 6.0824,
+      "num_input_tokens_seen": 55050240,
+      "step": 420
+    },
+    {
+      "epoch": 0.06941502864087877,
+      "grad_norm": 1.000831127166748,
+      "learning_rate": 0.00025474225605752297,
+      "loss": 6.0066,
+      "num_input_tokens_seen": 55443456,
+      "step": 423
+    },
+    {
+      "epoch": 0.06990733380854457,
+      "grad_norm": 1.0827128887176514,
+      "learning_rate": 0.00025384369150780535,
+      "loss": 6.0375,
+      "num_input_tokens_seen": 55836672,
+      "step": 426
+    },
+    {
+      "epoch": 0.07039963897621038,
+      "grad_norm": 0.8842981457710266,
+      "learning_rate": 0.00025295456901588867,
+      "loss": 6.031,
+      "num_input_tokens_seen": 56229888,
+      "step": 429
+    },
+    {
+      "epoch": 0.07089194414387619,
+      "grad_norm": 0.734250545501709,
+      "learning_rate": 0.00025207472437090286,
+      "loss": 6.0575,
+      "num_input_tokens_seen": 56623104,
+      "step": 432
+    },
+    {
+      "epoch": 0.071384249311542,
+      "grad_norm": 1.13999605178833,
+      "learning_rate": 0.0002512039973325704,
+      "loss": 6.0268,
+      "num_input_tokens_seen": 57016320,
+      "step": 435
+    },
+    {
+      "epoch": 0.0718765544792078,
+      "grad_norm": 1.5524462461471558,
+      "learning_rate": 0.0002503422315086136,
+      "loss": 6.0097,
+      "num_input_tokens_seen": 57409536,
+      "step": 438
+    },
+    {
+      "epoch": 0.0723688596468736,
+      "grad_norm": 1.8856819868087769,
+      "learning_rate": 0.0002494892742367568,
+      "loss": 6.0882,
+      "num_input_tokens_seen": 57802752,
+      "step": 441
+    },
+    {
+      "epoch": 0.07286116481453941,
+      "grad_norm": 1.1143220663070679,
+      "learning_rate": 0.000248644976471121,
+      "loss": 5.9855,
+      "num_input_tokens_seen": 58195968,
+      "step": 444
+    },
+    {
+      "epoch": 0.07335346998220522,
+      "grad_norm": 1.0051558017730713,
+      "learning_rate": 0.00024780919267281904,
+      "loss": 6.0529,
+      "num_input_tokens_seen": 58589184,
+      "step": 447
+    },
+    {
+      "epoch": 0.07384577514987102,
+      "grad_norm": 0.9793909788131714,
+      "learning_rate": 0.00024698178070456936,
+      "loss": 6.0412,
+      "num_input_tokens_seen": 58982400,
+      "step": 450
+    },
+    {
+      "epoch": 0.07433808031753683,
+      "grad_norm": 1.0926892757415771,
+      "learning_rate": 0.00024616260172915426,
+      "loss": 6.0872,
+      "num_input_tokens_seen": 59375616,
+      "step": 453
+    },
+    {
+      "epoch": 0.07483038548520264,
+      "grad_norm": 0.8595672845840454,
+      "learning_rate": 0.00024535152011155874,
+      "loss": 6.004,
+      "num_input_tokens_seen": 59768832,
+      "step": 456
+    },
+    {
+      "epoch": 0.07532269065286845,
+      "grad_norm": 0.7526682615280151,
+      "learning_rate": 0.00024454840332463316,
+      "loss": 6.0032,
+      "num_input_tokens_seen": 60162048,
+      "step": 459
+    },
+    {
+      "epoch": 0.07581499582053425,
+      "grad_norm": 1.0615681409835815,
+      "learning_rate": 0.00024375312185813004,
+      "loss": 6.0034,
+      "num_input_tokens_seen": 60555264,
+      "step": 462
+    },
+    {
+      "epoch": 0.07630730098820006,
+      "grad_norm": 1.3442180156707764,
+      "learning_rate": 0.00024296554913097476,
+      "loss": 6.0288,
+      "num_input_tokens_seen": 60948480,
+      "step": 465
+    },
+    {
+      "epoch": 0.07679960615586587,
+      "grad_norm": 1.1294552087783813,
+      "learning_rate": 0.00024218556140663327,
+      "loss": 5.9879,
+      "num_input_tokens_seen": 61341696,
+      "step": 468
+    },
+    {
+      "epoch": 0.07729191132353168,
+      "grad_norm": 1.0565030574798584,
+      "learning_rate": 0.00024141303771145015,
+      "loss": 5.9733,
+      "num_input_tokens_seen": 61734912,
+      "step": 471
+    },
+    {
+      "epoch": 0.07778421649119747,
+      "grad_norm": 0.8977353572845459,
+      "learning_rate": 0.00024064785975583342,
+      "loss": 5.9904,
+      "num_input_tokens_seen": 62128128,
+      "step": 474
+    },
+    {
+      "epoch": 0.07827652165886329,
+      "grad_norm": 1.3904660940170288,
+      "learning_rate": 0.00023988991185817037,
+      "loss": 6.0155,
+      "num_input_tokens_seen": 62521344,
+      "step": 477
+    },
+    {
+      "epoch": 0.0787688268265291,
+      "grad_norm": 1.113961935043335,
+      "learning_rate": 0.0002391390808713624,
+      "loss": 5.9818,
+      "num_input_tokens_seen": 62914560,
+      "step": 480
+    },
+    {
+      "epoch": 0.0792611319941949,
+      "grad_norm": 0.8899008631706238,
+      "learning_rate": 0.00023839525611187392,
+      "loss": 5.9606,
+      "num_input_tokens_seen": 63307776,
+      "step": 483
+    },
+    {
+      "epoch": 0.0797534371618607,
+      "grad_norm": 0.802760660648346,
+      "learning_rate": 0.00023765832929119373,
+      "loss": 5.9345,
+      "num_input_tokens_seen": 63700992,
+      "step": 486
+    },
+    {
+      "epoch": 0.08024574232952651,
+      "grad_norm": 0.7717273235321045,
+      "learning_rate": 0.00023692819444961244,
+      "loss": 6.0008,
+      "num_input_tokens_seen": 64094208,
+      "step": 489
+    },
+    {
+      "epoch": 0.08073804749719232,
+      "grad_norm": 0.7769405245780945,
+      "learning_rate": 0.00023620474789222436,
+      "loss": 5.9624,
+      "num_input_tokens_seen": 64487424,
+      "step": 492
+    },
+    {
+      "epoch": 0.08123035266485813,
+      "grad_norm": 0.8420572876930237,
+      "learning_rate": 0.00023548788812706575,
+      "loss": 5.9664,
+      "num_input_tokens_seen": 64880640,
+      "step": 495
+    },
+    {
+      "epoch": 0.08172265783252393,
+      "grad_norm": 0.6086786389350891,
+      "learning_rate": 0.00023477751580530627,
+      "loss": 5.952,
+      "num_input_tokens_seen": 65273856,
+      "step": 498
+    },
+    {
+      "epoch": 0.08221496300018974,
+      "grad_norm": 0.9240081906318665,
+      "learning_rate": 0.00023407353366341235,
+      "loss": 5.9447,
+      "num_input_tokens_seen": 65667072,
+      "step": 501
+    },
+    {
+      "epoch": 0.08270726816785555,
+      "grad_norm": 0.9431899189949036,
+      "learning_rate": 0.0002333758464672077,
+      "loss": 5.9525,
+      "num_input_tokens_seen": 66060288,
+      "step": 504
+    },
+    {
+      "epoch": 0.08319957333552136,
+      "grad_norm": 0.8989746570587158,
+      "learning_rate": 0.0002326843609577565,
+      "loss": 5.9485,
+      "num_input_tokens_seen": 66453504,
+      "step": 507
+    },
+    {
+      "epoch": 0.08369187850318717,
+      "grad_norm": 0.6996486783027649,
+      "learning_rate": 0.00023199898579900018,
+      "loss": 5.9664,
+      "num_input_tokens_seen": 66846720,
+      "step": 510
+    },
+    {
+      "epoch": 0.08418418367085297,
+      "grad_norm": 0.8220193386077881,
+      "learning_rate": 0.00023131963152708105,
+      "loss": 5.9256,
+      "num_input_tokens_seen": 67239936,
+      "step": 513
+    },
+    {
+      "epoch": 0.08467648883851878,
+      "grad_norm": 0.8553633093833923,
+      "learning_rate": 0.0002306462105012884,
+      "loss": 5.9676,
+      "num_input_tokens_seen": 67633152,
+      "step": 516
+    },
+    {
+      "epoch": 0.08516879400618459,
+      "grad_norm": 0.6745681166648865,
+      "learning_rate": 0.00022997863685656676,
+      "loss": 5.9794,
+      "num_input_tokens_seen": 68026368,
+      "step": 519
+    },
+    {
+      "epoch": 0.0856610991738504,
+      "grad_norm": 0.638691246509552,
+      "learning_rate": 0.00022931682645752736,
+      "loss": 5.9337,
+      "num_input_tokens_seen": 68419584,
+      "step": 522
+    },
+    {
+      "epoch": 0.0861534043415162,
+      "grad_norm": 0.7927042841911316,
+      "learning_rate": 0.00022866069685390685,
+      "loss": 5.933,
+      "num_input_tokens_seen": 68812800,
+      "step": 525
+    },
+    {
+      "epoch": 0.086645709509182,
+      "grad_norm": 1.2133303880691528,
+      "learning_rate": 0.00022801016723742026,
+      "loss": 5.9536,
+      "num_input_tokens_seen": 69206016,
+      "step": 528
+    },
+    {
+      "epoch": 0.08713801467684781,
+      "grad_norm": 0.8164889812469482,
+      "learning_rate": 0.00022736515839995644,
+      "loss": 5.9201,
+      "num_input_tokens_seen": 69599232,
+      "step": 531
+    },
+    {
+      "epoch": 0.08763031984451362,
+      "grad_norm": 1.072871208190918,
+      "learning_rate": 0.00022672559269306688,
+      "loss": 5.9214,
+      "num_input_tokens_seen": 69992448,
+      "step": 534
+    },
+    {
+      "epoch": 0.08812262501217942,
+      "grad_norm": 1.145111322402954,
+      "learning_rate": 0.00022609139398870132,
+      "loss": 5.9051,
+      "num_input_tokens_seen": 70385664,
+      "step": 537
+    },
+    {
+      "epoch": 0.08861493017984523,
+      "grad_norm": 1.074684739112854,
+      "learning_rate": 0.00022546248764114467,
+      "loss": 5.9146,
+      "num_input_tokens_seen": 70778880,
+      "step": 540
+    },
+    {
+      "epoch": 0.08910723534751104,
+      "grad_norm": 1.3423213958740234,
+      "learning_rate": 0.000224838800450112,
+      "loss": 5.8908,
+      "num_input_tokens_seen": 71172096,
+      "step": 543
+    },
+    {
+      "epoch": 0.08959954051517685,
+      "grad_norm": 0.901797354221344,
+      "learning_rate": 0.00022422026062496062,
+      "loss": 5.8608,
+      "num_input_tokens_seen": 71565312,
+      "step": 546
+    },
+    {
+      "epoch": 0.09009184568284265,
+      "grad_norm": 1.848212480545044,
+      "learning_rate": 0.00022360679774997895,
+      "loss": 5.9256,
+      "num_input_tokens_seen": 71958528,
+      "step": 549
+    },
+    {
+      "epoch": 0.09058415085050846,
+      "grad_norm": 1.1397554874420166,
+      "learning_rate": 0.00022299834275071466,
+      "loss": 5.9315,
+      "num_input_tokens_seen": 72351744,
+      "step": 552
+    },
+    {
+      "epoch": 0.09107645601817427,
+      "grad_norm": 1.3297491073608398,
+      "learning_rate": 0.00022239482786130492,
+      "loss": 5.8853,
+      "num_input_tokens_seen": 72744960,
+      "step": 555
+    },
+    {
+      "epoch": 0.09156876118584008,
+      "grad_norm": 1.0274791717529297,
+      "learning_rate": 0.00022179618659277431,
+      "loss": 5.9317,
+      "num_input_tokens_seen": 73138176,
+      "step": 558
+    },
+    {
+      "epoch": 0.09206106635350587,
+      "grad_norm": 0.9891603589057922,
+      "learning_rate": 0.00022120235370226617,
+      "loss": 5.8753,
+      "num_input_tokens_seen": 73531392,
+      "step": 561
+    },
+    {
+      "epoch": 0.09255337152117168,
+      "grad_norm": 1.1428287029266357,
+      "learning_rate": 0.00022061326516317517,
+      "loss": 5.8625,
+      "num_input_tokens_seen": 73924608,
+      "step": 564
+    },
+    {
+      "epoch": 0.0930456766888375,
+      "grad_norm": 1.1694408655166626,
+      "learning_rate": 0.00022002885813615086,
+      "loss": 5.882,
+      "num_input_tokens_seen": 74317824,
+      "step": 567
+    },
+    {
+      "epoch": 0.0935379818565033,
+      "grad_norm": 0.9096398949623108,
+      "learning_rate": 0.00021944907094094087,
+      "loss": 5.874,
+      "num_input_tokens_seen": 74711040,
+      "step": 570
+    },
+    {
+      "epoch": 0.0940302870241691,
+      "grad_norm": 0.8736022710800171,
+      "learning_rate": 0.00021887384302904644,
+      "loss": 5.9158,
+      "num_input_tokens_seen": 75104256,
+      "step": 573
+    },
+    {
+      "epoch": 0.09452259219183491,
+      "grad_norm": 0.6410363912582397,
+      "learning_rate": 0.00021830311495716224,
+      "loss": 5.8954,
+      "num_input_tokens_seen": 75497472,
+      "step": 576
+    },
+    {
+      "epoch": 0.09501489735950072,
+      "grad_norm": 0.8881443738937378,
+      "learning_rate": 0.00021773682836137405,
+      "loss": 5.8734,
+      "num_input_tokens_seen": 75890688,
+      "step": 579
+    },
+    {
+      "epoch": 0.09550720252716653,
+      "grad_norm": 0.7709605097770691,
+      "learning_rate": 0.00021717492593208875,
+      "loss": 5.8888,
+      "num_input_tokens_seen": 76283904,
+      "step": 582
+    },
+    {
+      "epoch": 0.09599950769483233,
+      "grad_norm": 0.6798803210258484,
+      "learning_rate": 0.00021661735138967265,
+      "loss": 5.8759,
+      "num_input_tokens_seen": 76677120,
+      "step": 585
+    },
+    {
+      "epoch": 0.09649181286249814,
+      "grad_norm": 0.5487522482872009,
+      "learning_rate": 0.0002160640494607739,
+      "loss": 5.8661,
+      "num_input_tokens_seen": 77070336,
+      "step": 588
+    },
+    {
+      "epoch": 0.09698411803016395,
+      "grad_norm": 0.6725640892982483,
+      "learning_rate": 0.00021551496585530715,
+      "loss": 5.8954,
+      "num_input_tokens_seen": 77463552,
+      "step": 591
+    },
+    {
+      "epoch": 0.09747642319782976,
+      "grad_norm": 0.6947436928749084,
+      "learning_rate": 0.00021497004724407818,
+      "loss": 5.8171,
+      "num_input_tokens_seen": 77856768,
+      "step": 594
+    },
+    {
+      "epoch": 0.09796872836549556,
+      "grad_norm": 0.9774389266967773,
+      "learning_rate": 0.00021442924123702773,
+      "loss": 5.8996,
+      "num_input_tokens_seen": 78249984,
+      "step": 597
+    },
+    {
+      "epoch": 0.09846103353316137,
+      "grad_norm": 1.041482925415039,
+      "learning_rate": 0.00021389249636207436,
+      "loss": 5.8522,
+      "num_input_tokens_seen": 78643200,
+      "step": 600
+    },
+    {
+      "epoch": 0.09895333870082718,
+      "grad_norm": 0.7540446519851685,
+      "learning_rate": 0.0002133597620445371,
+      "loss": 5.7978,
+      "num_input_tokens_seen": 79036416,
+      "step": 603
+    },
+    {
+      "epoch": 0.09944564386849299,
+      "grad_norm": 0.7133435606956482,
+      "learning_rate": 0.00021283098858711878,
+      "loss": 5.8597,
+      "num_input_tokens_seen": 79429632,
+      "step": 606
+    },
+    {
+      "epoch": 0.09993794903615878,
+      "grad_norm": 0.9211872220039368,
+      "learning_rate": 0.00021230612715043284,
+      "loss": 5.827,
+      "num_input_tokens_seen": 79822848,
+      "step": 609
+    },
+    {
+      "epoch": 0.10043025420382459,
+      "grad_norm": 0.9481165409088135,
+      "learning_rate": 0.00021178512973405518,
+      "loss": 5.8291,
+      "num_input_tokens_seen": 80216064,
+      "step": 612
+    },
+    {
+      "epoch": 0.1009225593714904,
+      "grad_norm": 0.9099063873291016,
+      "learning_rate": 0.00021126794915808552,
+      "loss": 5.8853,
+      "num_input_tokens_seen": 80609280,
+      "step": 615
+    },
+    {
+      "epoch": 0.10141486453915621,
+      "grad_norm": 1.089982032775879,
+      "learning_rate": 0.00021075453904520141,
+      "loss": 5.8042,
+      "num_input_tokens_seen": 81002496,
+      "step": 618
+    },
+    {
+      "epoch": 0.10190716970682201,
+      "grad_norm": 1.2257575988769531,
+      "learning_rate": 0.00021024485380318974,
+      "loss": 5.8316,
+      "num_input_tokens_seen": 81395712,
+      "step": 621
+    },
+    {
+      "epoch": 0.10239947487448782,
+      "grad_norm": 0.922932505607605,
+      "learning_rate": 0.00020973884860794057,
+      "loss": 5.8464,
+      "num_input_tokens_seen": 81788928,
+      "step": 624
+    },
+    {
+      "epoch": 0.10289178004215363,
+      "grad_norm": 0.9160616397857666,
+      "learning_rate": 0.00020923647938688914,
+      "loss": 5.8233,
+      "num_input_tokens_seen": 82182144,
+      "step": 627
+    },
+    {
+      "epoch": 0.10338408520981944,
+      "grad_norm": 1.4857609272003174,
+      "learning_rate": 0.00020873770280289224,
+      "loss": 5.8783,
+      "num_input_tokens_seen": 82575360,
+      "step": 630
+    },
+    {
+      "epoch": 0.10387639037748525,
+      "grad_norm": 0.8238282799720764,
+      "learning_rate": 0.00020824247623852486,
+      "loss": 5.8476,
+      "num_input_tokens_seen": 82968576,
+      "step": 633
+    },
+    {
+      "epoch": 0.10436869554515105,
+      "grad_norm": 0.9078794717788696,
+      "learning_rate": 0.0002077507577807854,
+      "loss": 5.8344,
+      "num_input_tokens_seen": 83361792,
+      "step": 636
+    },
+    {
+      "epoch": 0.10486100071281686,
+      "grad_norm": 0.8983127474784851,
+      "learning_rate": 0.0002072625062061955,
+      "loss": 5.8672,
+      "num_input_tokens_seen": 83755008,
+      "step": 639
+    },
+    {
+      "epoch": 0.10535330588048267,
+      "grad_norm": 0.9999786019325256,
+      "learning_rate": 0.00020677768096628412,
+      "loss": 5.8052,
+      "num_input_tokens_seen": 84148224,
+      "step": 642
+    },
+    {
+      "epoch": 0.10584561104814848,
+      "grad_norm": 0.9309632778167725,
+      "learning_rate": 0.0002062962421734427,
+      "loss": 5.8597,
+      "num_input_tokens_seen": 84541440,
+      "step": 645
+    },
+    {
+      "epoch": 0.10633791621581427,
+      "grad_norm": 0.9224460124969482,
+      "learning_rate": 0.00020581815058714115,
+      "loss": 5.8526,
+      "num_input_tokens_seen": 84934656,
+      "step": 648
+    },
+    {
+      "epoch": 0.10683022138348008,
+      "grad_norm": 0.7043266892433167,
+      "learning_rate": 0.00020534336760049378,
+      "loss": 5.7704,
+      "num_input_tokens_seen": 85327872,
+      "step": 651
+    },
+    {
+      "epoch": 0.1073225265511459,
+      "grad_norm": 0.7217307090759277,
+      "learning_rate": 0.00020487185522716434,
+      "loss": 5.8222,
+      "num_input_tokens_seen": 85721088,
+      "step": 654
+    },
+    {
+      "epoch": 0.1078148317188117,
+      "grad_norm": 0.7779012322425842,
+      "learning_rate": 0.0002044035760886003,
+      "loss": 5.8596,
+      "num_input_tokens_seen": 86114304,
+      "step": 657
+    },
+    {
+      "epoch": 0.1083071368864775,
+      "grad_norm": 0.7242317795753479,
+      "learning_rate": 0.00020393849340158684,
+      "loss": 5.8273,
+      "num_input_tokens_seen": 86507520,
+      "step": 660
+    },
+    {
+      "epoch": 0.10879944205414331,
+      "grad_norm": 0.8740100264549255,
+      "learning_rate": 0.00020347657096611072,
+      "loss": 5.7984,
+      "num_input_tokens_seen": 86900736,
+      "step": 663
+    },
+    {
+      "epoch": 0.10929174722180912,
+      "grad_norm": 0.9838325381278992,
+      "learning_rate": 0.0002030177731535252,
+      "loss": 5.851,
+      "num_input_tokens_seen": 87293952,
+      "step": 666
+    },
+    {
+      "epoch": 0.10978405238947493,
+      "grad_norm": 0.989205002784729,
+      "learning_rate": 0.0002025620648950073,
+      "loss": 5.8132,
+      "num_input_tokens_seen": 87687168,
+      "step": 669
+    },
+    {
+      "epoch": 0.11027635755714073,
+      "grad_norm": 1.137012004852295,
+      "learning_rate": 0.00020210941167029872,
+      "loss": 5.7769,
+      "num_input_tokens_seen": 88080384,
+      "step": 672
+    },
+    {
+      "epoch": 0.11076866272480654,
+      "grad_norm": 1.0733401775360107,
+      "learning_rate": 0.00020165977949672233,
+      "loss": 5.7877,
+      "num_input_tokens_seen": 88473600,
+      "step": 675
+    },
+    {
+      "epoch": 0.11126096789247235,
+      "grad_norm": 1.1243098974227905,
+      "learning_rate": 0.00020121313491846602,
+      "loss": 5.7843,
+      "num_input_tokens_seen": 88866816,
+      "step": 678
+    },
+    {
+      "epoch": 0.11175327306013816,
+      "grad_norm": 0.9308670163154602,
+      "learning_rate": 0.000200769444996127,
+      "loss": 5.8465,
+      "num_input_tokens_seen": 89260032,
+      "step": 681
+    },
+    {
+      "epoch": 0.11224557822780395,
+      "grad_norm": 1.067434549331665,
+      "learning_rate": 0.00020032867729650794,
+      "loss": 5.7991,
+      "num_input_tokens_seen": 89653248,
+      "step": 684
+    },
+    {
+      "epoch": 0.11273788339546977,
+      "grad_norm": 0.9211784601211548,
+      "learning_rate": 0.00019989079988265906,
+      "loss": 5.7806,
+      "num_input_tokens_seen": 90046464,
+      "step": 687
+    },
+    {
+      "epoch": 0.11323018856313558,
+      "grad_norm": 0.8653129935264587,
+      "learning_rate": 0.00019945578130415816,
+      "loss": 5.8365,
+      "num_input_tokens_seen": 90439680,
+      "step": 690
+    },
+    {
+      "epoch": 0.11372249373080139,
+      "grad_norm": 1.8020168542861938,
+      "learning_rate": 0.00019902359058762258,
+      "loss": 5.8169,
+      "num_input_tokens_seen": 90832896,
+      "step": 693
+    },
+    {
+      "epoch": 0.11421479889846718,
+      "grad_norm": 1.38017737865448,
+      "learning_rate": 0.00019859419722744617,
+      "loss": 5.7989,
+      "num_input_tokens_seen": 91226112,
+      "step": 696
+    },
+    {
+      "epoch": 0.11470710406613299,
+      "grad_norm": 1.2626079320907593,
+      "learning_rate": 0.0001981675711767554,
+      "loss": 5.7555,
+      "num_input_tokens_seen": 91619328,
+      "step": 699
+    },
+    {
+      "epoch": 0.1151994092337988,
+      "grad_norm": 0.9266194105148315,
+      "learning_rate": 0.00019774368283857792,
+      "loss": 5.8277,
+      "num_input_tokens_seen": 92012544,
+      "step": 702
+    },
+    {
+      "epoch": 0.11569171440146461,
+      "grad_norm": 1.0578334331512451,
+      "learning_rate": 0.00019732250305721835,
+      "loss": 5.7924,
+      "num_input_tokens_seen": 92405760,
+      "step": 705
+    },
+    {
+      "epoch": 0.11618401956913041,
+      "grad_norm": 0.6349201798439026,
+      "learning_rate": 0.00019690400310983514,
+      "loss": 5.7968,
+      "num_input_tokens_seen": 92798976,
+      "step": 708
+    },
+    {
+      "epoch": 0.11667632473679622,
+      "grad_norm": 0.6735845804214478,
+      "learning_rate": 0.0001964881546982129,
+      "loss": 5.7552,
+      "num_input_tokens_seen": 93192192,
+      "step": 711
+    },
+    {
+      "epoch": 0.11716862990446203,
+      "grad_norm": 0.706881582736969,
+      "learning_rate": 0.0001960749299407257,
+      "loss": 5.7713,
+      "num_input_tokens_seen": 93585408,
+      "step": 714
+    },
+    {
+      "epoch": 0.11766093507212784,
+      "grad_norm": 0.7729910612106323,
+      "learning_rate": 0.00019566430136448468,
+      "loss": 5.8106,
+      "num_input_tokens_seen": 93978624,
+      "step": 717
+    },
+    {
+      "epoch": 0.11815324023979364,
+      "grad_norm": 0.7226433157920837,
+      "learning_rate": 0.00019525624189766633,
+      "loss": 5.7983,
+      "num_input_tokens_seen": 94371840,
+      "step": 720
+    },
+    {
+      "epoch": 0.11864554540745945,
+      "grad_norm": 0.6069265604019165,
+      "learning_rate": 0.0001948507248620161,
+      "loss": 5.7355,
+      "num_input_tokens_seen": 94765056,
+      "step": 723
+    },
+    {
+      "epoch": 0.11913785057512526,
+      "grad_norm": 0.7411084175109863,
+      "learning_rate": 0.00019444772396552212,
+      "loss": 5.7673,
+      "num_input_tokens_seen": 95158272,
+      "step": 726
+    },
+    {
+      "epoch": 0.11963015574279107,
+      "grad_norm": 0.9155630469322205,
+      "learning_rate": 0.0001940472132952553,
+      "loss": 5.7936,
+      "num_input_tokens_seen": 95551488,
+      "step": 729
+    },
+    {
+      "epoch": 0.12012246091045686,
+      "grad_norm": 0.8562064170837402,
+      "learning_rate": 0.00019364916731037083,
+      "loss": 5.7453,
+      "num_input_tokens_seen": 95944704,
+      "step": 732
+    },
+    {
+      "epoch": 0.12061476607812267,
+      "grad_norm": 0.9439576268196106,
+      "learning_rate": 0.0001932535608352669,
+      "loss": 5.7545,
+      "num_input_tokens_seen": 96337920,
+      "step": 735
+    },
+    {
+      "epoch": 0.12110707124578848,
+      "grad_norm": 0.9385275840759277,
+      "learning_rate": 0.00019286036905289666,
+      "loss": 5.7467,
+      "num_input_tokens_seen": 96731136,
+      "step": 738
+    },
+    {
+      "epoch": 0.1215993764134543,
+      "grad_norm": 0.7505801320075989,
+      "learning_rate": 0.00019246956749822933,
+      "loss": 5.7087,
+      "num_input_tokens_seen": 97124352,
+      "step": 741
+    },
+    {
+      "epoch": 0.12209168158112009,
+      "grad_norm": 0.6224293112754822,
+      "learning_rate": 0.0001920811320518561,
+      "loss": 5.7848,
+      "num_input_tokens_seen": 97517568,
+      "step": 744
+    },
+    {
+      "epoch": 0.1225839867487859,
+      "grad_norm": 0.5917067527770996,
+      "learning_rate": 0.00019169503893373772,
+      "loss": 5.7164,
+      "num_input_tokens_seen": 97910784,
+      "step": 747
+    },
+    {
+      "epoch": 0.12307629191645171,
+      "grad_norm": 0.6723655462265015,
+      "learning_rate": 0.00019131126469708987,
+      "loss": 5.7364,
+      "num_input_tokens_seen": 98304000,
+      "step": 750
+    },
+    {
+      "epoch": 0.12356859708411752,
+      "grad_norm": 0.6307470202445984,
+      "learning_rate": 0.00019092978622240234,
+      "loss": 5.7312,
+      "num_input_tokens_seen": 98697216,
+      "step": 753
+    },
+    {
+      "epoch": 0.12406090225178333,
+      "grad_norm": 0.6715870499610901,
+      "learning_rate": 0.00019055058071158903,
+      "loss": 5.7555,
+      "num_input_tokens_seen": 99090432,
+      "step": 756
+    },
+    {
+      "epoch": 0.12455320741944913,
+      "grad_norm": 0.7827840447425842,
+      "learning_rate": 0.00019017362568226525,
+      "loss": 5.7412,
+      "num_input_tokens_seen": 99483648,
+      "step": 759
+    },
+    {
+      "epoch": 0.12504551258711494,
+      "grad_norm": 0.7358651757240295,
+      "learning_rate": 0.0001897988989621491,
+      "loss": 5.7403,
+      "num_input_tokens_seen": 99876864,
+      "step": 762
+    },
+    {
+      "epoch": 0.12553781775478073,
+      "grad_norm": 0.6782851219177246,
+      "learning_rate": 0.00018942637868358373,
+      "loss": 5.7582,
+      "num_input_tokens_seen": 100270080,
+      "step": 765
+    },
+    {
+      "epoch": 0.12603012292244656,
+      "grad_norm": 0.675395667552948,
+      "learning_rate": 0.00018905604327817716,
+      "loss": 5.7459,
+      "num_input_tokens_seen": 100663296,
+      "step": 768
+    },
+    {
+      "epoch": 0.12652242809011235,
+      "grad_norm": 0.6478227972984314,
+      "learning_rate": 0.0001886878714715573,
+      "loss": 5.7242,
+      "num_input_tokens_seen": 101056512,
+      "step": 771
+    },
+    {
+      "epoch": 0.12701473325777818,
+      "grad_norm": 0.5236144661903381,
+      "learning_rate": 0.00018832184227823856,
+      "loss": 5.7381,
+      "num_input_tokens_seen": 101449728,
+      "step": 774
+    },
+    {
+      "epoch": 0.12750703842544397,
+      "grad_norm": 0.6938157081604004,
+      "learning_rate": 0.0001879579349965979,
+      "loss": 5.7178,
+      "num_input_tokens_seen": 101842944,
+      "step": 777
+    },
+    {
+      "epoch": 0.12799934359310977,
+      "grad_norm": 0.735215425491333,
+      "learning_rate": 0.00018759612920395688,
+      "loss": 5.7341,
+      "num_input_tokens_seen": 102236160,
+      "step": 780
+    },
+    {
+      "epoch": 0.1284916487607756,
+      "grad_norm": 0.7181572914123535,
+      "learning_rate": 0.0001872364047517678,
+      "loss": 5.7262,
+      "num_input_tokens_seen": 102629376,
+      "step": 783
+    },
+    {
+      "epoch": 0.1289839539284414,
+      "grad_norm": 0.5627852082252502,
+      "learning_rate": 0.00018687874176090066,
+      "loss": 5.7525,
+      "num_input_tokens_seen": 103022592,
+      "step": 786
+    },
+    {
+      "epoch": 0.1294762590961072,
+      "grad_norm": 0.6639107465744019,
+      "learning_rate": 0.0001865231206170292,
+      "loss": 5.7159,
+      "num_input_tokens_seen": 103415808,
+      "step": 789
+    },
+    {
+      "epoch": 0.129968564263773,
+      "grad_norm": 0.6468229293823242,
+      "learning_rate": 0.00018616952196611267,
+      "loss": 5.7392,
+      "num_input_tokens_seen": 103809024,
+      "step": 792
+    },
+    {
+      "epoch": 0.1304608694314388,
+      "grad_norm": 0.657605767250061,
+      "learning_rate": 0.00018581792670997177,
+      "loss": 5.7256,
+      "num_input_tokens_seen": 104202240,
+      "step": 795
+    },
+    {
+      "epoch": 0.13095317459910463,
+      "grad_norm": 0.6212555766105652,
+      "learning_rate": 0.00018546831600195623,
+      "loss": 5.723,
+      "num_input_tokens_seen": 104595456,
+      "step": 798
+    },
+    {
+      "epoch": 0.13128137804421516,
+      "eval_accuracy": 0.18926233512457255,
+      "eval_loss": 5.911603927612305,
+      "eval_runtime": 111.0852,
+      "eval_samples_per_second": 2.701,
+      "eval_steps_per_second": 1.35,
+      "num_input_tokens_seen": 104857600,
+      "step": 800
+    },
+    {
+      "epoch": 0.13144547976677043,
+      "grad_norm": 0.6446166634559631,
+      "learning_rate": 0.00018512067124270133,
+      "loss": 5.7414,
+      "num_input_tokens_seen": 104988672,
+      "step": 801
+    },
+    {
+      "epoch": 0.13193778493443623,
+      "grad_norm": 0.7650989294052124,
+      "learning_rate": 0.00018477497407597197,
+      "loss": 5.7188,
+      "num_input_tokens_seen": 105381888,
+      "step": 804
+    },
+    {
+      "epoch": 0.13243009010210205,
+      "grad_norm": 0.6955244541168213,
+      "learning_rate": 0.00018443120638459164,
+      "loss": 5.7054,
+      "num_input_tokens_seen": 105775104,
+      "step": 807
+    },
+    {
+      "epoch": 0.13292239526976785,
+      "grad_norm": 0.7522872090339661,
+      "learning_rate": 0.00018408935028645438,
+      "loss": 5.7231,
+      "num_input_tokens_seen": 106168320,
+      "step": 810
+    },
+    {
+      "epoch": 0.13341470043743364,
+      "grad_norm": 0.7220354080200195,
+      "learning_rate": 0.00018374938813061763,
+      "loss": 5.7101,
+      "num_input_tokens_seen": 106561536,
+      "step": 813
+    },
+    {
+      "epoch": 0.13390700560509947,
+      "grad_norm": 0.7726113796234131,
+      "learning_rate": 0.00018341130249347484,
+      "loss": 5.78,
+      "num_input_tokens_seen": 106954752,
+      "step": 816
+    },
+    {
+      "epoch": 0.13439931077276526,
+      "grad_norm": 0.8380802869796753,
+      "learning_rate": 0.000183075076175004,
+      "loss": 5.7309,
+      "num_input_tokens_seen": 107347968,
+      "step": 819
+    },
+    {
+      "epoch": 0.1348916159404311,
+      "grad_norm": 0.871356725692749,
+      "learning_rate": 0.0001827406921950927,
+      "loss": 5.6989,
+      "num_input_tokens_seen": 107741184,
+      "step": 822
+    },
+    {
+      "epoch": 0.13538392110809688,
+      "grad_norm": 0.8443216681480408,
+      "learning_rate": 0.0001824081337899362,
+      "loss": 5.7078,
+      "num_input_tokens_seen": 108134400,
+      "step": 825
+    },
+    {
+      "epoch": 0.13587622627576268,
+      "grad_norm": 0.8105941414833069,
+      "learning_rate": 0.00018207738440850766,
+      "loss": 5.757,
+      "num_input_tokens_seen": 108527616,
+      "step": 828
+    },
+    {
+      "epoch": 0.1363685314434285,
+      "grad_norm": 0.9102969765663147,
+      "learning_rate": 0.00018174842770909803,
+      "loss": 5.6674,
+      "num_input_tokens_seen": 108920832,
+      "step": 831
+    },
+    {
+      "epoch": 0.1368608366110943,
+      "grad_norm": 0.9995675086975098,
+      "learning_rate": 0.00018142124755592492,
+      "loss": 5.7259,
+      "num_input_tokens_seen": 109314048,
+      "step": 834
+    },
+    {
+      "epoch": 0.1373531417787601,
+      "grad_norm": 0.9919428825378418,
+      "learning_rate": 0.00018109582801580817,
+      "loss": 5.7217,
+      "num_input_tokens_seen": 109707264,
+      "step": 837
+    },
+    {
+      "epoch": 0.13784544694642592,
+      "grad_norm": 0.7615512609481812,
+      "learning_rate": 0.0001807721533549109,
+      "loss": 5.66,
+      "num_input_tokens_seen": 110100480,
+      "step": 840
+    },
+    {
+      "epoch": 0.13833775211409172,
+      "grad_norm": 0.6651118397712708,
+      "learning_rate": 0.0001804502080355442,
+      "loss": 5.6899,
+      "num_input_tokens_seen": 110493696,
+      "step": 843
+    },
+    {
+      "epoch": 0.13883005728175754,
+      "grad_norm": 0.8231689929962158,
+      "learning_rate": 0.00018012997671303435,
+      "loss": 5.7131,
+      "num_input_tokens_seen": 110886912,
+      "step": 846
+    },
+    {
+      "epoch": 0.13932236244942334,
+      "grad_norm": 0.8676853179931641,
+      "learning_rate": 0.00017981144423265112,
+      "loss": 5.6746,
+      "num_input_tokens_seen": 111280128,
+      "step": 849
+    },
+    {
+      "epoch": 0.13981466761708913,
+      "grad_norm": 1.0326123237609863,
+      "learning_rate": 0.00017949459562659518,
+      "loss": 5.7139,
+      "num_input_tokens_seen": 111673344,
+      "step": 852
+    },
+    {
+      "epoch": 0.14030697278475496,
+      "grad_norm": 1.2157782316207886,
+      "learning_rate": 0.00017917941611104426,
+      "loss": 5.7368,
+      "num_input_tokens_seen": 112066560,
+      "step": 855
+    },
+    {
+      "epoch": 0.14079927795242075,
+      "grad_norm": 1.1053582429885864,
+      "learning_rate": 0.0001788658910832554,
+      "loss": 5.7108,
+      "num_input_tokens_seen": 112459776,
+      "step": 858
+    },
+    {
+      "epoch": 0.14129158312008655,
+      "grad_norm": 1.0346806049346924,
+      "learning_rate": 0.0001785540061187239,
+      "loss": 5.6973,
+      "num_input_tokens_seen": 112852992,
+      "step": 861
+    },
+    {
+      "epoch": 0.14178388828775237,
+      "grad_norm": 0.9909515976905823,
+      "learning_rate": 0.0001782437469683953,
+      "loss": 5.6803,
+      "num_input_tokens_seen": 113246208,
+      "step": 864
+    },
+    {
+      "epoch": 0.14227619345541817,
+      "grad_norm": 0.8728073835372925,
+      "learning_rate": 0.00017793509955593145,
+      "loss": 5.6894,
+      "num_input_tokens_seen": 113639424,
+      "step": 867
+    },
+    {
+      "epoch": 0.142768498623084,
+      "grad_norm": 0.8351977467536926,
+      "learning_rate": 0.00017762804997502798,
+      "loss": 5.6577,
+      "num_input_tokens_seen": 114032640,
+      "step": 870
+    },
+    {
+      "epoch": 0.1432608037907498,
+      "grad_norm": 0.8890995979309082,
+      "learning_rate": 0.00017732258448678262,
+      "loss": 5.7118,
+      "num_input_tokens_seen": 114425856,
+      "step": 873
+    },
+    {
+      "epoch": 0.1437531089584156,
+      "grad_norm": 0.6228556036949158,
+      "learning_rate": 0.0001770186895171133,
+      "loss": 5.6714,
+      "num_input_tokens_seen": 114819072,
+      "step": 876
+    },
+    {
+      "epoch": 0.1442454141260814,
+      "grad_norm": 0.6140576004981995,
+      "learning_rate": 0.00017671635165422445,
+      "loss": 5.6817,
+      "num_input_tokens_seen": 115212288,
+      "step": 879
+    },
+    {
+      "epoch": 0.1447377192937472,
+      "grad_norm": 0.5662062168121338,
+      "learning_rate": 0.00017641555764612098,
+      "loss": 5.713,
+      "num_input_tokens_seen": 115605504,
+      "step": 882
+    },
+    {
+      "epoch": 0.14523002446141303,
+      "grad_norm": 0.578356146812439,
+      "learning_rate": 0.00017611629439816853,
+      "loss": 5.69,
+      "num_input_tokens_seen": 115998720,
+      "step": 885
+    },
+    {
+      "epoch": 0.14572232962907883,
+      "grad_norm": 0.5570440888404846,
+      "learning_rate": 0.0001758185489706992,
+      "loss": 5.6513,
+      "num_input_tokens_seen": 116391936,
+      "step": 888
+    },
+    {
+      "epoch": 0.14621463479674462,
+      "grad_norm": 0.5778940320014954,
+      "learning_rate": 0.00017552230857666157,
+      "loss": 5.6607,
+      "num_input_tokens_seen": 116785152,
+      "step": 891
+    },
+    {
+      "epoch": 0.14670693996441045,
+      "grad_norm": 0.5713542699813843,
+      "learning_rate": 0.00017522756057931406,
+      "loss": 5.6619,
+      "num_input_tokens_seen": 117178368,
+      "step": 894
+    },
+    {
+      "epoch": 0.14719924513207625,
+      "grad_norm": 0.53873211145401,
+      "learning_rate": 0.00017493429248996095,
+      "loss": 5.6388,
+      "num_input_tokens_seen": 117571584,
+      "step": 897
+    },
+    {
+      "epoch": 0.14769155029974204,
+      "grad_norm": 0.6463914513587952,
+      "learning_rate": 0.0001746424919657298,
+      "loss": 5.6553,
+      "num_input_tokens_seen": 117964800,
+      "step": 900
+    },
+    {
+      "epoch": 0.14818385546740787,
+      "grad_norm": 0.5670692920684814,
+      "learning_rate": 0.00017435214680738953,
+      "loss": 5.6801,
+      "num_input_tokens_seen": 118358016,
+      "step": 903
+    },
+    {
+      "epoch": 0.14867616063507366,
+      "grad_norm": 0.6744683980941772,
+      "learning_rate": 0.00017406324495720832,
+      "loss": 5.6817,
+      "num_input_tokens_seen": 118751232,
+      "step": 906
+    },
+    {
+      "epoch": 0.14916846580273949,
+      "grad_norm": 0.6172819137573242,
+      "learning_rate": 0.0001737757744968504,
+      "loss": 5.6584,
+      "num_input_tokens_seen": 119144448,
+      "step": 909
+    },
+    {
+      "epoch": 0.14966077097040528,
+      "grad_norm": 0.7147213220596313,
+      "learning_rate": 0.0001734897236453108,
+      "loss": 5.6536,
+      "num_input_tokens_seen": 119537664,
+      "step": 912
+    },
+    {
+      "epoch": 0.15015307613807108,
+      "grad_norm": 0.5693617463111877,
+      "learning_rate": 0.00017320508075688773,
+      "loss": 5.6305,
+      "num_input_tokens_seen": 119930880,
+      "step": 915
+    },
+    {
+      "epoch": 0.1506453813057369,
+      "grad_norm": 0.7820476293563843,
+      "learning_rate": 0.00017292183431919094,
+      "loss": 5.6358,
+      "num_input_tokens_seen": 120324096,
+      "step": 918
+    },
+    {
+      "epoch": 0.1511376864734027,
+      "grad_norm": 0.7270592451095581,
+      "learning_rate": 0.00017263997295118624,
+      "loss": 5.6412,
+      "num_input_tokens_seen": 120717312,
+      "step": 921
+    },
+    {
+      "epoch": 0.1516299916410685,
+      "grad_norm": 0.6628842353820801,
+      "learning_rate": 0.00017235948540127462,
+      "loss": 5.6695,
+      "num_input_tokens_seen": 121110528,
+      "step": 924
+    },
+    {
+      "epoch": 0.15212229680873432,
+      "grad_norm": 0.7470645308494568,
+      "learning_rate": 0.00017208036054540591,
+      "loss": 5.6533,
+      "num_input_tokens_seen": 121503744,
+      "step": 927
+    },
+    {
+      "epoch": 0.15261460197640012,
+      "grad_norm": 0.6823071241378784,
+      "learning_rate": 0.00017180258738522556,
+      "loss": 5.6321,
+      "num_input_tokens_seen": 121896960,
+      "step": 930
+    },
+    {
+      "epoch": 0.15310690714406594,
+      "grad_norm": 0.6298269033432007,
+      "learning_rate": 0.0001715261550462546,
+      "loss": 5.69,
+      "num_input_tokens_seen": 122290176,
+      "step": 933
+    },
+    {
+      "epoch": 0.15359921231173174,
+      "grad_norm": 0.6772837042808533,
+      "learning_rate": 0.00017125105277610142,
+      "loss": 5.6243,
+      "num_input_tokens_seen": 122683392,
+      "step": 936
+    },
+    {
+      "epoch": 0.15409151747939753,
+      "grad_norm": 0.9537835121154785,
+      "learning_rate": 0.00017097726994270523,
+      "loss": 5.6595,
+      "num_input_tokens_seen": 123076608,
+      "step": 939
+    },
+    {
+      "epoch": 0.15458382264706336,
+      "grad_norm": 0.9860975742340088,
+      "learning_rate": 0.00017070479603261012,
+      "loss": 5.6391,
+      "num_input_tokens_seen": 123469824,
+      "step": 942
+    },
+    {
+      "epoch": 0.15507612781472915,
+      "grad_norm": 0.7856841683387756,
+      "learning_rate": 0.00017043362064926934,
+      "loss": 5.627,
+      "num_input_tokens_seen": 123863040,
+      "step": 945
+    },
+    {
+      "epoch": 0.15556843298239495,
+      "grad_norm": 0.8125623464584351,
+      "learning_rate": 0.00017016373351137908,
+      "loss": 5.6797,
+      "num_input_tokens_seen": 124256256,
+      "step": 948
+    },
+    {
+      "epoch": 0.15606073815006077,
+      "grad_norm": 0.8572544455528259,
+      "learning_rate": 0.0001698951244512415,
+      "loss": 5.6211,
+      "num_input_tokens_seen": 124649472,
+      "step": 951
+    },
+    {
+      "epoch": 0.15655304331772657,
+      "grad_norm": 0.6495417356491089,
+      "learning_rate": 0.0001696277834131554,
+      "loss": 5.6775,
+      "num_input_tokens_seen": 125042688,
+      "step": 954
+    },
+    {
+      "epoch": 0.1570453484853924,
+      "grad_norm": 0.7361529469490051,
+      "learning_rate": 0.00016936170045183562,
+      "loss": 5.6332,
+      "num_input_tokens_seen": 125435904,
+      "step": 957
+    },
+    {
+      "epoch": 0.1575376536530582,
+      "grad_norm": 0.6628533601760864,
+      "learning_rate": 0.0001690968657308585,
+      "loss": 5.6219,
+      "num_input_tokens_seen": 125829120,
+      "step": 960
+    },
+    {
+      "epoch": 0.158029958820724,
+      "grad_norm": 0.6602532267570496,
+      "learning_rate": 0.00016883326952113513,
+      "loss": 5.6377,
+      "num_input_tokens_seen": 126222336,
+      "step": 963
+    },
+    {
+      "epoch": 0.1585222639883898,
+      "grad_norm": 0.7363507151603699,
+      "learning_rate": 0.0001685709021994098,
+      "loss": 5.6573,
+      "num_input_tokens_seen": 126615552,
+      "step": 966
+    },
+    {
+      "epoch": 0.1590145691560556,
+      "grad_norm": 0.6305029988288879,
+      "learning_rate": 0.00016830975424678453,
+      "loss": 5.6764,
+      "num_input_tokens_seen": 127008768,
+      "step": 969
+    },
+    {
+      "epoch": 0.1595068743237214,
+      "grad_norm": 0.7641071081161499,
+      "learning_rate": 0.0001680498162472686,
+      "loss": 5.6473,
+      "num_input_tokens_seen": 127401984,
+      "step": 972
+    },
+    {
+      "epoch": 0.15999917949138723,
+      "grad_norm": 0.7363148331642151,
+      "learning_rate": 0.00016779107888635245,
+      "loss": 5.6144,
+      "num_input_tokens_seen": 127795200,
+      "step": 975
+    },
+    {
+      "epoch": 0.16049148465905302,
+      "grad_norm": 0.8142803311347961,
+      "learning_rate": 0.0001675335329496059,
+      "loss": 5.6206,
+      "num_input_tokens_seen": 128188416,
+      "step": 978
+    },
+    {
+      "epoch": 0.16098378982671885,
+      "grad_norm": 0.7114639282226562,
+      "learning_rate": 0.00016727716932129973,
+      "loss": 5.6103,
+      "num_input_tokens_seen": 128581632,
+      "step": 981
+    },
+    {
+      "epoch": 0.16147609499438464,
+      "grad_norm": 0.8105225563049316,
+      "learning_rate": 0.0001670219789830507,
+      "loss": 5.6128,
+      "num_input_tokens_seen": 128974848,
+      "step": 984
+    },
+    {
+      "epoch": 0.16196840016205044,
+      "grad_norm": 0.8542172312736511,
+      "learning_rate": 0.00016676795301248881,
+      "loss": 5.6622,
+      "num_input_tokens_seen": 129368064,
+      "step": 987
+    },
+    {
+      "epoch": 0.16246070532971627,
+      "grad_norm": 0.848969578742981,
+      "learning_rate": 0.00016651508258194728,
+      "loss": 5.644,
+      "num_input_tokens_seen": 129761280,
+      "step": 990
+    },
+    {
+      "epoch": 0.16295301049738206,
+      "grad_norm": 0.7594742774963379,
+      "learning_rate": 0.0001662633589571739,
+      "loss": 5.6315,
+      "num_input_tokens_seen": 130154496,
+      "step": 993
+    },
+    {
+      "epoch": 0.16344531566504786,
+      "grad_norm": 0.5903003811836243,
+      "learning_rate": 0.0001660127734960639,
+      "loss": 5.6235,
+      "num_input_tokens_seen": 130547712,
+      "step": 996
+    },
+    {
+      "epoch": 0.16393762083271368,
+      "grad_norm": 0.680317759513855,
+      "learning_rate": 0.00016576331764741402,
+      "loss": 5.6303,
+      "num_input_tokens_seen": 130940928,
+      "step": 999
+    },
+    {
+      "epoch": 0.16442992600037948,
+      "grad_norm": 0.6790865659713745,
+      "learning_rate": 0.00016551498294969648,
+      "loss": 5.6244,
+      "num_input_tokens_seen": 131334144,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1649222311680453,
+      "grad_norm": 0.77605140209198,
+      "learning_rate": 0.00016526776102985388,
+      "loss": 5.6269,
+      "num_input_tokens_seen": 131727360,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1654145363357111,
+      "grad_norm": 0.7890040874481201,
+      "learning_rate": 0.00016502164360211315,
+      "loss": 5.6388,
+      "num_input_tokens_seen": 132120576,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1659068415033769,
+      "grad_norm": 0.7248251438140869,
+      "learning_rate": 0.0001647766224668193,
+      "loss": 5.6087,
+      "num_input_tokens_seen": 132513792,
+      "step": 1011
+    },
+    {
+      "epoch": 0.16639914667104272,
+      "grad_norm": 0.7000757455825806,
+      "learning_rate": 0.00016453268950928797,
+      "loss": 5.61,
+      "num_input_tokens_seen": 132907008,
+      "step": 1014
+    },
+    {
+      "epoch": 0.16689145183870852,
+      "grad_norm": 0.6521958112716675,
+      "learning_rate": 0.00016428983669867676,
+      "loss": 5.6407,
+      "num_input_tokens_seen": 133300224,
+      "step": 1017
+    },
+    {
+      "epoch": 0.16738375700637434,
+      "grad_norm": 0.7919925451278687,
+      "learning_rate": 0.00016404805608687456,
+      "loss": 5.6145,
+      "num_input_tokens_seen": 133693440,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16787606217404014,
+      "grad_norm": 1.0986133813858032,
+      "learning_rate": 0.0001638073398074093,
+      "loss": 5.6294,
+      "num_input_tokens_seen": 134086656,
+      "step": 1023
+    },
+    {
+      "epoch": 0.16836836734170593,
+      "grad_norm": 0.9454318881034851,
+      "learning_rate": 0.0001635676800743725,
+      "loss": 5.6277,
+      "num_input_tokens_seen": 134479872,
+      "step": 1026
+    },
+    {
+      "epoch": 0.16886067250937176,
+      "grad_norm": 0.8960498571395874,
+      "learning_rate": 0.000163329069181362,
+      "loss": 5.5897,
+      "num_input_tokens_seen": 134873088,
+      "step": 1029
+    },
+    {
+      "epoch": 0.16935297767703755,
+      "grad_norm": 0.7676910758018494,
+      "learning_rate": 0.00016309149950044093,
+      "loss": 5.6505,
+      "num_input_tokens_seen": 135266304,
+      "step": 1032
+    },
+    {
+      "epoch": 0.16984528284470335,
+      "grad_norm": 0.6834097504615784,
+      "learning_rate": 0.0001628549634811134,
+      "loss": 5.6003,
+      "num_input_tokens_seen": 135659520,
+      "step": 1035
+    },
+    {
+      "epoch": 0.17033758801236917,
+      "grad_norm": 0.7149432301521301,
+      "learning_rate": 0.00016261945364931684,
+      "loss": 5.599,
+      "num_input_tokens_seen": 136052736,
+      "step": 1038
+    },
+    {
+      "epoch": 0.17082989318003497,
+      "grad_norm": 0.635908842086792,
+      "learning_rate": 0.00016238496260642988,
+      "loss": 5.5852,
+      "num_input_tokens_seen": 136445952,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1713221983477008,
+      "grad_norm": 0.7218221426010132,
+      "learning_rate": 0.0001621514830282963,
+      "loss": 5.6205,
+      "num_input_tokens_seen": 136839168,
+      "step": 1044
+    },
+    {
+      "epoch": 0.1718145035153666,
+      "grad_norm": 0.5910034775733948,
+      "learning_rate": 0.00016191900766426384,
+      "loss": 5.5992,
+      "num_input_tokens_seen": 137232384,
+      "step": 1047
+    },
+    {
+      "epoch": 0.1723068086830324,
+      "grad_norm": 0.729131281375885,
+      "learning_rate": 0.000161687529336239,
+      "loss": 5.6396,
+      "num_input_tokens_seen": 137625600,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1727991138506982,
+      "grad_norm": 0.5640085935592651,
+      "learning_rate": 0.00016145704093775551,
+      "loss": 5.5828,
+      "num_input_tokens_seen": 138018816,
+      "step": 1053
+    },
+    {
+      "epoch": 0.173291419018364,
+      "grad_norm": 0.6498822569847107,
+      "learning_rate": 0.00016122753543305863,
+      "loss": 5.6024,
+      "num_input_tokens_seen": 138412032,
+      "step": 1056
+    },
+    {
+      "epoch": 0.1737837241860298,
+      "grad_norm": 0.6818384528160095,
+      "learning_rate": 0.00016099900585620256,
+      "loss": 5.6227,
+      "num_input_tokens_seen": 138805248,
+      "step": 1059
+    },
+    {
+      "epoch": 0.17427602935369563,
+      "grad_norm": 0.6664714813232422,
+      "learning_rate": 0.00016077144531016272,
+      "loss": 5.5499,
+      "num_input_tokens_seen": 139198464,
+      "step": 1062
+    },
+    {
+      "epoch": 0.17476833452136142,
+      "grad_norm": 0.5847983956336975,
+      "learning_rate": 0.00016054484696596133,
+      "loss": 5.5775,
+      "num_input_tokens_seen": 139591680,
+      "step": 1065
+    },
+    {
+      "epoch": 0.17526063968902725,
+      "grad_norm": 0.7117697596549988,
+      "learning_rate": 0.00016031920406180673,
+      "loss": 5.5939,
+      "num_input_tokens_seen": 139984896,
+      "step": 1068
+    },
+    {
+      "epoch": 0.17575294485669304,
+      "grad_norm": 0.6359485983848572,
+      "learning_rate": 0.00016009450990224597,
+      "loss": 5.5776,
+      "num_input_tokens_seen": 140378112,
+      "step": 1071
+    },
+    {
+      "epoch": 0.17624525002435884,
+      "grad_norm": 0.7390355467796326,
+      "learning_rate": 0.00015987075785733017,
+      "loss": 5.5825,
+      "num_input_tokens_seen": 140771328,
+      "step": 1074
+    },
+    {
+      "epoch": 0.17673755519202466,
+      "grad_norm": 0.6650616526603699,
+      "learning_rate": 0.000159647941361793,
+      "loss": 5.5637,
+      "num_input_tokens_seen": 141164544,
+      "step": 1077
+    },
+    {
+      "epoch": 0.17722986035969046,
+      "grad_norm": 0.642583429813385,
+      "learning_rate": 0.0001594260539142416,
+      "loss": 5.5931,
+      "num_input_tokens_seen": 141557760,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17772216552735626,
+      "grad_norm": 0.7720558047294617,
+      "learning_rate": 0.0001592050890763597,
+      "loss": 5.6194,
+      "num_input_tokens_seen": 141950976,
+      "step": 1083
+    },
+    {
+      "epoch": 0.17821447069502208,
+      "grad_norm": 0.9935165047645569,
+      "learning_rate": 0.00015898504047212318,
+      "loss": 5.5858,
+      "num_input_tokens_seen": 142344192,
+      "step": 1086
+    },
+    {
+      "epoch": 0.17870677586268788,
+      "grad_norm": 1.5645318031311035,
+      "learning_rate": 0.00015876590178702708,
+      "loss": 5.5929,
+      "num_input_tokens_seen": 142737408,
+      "step": 1089
+    },
+    {
+      "epoch": 0.1791990810303537,
+      "grad_norm": 0.6664535403251648,
+      "learning_rate": 0.0001585476667673247,
+      "loss": 5.5807,
+      "num_input_tokens_seen": 143130624,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1796913861980195,
+      "grad_norm": 0.889173686504364,
+      "learning_rate": 0.0001583303292192779,
+      "loss": 5.6241,
+      "num_input_tokens_seen": 143523840,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1801836913656853,
+      "grad_norm": 0.9507172703742981,
+      "learning_rate": 0.00015811388300841897,
+      "loss": 5.6112,
+      "num_input_tokens_seen": 143917056,
+      "step": 1098
+    },
+    {
+      "epoch": 0.18067599653335112,
+      "grad_norm": 0.7204018831253052,
+      "learning_rate": 0.00015789832205882312,
+      "loss": 5.5713,
+      "num_input_tokens_seen": 144310272,
+      "step": 1101
+    },
+    {
+      "epoch": 0.18116830170101691,
+      "grad_norm": 0.7509574890136719,
+      "learning_rate": 0.0001576836403523923,
+      "loss": 5.5944,
+      "num_input_tokens_seen": 144703488,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1816606068686827,
+      "grad_norm": 0.9350889921188354,
+      "learning_rate": 0.00015746983192814956,
+      "loss": 5.5641,
+      "num_input_tokens_seen": 145096704,
+      "step": 1107
+    },
+    {
+      "epoch": 0.18215291203634854,
+      "grad_norm": 1.061766266822815,
+      "learning_rate": 0.00015725689088154365,
+      "loss": 5.5659,
+      "num_input_tokens_seen": 145489920,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18264521720401433,
+      "grad_norm": 0.8999969363212585,
+      "learning_rate": 0.00015704481136376432,
+      "loss": 5.5741,
+      "num_input_tokens_seen": 145883136,
+      "step": 1113
+    },
+    {
+      "epoch": 0.18313752237168016,
+      "grad_norm": 0.6309444308280945,
+      "learning_rate": 0.00015683358758106756,
+      "loss": 5.566,
+      "num_input_tokens_seen": 146276352,
+      "step": 1116
+    },
+    {
+      "epoch": 0.18362982753934595,
+      "grad_norm": 0.5857992768287659,
+      "learning_rate": 0.0001566232137941107,
+      "loss": 5.5473,
+      "num_input_tokens_seen": 146669568,
+      "step": 1119
+    },
+    {
+      "epoch": 0.18412213270701175,
+      "grad_norm": 0.6874331831932068,
+      "learning_rate": 0.0001564136843172976,
+      "loss": 5.5815,
+      "num_input_tokens_seen": 147062784,
+      "step": 1122
+    },
+    {
+      "epoch": 0.18461443787467757,
+      "grad_norm": 0.9051715135574341,
+      "learning_rate": 0.00015620499351813306,
+      "loss": 5.5711,
+      "num_input_tokens_seen": 147456000,
+      "step": 1125
+    },
+    {
+      "epoch": 0.18510674304234337,
+      "grad_norm": 0.77443528175354,
+      "learning_rate": 0.0001559971358165871,
+      "loss": 5.5898,
+      "num_input_tokens_seen": 147849216,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1855990482100092,
+      "grad_norm": 0.5776637196540833,
+      "learning_rate": 0.00015579010568446804,
+      "loss": 5.6116,
+      "num_input_tokens_seen": 148242432,
+      "step": 1131
+    },
+    {
+      "epoch": 0.186091353377675,
+      "grad_norm": 0.6061506271362305,
+      "learning_rate": 0.00015558389764480516,
+      "loss": 5.5514,
+      "num_input_tokens_seen": 148635648,
+      "step": 1134
+    },
+    {
+      "epoch": 0.18658365854534079,
+      "grad_norm": 0.7481436729431152,
+      "learning_rate": 0.0001553785062712401,
+      "loss": 5.5469,
+      "num_input_tokens_seen": 149028864,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1870759637130066,
+      "grad_norm": 0.7162047028541565,
+      "learning_rate": 0.00015517392618742703,
+      "loss": 5.5208,
+      "num_input_tokens_seen": 149422080,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1875682688806724,
+      "grad_norm": 0.6034492254257202,
+      "learning_rate": 0.00015497015206644168,
+      "loss": 5.5255,
+      "num_input_tokens_seen": 149815296,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1880605740483382,
+      "grad_norm": 0.7142524719238281,
+      "learning_rate": 0.00015476717863019868,
+      "loss": 5.5561,
+      "num_input_tokens_seen": 150208512,
+      "step": 1146
+    },
+    {
+      "epoch": 0.18855287921600403,
+      "grad_norm": 0.6338310241699219,
+      "learning_rate": 0.0001545650006488774,
+      "loss": 5.524,
+      "num_input_tokens_seen": 150601728,
+      "step": 1149
+    },
+    {
+      "epoch": 0.18904518438366982,
+      "grad_norm": 0.7229697108268738,
+      "learning_rate": 0.00015436361294035586,
+      "loss": 5.5738,
+      "num_input_tokens_seen": 150994944,
+      "step": 1152
+    },
+    {
+      "epoch": 0.18953748955133565,
+      "grad_norm": 0.8610914349555969,
+      "learning_rate": 0.00015416301036965307,
+      "loss": 5.5242,
+      "num_input_tokens_seen": 151388160,
+      "step": 1155
+    },
+    {
+      "epoch": 0.19002979471900144,
+      "grad_norm": 0.8543524742126465,
+      "learning_rate": 0.00015396318784837899,
+      "loss": 5.5251,
+      "num_input_tokens_seen": 151781376,
+      "step": 1158
+    },
+    {
+      "epoch": 0.19052209988666724,
+      "grad_norm": 0.9472544193267822,
+      "learning_rate": 0.00015376414033419227,
+      "loss": 5.5143,
+      "num_input_tokens_seen": 152174592,
+      "step": 1161
+    },
+    {
+      "epoch": 0.19101440505433306,
+      "grad_norm": 0.8483441472053528,
+      "learning_rate": 0.00015356586283026615,
+      "loss": 5.5312,
+      "num_input_tokens_seen": 152567808,
+      "step": 1164
+    },
+    {
+      "epoch": 0.19150671022199886,
+      "grad_norm": 0.71060711145401,
+      "learning_rate": 0.00015336835038476135,
+      "loss": 5.5537,
+      "num_input_tokens_seen": 152961024,
+      "step": 1167
+    },
+    {
+      "epoch": 0.19199901538966466,
+      "grad_norm": 0.5833298563957214,
+      "learning_rate": 0.00015317159809030676,
+      "loss": 5.5685,
+      "num_input_tokens_seen": 153354240,
+      "step": 1170
+    },
+    {
+      "epoch": 0.19249132055733048,
+      "grad_norm": 0.6280092597007751,
+      "learning_rate": 0.0001529756010834872,
+      "loss": 5.5238,
+      "num_input_tokens_seen": 153747456,
+      "step": 1173
+    },
+    {
+      "epoch": 0.19298362572499628,
+      "grad_norm": 0.6287539005279541,
+      "learning_rate": 0.00015278035454433883,
+      "loss": 5.5883,
+      "num_input_tokens_seen": 154140672,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1934759308926621,
+      "grad_norm": 0.5246036648750305,
+      "learning_rate": 0.00015258585369585086,
+      "loss": 5.4878,
+      "num_input_tokens_seen": 154533888,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1939682360603279,
+      "grad_norm": 0.5769249796867371,
+      "learning_rate": 0.00015239209380347492,
+      "loss": 5.5545,
+      "num_input_tokens_seen": 154927104,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1944605412279937,
+      "grad_norm": 0.5999164581298828,
+      "learning_rate": 0.00015219907017464103,
+      "loss": 5.5295,
+      "num_input_tokens_seen": 155320320,
+      "step": 1185
+    },
+    {
+      "epoch": 0.19495284639565952,
+      "grad_norm": 0.7166159749031067,
+      "learning_rate": 0.00015200677815828016,
+      "loss": 5.5673,
+      "num_input_tokens_seen": 155713536,
+      "step": 1188
+    },
+    {
+      "epoch": 0.19544515156332531,
+      "grad_norm": 0.6548975110054016,
+      "learning_rate": 0.0001518152131443535,
+      "loss": 5.5044,
+      "num_input_tokens_seen": 156106752,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1959374567309911,
+      "grad_norm": 0.6346669793128967,
+      "learning_rate": 0.00015162437056338838,
+      "loss": 5.5459,
+      "num_input_tokens_seen": 156499968,
+      "step": 1194
+    },
+    {
+      "epoch": 0.19642976189865693,
+      "grad_norm": 0.7255611419677734,
+      "learning_rate": 0.00015143424588602033,
+      "loss": 5.5296,
+      "num_input_tokens_seen": 156893184,
+      "step": 1197
+    },
+    {
+      "epoch": 0.19692206706632273,
+      "grad_norm": 0.7328153848648071,
+      "learning_rate": 0.0001512448346225417,
+      "loss": 5.5106,
+      "num_input_tokens_seen": 157286400,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19692206706632273,
+      "eval_accuracy": 0.1976241654453672,
+      "eval_loss": 5.751555919647217,
+      "eval_runtime": 110.6176,
+      "eval_samples_per_second": 2.712,
+      "eval_steps_per_second": 1.356,
+      "num_input_tokens_seen": 157286400,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19741437223398856,
+      "grad_norm": 0.6580715179443359,
+      "learning_rate": 0.00015105613232245638,
+      "loss": 5.545,
+      "num_input_tokens_seen": 157679616,
+      "step": 1203
+    },
+    {
+      "epoch": 0.19790667740165435,
+      "grad_norm": 0.5751286745071411,
+      "learning_rate": 0.00015086813457404033,
+      "loss": 5.5439,
+      "num_input_tokens_seen": 158072832,
+      "step": 1206
+    },
+    {
+      "epoch": 0.19839898256932015,
+      "grad_norm": 0.6533809900283813,
+      "learning_rate": 0.00015068083700390872,
+      "loss": 5.5193,
+      "num_input_tokens_seen": 158466048,
+      "step": 1209
+    },
+    {
+      "epoch": 0.19889128773698597,
+      "grad_norm": 0.6799106597900391,
+      "learning_rate": 0.0001504942352765884,
+      "loss": 5.5718,
+      "num_input_tokens_seen": 158859264,
+      "step": 1212
+    },
+    {
+      "epoch": 0.19938359290465177,
+      "grad_norm": 0.6222509145736694,
+      "learning_rate": 0.00015030832509409646,
+      "loss": 5.5177,
+      "num_input_tokens_seen": 159252480,
+      "step": 1215
+    },
+    {
+      "epoch": 0.19987589807231756,
+      "grad_norm": 0.6463878750801086,
+      "learning_rate": 0.00015012310219552445,
+      "loss": 5.519,
+      "num_input_tokens_seen": 159645696,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2003682032399834,
+      "grad_norm": 0.6764931678771973,
+      "learning_rate": 0.00014993856235662816,
+      "loss": 5.5437,
+      "num_input_tokens_seen": 160038912,
+      "step": 1221
+    },
+    {
+      "epoch": 0.20086050840764919,
+      "grad_norm": 0.6322119235992432,
+      "learning_rate": 0.00014975470138942312,
+      "loss": 5.5726,
+      "num_input_tokens_seen": 160432128,
+      "step": 1224
+    },
+    {
+      "epoch": 0.201352813575315,
+      "grad_norm": 0.7915647029876709,
+      "learning_rate": 0.00014957151514178522,
+      "loss": 5.5265,
+      "num_input_tokens_seen": 160825344,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2018451187429808,
+      "grad_norm": 0.9680132269859314,
+      "learning_rate": 0.00014938899949705703,
+      "loss": 5.5909,
+      "num_input_tokens_seen": 161218560,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2023374239106466,
+      "grad_norm": 0.910897433757782,
+      "learning_rate": 0.00014920715037365913,
+      "loss": 5.5585,
+      "num_input_tokens_seen": 161611776,
+      "step": 1233
+    },
+    {
+      "epoch": 0.20282972907831243,
+      "grad_norm": 0.7097917199134827,
+      "learning_rate": 0.00014902596372470695,
+      "loss": 5.5273,
+      "num_input_tokens_seen": 162004992,
+      "step": 1236
+    },
+    {
+      "epoch": 0.20332203424597822,
+      "grad_norm": 0.8189403414726257,
+      "learning_rate": 0.00014884543553763215,
+      "loss": 5.4965,
+      "num_input_tokens_seen": 162398208,
+      "step": 1239
+    },
+    {
+      "epoch": 0.20381433941364402,
+      "grad_norm": 0.6247571110725403,
+      "learning_rate": 0.00014866556183380976,
+      "loss": 5.4886,
+      "num_input_tokens_seen": 162791424,
+      "step": 1242
+    },
+    {
+      "epoch": 0.20430664458130984,
+      "grad_norm": 0.7881289124488831,
+      "learning_rate": 0.0001484863386681897,
+      "loss": 5.5525,
+      "num_input_tokens_seen": 163184640,
+      "step": 1245
+    },
+    {
+      "epoch": 0.20479894974897564,
+      "grad_norm": 0.8643481731414795,
+      "learning_rate": 0.00014830776212893345,
+      "loss": 5.5544,
+      "num_input_tokens_seen": 163577856,
+      "step": 1248
+    },
+    {
+      "epoch": 0.20529125491664146,
+      "grad_norm": 0.7672863602638245,
+      "learning_rate": 0.0001481298283370553,
+      "loss": 5.5501,
+      "num_input_tokens_seen": 163971072,
+      "step": 1251
+    },
+    {
+      "epoch": 0.20578356008430726,
+      "grad_norm": 0.7655563950538635,
+      "learning_rate": 0.0001479525334460686,
+      "loss": 5.5266,
+      "num_input_tokens_seen": 164364288,
+      "step": 1254
+    },
+    {
+      "epoch": 0.20627586525197306,
+      "grad_norm": 0.6450381278991699,
+      "learning_rate": 0.00014777587364163652,
+      "loss": 5.5056,
+      "num_input_tokens_seen": 164757504,
+      "step": 1257
+    },
+    {
+      "epoch": 0.20676817041963888,
+      "grad_norm": 0.7719587683677673,
+      "learning_rate": 0.00014759984514122729,
+      "loss": 5.4903,
+      "num_input_tokens_seen": 165150720,
+      "step": 1260
+    },
+    {
+      "epoch": 0.20726047558730468,
+      "grad_norm": 0.7485530972480774,
+      "learning_rate": 0.00014742444419377413,
+      "loss": 5.5165,
+      "num_input_tokens_seen": 165543936,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2077527807549705,
+      "grad_norm": 0.6037641763687134,
+      "learning_rate": 0.00014724966707933943,
+      "loss": 5.5313,
+      "num_input_tokens_seen": 165937152,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2082450859226363,
+      "grad_norm": 0.6211085915565491,
+      "learning_rate": 0.00014707551010878346,
+      "loss": 5.4773,
+      "num_input_tokens_seen": 166330368,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2087373910903021,
+      "grad_norm": 0.6303486824035645,
+      "learning_rate": 0.00014690196962343724,
+      "loss": 5.5021,
+      "num_input_tokens_seen": 166723584,
+      "step": 1272
+    },
+    {
+      "epoch": 0.20922969625796792,
+      "grad_norm": 0.5903136730194092,
+      "learning_rate": 0.00014672904199477987,
+      "loss": 5.5041,
+      "num_input_tokens_seen": 167116800,
+      "step": 1275
+    },
+    {
+      "epoch": 0.20972200142563371,
+      "grad_norm": 0.6164759397506714,
+      "learning_rate": 0.00014655672362411974,
+      "loss": 5.4536,
+      "num_input_tokens_seen": 167510016,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2102143065932995,
+      "grad_norm": 0.7181910872459412,
+      "learning_rate": 0.00014638501094227995,
+      "loss": 5.4979,
+      "num_input_tokens_seen": 167903232,
+      "step": 1281
+    },
+    {
+      "epoch": 0.21070661176096533,
+      "grad_norm": 0.8058461546897888,
+      "learning_rate": 0.000146213900409288,
+      "loss": 5.564,
+      "num_input_tokens_seen": 168296448,
+      "step": 1284
+    },
+    {
+      "epoch": 0.21119891692863113,
+      "grad_norm": 0.8044641613960266,
+      "learning_rate": 0.00014604338851406909,
+      "loss": 5.4864,
+      "num_input_tokens_seen": 168689664,
+      "step": 1287
+    },
+    {
+      "epoch": 0.21169122209629695,
+      "grad_norm": 0.5943960547447205,
+      "learning_rate": 0.00014587347177414357,
+      "loss": 5.4307,
+      "num_input_tokens_seen": 169082880,
+      "step": 1290
+    },
+    {
+      "epoch": 0.21218352726396275,
+      "grad_norm": 0.6272845268249512,
+      "learning_rate": 0.0001457041467353283,
+      "loss": 5.5211,
+      "num_input_tokens_seen": 169476096,
+      "step": 1293
+    },
+    {
+      "epoch": 0.21267583243162855,
+      "grad_norm": 0.6123335361480713,
+      "learning_rate": 0.0001455354099714415,
+      "loss": 5.5025,
+      "num_input_tokens_seen": 169869312,
+      "step": 1296
+    },
+    {
+      "epoch": 0.21316813759929437,
+      "grad_norm": 0.5472930669784546,
+      "learning_rate": 0.00014536725808401196,
+      "loss": 5.5156,
+      "num_input_tokens_seen": 170262528,
+      "step": 1299
+    },
+    {
+      "epoch": 0.21366044276696017,
+      "grad_norm": 0.6367286443710327,
+      "learning_rate": 0.00014519968770199113,
+      "loss": 5.4662,
+      "num_input_tokens_seen": 170655744,
+      "step": 1302
+    },
+    {
+      "epoch": 0.21415274793462596,
+      "grad_norm": 0.662023663520813,
+      "learning_rate": 0.0001450326954814696,
+      "loss": 5.5199,
+      "num_input_tokens_seen": 171048960,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2146450531022918,
+      "grad_norm": 0.8450511693954468,
+      "learning_rate": 0.00014486627810539652,
+      "loss": 5.4729,
+      "num_input_tokens_seen": 171442176,
+      "step": 1308
+    },
+    {
+      "epoch": 0.21513735826995758,
+      "grad_norm": 0.7732052803039551,
+      "learning_rate": 0.00014470043228330322,
+      "loss": 5.4919,
+      "num_input_tokens_seen": 171835392,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2156296634376234,
+      "grad_norm": 0.6616522073745728,
+      "learning_rate": 0.00014453515475102972,
+      "loss": 5.5101,
+      "num_input_tokens_seen": 172228608,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2161219686052892,
+      "grad_norm": 0.726275622844696,
+      "learning_rate": 0.00014437044227045488,
+      "loss": 5.5317,
+      "num_input_tokens_seen": 172621824,
+      "step": 1317
+    },
+    {
+      "epoch": 0.216614273772955,
+      "grad_norm": 0.8850323557853699,
+      "learning_rate": 0.00014420629162923004,
+      "loss": 5.4694,
+      "num_input_tokens_seen": 173015040,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21710657894062083,
+      "grad_norm": 0.8814117908477783,
+      "learning_rate": 0.00014404269964051592,
+      "loss": 5.5393,
+      "num_input_tokens_seen": 173408256,
+      "step": 1323
+    },
+    {
+      "epoch": 0.21759888410828662,
+      "grad_norm": 0.7925286889076233,
+      "learning_rate": 0.00014387966314272267,
+      "loss": 5.4806,
+      "num_input_tokens_seen": 173801472,
+      "step": 1326
+    },
+    {
+      "epoch": 0.21809118927595242,
+      "grad_norm": 0.5634031891822815,
+      "learning_rate": 0.00014371717899925318,
+      "loss": 5.5134,
+      "num_input_tokens_seen": 174194688,
+      "step": 1329
+    },
+    {
+      "epoch": 0.21858349444361824,
+      "grad_norm": 0.7209234237670898,
+      "learning_rate": 0.00014355524409824985,
+      "loss": 5.4813,
+      "num_input_tokens_seen": 174587904,
+      "step": 1332
+    },
+    {
+      "epoch": 0.21907579961128404,
+      "grad_norm": 0.891487717628479,
+      "learning_rate": 0.00014339385535234412,
+      "loss": 5.4888,
+      "num_input_tokens_seen": 174981120,
+      "step": 1335
+    },
+    {
+      "epoch": 0.21956810477894986,
+      "grad_norm": 0.7084378600120544,
+      "learning_rate": 0.00014323300969840914,
+      "loss": 5.4947,
+      "num_input_tokens_seen": 175374336,
+      "step": 1338
+    },
+    {
+      "epoch": 0.22006040994661566,
+      "grad_norm": 0.6455901265144348,
+      "learning_rate": 0.0001430727040973159,
+      "loss": 5.5063,
+      "num_input_tokens_seen": 175767552,
+      "step": 1341
+    },
+    {
+      "epoch": 0.22055271511428146,
+      "grad_norm": 0.72601717710495,
+      "learning_rate": 0.00014291293553369175,
+      "loss": 5.4397,
+      "num_input_tokens_seen": 176160768,
+      "step": 1344
+    },
+    {
+      "epoch": 0.22104502028194728,
+      "grad_norm": 0.8337487578392029,
+      "learning_rate": 0.00014275370101568235,
+      "loss": 5.491,
+      "num_input_tokens_seen": 176553984,
+      "step": 1347
+    },
+    {
+      "epoch": 0.22153732544961308,
+      "grad_norm": 0.6962984800338745,
+      "learning_rate": 0.00014259499757471626,
+      "loss": 5.514,
+      "num_input_tokens_seen": 176947200,
+      "step": 1350
+    },
+    {
+      "epoch": 0.22202963061727887,
+      "grad_norm": 0.6092506051063538,
+      "learning_rate": 0.00014243682226527246,
+      "loss": 5.44,
+      "num_input_tokens_seen": 177340416,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2225219357849447,
+      "grad_norm": 0.663375735282898,
+      "learning_rate": 0.000142279172164651,
+      "loss": 5.4647,
+      "num_input_tokens_seen": 177733632,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2230142409526105,
+      "grad_norm": 0.7483349442481995,
+      "learning_rate": 0.00014212204437274583,
+      "loss": 5.4908,
+      "num_input_tokens_seen": 178126848,
+      "step": 1359
+    },
+    {
+      "epoch": 0.22350654612027632,
+      "grad_norm": 0.5030447244644165,
+      "learning_rate": 0.00014196543601182097,
+      "loss": 5.4734,
+      "num_input_tokens_seen": 178520064,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2239988512879421,
+      "grad_norm": 0.6785706877708435,
+      "learning_rate": 0.00014180934422628892,
+      "loss": 5.5189,
+      "num_input_tokens_seen": 178913280,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2244911564556079,
+      "grad_norm": 0.6665587425231934,
+      "learning_rate": 0.00014165376618249234,
+      "loss": 5.5011,
+      "num_input_tokens_seen": 179306496,
+      "step": 1368
+    },
+    {
+      "epoch": 0.22498346162327373,
+      "grad_norm": 0.6461958885192871,
+      "learning_rate": 0.00014149869906848755,
+      "loss": 5.4507,
+      "num_input_tokens_seen": 179699712,
+      "step": 1371
+    },
+    {
+      "epoch": 0.22547576679093953,
+      "grad_norm": 0.8757199048995972,
+      "learning_rate": 0.00014134414009383135,
+      "loss": 5.4327,
+      "num_input_tokens_seen": 180092928,
+      "step": 1374
+    },
+    {
+      "epoch": 0.22596807195860535,
+      "grad_norm": 0.7910983562469482,
+      "learning_rate": 0.0001411900864893701,
+      "loss": 5.4511,
+      "num_input_tokens_seen": 180486144,
+      "step": 1377
+    },
+    {
+      "epoch": 0.22646037712627115,
+      "grad_norm": 0.6167607307434082,
+      "learning_rate": 0.00014103653550703125,
+      "loss": 5.4763,
+      "num_input_tokens_seen": 180879360,
+      "step": 1380
+    },
+    {
+      "epoch": 0.22695268229393695,
+      "grad_norm": 0.6035879850387573,
+      "learning_rate": 0.00014088348441961742,
+      "loss": 5.516,
+      "num_input_tokens_seen": 181272576,
+      "step": 1383
+    },
+    {
+      "epoch": 0.22744498746160277,
+      "grad_norm": 0.6464037299156189,
+      "learning_rate": 0.00014073093052060305,
+      "loss": 5.5,
+      "num_input_tokens_seen": 181665792,
+      "step": 1386
+    },
+    {
+      "epoch": 0.22793729262926857,
+      "grad_norm": 0.6020700335502625,
+      "learning_rate": 0.0001405788711239334,
+      "loss": 5.4977,
+      "num_input_tokens_seen": 182059008,
+      "step": 1389
+    },
+    {
+      "epoch": 0.22842959779693436,
+      "grad_norm": 0.5898153185844421,
+      "learning_rate": 0.00014042730356382584,
+      "loss": 5.4852,
+      "num_input_tokens_seen": 182452224,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2289219029646002,
+      "grad_norm": 0.6911790370941162,
+      "learning_rate": 0.00014027622519457354,
+      "loss": 5.4513,
+      "num_input_tokens_seen": 182845440,
+      "step": 1395
+    },
+    {
+      "epoch": 0.22941420813226598,
+      "grad_norm": 0.7918622493743896,
+      "learning_rate": 0.00014012563339035157,
+      "loss": 5.5059,
+      "num_input_tokens_seen": 183238656,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2299065132999318,
+      "grad_norm": 0.7177248597145081,
+      "learning_rate": 0.00013997552554502517,
+      "loss": 5.4782,
+      "num_input_tokens_seen": 183631872,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2303988184675976,
+      "grad_norm": 0.5545403361320496,
+      "learning_rate": 0.00013982589907196038,
+      "loss": 5.4598,
+      "num_input_tokens_seen": 184025088,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2308911236352634,
+      "grad_norm": 0.5726358294487,
+      "learning_rate": 0.00013967675140383676,
+      "loss": 5.4737,
+      "num_input_tokens_seen": 184418304,
+      "step": 1407
+    },
+    {
+      "epoch": 0.23138342880292923,
+      "grad_norm": 0.6597428917884827,
+      "learning_rate": 0.00013952807999246237,
+      "loss": 5.4761,
+      "num_input_tokens_seen": 184811520,
+      "step": 1410
+    },
+    {
+      "epoch": 0.23187573397059502,
+      "grad_norm": 0.7973009347915649,
+      "learning_rate": 0.000139379882308591,
+      "loss": 5.4572,
+      "num_input_tokens_seen": 185204736,
+      "step": 1413
+    },
+    {
+      "epoch": 0.23236803913826082,
+      "grad_norm": 0.797472357749939,
+      "learning_rate": 0.00013923215584174146,
+      "loss": 5.456,
+      "num_input_tokens_seen": 185597952,
+      "step": 1416
+    },
+    {
+      "epoch": 0.23286034430592664,
+      "grad_norm": 0.6406548023223877,
+      "learning_rate": 0.00013908489810001876,
+      "loss": 5.4899,
+      "num_input_tokens_seen": 185991168,
+      "step": 1419
+    },
+    {
+      "epoch": 0.23335264947359244,
+      "grad_norm": 0.6479483246803284,
+      "learning_rate": 0.00013893810660993777,
+      "loss": 5.451,
+      "num_input_tokens_seen": 186384384,
+      "step": 1422
+    },
+    {
+      "epoch": 0.23384495464125826,
+      "grad_norm": 0.7050771713256836,
+      "learning_rate": 0.00013879177891624862,
+      "loss": 5.4301,
+      "num_input_tokens_seen": 186777600,
+      "step": 1425
+    },
+    {
+      "epoch": 0.23433725980892406,
+      "grad_norm": 0.7287415862083435,
+      "learning_rate": 0.00013864591258176437,
+      "loss": 5.4478,
+      "num_input_tokens_seen": 187170816,
+      "step": 1428
+    },
+    {
+      "epoch": 0.23482956497658986,
+      "grad_norm": 0.7213960289955139,
+      "learning_rate": 0.00013850050518719026,
+      "loss": 5.4639,
+      "num_input_tokens_seen": 187564032,
+      "step": 1431
+    },
+    {
+      "epoch": 0.23532187014425568,
+      "grad_norm": 0.6890702843666077,
+      "learning_rate": 0.00013835555433095535,
+      "loss": 5.4457,
+      "num_input_tokens_seen": 187957248,
+      "step": 1434
+    },
+    {
+      "epoch": 0.23581417531192148,
+      "grad_norm": 0.7307390570640564,
+      "learning_rate": 0.0001382110576290459,
+      "loss": 5.4538,
+      "num_input_tokens_seen": 188350464,
+      "step": 1437
+    },
+    {
+      "epoch": 0.23630648047958727,
+      "grad_norm": 0.578571617603302,
+      "learning_rate": 0.00013806701271484075,
+      "loss": 5.4692,
+      "num_input_tokens_seen": 188743680,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2367987856472531,
+      "grad_norm": 0.7552468180656433,
+      "learning_rate": 0.0001379234172389483,
+      "loss": 5.4891,
+      "num_input_tokens_seen": 189136896,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2372910908149189,
+      "grad_norm": 1.0629295110702515,
+      "learning_rate": 0.00013778026886904584,
+      "loss": 5.3936,
+      "num_input_tokens_seen": 189530112,
+      "step": 1446
+    },
+    {
+      "epoch": 0.23778339598258472,
+      "grad_norm": 1.178006649017334,
+      "learning_rate": 0.00013763756528972017,
+      "loss": 5.4754,
+      "num_input_tokens_seen": 189923328,
+      "step": 1449
+    },
+    {
+      "epoch": 0.2382757011502505,
+      "grad_norm": 0.9626088738441467,
+      "learning_rate": 0.00013749530420231065,
+      "loss": 5.4676,
+      "num_input_tokens_seen": 190316544,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2387680063179163,
+      "grad_norm": 0.7882675528526306,
+      "learning_rate": 0.00013735348332475335,
+      "loss": 5.4157,
+      "num_input_tokens_seen": 190709760,
+      "step": 1455
+    },
+    {
+      "epoch": 0.23926031148558213,
+      "grad_norm": 0.6379857659339905,
+      "learning_rate": 0.0001372121003914274,
+      "loss": 5.4558,
+      "num_input_tokens_seen": 191102976,
+      "step": 1458
+    },
+    {
+      "epoch": 0.23975261665324793,
+      "grad_norm": 0.7245479822158813,
+      "learning_rate": 0.00013707115315300314,
+      "loss": 5.4405,
+      "num_input_tokens_seen": 191496192,
+      "step": 1461
+    },
+    {
+      "epoch": 0.24024492182091373,
+      "grad_norm": 0.7578230500221252,
+      "learning_rate": 0.00013693063937629153,
+      "loss": 5.4516,
+      "num_input_tokens_seen": 191889408,
+      "step": 1464
+    },
+    {
+      "epoch": 0.24073722698857955,
+      "grad_norm": 0.6436443328857422,
+      "learning_rate": 0.00013679055684409573,
+      "loss": 5.46,
+      "num_input_tokens_seen": 192282624,
+      "step": 1467
+    },
+    {
+      "epoch": 0.24122953215624535,
+      "grad_norm": 0.6767442226409912,
+      "learning_rate": 0.00013665090335506422,
+      "loss": 5.472,
+      "num_input_tokens_seen": 192675840,
+      "step": 1470
+    },
+    {
+      "epoch": 0.24172183732391117,
+      "grad_norm": 0.7930536270141602,
+      "learning_rate": 0.00013651167672354525,
+      "loss": 5.4666,
+      "num_input_tokens_seen": 193069056,
+      "step": 1473
+    },
+    {
+      "epoch": 0.24221414249157697,
+      "grad_norm": 0.8025949001312256,
+      "learning_rate": 0.0001363728747794434,
+      "loss": 5.423,
+      "num_input_tokens_seen": 193462272,
+      "step": 1476
+    },
+    {
+      "epoch": 0.24270644765924276,
+      "grad_norm": 0.79132479429245,
+      "learning_rate": 0.00013623449536807747,
+      "loss": 5.5005,
+      "num_input_tokens_seen": 193855488,
+      "step": 1479
+    },
+    {
+      "epoch": 0.2431987528269086,
+      "grad_norm": 0.7200051546096802,
+      "learning_rate": 0.00013609653635003992,
+      "loss": 5.4393,
+      "num_input_tokens_seen": 194248704,
+      "step": 1482
+    },
+    {
+      "epoch": 0.24369105799457438,
+      "grad_norm": 0.8778846263885498,
+      "learning_rate": 0.0001359589956010579,
+      "loss": 5.4721,
+      "num_input_tokens_seen": 194641920,
+      "step": 1485
+    },
+    {
+      "epoch": 0.24418336316224018,
+      "grad_norm": 0.7225459814071655,
+      "learning_rate": 0.00013582187101185615,
+      "loss": 5.4733,
+      "num_input_tokens_seen": 195035136,
+      "step": 1488
+    },
+    {
+      "epoch": 0.244675668329906,
+      "grad_norm": 0.7456029057502747,
+      "learning_rate": 0.00013568516048802077,
+      "loss": 5.4594,
+      "num_input_tokens_seen": 195428352,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2451679734975718,
+      "grad_norm": 0.7556261420249939,
+      "learning_rate": 0.0001355488619498652,
+      "loss": 5.4692,
+      "num_input_tokens_seen": 195821568,
+      "step": 1494
+    },
+    {
+      "epoch": 0.24566027866523762,
+      "grad_norm": 0.6962366104125977,
+      "learning_rate": 0.00013541297333229701,
+      "loss": 5.4188,
+      "num_input_tokens_seen": 196214784,
+      "step": 1497
+    },
+    {
+      "epoch": 0.24615258383290342,
+      "grad_norm": 0.5996794104576111,
+      "learning_rate": 0.00013527749258468682,
+      "loss": 5.483,
+      "num_input_tokens_seen": 196608000,
+      "step": 1500
+    },
+    {
+      "epoch": 0.24664488900056922,
+      "grad_norm": 0.8026195764541626,
+      "learning_rate": 0.00013514241767073804,
+      "loss": 5.4271,
+      "num_input_tokens_seen": 197001216,
+      "step": 1503
+    },
+    {
+      "epoch": 0.24713719416823504,
+      "grad_norm": 0.7598974704742432,
+      "learning_rate": 0.00013500774656835854,
+      "loss": 5.457,
+      "num_input_tokens_seen": 197394432,
+      "step": 1506
+    },
+    {
+      "epoch": 0.24762949933590084,
+      "grad_norm": 0.6312159895896912,
+      "learning_rate": 0.00013487347726953342,
+      "loss": 5.4753,
+      "num_input_tokens_seen": 197787648,
+      "step": 1509
+    },
+    {
+      "epoch": 0.24812180450356666,
+      "grad_norm": 0.7285313606262207,
+      "learning_rate": 0.00013473960778019915,
+      "loss": 5.4443,
+      "num_input_tokens_seen": 198180864,
+      "step": 1512
+    },
+    {
+      "epoch": 0.24861410967123246,
+      "grad_norm": 0.7714123725891113,
+      "learning_rate": 0.0001346061361201194,
+      "loss": 5.4059,
+      "num_input_tokens_seen": 198574080,
+      "step": 1515
+    },
+    {
+      "epoch": 0.24910641483889825,
+      "grad_norm": 0.790026843547821,
+      "learning_rate": 0.00013447306032276192,
+      "loss": 5.3898,
+      "num_input_tokens_seen": 198967296,
+      "step": 1518
+    },
+    {
+      "epoch": 0.24959872000656408,
+      "grad_norm": 0.6501436233520508,
+      "learning_rate": 0.00013434037843517677,
+      "loss": 5.4489,
+      "num_input_tokens_seen": 199360512,
+      "step": 1521
+    },
+    {
+      "epoch": 0.2500910251742299,
+      "grad_norm": 0.6346362829208374,
+      "learning_rate": 0.00013420808851787603,
+      "loss": 5.4448,
+      "num_input_tokens_seen": 199753728,
+      "step": 1524
+    },
+    {
+      "epoch": 0.25058333034189567,
+      "grad_norm": 0.6744604706764221,
+      "learning_rate": 0.00013407618864471469,
+      "loss": 5.434,
+      "num_input_tokens_seen": 200146944,
+      "step": 1527
+    },
+    {
+      "epoch": 0.25107563550956147,
+      "grad_norm": 0.7855573296546936,
+      "learning_rate": 0.00013394467690277295,
+      "loss": 5.4304,
+      "num_input_tokens_seen": 200540160,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2515679406772273,
+      "grad_norm": 0.5886304974555969,
+      "learning_rate": 0.0001338135513922395,
+      "loss": 5.428,
+      "num_input_tokens_seen": 200933376,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2520602458448931,
+      "grad_norm": 0.6847313046455383,
+      "learning_rate": 0.00013368281022629647,
+      "loss": 5.4176,
+      "num_input_tokens_seen": 201326592,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2525525510125589,
+      "grad_norm": 0.600709080696106,
+      "learning_rate": 0.0001335524515310053,
+      "loss": 5.4369,
+      "num_input_tokens_seen": 201719808,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2530448561802247,
+      "grad_norm": 0.5900317430496216,
+      "learning_rate": 0.00013342247344519384,
+      "loss": 5.4217,
+      "num_input_tokens_seen": 202113024,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2535371613478905,
+      "grad_norm": 0.6889588236808777,
+      "learning_rate": 0.00013329287412034498,
+      "loss": 5.4327,
+      "num_input_tokens_seen": 202506240,
+      "step": 1545
+    },
+    {
+      "epoch": 0.25402946651555636,
+      "grad_norm": 0.699284017086029,
+      "learning_rate": 0.00013316365172048595,
+      "loss": 5.4391,
+      "num_input_tokens_seen": 202899456,
+      "step": 1548
+    },
+    {
+      "epoch": 0.25452177168322215,
+      "grad_norm": 0.7852107286453247,
+      "learning_rate": 0.0001330348044220793,
+      "loss": 5.4022,
+      "num_input_tokens_seen": 203292672,
+      "step": 1551
+    },
+    {
+      "epoch": 0.25501407685088795,
+      "grad_norm": 0.9291219115257263,
+      "learning_rate": 0.00013290633041391467,
+      "loss": 5.4094,
+      "num_input_tokens_seen": 203685888,
+      "step": 1554
+    },
+    {
+      "epoch": 0.25550638201855375,
+      "grad_norm": 0.7158473134040833,
+      "learning_rate": 0.000132778227897002,
+      "loss": 5.397,
+      "num_input_tokens_seen": 204079104,
+      "step": 1557
+    },
+    {
+      "epoch": 0.25599868718621954,
+      "grad_norm": 0.761669933795929,
+      "learning_rate": 0.00013265049508446564,
+      "loss": 5.4425,
+      "num_input_tokens_seen": 204472320,
+      "step": 1560
+    },
+    {
+      "epoch": 0.25649099235388534,
+      "grad_norm": 0.7269110679626465,
+      "learning_rate": 0.0001325231302014396,
+      "loss": 5.4098,
+      "num_input_tokens_seen": 204865536,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2569832975215512,
+      "grad_norm": 0.8132617473602295,
+      "learning_rate": 0.0001323961314849641,
+      "loss": 5.4133,
+      "num_input_tokens_seen": 205258752,
+      "step": 1566
+    },
+    {
+      "epoch": 0.257475602689217,
+      "grad_norm": 0.7336274981498718,
+      "learning_rate": 0.00013226949718388306,
+      "loss": 5.4255,
+      "num_input_tokens_seen": 205651968,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2579679078568828,
+      "grad_norm": 0.6194922924041748,
+      "learning_rate": 0.0001321432255587425,
+      "loss": 5.4495,
+      "num_input_tokens_seen": 206045184,
+      "step": 1572
+    },
+    {
+      "epoch": 0.2584602130245486,
+      "grad_norm": 0.6153448820114136,
+      "learning_rate": 0.00013201731488169053,
+      "loss": 5.4278,
+      "num_input_tokens_seen": 206438400,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2589525181922144,
+      "grad_norm": 0.6616208553314209,
+      "learning_rate": 0.0001318917634363777,
+      "loss": 5.3997,
+      "num_input_tokens_seen": 206831616,
+      "step": 1578
+    },
+    {
+      "epoch": 0.25944482335988023,
+      "grad_norm": 0.670545756816864,
+      "learning_rate": 0.00013176656951785888,
+      "loss": 5.4721,
+      "num_input_tokens_seen": 207224832,
+      "step": 1581
+    },
+    {
+      "epoch": 0.259937128527546,
+      "grad_norm": 0.6420415639877319,
+      "learning_rate": 0.00013164173143249616,
+      "loss": 5.3955,
+      "num_input_tokens_seen": 207618048,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2604294336952118,
+      "grad_norm": 0.6072604060173035,
+      "learning_rate": 0.00013151724749786237,
+      "loss": 5.4327,
+      "num_input_tokens_seen": 208011264,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2609217388628776,
+      "grad_norm": 0.6046293377876282,
+      "learning_rate": 0.00013139311604264595,
+      "loss": 5.4234,
+      "num_input_tokens_seen": 208404480,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2614140440305434,
+      "grad_norm": 0.6301759481430054,
+      "learning_rate": 0.00013126933540655674,
+      "loss": 5.412,
+      "num_input_tokens_seen": 208797696,
+      "step": 1593
+    },
+    {
+      "epoch": 0.26190634919820927,
+      "grad_norm": 0.5329933762550354,
+      "learning_rate": 0.00013114590394023272,
+      "loss": 5.4476,
+      "num_input_tokens_seen": 209190912,
+      "step": 1596
+    },
+    {
+      "epoch": 0.26239865436587506,
+      "grad_norm": 0.6598846912384033,
+      "learning_rate": 0.0001310228200051476,
+      "loss": 5.455,
+      "num_input_tokens_seen": 209584128,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2625627560884303,
+      "eval_accuracy": 0.20324051457417358,
+      "eval_loss": 5.642736911773682,
+      "eval_runtime": 109.8552,
+      "eval_samples_per_second": 2.731,
+      "eval_steps_per_second": 1.365,
+      "num_input_tokens_seen": 209715200,
+      "step": 1600
+    },
+    {
+      "epoch": 0.26289095953354086,
+      "grad_norm": 0.5440322756767273,
+      "learning_rate": 0.00013090008197351962,
+      "loss": 5.4318,
+      "num_input_tokens_seen": 209977344,
+      "step": 1602
+    },
+    {
+      "epoch": 0.26338326470120665,
+      "grad_norm": 0.5659436583518982,
+      "learning_rate": 0.0001307776882282209,
+      "loss": 5.4397,
+      "num_input_tokens_seen": 210370560,
+      "step": 1605
+    },
+    {
+      "epoch": 0.26387556986887245,
+      "grad_norm": 0.6262729167938232,
+      "learning_rate": 0.0001306556371626883,
+      "loss": 5.3882,
+      "num_input_tokens_seen": 210763776,
+      "step": 1608
+    },
+    {
+      "epoch": 0.26436787503653825,
+      "grad_norm": 0.6548302173614502,
+      "learning_rate": 0.00013053392718083447,
+      "loss": 5.4368,
+      "num_input_tokens_seen": 211156992,
+      "step": 1611
+    },
+    {
+      "epoch": 0.2648601802042041,
+      "grad_norm": 0.6563411355018616,
+      "learning_rate": 0.00013041255669696042,
+      "loss": 5.3665,
+      "num_input_tokens_seen": 211550208,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2653524853718699,
+      "grad_norm": 0.6907044649124146,
+      "learning_rate": 0.00013029152413566872,
+      "loss": 5.4226,
+      "num_input_tokens_seen": 211943424,
+      "step": 1617
+    },
+    {
+      "epoch": 0.2658447905395357,
+      "grad_norm": 0.715350866317749,
+      "learning_rate": 0.00013017082793177756,
+      "loss": 5.4769,
+      "num_input_tokens_seen": 212336640,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2663370957072015,
+      "grad_norm": 0.7537038326263428,
+      "learning_rate": 0.0001300504665302358,
+      "loss": 5.4195,
+      "num_input_tokens_seen": 212729856,
+      "step": 1623
+    },
+    {
+      "epoch": 0.2668294008748673,
+      "grad_norm": 0.7719680070877075,
+      "learning_rate": 0.00012993043838603865,
+      "loss": 5.4237,
+      "num_input_tokens_seen": 213123072,
+      "step": 1626
+    },
+    {
+      "epoch": 0.26732170604253314,
+      "grad_norm": 0.6925044059753418,
+      "learning_rate": 0.00012981074196414472,
+      "loss": 5.3933,
+      "num_input_tokens_seen": 213516288,
+      "step": 1629
+    },
+    {
+      "epoch": 0.26781401121019893,
+      "grad_norm": 0.6691201329231262,
+      "learning_rate": 0.0001296913757393932,
+      "loss": 5.4641,
+      "num_input_tokens_seen": 213909504,
+      "step": 1632
+    },
+    {
+      "epoch": 0.26830631637786473,
+      "grad_norm": 0.6009002923965454,
+      "learning_rate": 0.00012957233819642244,
+      "loss": 5.3632,
+      "num_input_tokens_seen": 214302720,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2687986215455305,
+      "grad_norm": 0.5620663166046143,
+      "learning_rate": 0.00012945362782958907,
+      "loss": 5.4091,
+      "num_input_tokens_seen": 214695936,
+      "step": 1638
+    },
+    {
+      "epoch": 0.2692909267131963,
+      "grad_norm": 0.585953414440155,
+      "learning_rate": 0.0001293352431428881,
+      "loss": 5.3819,
+      "num_input_tokens_seen": 215089152,
+      "step": 1641
+    },
+    {
+      "epoch": 0.2697832318808622,
+      "grad_norm": 0.604200541973114,
+      "learning_rate": 0.00012921718264987363,
+      "loss": 5.4161,
+      "num_input_tokens_seen": 215482368,
+      "step": 1644
+    },
+    {
+      "epoch": 0.27027553704852797,
+      "grad_norm": 0.7634391188621521,
+      "learning_rate": 0.00012909944487358055,
+      "loss": 5.3987,
+      "num_input_tokens_seen": 215875584,
+      "step": 1647
+    },
+    {
+      "epoch": 0.27076784221619377,
+      "grad_norm": 0.6789970993995667,
+      "learning_rate": 0.0001289820283464469,
+      "loss": 5.3772,
+      "num_input_tokens_seen": 216268800,
+      "step": 1650
+    },
+    {
+      "epoch": 0.27126014738385956,
+      "grad_norm": 0.615770697593689,
+      "learning_rate": 0.00012886493161023702,
+      "loss": 5.403,
+      "num_input_tokens_seen": 216662016,
+      "step": 1653
+    },
+    {
+      "epoch": 0.27175245255152536,
+      "grad_norm": 0.821854829788208,
+      "learning_rate": 0.00012874815321596553,
+      "loss": 5.3783,
+      "num_input_tokens_seen": 217055232,
+      "step": 1656
+    },
+    {
+      "epoch": 0.2722447577191912,
+      "grad_norm": 0.7120591998100281,
+      "learning_rate": 0.00012863169172382195,
+      "loss": 5.4067,
+      "num_input_tokens_seen": 217448448,
+      "step": 1659
+    },
+    {
+      "epoch": 0.272737062886857,
+      "grad_norm": 0.561850368976593,
+      "learning_rate": 0.00012851554570309626,
+      "loss": 5.4273,
+      "num_input_tokens_seen": 217841664,
+      "step": 1662
+    },
+    {
+      "epoch": 0.2732293680545228,
+      "grad_norm": 0.6624509692192078,
+      "learning_rate": 0.0001283997137321049,
+      "loss": 5.4209,
+      "num_input_tokens_seen": 218234880,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2737216732221886,
+      "grad_norm": 0.6127106547355652,
+      "learning_rate": 0.00012828419439811785,
+      "loss": 5.3708,
+      "num_input_tokens_seen": 218628096,
+      "step": 1668
+    },
+    {
+      "epoch": 0.2742139783898544,
+      "grad_norm": 0.6014697551727295,
+      "learning_rate": 0.00012816898629728628,
+      "loss": 5.4795,
+      "num_input_tokens_seen": 219021312,
+      "step": 1671
+    },
+    {
+      "epoch": 0.2747062835575202,
+      "grad_norm": 0.5605149865150452,
+      "learning_rate": 0.0001280540880345707,
+      "loss": 5.3716,
+      "num_input_tokens_seen": 219414528,
+      "step": 1674
+    },
+    {
+      "epoch": 0.27519858872518604,
+      "grad_norm": 0.680549144744873,
+      "learning_rate": 0.00012793949822367017,
+      "loss": 5.4007,
+      "num_input_tokens_seen": 219807744,
+      "step": 1677
+    },
+    {
+      "epoch": 0.27569089389285184,
+      "grad_norm": 0.584894597530365,
+      "learning_rate": 0.000127825215486952,
+      "loss": 5.3907,
+      "num_input_tokens_seen": 220200960,
+      "step": 1680
+    },
+    {
+      "epoch": 0.27618319906051764,
+      "grad_norm": 0.5957493782043457,
+      "learning_rate": 0.00012771123845538215,
+      "loss": 5.4456,
+      "num_input_tokens_seen": 220594176,
+      "step": 1683
+    },
+    {
+      "epoch": 0.27667550422818343,
+      "grad_norm": 0.6230633854866028,
+      "learning_rate": 0.00012759756576845652,
+      "loss": 5.4259,
+      "num_input_tokens_seen": 220987392,
+      "step": 1686
+    },
+    {
+      "epoch": 0.27716780939584923,
+      "grad_norm": 0.6842202544212341,
+      "learning_rate": 0.00012748419607413246,
+      "loss": 5.3754,
+      "num_input_tokens_seen": 221380608,
+      "step": 1689
+    },
+    {
+      "epoch": 0.2776601145635151,
+      "grad_norm": 0.7493085861206055,
+      "learning_rate": 0.00012737112802876149,
+      "loss": 5.3766,
+      "num_input_tokens_seen": 221773824,
+      "step": 1692
+    },
+    {
+      "epoch": 0.2781524197311809,
+      "grad_norm": 0.8707619905471802,
+      "learning_rate": 0.00012725836029702222,
+      "loss": 5.376,
+      "num_input_tokens_seen": 222167040,
+      "step": 1695
+    },
+    {
+      "epoch": 0.2786447248988467,
+      "grad_norm": 0.9452652335166931,
+      "learning_rate": 0.00012714589155185432,
+      "loss": 5.4282,
+      "num_input_tokens_seen": 222560256,
+      "step": 1698
+    },
+    {
+      "epoch": 0.27913703006651247,
+      "grad_norm": 0.7085415124893188,
+      "learning_rate": 0.00012703372047439269,
+      "loss": 5.4535,
+      "num_input_tokens_seen": 222953472,
+      "step": 1701
+    },
+    {
+      "epoch": 0.27962933523417827,
+      "grad_norm": 0.6184890866279602,
+      "learning_rate": 0.00012692184575390268,
+      "loss": 5.3622,
+      "num_input_tokens_seen": 223346688,
+      "step": 1704
+    },
+    {
+      "epoch": 0.2801216404018441,
+      "grad_norm": 0.8037028312683105,
+      "learning_rate": 0.0001268102660877157,
+      "loss": 5.3098,
+      "num_input_tokens_seen": 223739904,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2806139455695099,
+      "grad_norm": 0.8075725436210632,
+      "learning_rate": 0.00012669898018116552,
+      "loss": 5.421,
+      "num_input_tokens_seen": 224133120,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2811062507371757,
+      "grad_norm": 0.8237155079841614,
+      "learning_rate": 0.0001265879867475251,
+      "loss": 5.3636,
+      "num_input_tokens_seen": 224526336,
+      "step": 1713
+    },
+    {
+      "epoch": 0.2815985559048415,
+      "grad_norm": 0.7250863313674927,
+      "learning_rate": 0.00012647728450794433,
+      "loss": 5.4245,
+      "num_input_tokens_seen": 224919552,
+      "step": 1716
+    },
+    {
+      "epoch": 0.2820908610725073,
+      "grad_norm": 0.5800032019615173,
+      "learning_rate": 0.00012636687219138784,
+      "loss": 5.3722,
+      "num_input_tokens_seen": 225312768,
+      "step": 1719
+    },
+    {
+      "epoch": 0.2825831662401731,
+      "grad_norm": 0.6191242933273315,
+      "learning_rate": 0.00012625674853457394,
+      "loss": 5.3959,
+      "num_input_tokens_seen": 225705984,
+      "step": 1722
+    },
+    {
+      "epoch": 0.28307547140783895,
+      "grad_norm": 0.6112586259841919,
+      "learning_rate": 0.00012614691228191385,
+      "loss": 5.386,
+      "num_input_tokens_seen": 226099200,
+      "step": 1725
+    },
+    {
+      "epoch": 0.28356777657550475,
+      "grad_norm": 0.5698121190071106,
+      "learning_rate": 0.00012603736218545143,
+      "loss": 5.3502,
+      "num_input_tokens_seen": 226492416,
+      "step": 1728
+    },
+    {
+      "epoch": 0.28406008174317054,
+      "grad_norm": 0.5981942415237427,
+      "learning_rate": 0.00012592809700480388,
+      "loss": 5.3705,
+      "num_input_tokens_seen": 226885632,
+      "step": 1731
+    },
+    {
+      "epoch": 0.28455238691083634,
+      "grad_norm": 0.5503790378570557,
+      "learning_rate": 0.00012581911550710255,
+      "loss": 5.364,
+      "num_input_tokens_seen": 227278848,
+      "step": 1734
+    },
+    {
+      "epoch": 0.28504469207850214,
+      "grad_norm": 0.7332367897033691,
+      "learning_rate": 0.00012571041646693466,
+      "loss": 5.3551,
+      "num_input_tokens_seen": 227672064,
+      "step": 1737
+    },
+    {
+      "epoch": 0.285536997246168,
+      "grad_norm": 0.6840054392814636,
+      "learning_rate": 0.0001256019986662852,
+      "loss": 5.3328,
+      "num_input_tokens_seen": 228065280,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2860293024138338,
+      "grad_norm": 0.6030425429344177,
+      "learning_rate": 0.00012549386089447998,
+      "loss": 5.3826,
+      "num_input_tokens_seen": 228458496,
+      "step": 1743
+    },
+    {
+      "epoch": 0.2865216075814996,
+      "grad_norm": 0.689045786857605,
+      "learning_rate": 0.0001253860019481285,
+      "loss": 5.3991,
+      "num_input_tokens_seen": 228851712,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2870139127491654,
+      "grad_norm": 0.6273123621940613,
+      "learning_rate": 0.0001252784206310678,
+      "loss": 5.3241,
+      "num_input_tokens_seen": 229244928,
+      "step": 1749
+    },
+    {
+      "epoch": 0.2875062179168312,
+      "grad_norm": 0.7061653137207031,
+      "learning_rate": 0.0001251711157543068,
+      "loss": 5.3797,
+      "num_input_tokens_seen": 229638144,
+      "step": 1752
+    },
+    {
+      "epoch": 0.287998523084497,
+      "grad_norm": 0.6948421001434326,
+      "learning_rate": 0.00012506408613597125,
+      "loss": 5.4058,
+      "num_input_tokens_seen": 230031360,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2884908282521628,
+      "grad_norm": 0.6483265161514282,
+      "learning_rate": 0.00012495733060124866,
+      "loss": 5.3982,
+      "num_input_tokens_seen": 230424576,
+      "step": 1758
+    },
+    {
+      "epoch": 0.2889831334198286,
+      "grad_norm": 0.6624945998191833,
+      "learning_rate": 0.00012485084798233452,
+      "loss": 5.3528,
+      "num_input_tokens_seen": 230817792,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2894754385874944,
+      "grad_norm": 0.7422268390655518,
+      "learning_rate": 0.0001247446371183784,
+      "loss": 5.3745,
+      "num_input_tokens_seen": 231211008,
+      "step": 1764
+    },
+    {
+      "epoch": 0.2899677437551602,
+      "grad_norm": 0.5802010893821716,
+      "learning_rate": 0.00012463869685543102,
+      "loss": 5.3446,
+      "num_input_tokens_seen": 231604224,
+      "step": 1767
+    },
+    {
+      "epoch": 0.29046004892282606,
+      "grad_norm": 0.6384250521659851,
+      "learning_rate": 0.00012453302604639133,
+      "loss": 5.3833,
+      "num_input_tokens_seen": 231997440,
+      "step": 1770
+    },
+    {
+      "epoch": 0.29095235409049186,
+      "grad_norm": 0.6214941740036011,
+      "learning_rate": 0.00012442762355095458,
+      "loss": 5.3835,
+      "num_input_tokens_seen": 232390656,
+      "step": 1773
+    },
+    {
+      "epoch": 0.29144465925815766,
+      "grad_norm": 0.6470031142234802,
+      "learning_rate": 0.0001243224882355605,
+      "loss": 5.3594,
+      "num_input_tokens_seen": 232783872,
+      "step": 1776
+    },
+    {
+      "epoch": 0.29193696442582345,
+      "grad_norm": 0.5898450016975403,
+      "learning_rate": 0.00012421761897334212,
+      "loss": 5.3479,
+      "num_input_tokens_seen": 233177088,
+      "step": 1779
+    },
+    {
+      "epoch": 0.29242926959348925,
+      "grad_norm": 0.625322699546814,
+      "learning_rate": 0.00012411301464407512,
+      "loss": 5.3949,
+      "num_input_tokens_seen": 233570304,
+      "step": 1782
+    },
+    {
+      "epoch": 0.29292157476115505,
+      "grad_norm": 0.7103887796401978,
+      "learning_rate": 0.0001240086741341274,
+      "loss": 5.3434,
+      "num_input_tokens_seen": 233963520,
+      "step": 1785
+    },
+    {
+      "epoch": 0.2934138799288209,
+      "grad_norm": 0.6434682011604309,
+      "learning_rate": 0.00012390459633640952,
+      "loss": 5.3855,
+      "num_input_tokens_seen": 234356736,
+      "step": 1788
+    },
+    {
+      "epoch": 0.2939061850964867,
+      "grad_norm": 0.6767953038215637,
+      "learning_rate": 0.00012380078015032517,
+      "loss": 5.3375,
+      "num_input_tokens_seen": 234749952,
+      "step": 1791
+    },
+    {
+      "epoch": 0.2943984902641525,
+      "grad_norm": 0.6974107623100281,
+      "learning_rate": 0.00012369722448172233,
+      "loss": 5.3822,
+      "num_input_tokens_seen": 235143168,
+      "step": 1794
+    },
+    {
+      "epoch": 0.2948907954318183,
+      "grad_norm": 0.5709061026573181,
+      "learning_rate": 0.0001235939282428449,
+      "loss": 5.355,
+      "num_input_tokens_seen": 235536384,
+      "step": 1797
+    },
+    {
+      "epoch": 0.2953831005994841,
+      "grad_norm": 0.5687236189842224,
+      "learning_rate": 0.00012349089035228468,
+      "loss": 5.3811,
+      "num_input_tokens_seen": 235929600,
+      "step": 1800
+    },
+    {
+      "epoch": 0.29587540576714993,
+      "grad_norm": 0.5439820885658264,
+      "learning_rate": 0.0001233881097349338,
+      "loss": 5.3716,
+      "num_input_tokens_seen": 236322816,
+      "step": 1803
+    },
+    {
+      "epoch": 0.29636771093481573,
+      "grad_norm": 0.5737081170082092,
+      "learning_rate": 0.0001232855853219376,
+      "loss": 5.3847,
+      "num_input_tokens_seen": 236716032,
+      "step": 1806
+    },
+    {
+      "epoch": 0.2968600161024815,
+      "grad_norm": 0.663943350315094,
+      "learning_rate": 0.000123183316050648,
+      "loss": 5.3875,
+      "num_input_tokens_seen": 237109248,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2973523212701473,
+      "grad_norm": 0.6561760902404785,
+      "learning_rate": 0.00012308130086457713,
+      "loss": 5.3778,
+      "num_input_tokens_seen": 237502464,
+      "step": 1812
+    },
+    {
+      "epoch": 0.2978446264378131,
+      "grad_norm": 0.6318316459655762,
+      "learning_rate": 0.00012297953871335165,
+      "loss": 5.3343,
+      "num_input_tokens_seen": 237895680,
+      "step": 1815
+    },
+    {
+      "epoch": 0.29833693160547897,
+      "grad_norm": 0.5666148066520691,
+      "learning_rate": 0.00012287802855266721,
+      "loss": 5.3317,
+      "num_input_tokens_seen": 238288896,
+      "step": 1818
+    },
+    {
+      "epoch": 0.29882923677314477,
+      "grad_norm": 0.7399240732192993,
+      "learning_rate": 0.00012277676934424343,
+      "loss": 5.3615,
+      "num_input_tokens_seen": 238682112,
+      "step": 1821
+    },
+    {
+      "epoch": 0.29932154194081056,
+      "grad_norm": 0.7315434813499451,
+      "learning_rate": 0.00012267576005577937,
+      "loss": 5.3207,
+      "num_input_tokens_seen": 239075328,
+      "step": 1824
+    },
+    {
+      "epoch": 0.29981384710847636,
+      "grad_norm": 0.6648403406143188,
+      "learning_rate": 0.00012257499966090933,
+      "loss": 5.387,
+      "num_input_tokens_seen": 239468544,
+      "step": 1827
+    },
+    {
+      "epoch": 0.30030615227614216,
+      "grad_norm": 0.6215511560440063,
+      "learning_rate": 0.00012247448713915892,
+      "loss": 5.3734,
+      "num_input_tokens_seen": 239861760,
+      "step": 1830
+    },
+    {
+      "epoch": 0.30079845744380795,
+      "grad_norm": 0.6501013040542603,
+      "learning_rate": 0.0001223742214759018,
+      "loss": 5.3226,
+      "num_input_tokens_seen": 240254976,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3012907626114738,
+      "grad_norm": 0.5968300104141235,
+      "learning_rate": 0.00012227420166231658,
+      "loss": 5.3354,
+      "num_input_tokens_seen": 240648192,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3017830677791396,
+      "grad_norm": 0.6109431385993958,
+      "learning_rate": 0.00012217442669534425,
+      "loss": 5.3623,
+      "num_input_tokens_seen": 241041408,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3022753729468054,
+      "grad_norm": 0.7393885254859924,
+      "learning_rate": 0.00012207489557764593,
+      "loss": 5.3629,
+      "num_input_tokens_seen": 241434624,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3027676781144712,
+      "grad_norm": 0.7628598213195801,
+      "learning_rate": 0.00012197560731756083,
+      "loss": 5.3368,
+      "num_input_tokens_seen": 241827840,
+      "step": 1845
+    },
+    {
+      "epoch": 0.303259983282137,
+      "grad_norm": 0.8370330929756165,
+      "learning_rate": 0.00012187656092906502,
+      "loss": 5.3438,
+      "num_input_tokens_seen": 242221056,
+      "step": 1848
+    },
+    {
+      "epoch": 0.30375228844980284,
+      "grad_norm": 0.7718897461891174,
+      "learning_rate": 0.0001217777554317301,
+      "loss": 5.376,
+      "num_input_tokens_seen": 242614272,
+      "step": 1851
+    },
+    {
+      "epoch": 0.30424459361746864,
+      "grad_norm": 0.6558852791786194,
+      "learning_rate": 0.00012167918985068255,
+      "loss": 5.3692,
+      "num_input_tokens_seen": 243007488,
+      "step": 1854
+    },
+    {
+      "epoch": 0.30473689878513444,
+      "grad_norm": 0.647803008556366,
+      "learning_rate": 0.00012158086321656318,
+      "loss": 5.3617,
+      "num_input_tokens_seen": 243400704,
+      "step": 1857
+    },
+    {
+      "epoch": 0.30522920395280023,
+      "grad_norm": 0.5871676206588745,
+      "learning_rate": 0.00012148277456548738,
+      "loss": 5.3675,
+      "num_input_tokens_seen": 243793920,
+      "step": 1860
+    },
+    {
+      "epoch": 0.30572150912046603,
+      "grad_norm": 0.690800130367279,
+      "learning_rate": 0.00012138492293900513,
+      "loss": 5.3202,
+      "num_input_tokens_seen": 244187136,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3062138142881319,
+      "grad_norm": 0.6760239005088806,
+      "learning_rate": 0.00012128730738406176,
+      "loss": 5.3634,
+      "num_input_tokens_seen": 244580352,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3067061194557977,
+      "grad_norm": 0.8584226369857788,
+      "learning_rate": 0.00012118992695295909,
+      "loss": 5.3221,
+      "num_input_tokens_seen": 244973568,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3071984246234635,
+      "grad_norm": 0.7716052532196045,
+      "learning_rate": 0.00012109278070331664,
+      "loss": 5.378,
+      "num_input_tokens_seen": 245366784,
+      "step": 1872
+    },
+    {
+      "epoch": 0.30769072979112927,
+      "grad_norm": 0.5956524610519409,
+      "learning_rate": 0.0001209958676980334,
+      "loss": 5.2991,
+      "num_input_tokens_seen": 245760000,
+      "step": 1875
+    },
+    {
+      "epoch": 0.30818303495879507,
+      "grad_norm": 0.6713026762008667,
+      "learning_rate": 0.00012089918700524974,
+      "loss": 5.3229,
+      "num_input_tokens_seen": 246153216,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3086753401264609,
+      "grad_norm": 0.8650434017181396,
+      "learning_rate": 0.00012080273769831004,
+      "loss": 5.3466,
+      "num_input_tokens_seen": 246546432,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3091676452941267,
+      "grad_norm": 0.7258638143539429,
+      "learning_rate": 0.00012070651885572507,
+      "loss": 5.3342,
+      "num_input_tokens_seen": 246939648,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3096599504617925,
+      "grad_norm": 0.6096934676170349,
+      "learning_rate": 0.00012061052956113527,
+      "loss": 5.327,
+      "num_input_tokens_seen": 247332864,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3101522556294583,
+      "grad_norm": 0.7494550943374634,
+      "learning_rate": 0.00012051476890327393,
+      "loss": 5.3402,
+      "num_input_tokens_seen": 247726080,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3106445607971241,
+      "grad_norm": 0.586797833442688,
+      "learning_rate": 0.00012041923597593093,
+      "loss": 5.3722,
+      "num_input_tokens_seen": 248119296,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3111368659647899,
+      "grad_norm": 0.6233952641487122,
+      "learning_rate": 0.00012032392987791671,
+      "loss": 5.3152,
+      "num_input_tokens_seen": 248512512,
+      "step": 1896
+    },
+    {
+      "epoch": 0.31162917113245575,
+      "grad_norm": 0.7030310034751892,
+      "learning_rate": 0.0001202288497130266,
+      "loss": 5.2975,
+      "num_input_tokens_seen": 248905728,
+      "step": 1899
+    },
+    {
+      "epoch": 0.31212147630012155,
+      "grad_norm": 0.603950560092926,
+      "learning_rate": 0.00012013399459000527,
+      "loss": 5.3444,
+      "num_input_tokens_seen": 249298944,
+      "step": 1902
+    },
+    {
+      "epoch": 0.31261378146778734,
+      "grad_norm": 0.5920315384864807,
+      "learning_rate": 0.00012003936362251192,
+      "loss": 5.3394,
+      "num_input_tokens_seen": 249692160,
+      "step": 1905
+    },
+    {
+      "epoch": 0.31310608663545314,
+      "grad_norm": 0.5622957944869995,
+      "learning_rate": 0.00011994495592908519,
+      "loss": 5.3122,
+      "num_input_tokens_seen": 250085376,
+      "step": 1908
+    },
+    {
+      "epoch": 0.31359839180311894,
+      "grad_norm": 0.5327597260475159,
+      "learning_rate": 0.0001198507706331089,
+      "loss": 5.3086,
+      "num_input_tokens_seen": 250478592,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3140906969707848,
+      "grad_norm": 0.7794554233551025,
+      "learning_rate": 0.00011975680686277773,
+      "loss": 5.3564,
+      "num_input_tokens_seen": 250871808,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3145830021384506,
+      "grad_norm": 0.7415952682495117,
+      "learning_rate": 0.00011966306375106347,
+      "loss": 5.3224,
+      "num_input_tokens_seen": 251265024,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3150753073061164,
+      "grad_norm": 0.7701389789581299,
+      "learning_rate": 0.0001195695404356812,
+      "loss": 5.3842,
+      "num_input_tokens_seen": 251658240,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3155676124737822,
+      "grad_norm": 0.6872450113296509,
+      "learning_rate": 0.00011947623605905617,
+      "loss": 5.3254,
+      "num_input_tokens_seen": 252051456,
+      "step": 1923
+    },
+    {
+      "epoch": 0.316059917641448,
+      "grad_norm": 0.7382259964942932,
+      "learning_rate": 0.0001193831497682907,
+      "loss": 5.3764,
+      "num_input_tokens_seen": 252444672,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3165522228091138,
+      "grad_norm": 0.7021927237510681,
+      "learning_rate": 0.00011929028071513144,
+      "loss": 5.3697,
+      "num_input_tokens_seen": 252837888,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3170445279767796,
+      "grad_norm": 0.5273075699806213,
+      "learning_rate": 0.00011919762805593696,
+      "loss": 5.3248,
+      "num_input_tokens_seen": 253231104,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3175368331444454,
+      "grad_norm": 0.6100038290023804,
+      "learning_rate": 0.00011910519095164537,
+      "loss": 5.3604,
+      "num_input_tokens_seen": 253624320,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3180291383121112,
+      "grad_norm": 0.608025848865509,
+      "learning_rate": 0.00011901296856774264,
+      "loss": 5.3192,
+      "num_input_tokens_seen": 254017536,
+      "step": 1938
+    },
+    {
+      "epoch": 0.318521443479777,
+      "grad_norm": 0.8330180048942566,
+      "learning_rate": 0.00011892096007423088,
+      "loss": 5.2825,
+      "num_input_tokens_seen": 254410752,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3190137486474428,
+      "grad_norm": 0.7766706347465515,
+      "learning_rate": 0.00011882916464559686,
+      "loss": 5.329,
+      "num_input_tokens_seen": 254803968,
+      "step": 1944
+    },
+    {
+      "epoch": 0.31950605381510866,
+      "grad_norm": 0.7442476749420166,
+      "learning_rate": 0.00011873758146078108,
+      "loss": 5.3182,
+      "num_input_tokens_seen": 255197184,
+      "step": 1947
+    },
+    {
+      "epoch": 0.31999835898277446,
+      "grad_norm": 0.7740505337715149,
+      "learning_rate": 0.00011864620970314674,
+      "loss": 5.3424,
+      "num_input_tokens_seen": 255590400,
+      "step": 1950
+    },
+    {
+      "epoch": 0.32049066415044025,
+      "grad_norm": 0.7652367949485779,
+      "learning_rate": 0.00011855504856044936,
+      "loss": 5.3002,
+      "num_input_tokens_seen": 255983616,
+      "step": 1953
+    },
+    {
+      "epoch": 0.32098296931810605,
+      "grad_norm": 0.683326780796051,
+      "learning_rate": 0.00011846409722480622,
+      "loss": 5.359,
+      "num_input_tokens_seen": 256376832,
+      "step": 1956
+    },
+    {
+      "epoch": 0.32147527448577184,
+      "grad_norm": 0.6694085597991943,
+      "learning_rate": 0.0001183733548926665,
+      "loss": 5.2737,
+      "num_input_tokens_seen": 256770048,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3219675796534377,
+      "grad_norm": 0.653356671333313,
+      "learning_rate": 0.00011828282076478137,
+      "loss": 5.3363,
+      "num_input_tokens_seen": 257163264,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3224598848211035,
+      "grad_norm": 0.6444686055183411,
+      "learning_rate": 0.00011819249404617434,
+      "loss": 5.3049,
+      "num_input_tokens_seen": 257556480,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3229521899887693,
+      "grad_norm": 0.6730981469154358,
+      "learning_rate": 0.00011810237394611218,
+      "loss": 5.3095,
+      "num_input_tokens_seen": 257949696,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3234444951564351,
+      "grad_norm": 0.5876457691192627,
+      "learning_rate": 0.00011801245967807553,
+      "loss": 5.301,
+      "num_input_tokens_seen": 258342912,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3239368003241009,
+      "grad_norm": 0.7190306782722473,
+      "learning_rate": 0.00011792275045973037,
+      "loss": 5.3266,
+      "num_input_tokens_seen": 258736128,
+      "step": 1974
+    },
+    {
+      "epoch": 0.32442910549176673,
+      "grad_norm": 0.9150959253311157,
+      "learning_rate": 0.00011783324551289922,
+      "loss": 5.3359,
+      "num_input_tokens_seen": 259129344,
+      "step": 1977
+    },
+    {
+      "epoch": 0.32492141065943253,
+      "grad_norm": 0.6939476728439331,
+      "learning_rate": 0.00011774394406353287,
+      "loss": 5.3251,
+      "num_input_tokens_seen": 259522560,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3254137158270983,
+      "grad_norm": 0.6032900810241699,
+      "learning_rate": 0.00011765484534168232,
+      "loss": 5.3072,
+      "num_input_tokens_seen": 259915776,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3259060209947641,
+      "grad_norm": 0.674830436706543,
+      "learning_rate": 0.00011756594858147075,
+      "loss": 5.3009,
+      "num_input_tokens_seen": 260308992,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3263983261624299,
+      "grad_norm": 0.5875729918479919,
+      "learning_rate": 0.000117477253021066,
+      "loss": 5.35,
+      "num_input_tokens_seen": 260702208,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3268906313300957,
+      "grad_norm": 0.7125598788261414,
+      "learning_rate": 0.00011738875790265313,
+      "loss": 5.3612,
+      "num_input_tokens_seen": 261095424,
+      "step": 1992
+    },
+    {
+      "epoch": 0.32738293649776157,
+      "grad_norm": 0.6010739207267761,
+      "learning_rate": 0.00011730046247240715,
+      "loss": 5.3347,
+      "num_input_tokens_seen": 261488640,
+      "step": 1995
+    },
+    {
+      "epoch": 0.32787524166542736,
+      "grad_norm": 0.6644861698150635,
+      "learning_rate": 0.00011721236598046614,
+      "loss": 5.3236,
+      "num_input_tokens_seen": 261881856,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3282034451105379,
+      "eval_accuracy": 0.2103419638495359,
+      "eval_loss": 5.556689262390137,
+      "eval_runtime": 109.9195,
+      "eval_samples_per_second": 2.729,
+      "eval_steps_per_second": 1.365,
+      "num_input_tokens_seen": 262144000,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32836754683309316,
+      "grad_norm": 0.6464180946350098,
+      "learning_rate": 0.00011712446768090445,
+      "loss": 5.3375,
+      "num_input_tokens_seen": 262275072,
+      "step": 2001
+    },
+    {
+      "epoch": 0.32885985200075896,
+      "grad_norm": 0.6498216390609741,
+      "learning_rate": 0.00011703676683170618,
+      "loss": 5.3931,
+      "num_input_tokens_seen": 262668288,
+      "step": 2004
+    },
+    {
+      "epoch": 0.32935215716842475,
+      "grad_norm": 0.6177734732627869,
+      "learning_rate": 0.00011694926269473891,
+      "loss": 5.3224,
+      "num_input_tokens_seen": 263061504,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3298444623360906,
+      "grad_norm": 0.6429308652877808,
+      "learning_rate": 0.00011686195453572751,
+      "loss": 5.3532,
+      "num_input_tokens_seen": 263454720,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3303367675037564,
+      "grad_norm": 0.5821199417114258,
+      "learning_rate": 0.00011677484162422844,
+      "loss": 5.357,
+      "num_input_tokens_seen": 263847936,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3308290726714222,
+      "grad_norm": 0.6923580169677734,
+      "learning_rate": 0.00011668792323360385,
+      "loss": 5.3397,
+      "num_input_tokens_seen": 264241152,
+      "step": 2016
+    },
+    {
+      "epoch": 0.331321377839088,
+      "grad_norm": 0.6655700206756592,
+      "learning_rate": 0.00011660119864099633,
+      "loss": 5.339,
+      "num_input_tokens_seen": 264634368,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3318136830067538,
+      "grad_norm": 0.5696009397506714,
+      "learning_rate": 0.00011651466712730354,
+      "loss": 5.332,
+      "num_input_tokens_seen": 265027584,
+      "step": 2022
+    },
+    {
+      "epoch": 0.33230598817441964,
+      "grad_norm": 0.6818378567695618,
+      "learning_rate": 0.0001164283279771532,
+      "loss": 5.2651,
+      "num_input_tokens_seen": 265420800,
+      "step": 2025
+    },
+    {
+      "epoch": 0.33279829334208544,
+      "grad_norm": 0.6137866377830505,
+      "learning_rate": 0.00011634218047887825,
+      "loss": 5.2808,
+      "num_input_tokens_seen": 265814016,
+      "step": 2028
+    },
+    {
+      "epoch": 0.33329059850975123,
+      "grad_norm": 0.9078963994979858,
+      "learning_rate": 0.00011625622392449224,
+      "loss": 5.3266,
+      "num_input_tokens_seen": 266207232,
+      "step": 2031
+    },
+    {
+      "epoch": 0.33378290367741703,
+      "grad_norm": 0.7636304497718811,
+      "learning_rate": 0.00011617045760966484,
+      "loss": 5.3276,
+      "num_input_tokens_seen": 266600448,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3342752088450828,
+      "grad_norm": 0.590513288974762,
+      "learning_rate": 0.00011608488083369763,
+      "loss": 5.3353,
+      "num_input_tokens_seen": 266993664,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3347675140127487,
+      "grad_norm": 0.6252188682556152,
+      "learning_rate": 0.00011599949289950009,
+      "loss": 5.329,
+      "num_input_tokens_seen": 267386880,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3352598191804145,
+      "grad_norm": 0.5734694600105286,
+      "learning_rate": 0.00011591429311356567,
+      "loss": 5.3274,
+      "num_input_tokens_seen": 267780096,
+      "step": 2043
+    },
+    {
+      "epoch": 0.33575212434808027,
+      "grad_norm": 0.5329940319061279,
+      "learning_rate": 0.00011582928078594821,
+      "loss": 5.3475,
+      "num_input_tokens_seen": 268173312,
+      "step": 2046
+    },
+    {
+      "epoch": 0.33624442951574607,
+      "grad_norm": 0.5748163461685181,
+      "learning_rate": 0.00011574445523023836,
+      "loss": 5.2668,
+      "num_input_tokens_seen": 268566528,
+      "step": 2049
+    },
+    {
+      "epoch": 0.33673673468341186,
+      "grad_norm": 0.5230370759963989,
+      "learning_rate": 0.00011565981576354052,
+      "loss": 5.3068,
+      "num_input_tokens_seen": 268959744,
+      "step": 2052
+    },
+    {
+      "epoch": 0.33722903985107766,
+      "grad_norm": 0.6396874189376831,
+      "learning_rate": 0.00011557536170644955,
+      "loss": 5.294,
+      "num_input_tokens_seen": 269352960,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3377213450187435,
+      "grad_norm": 0.5591505169868469,
+      "learning_rate": 0.00011549109238302785,
+      "loss": 5.3216,
+      "num_input_tokens_seen": 269746176,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3382136501864093,
+      "grad_norm": 0.6073426604270935,
+      "learning_rate": 0.00011540700712078282,
+      "loss": 5.3153,
+      "num_input_tokens_seen": 270139392,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3387059553540751,
+      "grad_norm": 0.5904248356819153,
+      "learning_rate": 0.0001153231052506442,
+      "loss": 5.2798,
+      "num_input_tokens_seen": 270532608,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3391982605217409,
+      "grad_norm": 0.570007860660553,
+      "learning_rate": 0.0001152393861069417,
+      "loss": 5.2725,
+      "num_input_tokens_seen": 270925824,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3396905656894067,
+      "grad_norm": 0.686757504940033,
+      "learning_rate": 0.00011515584902738283,
+      "loss": 5.3031,
+      "num_input_tokens_seen": 271319040,
+      "step": 2070
+    },
+    {
+      "epoch": 0.34018287085707255,
+      "grad_norm": 0.7036386132240295,
+      "learning_rate": 0.00011507249335303097,
+      "loss": 5.2997,
+      "num_input_tokens_seen": 271712256,
+      "step": 2073
+    },
+    {
+      "epoch": 0.34067517602473835,
+      "grad_norm": 0.5441813468933105,
+      "learning_rate": 0.00011498931842828338,
+      "loss": 5.3026,
+      "num_input_tokens_seen": 272105472,
+      "step": 2076
+    },
+    {
+      "epoch": 0.34116748119240414,
+      "grad_norm": 0.5833845138549805,
+      "learning_rate": 0.00011490632360084974,
+      "loss": 5.2784,
+      "num_input_tokens_seen": 272498688,
+      "step": 2079
+    },
+    {
+      "epoch": 0.34165978636006994,
+      "grad_norm": 0.6387719511985779,
+      "learning_rate": 0.00011482350822173052,
+      "loss": 5.32,
+      "num_input_tokens_seen": 272891904,
+      "step": 2082
+    },
+    {
+      "epoch": 0.34215209152773574,
+      "grad_norm": 0.530159056186676,
+      "learning_rate": 0.00011474087164519571,
+      "loss": 5.2776,
+      "num_input_tokens_seen": 273285120,
+      "step": 2085
+    },
+    {
+      "epoch": 0.3426443966954016,
+      "grad_norm": 0.5628234148025513,
+      "learning_rate": 0.00011465841322876368,
+      "loss": 5.3392,
+      "num_input_tokens_seen": 273678336,
+      "step": 2088
+    },
+    {
+      "epoch": 0.3431367018630674,
+      "grad_norm": 0.6385974884033203,
+      "learning_rate": 0.00011457613233318018,
+      "loss": 5.3113,
+      "num_input_tokens_seen": 274071552,
+      "step": 2091
+    },
+    {
+      "epoch": 0.3436290070307332,
+      "grad_norm": 0.5761107802391052,
+      "learning_rate": 0.00011449402832239752,
+      "loss": 5.3004,
+      "num_input_tokens_seen": 274464768,
+      "step": 2094
+    },
+    {
+      "epoch": 0.344121312198399,
+      "grad_norm": 0.6140743494033813,
+      "learning_rate": 0.00011441210056355405,
+      "loss": 5.3376,
+      "num_input_tokens_seen": 274857984,
+      "step": 2097
+    },
+    {
+      "epoch": 0.3446136173660648,
+      "grad_norm": 0.7067747116088867,
+      "learning_rate": 0.00011433034842695343,
+      "loss": 5.2842,
+      "num_input_tokens_seen": 275251200,
+      "step": 2100
+    },
+    {
+      "epoch": 0.34510592253373057,
+      "grad_norm": 0.5610840916633606,
+      "learning_rate": 0.00011424877128604445,
+      "loss": 5.2966,
+      "num_input_tokens_seen": 275644416,
+      "step": 2103
+    },
+    {
+      "epoch": 0.3455982277013964,
+      "grad_norm": 0.6917890310287476,
+      "learning_rate": 0.00011416736851740093,
+      "loss": 5.3038,
+      "num_input_tokens_seen": 276037632,
+      "step": 2106
+    },
+    {
+      "epoch": 0.3460905328690622,
+      "grad_norm": 0.7923979759216309,
+      "learning_rate": 0.00011408613950070158,
+      "loss": 5.2936,
+      "num_input_tokens_seen": 276430848,
+      "step": 2109
+    },
+    {
+      "epoch": 0.346582838036728,
+      "grad_norm": 0.6658058762550354,
+      "learning_rate": 0.00011400508361871013,
+      "loss": 5.284,
+      "num_input_tokens_seen": 276824064,
+      "step": 2112
+    },
+    {
+      "epoch": 0.3470751432043938,
+      "grad_norm": 0.5722377896308899,
+      "learning_rate": 0.00011392420025725577,
+      "loss": 5.3168,
+      "num_input_tokens_seen": 277217280,
+      "step": 2115
+    },
+    {
+      "epoch": 0.3475674483720596,
+      "grad_norm": 0.6443458199501038,
+      "learning_rate": 0.00011384348880521352,
+      "loss": 5.3339,
+      "num_input_tokens_seen": 277610496,
+      "step": 2118
+    },
+    {
+      "epoch": 0.34805975353972546,
+      "grad_norm": 0.6641397476196289,
+      "learning_rate": 0.00011376294865448479,
+      "loss": 5.3043,
+      "num_input_tokens_seen": 278003712,
+      "step": 2121
+    },
+    {
+      "epoch": 0.34855205870739125,
+      "grad_norm": 0.7289448976516724,
+      "learning_rate": 0.00011368257919997822,
+      "loss": 5.2867,
+      "num_input_tokens_seen": 278396928,
+      "step": 2124
+    },
+    {
+      "epoch": 0.34904436387505705,
+      "grad_norm": 0.7701984643936157,
+      "learning_rate": 0.00011360237983959058,
+      "loss": 5.2322,
+      "num_input_tokens_seen": 278790144,
+      "step": 2127
+    },
+    {
+      "epoch": 0.34953666904272285,
+      "grad_norm": 0.7202409505844116,
+      "learning_rate": 0.00011352234997418777,
+      "loss": 5.2508,
+      "num_input_tokens_seen": 279183360,
+      "step": 2130
+    },
+    {
+      "epoch": 0.35002897421038864,
+      "grad_norm": 0.6271010637283325,
+      "learning_rate": 0.00011344248900758605,
+      "loss": 5.2954,
+      "num_input_tokens_seen": 279576576,
+      "step": 2133
+    },
+    {
+      "epoch": 0.3505212793780545,
+      "grad_norm": 0.6194177865982056,
+      "learning_rate": 0.00011336279634653344,
+      "loss": 5.2592,
+      "num_input_tokens_seen": 279969792,
+      "step": 2136
+    },
+    {
+      "epoch": 0.3510135845457203,
+      "grad_norm": 0.6322209239006042,
+      "learning_rate": 0.00011328327140069108,
+      "loss": 5.2867,
+      "num_input_tokens_seen": 280363008,
+      "step": 2139
+    },
+    {
+      "epoch": 0.3515058897133861,
+      "grad_norm": 0.6329113245010376,
+      "learning_rate": 0.000113203913582615,
+      "loss": 5.3342,
+      "num_input_tokens_seen": 280756224,
+      "step": 2142
+    },
+    {
+      "epoch": 0.3519981948810519,
+      "grad_norm": 0.6119714379310608,
+      "learning_rate": 0.00011312472230773781,
+      "loss": 5.2948,
+      "num_input_tokens_seen": 281149440,
+      "step": 2145
+    },
+    {
+      "epoch": 0.3524905000487177,
+      "grad_norm": 0.6921229362487793,
+      "learning_rate": 0.00011304569699435066,
+      "loss": 5.2664,
+      "num_input_tokens_seen": 281542656,
+      "step": 2148
+    },
+    {
+      "epoch": 0.35298280521638353,
+      "grad_norm": 0.6127797961235046,
+      "learning_rate": 0.00011296683706358528,
+      "loss": 5.3308,
+      "num_input_tokens_seen": 281935872,
+      "step": 2151
+    },
+    {
+      "epoch": 0.35347511038404933,
+      "grad_norm": 0.6175393462181091,
+      "learning_rate": 0.00011288814193939612,
+      "loss": 5.3327,
+      "num_input_tokens_seen": 282329088,
+      "step": 2154
+    },
+    {
+      "epoch": 0.3539674155517151,
+      "grad_norm": 0.7570388317108154,
+      "learning_rate": 0.00011280961104854276,
+      "loss": 5.2763,
+      "num_input_tokens_seen": 282722304,
+      "step": 2157
+    },
+    {
+      "epoch": 0.3544597207193809,
+      "grad_norm": 0.6202073693275452,
+      "learning_rate": 0.00011273124382057234,
+      "loss": 5.2544,
+      "num_input_tokens_seen": 283115520,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3549520258870467,
+      "grad_norm": 0.6008228063583374,
+      "learning_rate": 0.00011265303968780214,
+      "loss": 5.2426,
+      "num_input_tokens_seen": 283508736,
+      "step": 2163
+    },
+    {
+      "epoch": 0.3554443310547125,
+      "grad_norm": 0.6557415127754211,
+      "learning_rate": 0.00011257499808530228,
+      "loss": 5.2976,
+      "num_input_tokens_seen": 283901952,
+      "step": 2166
+    },
+    {
+      "epoch": 0.35593663622237837,
+      "grad_norm": 0.7166002988815308,
+      "learning_rate": 0.00011249711845087871,
+      "loss": 5.3111,
+      "num_input_tokens_seen": 284295168,
+      "step": 2169
+    },
+    {
+      "epoch": 0.35642894139004416,
+      "grad_norm": 0.5834003686904907,
+      "learning_rate": 0.000112419400225056,
+      "loss": 5.2879,
+      "num_input_tokens_seen": 284688384,
+      "step": 2172
+    },
+    {
+      "epoch": 0.35692124655770996,
+      "grad_norm": 0.6676968932151794,
+      "learning_rate": 0.00011234184285106067,
+      "loss": 5.3055,
+      "num_input_tokens_seen": 285081600,
+      "step": 2175
+    },
+    {
+      "epoch": 0.35741355172537576,
+      "grad_norm": 0.6959543824195862,
+      "learning_rate": 0.00011226444577480424,
+      "loss": 5.2653,
+      "num_input_tokens_seen": 285474816,
+      "step": 2178
+    },
+    {
+      "epoch": 0.35790585689304155,
+      "grad_norm": 0.7726870775222778,
+      "learning_rate": 0.00011218720844486681,
+      "loss": 5.3342,
+      "num_input_tokens_seen": 285868032,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3583981620607074,
+      "grad_norm": 0.7076993584632874,
+      "learning_rate": 0.00011211013031248031,
+      "loss": 5.2969,
+      "num_input_tokens_seen": 286261248,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3588904672283732,
+      "grad_norm": 0.6197788715362549,
+      "learning_rate": 0.0001120332108315124,
+      "loss": 5.2731,
+      "num_input_tokens_seen": 286654464,
+      "step": 2187
+    },
+    {
+      "epoch": 0.359382772396039,
+      "grad_norm": 0.6030928492546082,
+      "learning_rate": 0.00011195644945844996,
+      "loss": 5.303,
+      "num_input_tokens_seen": 287047680,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3598750775637048,
+      "grad_norm": 0.6310069561004639,
+      "learning_rate": 0.00011187984565238322,
+      "loss": 5.2488,
+      "num_input_tokens_seen": 287440896,
+      "step": 2193
+    },
+    {
+      "epoch": 0.3603673827313706,
+      "grad_norm": 0.6047622561454773,
+      "learning_rate": 0.00011180339887498948,
+      "loss": 5.2663,
+      "num_input_tokens_seen": 287834112,
+      "step": 2196
+    },
+    {
+      "epoch": 0.36085968789903644,
+      "grad_norm": 0.551209568977356,
+      "learning_rate": 0.0001117271085905174,
+      "loss": 5.2805,
+      "num_input_tokens_seen": 288227328,
+      "step": 2199
+    },
+    {
+      "epoch": 0.36135199306670224,
+      "grad_norm": 0.6387552618980408,
+      "learning_rate": 0.00011165097426577122,
+      "loss": 5.3208,
+      "num_input_tokens_seen": 288620544,
+      "step": 2202
+    },
+    {
+      "epoch": 0.36184429823436803,
+      "grad_norm": 0.5512715578079224,
+      "learning_rate": 0.00011157499537009505,
+      "loss": 5.29,
+      "num_input_tokens_seen": 289013760,
+      "step": 2205
+    },
+    {
+      "epoch": 0.36233660340203383,
+      "grad_norm": 0.6393409967422485,
+      "learning_rate": 0.00011149917137535733,
+      "loss": 5.2846,
+      "num_input_tokens_seen": 289406976,
+      "step": 2208
+    },
+    {
+      "epoch": 0.3628289085696996,
+      "grad_norm": 0.6270577907562256,
+      "learning_rate": 0.00011142350175593546,
+      "loss": 5.2922,
+      "num_input_tokens_seen": 289800192,
+      "step": 2211
+    },
+    {
+      "epoch": 0.3633212137373654,
+      "grad_norm": 0.6736873388290405,
+      "learning_rate": 0.00011134798598870045,
+      "loss": 5.2724,
+      "num_input_tokens_seen": 290193408,
+      "step": 2214
+    },
+    {
+      "epoch": 0.3638135189050313,
+      "grad_norm": 0.5813255906105042,
+      "learning_rate": 0.00011127262355300185,
+      "loss": 5.2424,
+      "num_input_tokens_seen": 290586624,
+      "step": 2217
+    },
+    {
+      "epoch": 0.36430582407269707,
+      "grad_norm": 0.6181595325469971,
+      "learning_rate": 0.00011119741393065246,
+      "loss": 5.3019,
+      "num_input_tokens_seen": 290979840,
+      "step": 2220
+    },
+    {
+      "epoch": 0.36479812924036287,
+      "grad_norm": 0.6010147929191589,
+      "learning_rate": 0.00011112235660591355,
+      "loss": 5.2262,
+      "num_input_tokens_seen": 291373056,
+      "step": 2223
+    },
+    {
+      "epoch": 0.36529043440802866,
+      "grad_norm": 0.5609191656112671,
+      "learning_rate": 0.00011104745106547993,
+      "loss": 5.3074,
+      "num_input_tokens_seen": 291766272,
+      "step": 2226
+    },
+    {
+      "epoch": 0.36578273957569446,
+      "grad_norm": 0.5455970168113708,
+      "learning_rate": 0.0001109726967984652,
+      "loss": 5.2696,
+      "num_input_tokens_seen": 292159488,
+      "step": 2229
+    },
+    {
+      "epoch": 0.3662750447433603,
+      "grad_norm": 0.6548165082931519,
+      "learning_rate": 0.00011089809329638716,
+      "loss": 5.2895,
+      "num_input_tokens_seen": 292552704,
+      "step": 2232
+    },
+    {
+      "epoch": 0.3667673499110261,
+      "grad_norm": 0.6180770397186279,
+      "learning_rate": 0.00011082364005315322,
+      "loss": 5.2736,
+      "num_input_tokens_seen": 292945920,
+      "step": 2235
+    },
+    {
+      "epoch": 0.3672596550786919,
+      "grad_norm": 0.6706951856613159,
+      "learning_rate": 0.00011074933656504608,
+      "loss": 5.3579,
+      "num_input_tokens_seen": 293339136,
+      "step": 2238
+    },
+    {
+      "epoch": 0.3677519602463577,
+      "grad_norm": 0.6596742868423462,
+      "learning_rate": 0.00011067518233070927,
+      "loss": 5.2774,
+      "num_input_tokens_seen": 293732352,
+      "step": 2241
+    },
+    {
+      "epoch": 0.3682442654140235,
+      "grad_norm": 0.6398391723632812,
+      "learning_rate": 0.00011060117685113308,
+      "loss": 5.2773,
+      "num_input_tokens_seen": 294125568,
+      "step": 2244
+    },
+    {
+      "epoch": 0.36873657058168935,
+      "grad_norm": 0.6828597784042358,
+      "learning_rate": 0.00011052731962964036,
+      "loss": 5.2476,
+      "num_input_tokens_seen": 294518784,
+      "step": 2247
+    },
+    {
+      "epoch": 0.36922887574935515,
+      "grad_norm": 0.6363155245780945,
+      "learning_rate": 0.0001104536101718726,
+      "loss": 5.3068,
+      "num_input_tokens_seen": 294912000,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36972118091702094,
+      "grad_norm": 0.6510285139083862,
+      "learning_rate": 0.00011038004798577598,
+      "loss": 5.3018,
+      "num_input_tokens_seen": 295305216,
+      "step": 2253
+    },
+    {
+      "epoch": 0.37021348608468674,
+      "grad_norm": 0.6058257222175598,
+      "learning_rate": 0.00011030663258158759,
+      "loss": 5.2721,
+      "num_input_tokens_seen": 295698432,
+      "step": 2256
+    },
+    {
+      "epoch": 0.37070579125235253,
+      "grad_norm": 0.6524608731269836,
+      "learning_rate": 0.0001102333634718217,
+      "loss": 5.3041,
+      "num_input_tokens_seen": 296091648,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3711980964200184,
+      "grad_norm": 0.6795722842216492,
+      "learning_rate": 0.00011016024017125623,
+      "loss": 5.2912,
+      "num_input_tokens_seen": 296484864,
+      "step": 2262
+    },
+    {
+      "epoch": 0.3716904015876842,
+      "grad_norm": 0.5707486867904663,
+      "learning_rate": 0.00011008726219691922,
+      "loss": 5.3142,
+      "num_input_tokens_seen": 296878080,
+      "step": 2265
+    },
+    {
+      "epoch": 0.37218270675535,
+      "grad_norm": 0.5843520760536194,
+      "learning_rate": 0.00011001442906807543,
+      "loss": 5.2925,
+      "num_input_tokens_seen": 297271296,
+      "step": 2268
+    },
+    {
+      "epoch": 0.3726750119230158,
+      "grad_norm": 0.5435792803764343,
+      "learning_rate": 0.00010994174030621302,
+      "loss": 5.254,
+      "num_input_tokens_seen": 297664512,
+      "step": 2271
+    },
+    {
+      "epoch": 0.37316731709068157,
+      "grad_norm": 0.5673099160194397,
+      "learning_rate": 0.00010986919543503034,
+      "loss": 5.2462,
+      "num_input_tokens_seen": 298057728,
+      "step": 2274
+    },
+    {
+      "epoch": 0.37365962225834737,
+      "grad_norm": 0.6960703134536743,
+      "learning_rate": 0.00010979679398042297,
+      "loss": 5.2829,
+      "num_input_tokens_seen": 298450944,
+      "step": 2277
+    },
+    {
+      "epoch": 0.3741519274260132,
+      "grad_norm": 0.6623157858848572,
+      "learning_rate": 0.00010972453547047044,
+      "loss": 5.259,
+      "num_input_tokens_seen": 298844160,
+      "step": 2280
+    },
+    {
+      "epoch": 0.374644232593679,
+      "grad_norm": 0.5777830481529236,
+      "learning_rate": 0.00010965241943542353,
+      "loss": 5.2606,
+      "num_input_tokens_seen": 299237376,
+      "step": 2283
+    },
+    {
+      "epoch": 0.3751365377613448,
+      "grad_norm": 0.6100009679794312,
+      "learning_rate": 0.00010958044540769138,
+      "loss": 5.2787,
+      "num_input_tokens_seen": 299630592,
+      "step": 2286
+    },
+    {
+      "epoch": 0.3756288429290106,
+      "grad_norm": 0.6693081259727478,
+      "learning_rate": 0.0001095086129218287,
+      "loss": 5.2721,
+      "num_input_tokens_seen": 300023808,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3761211480966764,
+      "grad_norm": 0.7031441330909729,
+      "learning_rate": 0.00010943692151452322,
+      "loss": 5.265,
+      "num_input_tokens_seen": 300417024,
+      "step": 2292
+    },
+    {
+      "epoch": 0.37661345326434226,
+      "grad_norm": 0.6610632538795471,
+      "learning_rate": 0.00010936537072458307,
+      "loss": 5.2384,
+      "num_input_tokens_seen": 300810240,
+      "step": 2295
+    },
+    {
+      "epoch": 0.37710575843200805,
+      "grad_norm": 0.5596596002578735,
+      "learning_rate": 0.00010929396009292432,
+      "loss": 5.2726,
+      "num_input_tokens_seen": 301203456,
+      "step": 2298
+    },
+    {
+      "epoch": 0.37759806359967385,
+      "grad_norm": 0.599071741104126,
+      "learning_rate": 0.00010922268916255866,
+      "loss": 5.256,
+      "num_input_tokens_seen": 301596672,
+      "step": 2301
+    },
+    {
+      "epoch": 0.37809036876733965,
+      "grad_norm": 0.5803152322769165,
+      "learning_rate": 0.00010915155747858112,
+      "loss": 5.2565,
+      "num_input_tokens_seen": 301989888,
+      "step": 2304
+    },
+    {
+      "epoch": 0.37858267393500544,
+      "grad_norm": 0.6008164882659912,
+      "learning_rate": 0.00010908056458815778,
+      "loss": 5.2591,
+      "num_input_tokens_seen": 302383104,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3790749791026713,
+      "grad_norm": 0.6585036516189575,
+      "learning_rate": 0.00010900971004051374,
+      "loss": 5.2829,
+      "num_input_tokens_seen": 302776320,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3795672842703371,
+      "grad_norm": 0.628734290599823,
+      "learning_rate": 0.00010893899338692111,
+      "loss": 5.2621,
+      "num_input_tokens_seen": 303169536,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3800595894380029,
+      "grad_norm": 0.6740299463272095,
+      "learning_rate": 0.00010886841418068702,
+      "loss": 5.2701,
+      "num_input_tokens_seen": 303562752,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3805518946056687,
+      "grad_norm": 0.7948876023292542,
+      "learning_rate": 0.00010879797197714182,
+      "loss": 5.2949,
+      "num_input_tokens_seen": 303955968,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3810441997733345,
+      "grad_norm": 0.8644827008247375,
+      "learning_rate": 0.00010872766633362728,
+      "loss": 5.2677,
+      "num_input_tokens_seen": 304349184,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3815365049410003,
+      "grad_norm": 0.8210058808326721,
+      "learning_rate": 0.00010865749680948491,
+      "loss": 5.2736,
+      "num_input_tokens_seen": 304742400,
+      "step": 2325
+    },
+    {
+      "epoch": 0.38202881010866613,
+      "grad_norm": 0.7495512366294861,
+      "learning_rate": 0.00010858746296604438,
+      "loss": 5.2665,
+      "num_input_tokens_seen": 305135616,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3825211152763319,
+      "grad_norm": 0.7410356402397156,
+      "learning_rate": 0.00010851756436661199,
+      "loss": 5.2467,
+      "num_input_tokens_seen": 305528832,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3830134204439977,
+      "grad_norm": 0.6325095891952515,
+      "learning_rate": 0.0001084478005764592,
+      "loss": 5.2277,
+      "num_input_tokens_seen": 305922048,
+      "step": 2334
+    },
+    {
+      "epoch": 0.3835057256116635,
+      "grad_norm": 0.6482600569725037,
+      "learning_rate": 0.00010837817116281135,
+      "loss": 5.2298,
+      "num_input_tokens_seen": 306315264,
+      "step": 2337
+    },
+    {
+      "epoch": 0.3839980307793293,
+      "grad_norm": 0.6560254693031311,
+      "learning_rate": 0.00010830867569483633,
+      "loss": 5.2803,
+      "num_input_tokens_seen": 306708480,
+      "step": 2340
+    },
+    {
+      "epoch": 0.38449033594699517,
+      "grad_norm": 0.5692320466041565,
+      "learning_rate": 0.00010823931374363337,
+      "loss": 5.2488,
+      "num_input_tokens_seen": 307101696,
+      "step": 2343
+    },
+    {
+      "epoch": 0.38498264111466096,
+      "grad_norm": 0.5976374745368958,
+      "learning_rate": 0.00010817008488222198,
+      "loss": 5.3247,
+      "num_input_tokens_seen": 307494912,
+      "step": 2346
+    },
+    {
+      "epoch": 0.38547494628232676,
+      "grad_norm": 0.663112998008728,
+      "learning_rate": 0.00010810098868553085,
+      "loss": 5.2795,
+      "num_input_tokens_seen": 307888128,
+      "step": 2349
+    },
+    {
+      "epoch": 0.38596725144999255,
+      "grad_norm": 0.6092638969421387,
+      "learning_rate": 0.00010803202473038695,
+      "loss": 5.279,
+      "num_input_tokens_seen": 308281344,
+      "step": 2352
+    },
+    {
+      "epoch": 0.38645955661765835,
+      "grad_norm": 0.5389038920402527,
+      "learning_rate": 0.00010796319259550458,
+      "loss": 5.2489,
+      "num_input_tokens_seen": 308674560,
+      "step": 2355
+    },
+    {
+      "epoch": 0.3869518617853242,
+      "grad_norm": 0.5788781046867371,
+      "learning_rate": 0.00010789449186147456,
+      "loss": 5.2873,
+      "num_input_tokens_seen": 309067776,
+      "step": 2358
+    },
+    {
+      "epoch": 0.38744416695299,
+      "grad_norm": 0.6551850438117981,
+      "learning_rate": 0.0001078259221107536,
+      "loss": 5.26,
+      "num_input_tokens_seen": 309460992,
+      "step": 2361
+    },
+    {
+      "epoch": 0.3879364721206558,
+      "grad_norm": 0.7511112093925476,
+      "learning_rate": 0.00010775748292765357,
+      "loss": 5.236,
+      "num_input_tokens_seen": 309854208,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3884287772883216,
+      "grad_norm": 0.5426856875419617,
+      "learning_rate": 0.00010768917389833085,
+      "loss": 5.2236,
+      "num_input_tokens_seen": 310247424,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3889210824559874,
+      "grad_norm": 0.6398828029632568,
+      "learning_rate": 0.00010762099461077592,
+      "loss": 5.2549,
+      "num_input_tokens_seen": 310640640,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3894133876236532,
+      "grad_norm": 0.5986840128898621,
+      "learning_rate": 0.00010755294465480287,
+      "loss": 5.2431,
+      "num_input_tokens_seen": 311033856,
+      "step": 2373
+    },
+    {
+      "epoch": 0.38990569279131904,
+      "grad_norm": 0.5876027941703796,
+      "learning_rate": 0.00010748502362203909,
+      "loss": 5.1995,
+      "num_input_tokens_seen": 311427072,
+      "step": 2376
+    },
+    {
+      "epoch": 0.39039799795898483,
+      "grad_norm": 0.5282983183860779,
+      "learning_rate": 0.00010741723110591491,
+      "loss": 5.2399,
+      "num_input_tokens_seen": 311820288,
+      "step": 2379
+    },
+    {
+      "epoch": 0.39089030312665063,
+      "grad_norm": 0.5736381411552429,
+      "learning_rate": 0.00010734956670165345,
+      "loss": 5.2859,
+      "num_input_tokens_seen": 312213504,
+      "step": 2382
+    },
+    {
+      "epoch": 0.3913826082943164,
+      "grad_norm": 0.5055151581764221,
+      "learning_rate": 0.00010728203000626037,
+      "loss": 5.269,
+      "num_input_tokens_seen": 312606720,
+      "step": 2385
+    },
+    {
+      "epoch": 0.3918749134619822,
+      "grad_norm": 0.5718653202056885,
+      "learning_rate": 0.00010721462061851386,
+      "loss": 5.2439,
+      "num_input_tokens_seen": 312999936,
+      "step": 2388
+    },
+    {
+      "epoch": 0.3923672186296481,
+      "grad_norm": 0.5438402891159058,
+      "learning_rate": 0.00010714733813895464,
+      "loss": 5.2659,
+      "num_input_tokens_seen": 313393152,
+      "step": 2391
+    },
+    {
+      "epoch": 0.39285952379731387,
+      "grad_norm": 0.550797700881958,
+      "learning_rate": 0.00010708018216987601,
+      "loss": 5.2214,
+      "num_input_tokens_seen": 313786368,
+      "step": 2394
+    },
+    {
+      "epoch": 0.39335182896497967,
+      "grad_norm": 0.6071121096611023,
+      "learning_rate": 0.00010701315231531391,
+      "loss": 5.2348,
+      "num_input_tokens_seen": 314179584,
+      "step": 2397
+    },
+    {
+      "epoch": 0.39384413413264546,
+      "grad_norm": 0.5938403606414795,
+      "learning_rate": 0.00010694624818103718,
+      "loss": 5.2764,
+      "num_input_tokens_seen": 314572800,
+      "step": 2400
+    },
+    {
+      "epoch": 0.39384413413264546,
+      "eval_accuracy": 0.21509688975736851,
+      "eval_loss": 5.491861343383789,
+      "eval_runtime": 110.2644,
+      "eval_samples_per_second": 2.721,
+      "eval_steps_per_second": 1.36,
+      "num_input_tokens_seen": 314572800,
+      "step": 2400
+    },
+    {
+      "epoch": 0.39433643930031126,
+      "grad_norm": 0.6181114912033081,
+      "learning_rate": 0.00010687946937453784,
+      "loss": 5.2461,
+      "num_input_tokens_seen": 314966016,
+      "step": 2403
+    },
+    {
+      "epoch": 0.3948287444679771,
+      "grad_norm": 0.8083803653717041,
+      "learning_rate": 0.00010681281550502132,
+      "loss": 5.2669,
+      "num_input_tokens_seen": 315359232,
+      "step": 2406
+    },
+    {
+      "epoch": 0.3953210496356429,
+      "grad_norm": 0.6575871109962463,
+      "learning_rate": 0.00010674628618339699,
+      "loss": 5.2582,
+      "num_input_tokens_seen": 315752448,
+      "step": 2409
+    },
+    {
+      "epoch": 0.3958133548033087,
+      "grad_norm": 0.6498743891716003,
+      "learning_rate": 0.00010667988102226855,
+      "loss": 5.2495,
+      "num_input_tokens_seen": 316145664,
+      "step": 2412
+    },
+    {
+      "epoch": 0.3963056599709745,
+      "grad_norm": 0.6563365459442139,
+      "learning_rate": 0.00010661359963592445,
+      "loss": 5.2312,
+      "num_input_tokens_seen": 316538880,
+      "step": 2415
+    },
+    {
+      "epoch": 0.3967979651386403,
+      "grad_norm": 0.6259288191795349,
+      "learning_rate": 0.00010654744164032871,
+      "loss": 5.2296,
+      "num_input_tokens_seen": 316932096,
+      "step": 2418
+    },
+    {
+      "epoch": 0.39729027030630615,
+      "grad_norm": 0.6058263778686523,
+      "learning_rate": 0.00010648140665311141,
+      "loss": 5.2278,
+      "num_input_tokens_seen": 317325312,
+      "step": 2421
+    },
+    {
+      "epoch": 0.39778257547397194,
+      "grad_norm": 0.6280511617660522,
+      "learning_rate": 0.00010641549429355939,
+      "loss": 5.2237,
+      "num_input_tokens_seen": 317718528,
+      "step": 2424
+    },
+    {
+      "epoch": 0.39827488064163774,
+      "grad_norm": 0.7042283415794373,
+      "learning_rate": 0.00010634970418260718,
+      "loss": 5.2665,
+      "num_input_tokens_seen": 318111744,
+      "step": 2427
+    },
+    {
+      "epoch": 0.39876718580930354,
+      "grad_norm": 0.673143208026886,
+      "learning_rate": 0.00010628403594282772,
+      "loss": 5.2682,
+      "num_input_tokens_seen": 318504960,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39925949097696933,
+      "grad_norm": 0.6177608370780945,
+      "learning_rate": 0.00010621848919842326,
+      "loss": 5.2381,
+      "num_input_tokens_seen": 318898176,
+      "step": 2433
+    },
+    {
+      "epoch": 0.39975179614463513,
+      "grad_norm": 0.5575879812240601,
+      "learning_rate": 0.00010615306357521642,
+      "loss": 5.2435,
+      "num_input_tokens_seen": 319291392,
+      "step": 2436
+    },
+    {
+      "epoch": 0.400244101312301,
+      "grad_norm": 0.6958820223808289,
+      "learning_rate": 0.00010608775870064112,
+      "loss": 5.2666,
+      "num_input_tokens_seen": 319684608,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4007364064799668,
+      "grad_norm": 0.7199931144714355,
+      "learning_rate": 0.00010602257420373379,
+      "loss": 5.2485,
+      "num_input_tokens_seen": 320077824,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4012287116476326,
+      "grad_norm": 0.6789767146110535,
+      "learning_rate": 0.00010595750971512437,
+      "loss": 5.2245,
+      "num_input_tokens_seen": 320471040,
+      "step": 2445
+    },
+    {
+      "epoch": 0.40172101681529837,
+      "grad_norm": 0.5763687491416931,
+      "learning_rate": 0.00010589256486702759,
+      "loss": 5.2227,
+      "num_input_tokens_seen": 320864256,
+      "step": 2448
+    },
+    {
+      "epoch": 0.40221332198296417,
+      "grad_norm": 0.5604720115661621,
+      "learning_rate": 0.0001058277392932343,
+      "loss": 5.2863,
+      "num_input_tokens_seen": 321257472,
+      "step": 2451
+    },
+    {
+      "epoch": 0.40270562715063,
+      "grad_norm": 0.5740640163421631,
+      "learning_rate": 0.00010576303262910272,
+      "loss": 5.2624,
+      "num_input_tokens_seen": 321650688,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4031979323182958,
+      "grad_norm": 0.5708188414573669,
+      "learning_rate": 0.00010569844451154979,
+      "loss": 5.2093,
+      "num_input_tokens_seen": 322043904,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4036902374859616,
+      "grad_norm": 0.5198714137077332,
+      "learning_rate": 0.00010563397457904276,
+      "loss": 5.2506,
+      "num_input_tokens_seen": 322437120,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4041825426536274,
+      "grad_norm": 0.7625793814659119,
+      "learning_rate": 0.00010556962247159053,
+      "loss": 5.2606,
+      "num_input_tokens_seen": 322830336,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4046748478212932,
+      "grad_norm": 0.6418973803520203,
+      "learning_rate": 0.00010550538783073529,
+      "loss": 5.2228,
+      "num_input_tokens_seen": 323223552,
+      "step": 2466
+    },
+    {
+      "epoch": 0.40516715298895906,
+      "grad_norm": 0.5481504797935486,
+      "learning_rate": 0.00010544127029954414,
+      "loss": 5.2283,
+      "num_input_tokens_seen": 323616768,
+      "step": 2469
+    },
+    {
+      "epoch": 0.40565945815662485,
+      "grad_norm": 0.5938975214958191,
+      "learning_rate": 0.00010537726952260071,
+      "loss": 5.2308,
+      "num_input_tokens_seen": 324009984,
+      "step": 2472
+    },
+    {
+      "epoch": 0.40615176332429065,
+      "grad_norm": 0.6005613803863525,
+      "learning_rate": 0.00010531338514599695,
+      "loss": 5.2342,
+      "num_input_tokens_seen": 324403200,
+      "step": 2475
+    },
+    {
+      "epoch": 0.40664406849195645,
+      "grad_norm": 0.6536208987236023,
+      "learning_rate": 0.00010524961681732482,
+      "loss": 5.2289,
+      "num_input_tokens_seen": 324796416,
+      "step": 2478
+    },
+    {
+      "epoch": 0.40713637365962224,
+      "grad_norm": 0.5923652052879333,
+      "learning_rate": 0.00010518596418566824,
+      "loss": 5.2962,
+      "num_input_tokens_seen": 325189632,
+      "step": 2481
+    },
+    {
+      "epoch": 0.40762867882728804,
+      "grad_norm": 0.6256945133209229,
+      "learning_rate": 0.00010512242690159487,
+      "loss": 5.2679,
+      "num_input_tokens_seen": 325582848,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4081209839949539,
+      "grad_norm": 0.6368097066879272,
+      "learning_rate": 0.00010505900461714815,
+      "loss": 5.1868,
+      "num_input_tokens_seen": 325976064,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4086132891626197,
+      "grad_norm": 0.7050658464431763,
+      "learning_rate": 0.00010499569698583921,
+      "loss": 5.2024,
+      "num_input_tokens_seen": 326369280,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4091055943302855,
+      "grad_norm": 0.6694772839546204,
+      "learning_rate": 0.000104932503662639,
+      "loss": 5.222,
+      "num_input_tokens_seen": 326762496,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4095978994979513,
+      "grad_norm": 0.5945850014686584,
+      "learning_rate": 0.00010486942430397028,
+      "loss": 5.2364,
+      "num_input_tokens_seen": 327155712,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4100902046656171,
+      "grad_norm": 0.5960330367088318,
+      "learning_rate": 0.00010480645856769992,
+      "loss": 5.2292,
+      "num_input_tokens_seen": 327548928,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4105825098332829,
+      "grad_norm": 0.550009548664093,
+      "learning_rate": 0.00010474360611313098,
+      "loss": 5.2305,
+      "num_input_tokens_seen": 327942144,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4110748150009487,
+      "grad_norm": 0.6489304900169373,
+      "learning_rate": 0.00010468086660099509,
+      "loss": 5.2478,
+      "num_input_tokens_seen": 328335360,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4115671201686145,
+      "grad_norm": 0.6759921312332153,
+      "learning_rate": 0.00010461823969344457,
+      "loss": 5.2794,
+      "num_input_tokens_seen": 328728576,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4120594253362803,
+      "grad_norm": 0.7441422939300537,
+      "learning_rate": 0.00010455572505404502,
+      "loss": 5.2267,
+      "num_input_tokens_seen": 329121792,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4125517305039461,
+      "grad_norm": 0.6741816997528076,
+      "learning_rate": 0.00010449332234776757,
+      "loss": 5.2342,
+      "num_input_tokens_seen": 329515008,
+      "step": 2514
+    },
+    {
+      "epoch": 0.41304403567161196,
+      "grad_norm": 0.6322781443595886,
+      "learning_rate": 0.00010443103124098138,
+      "loss": 5.2423,
+      "num_input_tokens_seen": 329908224,
+      "step": 2517
+    },
+    {
+      "epoch": 0.41353634083927776,
+      "grad_norm": 0.6095474362373352,
+      "learning_rate": 0.00010436885140144612,
+      "loss": 5.2405,
+      "num_input_tokens_seen": 330301440,
+      "step": 2520
+    },
+    {
+      "epoch": 0.41402864600694356,
+      "grad_norm": 0.6490298509597778,
+      "learning_rate": 0.00010430678249830464,
+      "loss": 5.2183,
+      "num_input_tokens_seen": 330694656,
+      "step": 2523
+    },
+    {
+      "epoch": 0.41452095117460935,
+      "grad_norm": 0.6782544255256653,
+      "learning_rate": 0.00010424482420207543,
+      "loss": 5.2436,
+      "num_input_tokens_seen": 331087872,
+      "step": 2526
+    },
+    {
+      "epoch": 0.41501325634227515,
+      "grad_norm": 0.6933419108390808,
+      "learning_rate": 0.00010418297618464539,
+      "loss": 5.2103,
+      "num_input_tokens_seen": 331481088,
+      "step": 2529
+    },
+    {
+      "epoch": 0.415505561509941,
+      "grad_norm": 0.7551579475402832,
+      "learning_rate": 0.00010412123811926243,
+      "loss": 5.2309,
+      "num_input_tokens_seen": 331874304,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4159978666776068,
+      "grad_norm": 0.5901357531547546,
+      "learning_rate": 0.00010405960968052833,
+      "loss": 5.2296,
+      "num_input_tokens_seen": 332267520,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4164901718452726,
+      "grad_norm": 0.6693659424781799,
+      "learning_rate": 0.0001039980905443914,
+      "loss": 5.2299,
+      "num_input_tokens_seen": 332660736,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4169824770129384,
+      "grad_norm": 0.6914536356925964,
+      "learning_rate": 0.00010393668038813947,
+      "loss": 5.2062,
+      "num_input_tokens_seen": 333053952,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4174747821806042,
+      "grad_norm": 0.8672217130661011,
+      "learning_rate": 0.0001038753788903927,
+      "loss": 5.2562,
+      "num_input_tokens_seen": 333447168,
+      "step": 2544
+    },
+    {
+      "epoch": 0.41796708734827,
+      "grad_norm": 0.7356572151184082,
+      "learning_rate": 0.0001038141857310965,
+      "loss": 5.2203,
+      "num_input_tokens_seen": 333840384,
+      "step": 2547
+    },
+    {
+      "epoch": 0.41845939251593584,
+      "grad_norm": 0.5476716756820679,
+      "learning_rate": 0.00010375310059151456,
+      "loss": 5.2411,
+      "num_input_tokens_seen": 334233600,
+      "step": 2550
+    },
+    {
+      "epoch": 0.41895169768360163,
+      "grad_norm": 0.8849159479141235,
+      "learning_rate": 0.00010369212315422186,
+      "loss": 5.2273,
+      "num_input_tokens_seen": 334626816,
+      "step": 2553
+    },
+    {
+      "epoch": 0.41944400285126743,
+      "grad_norm": 0.8213227987289429,
+      "learning_rate": 0.00010363125310309775,
+      "loss": 5.1921,
+      "num_input_tokens_seen": 335020032,
+      "step": 2556
+    },
+    {
+      "epoch": 0.4199363080189332,
+      "grad_norm": 0.806098461151123,
+      "learning_rate": 0.00010357049012331902,
+      "loss": 5.2061,
+      "num_input_tokens_seen": 335413248,
+      "step": 2559
+    },
+    {
+      "epoch": 0.420428613186599,
+      "grad_norm": 0.8755928874015808,
+      "learning_rate": 0.00010350983390135311,
+      "loss": 5.228,
+      "num_input_tokens_seen": 335806464,
+      "step": 2562
+    },
+    {
+      "epoch": 0.4209209183542649,
+      "grad_norm": 0.7494667172431946,
+      "learning_rate": 0.00010344928412495135,
+      "loss": 5.1867,
+      "num_input_tokens_seen": 336199680,
+      "step": 2565
+    },
+    {
+      "epoch": 0.42141322352193067,
+      "grad_norm": 0.7980634570121765,
+      "learning_rate": 0.00010338884048314206,
+      "loss": 5.2597,
+      "num_input_tokens_seen": 336592896,
+      "step": 2568
+    },
+    {
+      "epoch": 0.42190552868959647,
+      "grad_norm": 0.606177568435669,
+      "learning_rate": 0.00010332850266622407,
+      "loss": 5.2364,
+      "num_input_tokens_seen": 336986112,
+      "step": 2571
+    },
+    {
+      "epoch": 0.42239783385726226,
+      "grad_norm": 0.59128737449646,
+      "learning_rate": 0.0001032682703657598,
+      "loss": 5.2291,
+      "num_input_tokens_seen": 337379328,
+      "step": 2574
+    },
+    {
+      "epoch": 0.42289013902492806,
+      "grad_norm": 0.8499715328216553,
+      "learning_rate": 0.00010320814327456885,
+      "loss": 5.2151,
+      "num_input_tokens_seen": 337772544,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4233824441925939,
+      "grad_norm": 0.5828485488891602,
+      "learning_rate": 0.00010314812108672135,
+      "loss": 5.1643,
+      "num_input_tokens_seen": 338165760,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4238747493602597,
+      "grad_norm": 0.5701519250869751,
+      "learning_rate": 0.00010308820349753134,
+      "loss": 5.1919,
+      "num_input_tokens_seen": 338558976,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4243670545279255,
+      "grad_norm": 0.6905439496040344,
+      "learning_rate": 0.00010302839020355037,
+      "loss": 5.2067,
+      "num_input_tokens_seen": 338952192,
+      "step": 2586
+    },
+    {
+      "epoch": 0.4248593596955913,
+      "grad_norm": 0.6845247149467468,
+      "learning_rate": 0.00010296868090256107,
+      "loss": 5.1829,
+      "num_input_tokens_seen": 339345408,
+      "step": 2589
+    },
+    {
+      "epoch": 0.4253516648632571,
+      "grad_norm": 0.5913321375846863,
+      "learning_rate": 0.00010290907529357057,
+      "loss": 5.2025,
+      "num_input_tokens_seen": 339738624,
+      "step": 2592
+    },
+    {
+      "epoch": 0.4258439700309229,
+      "grad_norm": 0.627149760723114,
+      "learning_rate": 0.00010284957307680437,
+      "loss": 5.1632,
+      "num_input_tokens_seen": 340131840,
+      "step": 2595
+    },
+    {
+      "epoch": 0.42633627519858874,
+      "grad_norm": 0.574410617351532,
+      "learning_rate": 0.0001027901739536998,
+      "loss": 5.2493,
+      "num_input_tokens_seen": 340525056,
+      "step": 2598
+    },
+    {
+      "epoch": 0.42682858036625454,
+      "grad_norm": 0.5851385593414307,
+      "learning_rate": 0.00010273087762689989,
+      "loss": 5.2052,
+      "num_input_tokens_seen": 340918272,
+      "step": 2601
+    },
+    {
+      "epoch": 0.42732088553392034,
+      "grad_norm": 0.7069036960601807,
+      "learning_rate": 0.00010267168380024689,
+      "loss": 5.2107,
+      "num_input_tokens_seen": 341311488,
+      "step": 2604
+    },
+    {
+      "epoch": 0.42781319070158613,
+      "grad_norm": 0.5939748883247375,
+      "learning_rate": 0.00010261259217877632,
+      "loss": 5.2134,
+      "num_input_tokens_seen": 341704704,
+      "step": 2607
+    },
+    {
+      "epoch": 0.42830549586925193,
+      "grad_norm": 0.7017332315444946,
+      "learning_rate": 0.0001025536024687107,
+      "loss": 5.1983,
+      "num_input_tokens_seen": 342097920,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4287978010369178,
+      "grad_norm": 0.7420720458030701,
+      "learning_rate": 0.00010249471437745328,
+      "loss": 5.2393,
+      "num_input_tokens_seen": 342491136,
+      "step": 2613
+    },
+    {
+      "epoch": 0.4292901062045836,
+      "grad_norm": 0.6406148672103882,
+      "learning_rate": 0.00010243592761358217,
+      "loss": 5.2169,
+      "num_input_tokens_seen": 342884352,
+      "step": 2616
+    },
+    {
+      "epoch": 0.4297824113722494,
+      "grad_norm": 0.6864999532699585,
+      "learning_rate": 0.00010237724188684409,
+      "loss": 5.22,
+      "num_input_tokens_seen": 343277568,
+      "step": 2619
+    },
+    {
+      "epoch": 0.43027471653991517,
+      "grad_norm": 0.7687135934829712,
+      "learning_rate": 0.00010231865690814853,
+      "loss": 5.2909,
+      "num_input_tokens_seen": 343670784,
+      "step": 2622
+    },
+    {
+      "epoch": 0.43076702170758097,
+      "grad_norm": 0.6317099332809448,
+      "learning_rate": 0.0001022601723895616,
+      "loss": 5.2036,
+      "num_input_tokens_seen": 344064000,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4312593268752468,
+      "grad_norm": 0.6858761310577393,
+      "learning_rate": 0.00010220178804430015,
+      "loss": 5.1692,
+      "num_input_tokens_seen": 344457216,
+      "step": 2628
+    },
+    {
+      "epoch": 0.4317516320429126,
+      "grad_norm": 0.7586376667022705,
+      "learning_rate": 0.00010214350358672594,
+      "loss": 5.1816,
+      "num_input_tokens_seen": 344850432,
+      "step": 2631
+    },
+    {
+      "epoch": 0.4322439372105784,
+      "grad_norm": 0.643675684928894,
+      "learning_rate": 0.00010208531873233962,
+      "loss": 5.2572,
+      "num_input_tokens_seen": 345243648,
+      "step": 2634
+    },
+    {
+      "epoch": 0.4327362423782442,
+      "grad_norm": 0.5361793637275696,
+      "learning_rate": 0.00010202723319777505,
+      "loss": 5.2143,
+      "num_input_tokens_seen": 345636864,
+      "step": 2637
+    },
+    {
+      "epoch": 0.43322854754591,
+      "grad_norm": 0.597233772277832,
+      "learning_rate": 0.00010196924670079342,
+      "loss": 5.2139,
+      "num_input_tokens_seen": 346030080,
+      "step": 2640
+    },
+    {
+      "epoch": 0.43372085271357586,
+      "grad_norm": 0.6196411848068237,
+      "learning_rate": 0.00010191135896027748,
+      "loss": 5.2171,
+      "num_input_tokens_seen": 346423296,
+      "step": 2643
+    },
+    {
+      "epoch": 0.43421315788124165,
+      "grad_norm": 0.5964427590370178,
+      "learning_rate": 0.00010185356969622588,
+      "loss": 5.2317,
+      "num_input_tokens_seen": 346816512,
+      "step": 2646
+    },
+    {
+      "epoch": 0.43470546304890745,
+      "grad_norm": 0.5954868197441101,
+      "learning_rate": 0.00010179587862974739,
+      "loss": 5.1779,
+      "num_input_tokens_seen": 347209728,
+      "step": 2649
+    },
+    {
+      "epoch": 0.43519776821657324,
+      "grad_norm": 0.6737942099571228,
+      "learning_rate": 0.00010173828548305536,
+      "loss": 5.2293,
+      "num_input_tokens_seen": 347602944,
+      "step": 2652
+    },
+    {
+      "epoch": 0.43569007338423904,
+      "grad_norm": 0.6102513074874878,
+      "learning_rate": 0.00010168078997946198,
+      "loss": 5.1969,
+      "num_input_tokens_seen": 347996160,
+      "step": 2655
+    },
+    {
+      "epoch": 0.43618237855190484,
+      "grad_norm": 0.5308377742767334,
+      "learning_rate": 0.00010162339184337281,
+      "loss": 5.1909,
+      "num_input_tokens_seen": 348389376,
+      "step": 2658
+    },
+    {
+      "epoch": 0.4366746837195707,
+      "grad_norm": 0.6343761682510376,
+      "learning_rate": 0.00010156609080028115,
+      "loss": 5.1856,
+      "num_input_tokens_seen": 348782592,
+      "step": 2661
+    },
+    {
+      "epoch": 0.4371669888872365,
+      "grad_norm": 0.666679322719574,
+      "learning_rate": 0.0001015088865767626,
+      "loss": 5.2063,
+      "num_input_tokens_seen": 349175808,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4376592940549023,
+      "grad_norm": 0.5723670125007629,
+      "learning_rate": 0.00010145177890046946,
+      "loss": 5.1875,
+      "num_input_tokens_seen": 349569024,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4381515992225681,
+      "grad_norm": 0.6312902569770813,
+      "learning_rate": 0.00010139476750012542,
+      "loss": 5.2351,
+      "num_input_tokens_seen": 349962240,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4386439043902339,
+      "grad_norm": 0.6692864298820496,
+      "learning_rate": 0.00010133785210552012,
+      "loss": 5.2162,
+      "num_input_tokens_seen": 350355456,
+      "step": 2673
+    },
+    {
+      "epoch": 0.4391362095578997,
+      "grad_norm": 0.5867645740509033,
+      "learning_rate": 0.00010128103244750365,
+      "loss": 5.2184,
+      "num_input_tokens_seen": 350748672,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4396285147255655,
+      "grad_norm": 0.5672877430915833,
+      "learning_rate": 0.00010122430825798135,
+      "loss": 5.2469,
+      "num_input_tokens_seen": 351141888,
+      "step": 2679
+    },
+    {
+      "epoch": 0.4401208198932313,
+      "grad_norm": 0.7089299559593201,
+      "learning_rate": 0.00010116767926990843,
+      "loss": 5.1769,
+      "num_input_tokens_seen": 351535104,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4406131250608971,
+      "grad_norm": 0.677344560623169,
+      "learning_rate": 0.0001011111452172847,
+      "loss": 5.2314,
+      "num_input_tokens_seen": 351928320,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4411054302285629,
+      "grad_norm": 0.5708670616149902,
+      "learning_rate": 0.00010105470583514936,
+      "loss": 5.1938,
+      "num_input_tokens_seen": 352321536,
+      "step": 2688
+    },
+    {
+      "epoch": 0.44159773539622876,
+      "grad_norm": 0.5892400145530701,
+      "learning_rate": 0.00010099836085957568,
+      "loss": 5.177,
+      "num_input_tokens_seen": 352714752,
+      "step": 2691
+    },
+    {
+      "epoch": 0.44209004056389456,
+      "grad_norm": 0.6531316041946411,
+      "learning_rate": 0.00010094211002766593,
+      "loss": 5.2086,
+      "num_input_tokens_seen": 353107968,
+      "step": 2694
+    },
+    {
+      "epoch": 0.44258234573156036,
+      "grad_norm": 0.5908800363540649,
+      "learning_rate": 0.00010088595307754617,
+      "loss": 5.1706,
+      "num_input_tokens_seen": 353501184,
+      "step": 2697
+    },
+    {
+      "epoch": 0.44307465089922615,
+      "grad_norm": 0.5883475542068481,
+      "learning_rate": 0.00010082988974836116,
+      "loss": 5.1982,
+      "num_input_tokens_seen": 353894400,
+      "step": 2700
+    },
+    {
+      "epoch": 0.44356695606689195,
+      "grad_norm": 0.6906709671020508,
+      "learning_rate": 0.0001007739197802692,
+      "loss": 5.1891,
+      "num_input_tokens_seen": 354287616,
+      "step": 2703
+    },
+    {
+      "epoch": 0.44405926123455775,
+      "grad_norm": 0.7396822571754456,
+      "learning_rate": 0.00010071804291443717,
+      "loss": 5.2075,
+      "num_input_tokens_seen": 354680832,
+      "step": 2706
+    },
+    {
+      "epoch": 0.4445515664022236,
+      "grad_norm": 0.5959869027137756,
+      "learning_rate": 0.00010066225889303549,
+      "loss": 5.1908,
+      "num_input_tokens_seen": 355074048,
+      "step": 2709
+    },
+    {
+      "epoch": 0.4450438715698894,
+      "grad_norm": 0.5678725242614746,
+      "learning_rate": 0.00010060656745923301,
+      "loss": 5.2184,
+      "num_input_tokens_seen": 355467264,
+      "step": 2712
+    },
+    {
+      "epoch": 0.4455361767375552,
+      "grad_norm": 0.693349301815033,
+      "learning_rate": 0.00010055096835719215,
+      "loss": 5.2408,
+      "num_input_tokens_seen": 355860480,
+      "step": 2715
+    },
+    {
+      "epoch": 0.446028481905221,
+      "grad_norm": 0.6372326016426086,
+      "learning_rate": 0.000100495461332064,
+      "loss": 5.2121,
+      "num_input_tokens_seen": 356253696,
+      "step": 2718
+    },
+    {
+      "epoch": 0.4465207870728868,
+      "grad_norm": 0.6225445866584778,
+      "learning_rate": 0.00010044004612998325,
+      "loss": 5.1776,
+      "num_input_tokens_seen": 356646912,
+      "step": 2721
+    },
+    {
+      "epoch": 0.44701309224055263,
+      "grad_norm": 0.6060691475868225,
+      "learning_rate": 0.0001003847224980635,
+      "loss": 5.215,
+      "num_input_tokens_seen": 357040128,
+      "step": 2724
+    },
+    {
+      "epoch": 0.44750539740821843,
+      "grad_norm": 0.5323381423950195,
+      "learning_rate": 0.00010032949018439226,
+      "loss": 5.1929,
+      "num_input_tokens_seen": 357433344,
+      "step": 2727
+    },
+    {
+      "epoch": 0.4479977025758842,
+      "grad_norm": 0.5871007442474365,
+      "learning_rate": 0.00010027434893802628,
+      "loss": 5.1655,
+      "num_input_tokens_seen": 357826560,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44849000774355,
+      "grad_norm": 0.5575119853019714,
+      "learning_rate": 0.00010021929850898662,
+      "loss": 5.2098,
+      "num_input_tokens_seen": 358219776,
+      "step": 2733
+    },
+    {
+      "epoch": 0.4489823129112158,
+      "grad_norm": 0.6533513069152832,
+      "learning_rate": 0.00010016433864825397,
+      "loss": 5.2318,
+      "num_input_tokens_seen": 358612992,
+      "step": 2736
+    },
+    {
+      "epoch": 0.44947461807888167,
+      "grad_norm": 0.6353456377983093,
+      "learning_rate": 0.00010010946910776388,
+      "loss": 5.1966,
+      "num_input_tokens_seen": 359006208,
+      "step": 2739
+    },
+    {
+      "epoch": 0.44996692324654747,
+      "grad_norm": 0.5887168049812317,
+      "learning_rate": 0.00010005468964040215,
+      "loss": 5.1984,
+      "num_input_tokens_seen": 359399424,
+      "step": 2742
+    },
+    {
+      "epoch": 0.45045922841421326,
+      "grad_norm": 0.5238946080207825,
+      "learning_rate": 9.999999999999999e-05,
+      "loss": 5.1752,
+      "num_input_tokens_seen": 359792640,
+      "step": 2745
+    },
+    {
+      "epoch": 0.45095153358187906,
+      "grad_norm": 0.5709421038627625,
+      "learning_rate": 9.994539994132953e-05,
+      "loss": 5.2082,
+      "num_input_tokens_seen": 360185856,
+      "step": 2748
+    },
+    {
+      "epoch": 0.45144383874954486,
+      "grad_norm": 0.5375566482543945,
+      "learning_rate": 9.989088922009912e-05,
+      "loss": 5.1813,
+      "num_input_tokens_seen": 360579072,
+      "step": 2751
+    },
+    {
+      "epoch": 0.4519361439172107,
+      "grad_norm": 0.6022065877914429,
+      "learning_rate": 9.983646759294876e-05,
+      "loss": 5.2037,
+      "num_input_tokens_seen": 360972288,
+      "step": 2754
+    },
+    {
+      "epoch": 0.4524284490848765,
+      "grad_norm": 0.6292663216590881,
+      "learning_rate": 9.978213481744552e-05,
+      "loss": 5.2479,
+      "num_input_tokens_seen": 361365504,
+      "step": 2757
+    },
+    {
+      "epoch": 0.4529207542525423,
+      "grad_norm": 0.7360475659370422,
+      "learning_rate": 9.972789065207908e-05,
+      "loss": 5.1351,
+      "num_input_tokens_seen": 361758720,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4534130594202081,
+      "grad_norm": 0.5845218896865845,
+      "learning_rate": 9.967373485625708e-05,
+      "loss": 5.1405,
+      "num_input_tokens_seen": 362151936,
+      "step": 2763
+    },
+    {
+      "epoch": 0.4539053645878739,
+      "grad_norm": 0.5955424308776855,
+      "learning_rate": 9.961966719030078e-05,
+      "loss": 5.1566,
+      "num_input_tokens_seen": 362545152,
+      "step": 2766
+    },
+    {
+      "epoch": 0.4543976697555397,
+      "grad_norm": 0.568569004535675,
+      "learning_rate": 9.95656874154405e-05,
+      "loss": 5.1994,
+      "num_input_tokens_seen": 362938368,
+      "step": 2769
+    },
+    {
+      "epoch": 0.45488997492320554,
+      "grad_norm": 0.5579524636268616,
+      "learning_rate": 9.951179529381129e-05,
+      "loss": 5.2114,
+      "num_input_tokens_seen": 363331584,
+      "step": 2772
+    },
+    {
+      "epoch": 0.45538228009087134,
+      "grad_norm": 0.6336788535118103,
+      "learning_rate": 9.945799058844839e-05,
+      "loss": 5.1544,
+      "num_input_tokens_seen": 363724800,
+      "step": 2775
+    },
+    {
+      "epoch": 0.45587458525853713,
+      "grad_norm": 0.5949918627738953,
+      "learning_rate": 9.940427306328304e-05,
+      "loss": 5.2101,
+      "num_input_tokens_seen": 364118016,
+      "step": 2778
+    },
+    {
+      "epoch": 0.45636689042620293,
+      "grad_norm": 0.6075280904769897,
+      "learning_rate": 9.935064248313794e-05,
+      "loss": 5.201,
+      "num_input_tokens_seen": 364511232,
+      "step": 2781
+    },
+    {
+      "epoch": 0.4568591955938687,
+      "grad_norm": 0.580106794834137,
+      "learning_rate": 9.929709861372308e-05,
+      "loss": 5.1616,
+      "num_input_tokens_seen": 364904448,
+      "step": 2784
+    },
+    {
+      "epoch": 0.4573515007615346,
+      "grad_norm": 0.6350911855697632,
+      "learning_rate": 9.924364122163132e-05,
+      "loss": 5.2079,
+      "num_input_tokens_seen": 365297664,
+      "step": 2787
+    },
+    {
+      "epoch": 0.4578438059292004,
+      "grad_norm": 0.6221060156822205,
+      "learning_rate": 9.919027007433417e-05,
+      "loss": 5.211,
+      "num_input_tokens_seen": 365690880,
+      "step": 2790
+    },
+    {
+      "epoch": 0.45833611109686617,
+      "grad_norm": 0.6097714304924011,
+      "learning_rate": 9.913698494017759e-05,
+      "loss": 5.1863,
+      "num_input_tokens_seen": 366084096,
+      "step": 2793
+    },
+    {
+      "epoch": 0.45882841626453197,
+      "grad_norm": 0.6037392020225525,
+      "learning_rate": 9.90837855883777e-05,
+      "loss": 5.1546,
+      "num_input_tokens_seen": 366477312,
+      "step": 2796
+    },
+    {
+      "epoch": 0.45932072143219776,
+      "grad_norm": 0.6046000719070435,
+      "learning_rate": 9.903067178901658e-05,
+      "loss": 5.1625,
+      "num_input_tokens_seen": 366870528,
+      "step": 2799
+    },
+    {
+      "epoch": 0.45948482315475303,
+      "eval_accuracy": 0.21759485425826414,
+      "eval_loss": 5.443637371063232,
+      "eval_runtime": 107.9801,
+      "eval_samples_per_second": 2.778,
+      "eval_steps_per_second": 1.389,
+      "num_input_tokens_seen": 367001600,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4598130265998636,
+      "grad_norm": 0.623935878276825,
+      "learning_rate": 9.89776433130381e-05,
+      "loss": 5.1864,
+      "num_input_tokens_seen": 367263744,
+      "step": 2802
+    },
+    {
+      "epoch": 0.4603053317675294,
+      "grad_norm": 0.6677658557891846,
+      "learning_rate": 9.892469993224388e-05,
+      "loss": 5.2048,
+      "num_input_tokens_seen": 367656960,
+      "step": 2805
+    },
+    {
+      "epoch": 0.4607976369351952,
+      "grad_norm": 0.7684961557388306,
+      "learning_rate": 9.887184141928896e-05,
+      "loss": 5.1797,
+      "num_input_tokens_seen": 368050176,
+      "step": 2808
+    },
+    {
+      "epoch": 0.461289942102861,
+      "grad_norm": 0.7386724352836609,
+      "learning_rate": 9.881906754767789e-05,
+      "loss": 5.169,
+      "num_input_tokens_seen": 368443392,
+      "step": 2811
+    },
+    {
+      "epoch": 0.4617822472705268,
+      "grad_norm": 0.6707373261451721,
+      "learning_rate": 9.876637809176057e-05,
+      "loss": 5.2031,
+      "num_input_tokens_seen": 368836608,
+      "step": 2814
+    },
+    {
+      "epoch": 0.4622745524381926,
+      "grad_norm": 0.6488140821456909,
+      "learning_rate": 9.871377282672818e-05,
+      "loss": 5.204,
+      "num_input_tokens_seen": 369229824,
+      "step": 2817
+    },
+    {
+      "epoch": 0.46276685760585845,
+      "grad_norm": 0.5540419816970825,
+      "learning_rate": 9.866125152860918e-05,
+      "loss": 5.1672,
+      "num_input_tokens_seen": 369623040,
+      "step": 2820
+    },
+    {
+      "epoch": 0.46325916277352425,
+      "grad_norm": 0.753920316696167,
+      "learning_rate": 9.860881397426531e-05,
+      "loss": 5.2168,
+      "num_input_tokens_seen": 370016256,
+      "step": 2823
+    },
+    {
+      "epoch": 0.46375146794119004,
+      "grad_norm": 0.7269083857536316,
+      "learning_rate": 9.855645994138763e-05,
+      "loss": 5.1643,
+      "num_input_tokens_seen": 370409472,
+      "step": 2826
+    },
+    {
+      "epoch": 0.46424377310885584,
+      "grad_norm": 0.7552649974822998,
+      "learning_rate": 9.850418920849244e-05,
+      "loss": 5.1711,
+      "num_input_tokens_seen": 370802688,
+      "step": 2829
+    },
+    {
+      "epoch": 0.46473607827652164,
+      "grad_norm": 0.7487272024154663,
+      "learning_rate": 9.845200155491757e-05,
+      "loss": 5.2207,
+      "num_input_tokens_seen": 371195904,
+      "step": 2832
+    },
+    {
+      "epoch": 0.4652283834441875,
+      "grad_norm": 0.7448633909225464,
+      "learning_rate": 9.839989676081821e-05,
+      "loss": 5.2232,
+      "num_input_tokens_seen": 371589120,
+      "step": 2835
+    },
+    {
+      "epoch": 0.4657206886118533,
+      "grad_norm": 0.5762555003166199,
+      "learning_rate": 9.834787460716322e-05,
+      "loss": 5.1584,
+      "num_input_tokens_seen": 371982336,
+      "step": 2838
+    },
+    {
+      "epoch": 0.4662129937795191,
+      "grad_norm": 0.6208467483520508,
+      "learning_rate": 9.829593487573116e-05,
+      "loss": 5.2084,
+      "num_input_tokens_seen": 372375552,
+      "step": 2841
+    },
+    {
+      "epoch": 0.4667052989471849,
+      "grad_norm": 0.5821816325187683,
+      "learning_rate": 9.824407734910645e-05,
+      "loss": 5.1722,
+      "num_input_tokens_seen": 372768768,
+      "step": 2844
+    },
+    {
+      "epoch": 0.4671976041148507,
+      "grad_norm": 0.5190629363059998,
+      "learning_rate": 9.819230181067567e-05,
+      "loss": 5.1523,
+      "num_input_tokens_seen": 373161984,
+      "step": 2847
+    },
+    {
+      "epoch": 0.4676899092825165,
+      "grad_norm": 0.6430248618125916,
+      "learning_rate": 9.814060804462351e-05,
+      "loss": 5.2025,
+      "num_input_tokens_seen": 373555200,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4681822144501823,
+      "grad_norm": 0.6955888271331787,
+      "learning_rate": 9.808899583592925e-05,
+      "loss": 5.1616,
+      "num_input_tokens_seen": 373948416,
+      "step": 2853
+    },
+    {
+      "epoch": 0.4686745196178481,
+      "grad_norm": 0.6991474032402039,
+      "learning_rate": 9.803746497036285e-05,
+      "loss": 5.1637,
+      "num_input_tokens_seen": 374341632,
+      "step": 2856
+    },
+    {
+      "epoch": 0.4691668247855139,
+      "grad_norm": 0.7030759453773499,
+      "learning_rate": 9.798601523448131e-05,
+      "loss": 5.1855,
+      "num_input_tokens_seen": 374734848,
+      "step": 2859
+    },
+    {
+      "epoch": 0.4696591299531797,
+      "grad_norm": 0.6420966982841492,
+      "learning_rate": 9.793464641562482e-05,
+      "loss": 5.213,
+      "num_input_tokens_seen": 375128064,
+      "step": 2862
+    },
+    {
+      "epoch": 0.4701514351208455,
+      "grad_norm": 0.6913998126983643,
+      "learning_rate": 9.788335830191324e-05,
+      "loss": 5.2171,
+      "num_input_tokens_seen": 375521280,
+      "step": 2865
+    },
+    {
+      "epoch": 0.47064374028851136,
+      "grad_norm": 0.5875546336174011,
+      "learning_rate": 9.783215068224234e-05,
+      "loss": 5.1719,
+      "num_input_tokens_seen": 375914496,
+      "step": 2868
+    },
+    {
+      "epoch": 0.47113604545617715,
+      "grad_norm": 0.6099095344543457,
+      "learning_rate": 9.778102334628006e-05,
+      "loss": 5.1223,
+      "num_input_tokens_seen": 376307712,
+      "step": 2871
+    },
+    {
+      "epoch": 0.47162835062384295,
+      "grad_norm": 0.5603676438331604,
+      "learning_rate": 9.772997608446309e-05,
+      "loss": 5.1571,
+      "num_input_tokens_seen": 376700928,
+      "step": 2874
+    },
+    {
+      "epoch": 0.47212065579150875,
+      "grad_norm": 0.6761665940284729,
+      "learning_rate": 9.767900868799307e-05,
+      "loss": 5.1816,
+      "num_input_tokens_seen": 377094144,
+      "step": 2877
+    },
+    {
+      "epoch": 0.47261296095917454,
+      "grad_norm": 0.5453407764434814,
+      "learning_rate": 9.762812094883316e-05,
+      "loss": 5.1818,
+      "num_input_tokens_seen": 377487360,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4731052661268404,
+      "grad_norm": 0.5907162427902222,
+      "learning_rate": 9.757731265970434e-05,
+      "loss": 5.1711,
+      "num_input_tokens_seen": 377880576,
+      "step": 2883
+    },
+    {
+      "epoch": 0.4735975712945062,
+      "grad_norm": 0.6772825121879578,
+      "learning_rate": 9.752658361408191e-05,
+      "loss": 5.1518,
+      "num_input_tokens_seen": 378273792,
+      "step": 2886
+    },
+    {
+      "epoch": 0.474089876462172,
+      "grad_norm": 0.6370013356208801,
+      "learning_rate": 9.7475933606192e-05,
+      "loss": 5.1785,
+      "num_input_tokens_seen": 378667008,
+      "step": 2889
+    },
+    {
+      "epoch": 0.4745821816298378,
+      "grad_norm": 0.6495605707168579,
+      "learning_rate": 9.742536243100805e-05,
+      "loss": 5.1962,
+      "num_input_tokens_seen": 379060224,
+      "step": 2892
+    },
+    {
+      "epoch": 0.4750744867975036,
+      "grad_norm": 0.5345600843429565,
+      "learning_rate": 9.737486988424731e-05,
+      "loss": 5.2161,
+      "num_input_tokens_seen": 379453440,
+      "step": 2895
+    },
+    {
+      "epoch": 0.47556679196516943,
+      "grad_norm": 0.5919015407562256,
+      "learning_rate": 9.73244557623673e-05,
+      "loss": 5.148,
+      "num_input_tokens_seen": 379846656,
+      "step": 2898
+    },
+    {
+      "epoch": 0.47605909713283523,
+      "grad_norm": 0.6196090579032898,
+      "learning_rate": 9.727411986256259e-05,
+      "loss": 5.1709,
+      "num_input_tokens_seen": 380239872,
+      "step": 2901
+    },
+    {
+      "epoch": 0.476551402300501,
+      "grad_norm": 0.6973289847373962,
+      "learning_rate": 9.722386198276106e-05,
+      "loss": 5.2245,
+      "num_input_tokens_seen": 380633088,
+      "step": 2904
+    },
+    {
+      "epoch": 0.4770437074681668,
+      "grad_norm": 0.5490657687187195,
+      "learning_rate": 9.717368192162079e-05,
+      "loss": 5.1915,
+      "num_input_tokens_seen": 381026304,
+      "step": 2907
+    },
+    {
+      "epoch": 0.4775360126358326,
+      "grad_norm": 0.5101742148399353,
+      "learning_rate": 9.712357947852647e-05,
+      "loss": 5.1612,
+      "num_input_tokens_seen": 381419520,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47802831780349847,
+      "grad_norm": 0.5948254466056824,
+      "learning_rate": 9.707355445358611e-05,
+      "loss": 5.1209,
+      "num_input_tokens_seen": 381812736,
+      "step": 2913
+    },
+    {
+      "epoch": 0.47852062297116427,
+      "grad_norm": 0.6882492303848267,
+      "learning_rate": 9.702360664762765e-05,
+      "loss": 5.1987,
+      "num_input_tokens_seen": 382205952,
+      "step": 2916
+    },
+    {
+      "epoch": 0.47901292813883006,
+      "grad_norm": 0.6364412903785706,
+      "learning_rate": 9.697373586219577e-05,
+      "loss": 5.2198,
+      "num_input_tokens_seen": 382599168,
+      "step": 2919
+    },
+    {
+      "epoch": 0.47950523330649586,
+      "grad_norm": 0.5922911167144775,
+      "learning_rate": 9.692394189954834e-05,
+      "loss": 5.1431,
+      "num_input_tokens_seen": 382992384,
+      "step": 2922
+    },
+    {
+      "epoch": 0.47999753847416166,
+      "grad_norm": 0.7349563241004944,
+      "learning_rate": 9.687422456265331e-05,
+      "loss": 5.1448,
+      "num_input_tokens_seen": 383385600,
+      "step": 2925
+    },
+    {
+      "epoch": 0.48048984364182745,
+      "grad_norm": 0.5603036284446716,
+      "learning_rate": 9.682458365518541e-05,
+      "loss": 5.1633,
+      "num_input_tokens_seen": 383778816,
+      "step": 2928
+    },
+    {
+      "epoch": 0.4809821488094933,
+      "grad_norm": 0.5058417320251465,
+      "learning_rate": 9.677501898152282e-05,
+      "loss": 5.14,
+      "num_input_tokens_seen": 384172032,
+      "step": 2931
+    },
+    {
+      "epoch": 0.4814744539771591,
+      "grad_norm": 0.5244634747505188,
+      "learning_rate": 9.6725530346744e-05,
+      "loss": 5.1999,
+      "num_input_tokens_seen": 384565248,
+      "step": 2934
+    },
+    {
+      "epoch": 0.4819667591448249,
+      "grad_norm": 0.5387030243873596,
+      "learning_rate": 9.667611755662445e-05,
+      "loss": 5.151,
+      "num_input_tokens_seen": 384958464,
+      "step": 2937
+    },
+    {
+      "epoch": 0.4824590643124907,
+      "grad_norm": 0.5727872848510742,
+      "learning_rate": 9.662678041763345e-05,
+      "loss": 5.1763,
+      "num_input_tokens_seen": 385351680,
+      "step": 2940
+    },
+    {
+      "epoch": 0.4829513694801565,
+      "grad_norm": 0.6338843703269958,
+      "learning_rate": 9.657751873693102e-05,
+      "loss": 5.1369,
+      "num_input_tokens_seen": 385744896,
+      "step": 2943
+    },
+    {
+      "epoch": 0.48344367464782234,
+      "grad_norm": 0.6214335560798645,
+      "learning_rate": 9.652833232236462e-05,
+      "loss": 5.1523,
+      "num_input_tokens_seen": 386138112,
+      "step": 2946
+    },
+    {
+      "epoch": 0.48393597981548814,
+      "grad_norm": 0.6365739703178406,
+      "learning_rate": 9.647922098246606e-05,
+      "loss": 5.1471,
+      "num_input_tokens_seen": 386531328,
+      "step": 2949
+    },
+    {
+      "epoch": 0.48442828498315393,
+      "grad_norm": 0.6559484601020813,
+      "learning_rate": 9.643018452644833e-05,
+      "loss": 5.1941,
+      "num_input_tokens_seen": 386924544,
+      "step": 2952
+    },
+    {
+      "epoch": 0.48492059015081973,
+      "grad_norm": 0.6113404035568237,
+      "learning_rate": 9.638122276420258e-05,
+      "loss": 5.1769,
+      "num_input_tokens_seen": 387317760,
+      "step": 2955
+    },
+    {
+      "epoch": 0.4854128953184855,
+      "grad_norm": 0.6409229636192322,
+      "learning_rate": 9.633233550629488e-05,
+      "loss": 5.2046,
+      "num_input_tokens_seen": 387710976,
+      "step": 2958
+    },
+    {
+      "epoch": 0.4859052004861514,
+      "grad_norm": 0.5778577327728271,
+      "learning_rate": 9.628352256396328e-05,
+      "loss": 5.2029,
+      "num_input_tokens_seen": 388104192,
+      "step": 2961
+    },
+    {
+      "epoch": 0.4863975056538172,
+      "grad_norm": 0.5613879561424255,
+      "learning_rate": 9.623478374911467e-05,
+      "loss": 5.19,
+      "num_input_tokens_seen": 388497408,
+      "step": 2964
+    },
+    {
+      "epoch": 0.48688981082148297,
+      "grad_norm": 0.5774136781692505,
+      "learning_rate": 9.618611887432175e-05,
+      "loss": 5.1534,
+      "num_input_tokens_seen": 388890624,
+      "step": 2967
+    },
+    {
+      "epoch": 0.48738211598914877,
+      "grad_norm": 0.6295081377029419,
+      "learning_rate": 9.613752775282003e-05,
+      "loss": 5.1471,
+      "num_input_tokens_seen": 389283840,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48787442115681456,
+      "grad_norm": 0.5557025074958801,
+      "learning_rate": 9.608901019850477e-05,
+      "loss": 5.1841,
+      "num_input_tokens_seen": 389677056,
+      "step": 2973
+    },
+    {
+      "epoch": 0.48836672632448036,
+      "grad_norm": 0.6536008715629578,
+      "learning_rate": 9.604056602592805e-05,
+      "loss": 5.1354,
+      "num_input_tokens_seen": 390070272,
+      "step": 2976
+    },
+    {
+      "epoch": 0.4888590314921462,
+      "grad_norm": 0.6432478427886963,
+      "learning_rate": 9.599219505029571e-05,
+      "loss": 5.174,
+      "num_input_tokens_seen": 390463488,
+      "step": 2979
+    },
+    {
+      "epoch": 0.489351336659812,
+      "grad_norm": 0.6222785711288452,
+      "learning_rate": 9.594389708746449e-05,
+      "loss": 5.11,
+      "num_input_tokens_seen": 390856704,
+      "step": 2982
+    },
+    {
+      "epoch": 0.4898436418274778,
+      "grad_norm": 0.5929418206214905,
+      "learning_rate": 9.589567195393901e-05,
+      "loss": 5.1108,
+      "num_input_tokens_seen": 391249920,
+      "step": 2985
+    },
+    {
+      "epoch": 0.4903359469951436,
+      "grad_norm": 0.629441499710083,
+      "learning_rate": 9.584751946686886e-05,
+      "loss": 5.1573,
+      "num_input_tokens_seen": 391643136,
+      "step": 2988
+    },
+    {
+      "epoch": 0.4908282521628094,
+      "grad_norm": 0.5835034251213074,
+      "learning_rate": 9.57994394440457e-05,
+      "loss": 5.1485,
+      "num_input_tokens_seen": 392036352,
+      "step": 2991
+    },
+    {
+      "epoch": 0.49132055733047525,
+      "grad_norm": 0.5538569092750549,
+      "learning_rate": 9.575143170390034e-05,
+      "loss": 5.1571,
+      "num_input_tokens_seen": 392429568,
+      "step": 2994
+    },
+    {
+      "epoch": 0.49181286249814105,
+      "grad_norm": 0.7033787369728088,
+      "learning_rate": 9.57034960654999e-05,
+      "loss": 5.1544,
+      "num_input_tokens_seen": 392822784,
+      "step": 2997
+    },
+    {
+      "epoch": 0.49230516766580684,
+      "grad_norm": 0.6497957706451416,
+      "learning_rate": 9.565563234854494e-05,
+      "loss": 5.1979,
+      "num_input_tokens_seen": 393216000,
+      "step": 3000
+    },
+    {
+      "epoch": 0.49279747283347264,
+      "grad_norm": 0.6436689496040344,
+      "learning_rate": 9.560784037336655e-05,
+      "loss": 5.1928,
+      "num_input_tokens_seen": 393609216,
+      "step": 3003
+    },
+    {
+      "epoch": 0.49328977800113843,
+      "grad_norm": 0.6518992781639099,
+      "learning_rate": 9.556011996092359e-05,
+      "loss": 5.1835,
+      "num_input_tokens_seen": 394002432,
+      "step": 3006
+    },
+    {
+      "epoch": 0.4937820831688043,
+      "grad_norm": 0.6519846320152283,
+      "learning_rate": 9.551247093279984e-05,
+      "loss": 5.1321,
+      "num_input_tokens_seen": 394395648,
+      "step": 3009
+    },
+    {
+      "epoch": 0.4942743883364701,
+      "grad_norm": 0.5854218006134033,
+      "learning_rate": 9.546489311120117e-05,
+      "loss": 5.1535,
+      "num_input_tokens_seen": 394788864,
+      "step": 3012
+    },
+    {
+      "epoch": 0.4947666935041359,
+      "grad_norm": 0.6409735083580017,
+      "learning_rate": 9.541738631895289e-05,
+      "loss": 5.1168,
+      "num_input_tokens_seen": 395182080,
+      "step": 3015
+    },
+    {
+      "epoch": 0.4952589986718017,
+      "grad_norm": 0.6534507870674133,
+      "learning_rate": 9.536995037949675e-05,
+      "loss": 5.1957,
+      "num_input_tokens_seen": 395575296,
+      "step": 3018
+    },
+    {
+      "epoch": 0.49575130383946747,
+      "grad_norm": 0.5633962154388428,
+      "learning_rate": 9.53225851168884e-05,
+      "loss": 5.1291,
+      "num_input_tokens_seen": 395968512,
+      "step": 3021
+    },
+    {
+      "epoch": 0.4962436090071333,
+      "grad_norm": 0.646615207195282,
+      "learning_rate": 9.527529035579451e-05,
+      "loss": 5.14,
+      "num_input_tokens_seen": 396361728,
+      "step": 3024
+    },
+    {
+      "epoch": 0.4967359141747991,
+      "grad_norm": 0.6950668096542358,
+      "learning_rate": 9.522806592149013e-05,
+      "loss": 5.153,
+      "num_input_tokens_seen": 396754944,
+      "step": 3027
+    },
+    {
+      "epoch": 0.4972282193424649,
+      "grad_norm": 0.5976789593696594,
+      "learning_rate": 9.518091163985591e-05,
+      "loss": 5.1526,
+      "num_input_tokens_seen": 397148160,
+      "step": 3030
+    },
+    {
+      "epoch": 0.4977205245101307,
+      "grad_norm": 0.7048370838165283,
+      "learning_rate": 9.513382733737545e-05,
+      "loss": 5.0725,
+      "num_input_tokens_seen": 397541376,
+      "step": 3033
+    },
+    {
+      "epoch": 0.4982128296777965,
+      "grad_norm": 0.6047758460044861,
+      "learning_rate": 9.508681284113262e-05,
+      "loss": 5.1815,
+      "num_input_tokens_seen": 397934592,
+      "step": 3036
+    },
+    {
+      "epoch": 0.4987051348454623,
+      "grad_norm": 0.581046462059021,
+      "learning_rate": 9.503986797880886e-05,
+      "loss": 5.1895,
+      "num_input_tokens_seen": 398327808,
+      "step": 3039
+    },
+    {
+      "epoch": 0.49919744001312816,
+      "grad_norm": 0.5836671590805054,
+      "learning_rate": 9.499299257868052e-05,
+      "loss": 5.1375,
+      "num_input_tokens_seen": 398721024,
+      "step": 3042
+    },
+    {
+      "epoch": 0.49968974518079395,
+      "grad_norm": 0.6080839037895203,
+      "learning_rate": 9.494618646961631e-05,
+      "loss": 5.1807,
+      "num_input_tokens_seen": 399114240,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5001820503484598,
+      "grad_norm": 0.603197455406189,
+      "learning_rate": 9.489944948107455e-05,
+      "loss": 5.1038,
+      "num_input_tokens_seen": 399507456,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5006743555161256,
+      "grad_norm": 0.5664237141609192,
+      "learning_rate": 9.485278144310068e-05,
+      "loss": 5.132,
+      "num_input_tokens_seen": 399900672,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5011666606837913,
+      "grad_norm": 0.6530909538269043,
+      "learning_rate": 9.480618218632454e-05,
+      "loss": 5.1332,
+      "num_input_tokens_seen": 400293888,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5016589658514572,
+      "grad_norm": 0.6573035717010498,
+      "learning_rate": 9.475965154195791e-05,
+      "loss": 5.2157,
+      "num_input_tokens_seen": 400687104,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5021512710191229,
+      "grad_norm": 0.6319076418876648,
+      "learning_rate": 9.471318934179186e-05,
+      "loss": 5.1244,
+      "num_input_tokens_seen": 401080320,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5026435761867888,
+      "grad_norm": 0.6034519076347351,
+      "learning_rate": 9.466679541819426e-05,
+      "loss": 5.1375,
+      "num_input_tokens_seen": 401473536,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5031358813544546,
+      "grad_norm": 0.6358224749565125,
+      "learning_rate": 9.462046960410713e-05,
+      "loss": 5.1482,
+      "num_input_tokens_seen": 401866752,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5036281865221204,
+      "grad_norm": 0.5579473376274109,
+      "learning_rate": 9.457421173304426e-05,
+      "loss": 5.1301,
+      "num_input_tokens_seen": 402259968,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5041204916897862,
+      "grad_norm": 0.582528829574585,
+      "learning_rate": 9.452802163908858e-05,
+      "loss": 5.1343,
+      "num_input_tokens_seen": 402653184,
+      "step": 3072
+    },
+    {
+      "epoch": 0.504612796857452,
+      "grad_norm": 0.5620675683021545,
+      "learning_rate": 9.448189915688972e-05,
+      "loss": 5.1392,
+      "num_input_tokens_seen": 403046400,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5051051020251178,
+      "grad_norm": 0.5927409529685974,
+      "learning_rate": 9.443584412166155e-05,
+      "loss": 5.1243,
+      "num_input_tokens_seen": 403439616,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5055974071927837,
+      "grad_norm": 0.6298641562461853,
+      "learning_rate": 9.438985636917958e-05,
+      "loss": 5.1756,
+      "num_input_tokens_seen": 403832832,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5060897123604494,
+      "grad_norm": 0.5789916515350342,
+      "learning_rate": 9.434393573577864e-05,
+      "loss": 5.0964,
+      "num_input_tokens_seen": 404226048,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5065820175281153,
+      "grad_norm": 0.6592864394187927,
+      "learning_rate": 9.429808205835036e-05,
+      "loss": 5.1169,
+      "num_input_tokens_seen": 404619264,
+      "step": 3087
+    },
+    {
+      "epoch": 0.507074322695781,
+      "grad_norm": 0.5415318608283997,
+      "learning_rate": 9.42522951743408e-05,
+      "loss": 5.0733,
+      "num_input_tokens_seen": 405012480,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5075666278634469,
+      "grad_norm": 0.6288437247276306,
+      "learning_rate": 9.420657492174793e-05,
+      "loss": 5.1799,
+      "num_input_tokens_seen": 405405696,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5080589330311127,
+      "grad_norm": 0.5948610901832581,
+      "learning_rate": 9.416092113911928e-05,
+      "loss": 5.1405,
+      "num_input_tokens_seen": 405798912,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5085512381987785,
+      "grad_norm": 0.5589770078659058,
+      "learning_rate": 9.411533366554959e-05,
+      "loss": 5.1584,
+      "num_input_tokens_seen": 406192128,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5090435433664443,
+      "grad_norm": 0.6298828721046448,
+      "learning_rate": 9.406981234067836e-05,
+      "loss": 5.1886,
+      "num_input_tokens_seen": 406585344,
+      "step": 3102
+    },
+    {
+      "epoch": 0.50953584853411,
+      "grad_norm": 0.5577113628387451,
+      "learning_rate": 9.40243570046875e-05,
+      "loss": 5.145,
+      "num_input_tokens_seen": 406978560,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5100281537017759,
+      "grad_norm": 0.6395171284675598,
+      "learning_rate": 9.397896749829895e-05,
+      "loss": 5.0892,
+      "num_input_tokens_seen": 407371776,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5105204588694416,
+      "grad_norm": 0.7225360870361328,
+      "learning_rate": 9.393364366277242e-05,
+      "loss": 5.126,
+      "num_input_tokens_seen": 407764992,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5110127640371075,
+      "grad_norm": 0.7004207968711853,
+      "learning_rate": 9.388838533990295e-05,
+      "loss": 5.148,
+      "num_input_tokens_seen": 408158208,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5115050692047733,
+      "grad_norm": 0.705188512802124,
+      "learning_rate": 9.384319237201867e-05,
+      "loss": 5.1684,
+      "num_input_tokens_seen": 408551424,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5119973743724391,
+      "grad_norm": 0.6463566422462463,
+      "learning_rate": 9.379806460197844e-05,
+      "loss": 5.1332,
+      "num_input_tokens_seen": 408944640,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5124896795401049,
+      "grad_norm": 0.8040825724601746,
+      "learning_rate": 9.375300187316961e-05,
+      "loss": 5.1332,
+      "num_input_tokens_seen": 409337856,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5129819847077707,
+      "grad_norm": 0.6625365018844604,
+      "learning_rate": 9.37080040295057e-05,
+      "loss": 5.1883,
+      "num_input_tokens_seen": 409731072,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5134742898754365,
+      "grad_norm": 0.5686776041984558,
+      "learning_rate": 9.36630709154241e-05,
+      "loss": 5.108,
+      "num_input_tokens_seen": 410124288,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5139665950431024,
+      "grad_norm": 0.6858258247375488,
+      "learning_rate": 9.36182023758839e-05,
+      "loss": 5.1547,
+      "num_input_tokens_seen": 410517504,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5144589002107681,
+      "grad_norm": 0.7715529203414917,
+      "learning_rate": 9.357339825636354e-05,
+      "loss": 5.1571,
+      "num_input_tokens_seen": 410910720,
+      "step": 3135
+    },
+    {
+      "epoch": 0.514951205378434,
+      "grad_norm": 0.6023756265640259,
+      "learning_rate": 9.352865840285866e-05,
+      "loss": 5.1242,
+      "num_input_tokens_seen": 411303936,
+      "step": 3138
+    },
+    {
+      "epoch": 0.5154435105460997,
+      "grad_norm": 0.5971417427062988,
+      "learning_rate": 9.348398266187983e-05,
+      "loss": 5.1568,
+      "num_input_tokens_seen": 411697152,
+      "step": 3141
+    },
+    {
+      "epoch": 0.5159358157137656,
+      "grad_norm": 0.6816672682762146,
+      "learning_rate": 9.343937088045033e-05,
+      "loss": 5.0748,
+      "num_input_tokens_seen": 412090368,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5164281208814314,
+      "grad_norm": 0.6150919795036316,
+      "learning_rate": 9.339482290610404e-05,
+      "loss": 5.1536,
+      "num_input_tokens_seen": 412483584,
+      "step": 3147
+    },
+    {
+      "epoch": 0.5169204260490972,
+      "grad_norm": 0.5576636791229248,
+      "learning_rate": 9.335033858688308e-05,
+      "loss": 5.1204,
+      "num_input_tokens_seen": 412876800,
+      "step": 3150
+    },
+    {
+      "epoch": 0.517412731216763,
+      "grad_norm": 0.6487387418746948,
+      "learning_rate": 9.330591777133583e-05,
+      "loss": 5.1259,
+      "num_input_tokens_seen": 413270016,
+      "step": 3153
+    },
+    {
+      "epoch": 0.5179050363844288,
+      "grad_norm": 0.5932300090789795,
+      "learning_rate": 9.32615603085146e-05,
+      "loss": 5.142,
+      "num_input_tokens_seen": 413663232,
+      "step": 3156
+    },
+    {
+      "epoch": 0.5183973415520946,
+      "grad_norm": 0.595696210861206,
+      "learning_rate": 9.321726604797357e-05,
+      "loss": 5.1505,
+      "num_input_tokens_seen": 414056448,
+      "step": 3159
+    },
+    {
+      "epoch": 0.5188896467197605,
+      "grad_norm": 0.5437079668045044,
+      "learning_rate": 9.317303483976665e-05,
+      "loss": 5.0969,
+      "num_input_tokens_seen": 414449664,
+      "step": 3162
+    },
+    {
+      "epoch": 0.5193819518874262,
+      "grad_norm": 0.6077457666397095,
+      "learning_rate": 9.312886653444527e-05,
+      "loss": 5.175,
+      "num_input_tokens_seen": 414842880,
+      "step": 3165
+    },
+    {
+      "epoch": 0.519874257055092,
+      "grad_norm": 0.5742005109786987,
+      "learning_rate": 9.308476098305633e-05,
+      "loss": 5.1699,
+      "num_input_tokens_seen": 415236096,
+      "step": 3168
+    },
+    {
+      "epoch": 0.5203665622227578,
+      "grad_norm": 0.5593070983886719,
+      "learning_rate": 9.304071803714007e-05,
+      "loss": 5.1766,
+      "num_input_tokens_seen": 415629312,
+      "step": 3171
+    },
+    {
+      "epoch": 0.5208588673904236,
+      "grad_norm": 0.6762742400169373,
+      "learning_rate": 9.299673754872799e-05,
+      "loss": 5.1687,
+      "num_input_tokens_seen": 416022528,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5213511725580895,
+      "grad_norm": 0.6270245909690857,
+      "learning_rate": 9.295281937034069e-05,
+      "loss": 5.142,
+      "num_input_tokens_seen": 416415744,
+      "step": 3177
+    },
+    {
+      "epoch": 0.5218434777257552,
+      "grad_norm": 0.6227966547012329,
+      "learning_rate": 9.290896335498588e-05,
+      "loss": 5.1429,
+      "num_input_tokens_seen": 416808960,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5223357828934211,
+      "grad_norm": 0.6371598839759827,
+      "learning_rate": 9.286516935615632e-05,
+      "loss": 5.1405,
+      "num_input_tokens_seen": 417202176,
+      "step": 3183
+    },
+    {
+      "epoch": 0.5228280880610868,
+      "grad_norm": 0.5920236706733704,
+      "learning_rate": 9.282143722782764e-05,
+      "loss": 5.1078,
+      "num_input_tokens_seen": 417595392,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5233203932287527,
+      "grad_norm": 0.715696394443512,
+      "learning_rate": 9.277776682445643e-05,
+      "loss": 5.1564,
+      "num_input_tokens_seen": 417988608,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5238126983964185,
+      "grad_norm": 0.7197597026824951,
+      "learning_rate": 9.273415800097812e-05,
+      "loss": 5.1441,
+      "num_input_tokens_seen": 418381824,
+      "step": 3192
+    },
+    {
+      "epoch": 0.5243050035640843,
+      "grad_norm": 0.6280763745307922,
+      "learning_rate": 9.269061061280504e-05,
+      "loss": 5.1577,
+      "num_input_tokens_seen": 418775040,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5247973087317501,
+      "grad_norm": 0.5797785520553589,
+      "learning_rate": 9.264712451582432e-05,
+      "loss": 5.1851,
+      "num_input_tokens_seen": 419168256,
+      "step": 3198
+    },
+    {
+      "epoch": 0.5251255121768607,
+      "eval_accuracy": 0.22059110893991207,
+      "eval_loss": 5.397455215454102,
+      "eval_runtime": 110.5008,
+      "eval_samples_per_second": 2.715,
+      "eval_steps_per_second": 1.357,
+      "num_input_tokens_seen": 419430400,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5252896138994159,
+      "grad_norm": 0.6440132260322571,
+      "learning_rate": 9.260369956639594e-05,
+      "loss": 5.1808,
+      "num_input_tokens_seen": 419561472,
+      "step": 3201
+    },
+    {
+      "epoch": 0.5257819190670817,
+      "grad_norm": 0.563518762588501,
+      "learning_rate": 9.256033562135067e-05,
+      "loss": 5.1735,
+      "num_input_tokens_seen": 419954688,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5262742242347476,
+      "grad_norm": 0.6111391186714172,
+      "learning_rate": 9.251703253798821e-05,
+      "loss": 5.1588,
+      "num_input_tokens_seen": 420347904,
+      "step": 3207
+    },
+    {
+      "epoch": 0.5267665294024133,
+      "grad_norm": 0.723850429058075,
+      "learning_rate": 9.247379017407515e-05,
+      "loss": 5.1685,
+      "num_input_tokens_seen": 420741120,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5272588345700792,
+      "grad_norm": 0.6539550423622131,
+      "learning_rate": 9.24306083878429e-05,
+      "loss": 5.1203,
+      "num_input_tokens_seen": 421134336,
+      "step": 3213
+    },
+    {
+      "epoch": 0.5277511397377449,
+      "grad_norm": 0.6335772275924683,
+      "learning_rate": 9.238748703798599e-05,
+      "loss": 5.1889,
+      "num_input_tokens_seen": 421527552,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5282434449054108,
+      "grad_norm": 0.6394396424293518,
+      "learning_rate": 9.234442598365984e-05,
+      "loss": 5.0933,
+      "num_input_tokens_seen": 421920768,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5287357500730765,
+      "grad_norm": 0.5733250975608826,
+      "learning_rate": 9.230142508447905e-05,
+      "loss": 5.1459,
+      "num_input_tokens_seen": 422313984,
+      "step": 3222
+    },
+    {
+      "epoch": 0.5292280552407423,
+      "grad_norm": 0.6153180003166199,
+      "learning_rate": 9.225848420051536e-05,
+      "loss": 5.1234,
+      "num_input_tokens_seen": 422707200,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5297203604084082,
+      "grad_norm": 0.5874418616294861,
+      "learning_rate": 9.221560319229582e-05,
+      "loss": 5.1364,
+      "num_input_tokens_seen": 423100416,
+      "step": 3228
+    },
+    {
+      "epoch": 0.5302126655760739,
+      "grad_norm": 0.5672056674957275,
+      "learning_rate": 9.217278192080077e-05,
+      "loss": 5.1172,
+      "num_input_tokens_seen": 423493632,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5307049707437398,
+      "grad_norm": 0.5874174237251282,
+      "learning_rate": 9.213002024746207e-05,
+      "loss": 5.1071,
+      "num_input_tokens_seen": 423886848,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5311972759114055,
+      "grad_norm": 0.5822425484657288,
+      "learning_rate": 9.208731803416115e-05,
+      "loss": 5.1241,
+      "num_input_tokens_seen": 424280064,
+      "step": 3237
+    },
+    {
+      "epoch": 0.5316895810790714,
+      "grad_norm": 0.7283825278282166,
+      "learning_rate": 9.204467514322719e-05,
+      "loss": 5.1217,
+      "num_input_tokens_seen": 424673280,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5321818862467372,
+      "grad_norm": 0.6498001217842102,
+      "learning_rate": 9.200209143743507e-05,
+      "loss": 5.0671,
+      "num_input_tokens_seen": 425066496,
+      "step": 3243
+    },
+    {
+      "epoch": 0.532674191414403,
+      "grad_norm": 0.6793512105941772,
+      "learning_rate": 9.195956678000385e-05,
+      "loss": 5.1119,
+      "num_input_tokens_seen": 425459712,
+      "step": 3246
+    },
+    {
+      "epoch": 0.5331664965820688,
+      "grad_norm": 0.5778270959854126,
+      "learning_rate": 9.191710103459461e-05,
+      "loss": 5.1512,
+      "num_input_tokens_seen": 425852928,
+      "step": 3249
+    },
+    {
+      "epoch": 0.5336588017497346,
+      "grad_norm": 0.6366815567016602,
+      "learning_rate": 9.187469406530882e-05,
+      "loss": 5.1113,
+      "num_input_tokens_seen": 426246144,
+      "step": 3252
+    },
+    {
+      "epoch": 0.5341511069174004,
+      "grad_norm": 0.5294430255889893,
+      "learning_rate": 9.183234573668638e-05,
+      "loss": 5.1288,
+      "num_input_tokens_seen": 426639360,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5346434120850663,
+      "grad_norm": 0.5754626989364624,
+      "learning_rate": 9.179005591370386e-05,
+      "loss": 5.139,
+      "num_input_tokens_seen": 427032576,
+      "step": 3258
+    },
+    {
+      "epoch": 0.535135717252732,
+      "grad_norm": 0.6076295375823975,
+      "learning_rate": 9.174782446177271e-05,
+      "loss": 5.1094,
+      "num_input_tokens_seen": 427425792,
+      "step": 3261
+    },
+    {
+      "epoch": 0.5356280224203979,
+      "grad_norm": 0.6213559508323669,
+      "learning_rate": 9.170565124673742e-05,
+      "loss": 5.1228,
+      "num_input_tokens_seen": 427819008,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5361203275880636,
+      "grad_norm": 0.5182236433029175,
+      "learning_rate": 9.166353613487377e-05,
+      "loss": 5.1282,
+      "num_input_tokens_seen": 428212224,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5366126327557295,
+      "grad_norm": 0.5979652404785156,
+      "learning_rate": 9.162147899288702e-05,
+      "loss": 5.1215,
+      "num_input_tokens_seen": 428605440,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5371049379233953,
+      "grad_norm": 0.5665757060050964,
+      "learning_rate": 9.15794796879101e-05,
+      "loss": 5.1216,
+      "num_input_tokens_seen": 428998656,
+      "step": 3273
+    },
+    {
+      "epoch": 0.537597243091061,
+      "grad_norm": 0.7053664922714233,
+      "learning_rate": 9.1537538087502e-05,
+      "loss": 5.1368,
+      "num_input_tokens_seen": 429391872,
+      "step": 3276
+    },
+    {
+      "epoch": 0.5380895482587269,
+      "grad_norm": 0.6297977566719055,
+      "learning_rate": 9.149565405964579e-05,
+      "loss": 5.1418,
+      "num_input_tokens_seen": 429785088,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5385818534263926,
+      "grad_norm": 0.6866833567619324,
+      "learning_rate": 9.145382747274708e-05,
+      "loss": 5.0998,
+      "num_input_tokens_seen": 430178304,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5390741585940585,
+      "grad_norm": 0.5878569483757019,
+      "learning_rate": 9.141205819563218e-05,
+      "loss": 5.1253,
+      "num_input_tokens_seen": 430571520,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5395664637617243,
+      "grad_norm": 0.5595548748970032,
+      "learning_rate": 9.137034609754635e-05,
+      "loss": 5.1524,
+      "num_input_tokens_seen": 430964736,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5400587689293901,
+      "grad_norm": 0.5856724381446838,
+      "learning_rate": 9.132869104815211e-05,
+      "loss": 5.1267,
+      "num_input_tokens_seen": 431357952,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5405510740970559,
+      "grad_norm": 0.5340684652328491,
+      "learning_rate": 9.128709291752767e-05,
+      "loss": 5.1085,
+      "num_input_tokens_seen": 431751168,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5410433792647217,
+      "grad_norm": 0.5203419923782349,
+      "learning_rate": 9.124555157616496e-05,
+      "loss": 5.061,
+      "num_input_tokens_seen": 432144384,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5415356844323875,
+      "grad_norm": 0.6170658469200134,
+      "learning_rate": 9.12040668949681e-05,
+      "loss": 5.098,
+      "num_input_tokens_seen": 432537600,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5420279896000534,
+      "grad_norm": 0.6464248299598694,
+      "learning_rate": 9.116263874525175e-05,
+      "loss": 5.0687,
+      "num_input_tokens_seen": 432930816,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5425202947677191,
+      "grad_norm": 0.6081656217575073,
+      "learning_rate": 9.112126699873929e-05,
+      "loss": 5.1405,
+      "num_input_tokens_seen": 433324032,
+      "step": 3306
+    },
+    {
+      "epoch": 0.543012599935385,
+      "grad_norm": 0.7273305058479309,
+      "learning_rate": 9.10799515275613e-05,
+      "loss": 5.134,
+      "num_input_tokens_seen": 433717248,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5435049051030507,
+      "grad_norm": 0.5423167943954468,
+      "learning_rate": 9.103869220425383e-05,
+      "loss": 5.0982,
+      "num_input_tokens_seen": 434110464,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5439972102707166,
+      "grad_norm": 0.6296489238739014,
+      "learning_rate": 9.099748890175672e-05,
+      "loss": 5.1654,
+      "num_input_tokens_seen": 434503680,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5444895154383824,
+      "grad_norm": 0.5735786557197571,
+      "learning_rate": 9.0956341493412e-05,
+      "loss": 5.1322,
+      "num_input_tokens_seen": 434896896,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5449818206060482,
+      "grad_norm": 0.5762762427330017,
+      "learning_rate": 9.091524985296227e-05,
+      "loss": 5.1361,
+      "num_input_tokens_seen": 435290112,
+      "step": 3321
+    },
+    {
+      "epoch": 0.545474125773714,
+      "grad_norm": 0.5767176747322083,
+      "learning_rate": 9.087421385454902e-05,
+      "loss": 5.1571,
+      "num_input_tokens_seen": 435683328,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5459664309413798,
+      "grad_norm": 0.579581081867218,
+      "learning_rate": 9.083323337271104e-05,
+      "loss": 5.1392,
+      "num_input_tokens_seen": 436076544,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5464587361090456,
+      "grad_norm": 0.6044655442237854,
+      "learning_rate": 9.079230828238284e-05,
+      "loss": 5.1106,
+      "num_input_tokens_seen": 436469760,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5469510412767113,
+      "grad_norm": 0.5802748203277588,
+      "learning_rate": 9.075143845889296e-05,
+      "loss": 5.1174,
+      "num_input_tokens_seen": 436862976,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5474433464443772,
+      "grad_norm": 0.6058889031410217,
+      "learning_rate": 9.071062377796246e-05,
+      "loss": 5.0931,
+      "num_input_tokens_seen": 437256192,
+      "step": 3336
+    },
+    {
+      "epoch": 0.547935651612043,
+      "grad_norm": 0.5531756281852722,
+      "learning_rate": 9.066986411570333e-05,
+      "loss": 5.1264,
+      "num_input_tokens_seen": 437649408,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5484279567797088,
+      "grad_norm": 0.5685104131698608,
+      "learning_rate": 9.062915934861684e-05,
+      "loss": 5.0803,
+      "num_input_tokens_seen": 438042624,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5489202619473746,
+      "grad_norm": 0.6249758005142212,
+      "learning_rate": 9.058850935359201e-05,
+      "loss": 5.1444,
+      "num_input_tokens_seen": 438435840,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5494125671150404,
+      "grad_norm": 0.6605463624000549,
+      "learning_rate": 9.054791400790408e-05,
+      "loss": 5.1445,
+      "num_input_tokens_seen": 438829056,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5499048722827062,
+      "grad_norm": 0.5551895499229431,
+      "learning_rate": 9.050737318921291e-05,
+      "loss": 5.1388,
+      "num_input_tokens_seen": 439222272,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5503971774503721,
+      "grad_norm": 0.6781189441680908,
+      "learning_rate": 9.046688677556144e-05,
+      "loss": 5.1199,
+      "num_input_tokens_seen": 439615488,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5508894826180378,
+      "grad_norm": 0.7572456002235413,
+      "learning_rate": 9.042645464537411e-05,
+      "loss": 5.0652,
+      "num_input_tokens_seen": 440008704,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5513817877857037,
+      "grad_norm": 0.642903208732605,
+      "learning_rate": 9.038607667745545e-05,
+      "loss": 5.1682,
+      "num_input_tokens_seen": 440401920,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5518740929533694,
+      "grad_norm": 0.6645106673240662,
+      "learning_rate": 9.03457527509884e-05,
+      "loss": 5.1355,
+      "num_input_tokens_seen": 440795136,
+      "step": 3363
+    },
+    {
+      "epoch": 0.5523663981210353,
+      "grad_norm": 0.7782042026519775,
+      "learning_rate": 9.03054827455329e-05,
+      "loss": 5.0755,
+      "num_input_tokens_seen": 441188352,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5528587032887011,
+      "grad_norm": 0.6609596014022827,
+      "learning_rate": 9.026526654102436e-05,
+      "loss": 5.1567,
+      "num_input_tokens_seen": 441581568,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5533510084563669,
+      "grad_norm": 0.6382880210876465,
+      "learning_rate": 9.02251040177721e-05,
+      "loss": 5.0827,
+      "num_input_tokens_seen": 441974784,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5538433136240327,
+      "grad_norm": 0.6710590720176697,
+      "learning_rate": 9.018499505645787e-05,
+      "loss": 5.0826,
+      "num_input_tokens_seen": 442368000,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5543356187916985,
+      "grad_norm": 0.6333158612251282,
+      "learning_rate": 9.014493953813449e-05,
+      "loss": 5.1026,
+      "num_input_tokens_seen": 442761216,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5548279239593643,
+      "grad_norm": 0.6261973977088928,
+      "learning_rate": 9.010493734422417e-05,
+      "loss": 5.1521,
+      "num_input_tokens_seen": 443154432,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5553202291270302,
+      "grad_norm": 0.6474428772926331,
+      "learning_rate": 9.006498835651718e-05,
+      "loss": 5.1015,
+      "num_input_tokens_seen": 443547648,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5558125342946959,
+      "grad_norm": 0.606153666973114,
+      "learning_rate": 9.002509245717025e-05,
+      "loss": 5.0685,
+      "num_input_tokens_seen": 443940864,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5563048394623618,
+      "grad_norm": 0.6537392139434814,
+      "learning_rate": 8.998524952870532e-05,
+      "loss": 5.1488,
+      "num_input_tokens_seen": 444334080,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5567971446300275,
+      "grad_norm": 0.5912271738052368,
+      "learning_rate": 8.994545945400785e-05,
+      "loss": 5.1406,
+      "num_input_tokens_seen": 444727296,
+      "step": 3393
+    },
+    {
+      "epoch": 0.5572894497976933,
+      "grad_norm": 0.5567827820777893,
+      "learning_rate": 8.990572211632556e-05,
+      "loss": 5.0851,
+      "num_input_tokens_seen": 445120512,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5577817549653592,
+      "grad_norm": 0.5647953152656555,
+      "learning_rate": 8.986603739926683e-05,
+      "loss": 5.1124,
+      "num_input_tokens_seen": 445513728,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5582740601330249,
+      "grad_norm": 0.5637276768684387,
+      "learning_rate": 8.982640518679943e-05,
+      "loss": 5.0972,
+      "num_input_tokens_seen": 445906944,
+      "step": 3402
+    },
+    {
+      "epoch": 0.5587663653006908,
+      "grad_norm": 0.5551077127456665,
+      "learning_rate": 8.978682536324898e-05,
+      "loss": 5.1073,
+      "num_input_tokens_seen": 446300160,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5592586704683565,
+      "grad_norm": 0.5078896880149841,
+      "learning_rate": 8.974729781329759e-05,
+      "loss": 5.0471,
+      "num_input_tokens_seen": 446693376,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5597509756360224,
+      "grad_norm": 0.5744684338569641,
+      "learning_rate": 8.970782242198242e-05,
+      "loss": 5.0684,
+      "num_input_tokens_seen": 447086592,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5602432808036882,
+      "grad_norm": 0.587356448173523,
+      "learning_rate": 8.966839907469425e-05,
+      "loss": 5.1192,
+      "num_input_tokens_seen": 447479808,
+      "step": 3414
+    },
+    {
+      "epoch": 0.560735585971354,
+      "grad_norm": 0.5351012349128723,
+      "learning_rate": 8.962902765717617e-05,
+      "loss": 5.0918,
+      "num_input_tokens_seen": 447873024,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5612278911390198,
+      "grad_norm": 0.5939494967460632,
+      "learning_rate": 8.958970805552213e-05,
+      "loss": 5.096,
+      "num_input_tokens_seen": 448266240,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5617201963066856,
+      "grad_norm": 0.6274368166923523,
+      "learning_rate": 8.955044015617547e-05,
+      "loss": 5.0876,
+      "num_input_tokens_seen": 448659456,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5622125014743514,
+      "grad_norm": 0.5798152089118958,
+      "learning_rate": 8.951122384592781e-05,
+      "loss": 5.0891,
+      "num_input_tokens_seen": 449052672,
+      "step": 3426
+    },
+    {
+      "epoch": 0.5627048066420173,
+      "grad_norm": 0.652629017829895,
+      "learning_rate": 8.947205901191733e-05,
+      "loss": 5.1373,
+      "num_input_tokens_seen": 449445888,
+      "step": 3429
+    },
+    {
+      "epoch": 0.563197111809683,
+      "grad_norm": 0.5612648725509644,
+      "learning_rate": 8.94329455416277e-05,
+      "loss": 5.1606,
+      "num_input_tokens_seen": 449839104,
+      "step": 3432
+    },
+    {
+      "epoch": 0.5636894169773489,
+      "grad_norm": 0.6374395489692688,
+      "learning_rate": 8.939388332288653e-05,
+      "loss": 5.1123,
+      "num_input_tokens_seen": 450232320,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5641817221450146,
+      "grad_norm": 0.5904387831687927,
+      "learning_rate": 8.93548722438641e-05,
+      "loss": 5.0859,
+      "num_input_tokens_seen": 450625536,
+      "step": 3438
+    },
+    {
+      "epoch": 0.5646740273126805,
+      "grad_norm": 0.6425766944885254,
+      "learning_rate": 8.931591219307205e-05,
+      "loss": 5.0884,
+      "num_input_tokens_seen": 451018752,
+      "step": 3441
+    },
+    {
+      "epoch": 0.5651663324803462,
+      "grad_norm": 0.6238811612129211,
+      "learning_rate": 8.927700305936195e-05,
+      "loss": 5.0994,
+      "num_input_tokens_seen": 451411968,
+      "step": 3444
+    },
+    {
+      "epoch": 0.565658637648012,
+      "grad_norm": 0.6211217045783997,
+      "learning_rate": 8.923814473192402e-05,
+      "loss": 5.0887,
+      "num_input_tokens_seen": 451805184,
+      "step": 3447
+    },
+    {
+      "epoch": 0.5661509428156779,
+      "grad_norm": 0.6308383941650391,
+      "learning_rate": 8.919933710028586e-05,
+      "loss": 5.1025,
+      "num_input_tokens_seen": 452198400,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5666432479833436,
+      "grad_norm": 0.6414578557014465,
+      "learning_rate": 8.916058005431099e-05,
+      "loss": 5.1159,
+      "num_input_tokens_seen": 452591616,
+      "step": 3453
+    },
+    {
+      "epoch": 0.5671355531510095,
+      "grad_norm": 0.6173880100250244,
+      "learning_rate": 8.912187348419765e-05,
+      "loss": 5.0787,
+      "num_input_tokens_seen": 452984832,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5676278583186752,
+      "grad_norm": 0.567276656627655,
+      "learning_rate": 8.908321728047749e-05,
+      "loss": 5.0937,
+      "num_input_tokens_seen": 453378048,
+      "step": 3459
+    },
+    {
+      "epoch": 0.5681201634863411,
+      "grad_norm": 0.7052215337753296,
+      "learning_rate": 8.904461133401418e-05,
+      "loss": 5.1027,
+      "num_input_tokens_seen": 453771264,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5686124686540069,
+      "grad_norm": 0.6379631161689758,
+      "learning_rate": 8.900605553600226e-05,
+      "loss": 5.1089,
+      "num_input_tokens_seen": 454164480,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5691047738216727,
+      "grad_norm": 0.6003096699714661,
+      "learning_rate": 8.896754977796572e-05,
+      "loss": 5.1104,
+      "num_input_tokens_seen": 454557696,
+      "step": 3468
+    },
+    {
+      "epoch": 0.5695970789893385,
+      "grad_norm": 0.798956036567688,
+      "learning_rate": 8.892909395175676e-05,
+      "loss": 5.1038,
+      "num_input_tokens_seen": 454950912,
+      "step": 3471
+    },
+    {
+      "epoch": 0.5700893841570043,
+      "grad_norm": 0.6860202550888062,
+      "learning_rate": 8.889068794955451e-05,
+      "loss": 5.0823,
+      "num_input_tokens_seen": 455344128,
+      "step": 3474
+    },
+    {
+      "epoch": 0.5705816893246701,
+      "grad_norm": 0.6046934723854065,
+      "learning_rate": 8.885233166386384e-05,
+      "loss": 5.0855,
+      "num_input_tokens_seen": 455737344,
+      "step": 3477
+    },
+    {
+      "epoch": 0.571073994492336,
+      "grad_norm": 0.5969598293304443,
+      "learning_rate": 8.881402498751399e-05,
+      "loss": 5.0868,
+      "num_input_tokens_seen": 456130560,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5715662996600017,
+      "grad_norm": 0.6891276240348816,
+      "learning_rate": 8.877576781365732e-05,
+      "loss": 5.1207,
+      "num_input_tokens_seen": 456523776,
+      "step": 3483
+    },
+    {
+      "epoch": 0.5720586048276676,
+      "grad_norm": 0.6835845112800598,
+      "learning_rate": 8.87375600357681e-05,
+      "loss": 5.0923,
+      "num_input_tokens_seen": 456916992,
+      "step": 3486
+    },
+    {
+      "epoch": 0.5725509099953333,
+      "grad_norm": 0.6710256338119507,
+      "learning_rate": 8.869940154764131e-05,
+      "loss": 5.105,
+      "num_input_tokens_seen": 457310208,
+      "step": 3489
+    },
+    {
+      "epoch": 0.5730432151629992,
+      "grad_norm": 0.5943020582199097,
+      "learning_rate": 8.866129224339131e-05,
+      "loss": 5.1724,
+      "num_input_tokens_seen": 457703424,
+      "step": 3492
+    },
+    {
+      "epoch": 0.573535520330665,
+      "grad_norm": 0.6241229176521301,
+      "learning_rate": 8.862323201745062e-05,
+      "loss": 5.0805,
+      "num_input_tokens_seen": 458096640,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5740278254983308,
+      "grad_norm": 0.6597952842712402,
+      "learning_rate": 8.85852207645687e-05,
+      "loss": 5.051,
+      "num_input_tokens_seen": 458489856,
+      "step": 3498
+    },
+    {
+      "epoch": 0.5745201306659966,
+      "grad_norm": 0.5514165759086609,
+      "learning_rate": 8.854725837981081e-05,
+      "loss": 5.1251,
+      "num_input_tokens_seen": 458883072,
+      "step": 3501
+    },
+    {
+      "epoch": 0.5750124358336623,
+      "grad_norm": 0.5667795538902283,
+      "learning_rate": 8.850934475855665e-05,
+      "loss": 5.0993,
+      "num_input_tokens_seen": 459276288,
+      "step": 3504
+    },
+    {
+      "epoch": 0.5755047410013282,
+      "grad_norm": 0.5708199739456177,
+      "learning_rate": 8.847147979649926e-05,
+      "loss": 5.1099,
+      "num_input_tokens_seen": 459669504,
+      "step": 3507
+    },
+    {
+      "epoch": 0.575997046168994,
+      "grad_norm": 0.584382176399231,
+      "learning_rate": 8.843366338964375e-05,
+      "loss": 5.0977,
+      "num_input_tokens_seen": 460062720,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5764893513366598,
+      "grad_norm": 0.5634648203849792,
+      "learning_rate": 8.839589543430617e-05,
+      "loss": 5.0354,
+      "num_input_tokens_seen": 460455936,
+      "step": 3513
+    },
+    {
+      "epoch": 0.5769816565043256,
+      "grad_norm": 0.7195287942886353,
+      "learning_rate": 8.835817582711223e-05,
+      "loss": 5.1015,
+      "num_input_tokens_seen": 460849152,
+      "step": 3516
+    },
+    {
+      "epoch": 0.5774739616719914,
+      "grad_norm": 0.5957933664321899,
+      "learning_rate": 8.832050446499615e-05,
+      "loss": 5.0715,
+      "num_input_tokens_seen": 461242368,
+      "step": 3519
+    },
+    {
+      "epoch": 0.5779662668396572,
+      "grad_norm": 0.6452832818031311,
+      "learning_rate": 8.828288124519953e-05,
+      "loss": 5.1224,
+      "num_input_tokens_seen": 461635584,
+      "step": 3522
+    },
+    {
+      "epoch": 0.5784585720073231,
+      "grad_norm": 0.6192216277122498,
+      "learning_rate": 8.824530606527006e-05,
+      "loss": 5.0501,
+      "num_input_tokens_seen": 462028800,
+      "step": 3525
+    },
+    {
+      "epoch": 0.5789508771749888,
+      "grad_norm": 0.6137304902076721,
+      "learning_rate": 8.820777882306049e-05,
+      "loss": 5.1225,
+      "num_input_tokens_seen": 462422016,
+      "step": 3528
+    },
+    {
+      "epoch": 0.5794431823426547,
+      "grad_norm": 0.5840086936950684,
+      "learning_rate": 8.81702994167273e-05,
+      "loss": 5.095,
+      "num_input_tokens_seen": 462815232,
+      "step": 3531
+    },
+    {
+      "epoch": 0.5799354875103204,
+      "grad_norm": 0.6205256581306458,
+      "learning_rate": 8.81328677447297e-05,
+      "loss": 5.094,
+      "num_input_tokens_seen": 463208448,
+      "step": 3534
+    },
+    {
+      "epoch": 0.5804277926779863,
+      "grad_norm": 0.5789561867713928,
+      "learning_rate": 8.809548370582834e-05,
+      "loss": 5.127,
+      "num_input_tokens_seen": 463601664,
+      "step": 3537
+    },
+    {
+      "epoch": 0.5809200978456521,
+      "grad_norm": 0.5926523208618164,
+      "learning_rate": 8.805814719908426e-05,
+      "loss": 5.0785,
+      "num_input_tokens_seen": 463994880,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5814124030133179,
+      "grad_norm": 0.5335286855697632,
+      "learning_rate": 8.80208581238577e-05,
+      "loss": 5.1154,
+      "num_input_tokens_seen": 464388096,
+      "step": 3543
+    },
+    {
+      "epoch": 0.5819047081809837,
+      "grad_norm": 0.4992382824420929,
+      "learning_rate": 8.798361637980696e-05,
+      "loss": 5.0659,
+      "num_input_tokens_seen": 464781312,
+      "step": 3546
+    },
+    {
+      "epoch": 0.5823970133486495,
+      "grad_norm": 0.5407050251960754,
+      "learning_rate": 8.794642186688725e-05,
+      "loss": 5.0761,
+      "num_input_tokens_seen": 465174528,
+      "step": 3549
+    },
+    {
+      "epoch": 0.5828893185163153,
+      "grad_norm": 0.5841310024261475,
+      "learning_rate": 8.79092744853496e-05,
+      "loss": 5.0825,
+      "num_input_tokens_seen": 465567744,
+      "step": 3552
+    },
+    {
+      "epoch": 0.583381623683981,
+      "grad_norm": 0.5791674852371216,
+      "learning_rate": 8.787217413573975e-05,
+      "loss": 5.1143,
+      "num_input_tokens_seen": 465960960,
+      "step": 3555
+    },
+    {
+      "epoch": 0.5838739288516469,
+      "grad_norm": 0.6036920547485352,
+      "learning_rate": 8.783512071889697e-05,
+      "loss": 5.1271,
+      "num_input_tokens_seen": 466354176,
+      "step": 3558
+    },
+    {
+      "epoch": 0.5843662340193128,
+      "grad_norm": 0.5951531529426575,
+      "learning_rate": 8.779811413595294e-05,
+      "loss": 5.0859,
+      "num_input_tokens_seen": 466747392,
+      "step": 3561
+    },
+    {
+      "epoch": 0.5848585391869785,
+      "grad_norm": 0.7706826329231262,
+      "learning_rate": 8.776115428833078e-05,
+      "loss": 5.0744,
+      "num_input_tokens_seen": 467140608,
+      "step": 3564
+    },
+    {
+      "epoch": 0.5853508443546444,
+      "grad_norm": 0.6318513751029968,
+      "learning_rate": 8.772424107774375e-05,
+      "loss": 5.0896,
+      "num_input_tokens_seen": 467533824,
+      "step": 3567
+    },
+    {
+      "epoch": 0.5858431495223101,
+      "grad_norm": 0.6089410185813904,
+      "learning_rate": 8.768737440619431e-05,
+      "loss": 5.0848,
+      "num_input_tokens_seen": 467927040,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5863354546899759,
+      "grad_norm": 0.7453569173812866,
+      "learning_rate": 8.765055417597291e-05,
+      "loss": 5.0298,
+      "num_input_tokens_seen": 468320256,
+      "step": 3573
+    },
+    {
+      "epoch": 0.5868277598576418,
+      "grad_norm": 0.6902857422828674,
+      "learning_rate": 8.761378028965703e-05,
+      "loss": 5.0966,
+      "num_input_tokens_seen": 468713472,
+      "step": 3576
+    },
+    {
+      "epoch": 0.5873200650253075,
+      "grad_norm": 0.5212551951408386,
+      "learning_rate": 8.757705265010996e-05,
+      "loss": 5.1311,
+      "num_input_tokens_seen": 469106688,
+      "step": 3579
+    },
+    {
+      "epoch": 0.5878123701929734,
+      "grad_norm": 0.679226815700531,
+      "learning_rate": 8.754037116047984e-05,
+      "loss": 5.0775,
+      "num_input_tokens_seen": 469499904,
+      "step": 3582
+    },
+    {
+      "epoch": 0.5883046753606391,
+      "grad_norm": 0.600480318069458,
+      "learning_rate": 8.750373572419852e-05,
+      "loss": 5.0304,
+      "num_input_tokens_seen": 469893120,
+      "step": 3585
+    },
+    {
+      "epoch": 0.588796980528305,
+      "grad_norm": 0.557918906211853,
+      "learning_rate": 8.746714624498048e-05,
+      "loss": 5.082,
+      "num_input_tokens_seen": 470286336,
+      "step": 3588
+    },
+    {
+      "epoch": 0.5892892856959708,
+      "grad_norm": 0.6487769484519958,
+      "learning_rate": 8.743060262682181e-05,
+      "loss": 5.0953,
+      "num_input_tokens_seen": 470679552,
+      "step": 3591
+    },
+    {
+      "epoch": 0.5897815908636366,
+      "grad_norm": 0.5848097801208496,
+      "learning_rate": 8.739410477399918e-05,
+      "loss": 5.1032,
+      "num_input_tokens_seen": 471072768,
+      "step": 3594
+    },
+    {
+      "epoch": 0.5902738960313024,
+      "grad_norm": 0.6712628602981567,
+      "learning_rate": 8.735765259106869e-05,
+      "loss": 5.104,
+      "num_input_tokens_seen": 471465984,
+      "step": 3597
+    },
+    {
+      "epoch": 0.5907662011989682,
+      "grad_norm": 0.7293500304222107,
+      "learning_rate": 8.73212459828649e-05,
+      "loss": 5.0618,
+      "num_input_tokens_seen": 471859200,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5907662011989682,
+      "eval_accuracy": 0.21986158606090214,
+      "eval_loss": 5.362356185913086,
+      "eval_runtime": 110.5722,
+      "eval_samples_per_second": 2.713,
+      "eval_steps_per_second": 1.357,
+      "num_input_tokens_seen": 471859200,
+      "step": 3600
+    },
+    {
+      "epoch": 0.591258506366634,
+      "grad_norm": 0.6076777577400208,
+      "learning_rate": 8.728488485449973e-05,
+      "loss": 5.1032,
+      "num_input_tokens_seen": 472252416,
+      "step": 3603
+    },
+    {
+      "epoch": 0.5917508115342999,
+      "grad_norm": 0.5898135304450989,
+      "learning_rate": 8.724856911136155e-05,
+      "loss": 5.1125,
+      "num_input_tokens_seen": 472645632,
+      "step": 3606
+    },
+    {
+      "epoch": 0.5922431167019656,
+      "grad_norm": 0.6582708358764648,
+      "learning_rate": 8.721229865911391e-05,
+      "loss": 5.0623,
+      "num_input_tokens_seen": 473038848,
+      "step": 3609
+    },
+    {
+      "epoch": 0.5927354218696315,
+      "grad_norm": 0.6800294518470764,
+      "learning_rate": 8.717607340369476e-05,
+      "loss": 5.0556,
+      "num_input_tokens_seen": 473432064,
+      "step": 3612
+    },
+    {
+      "epoch": 0.5932277270372972,
+      "grad_norm": 0.6523749828338623,
+      "learning_rate": 8.713989325131527e-05,
+      "loss": 5.0741,
+      "num_input_tokens_seen": 473825280,
+      "step": 3615
+    },
+    {
+      "epoch": 0.593720032204963,
+      "grad_norm": 0.759653627872467,
+      "learning_rate": 8.710375810845887e-05,
+      "loss": 5.0508,
+      "num_input_tokens_seen": 474218496,
+      "step": 3618
+    },
+    {
+      "epoch": 0.5942123373726289,
+      "grad_norm": 0.6974580883979797,
+      "learning_rate": 8.706766788188021e-05,
+      "loss": 5.0766,
+      "num_input_tokens_seen": 474611712,
+      "step": 3621
+    },
+    {
+      "epoch": 0.5947046425402946,
+      "grad_norm": 0.5986417531967163,
+      "learning_rate": 8.703162247860416e-05,
+      "loss": 5.0783,
+      "num_input_tokens_seen": 475004928,
+      "step": 3624
+    },
+    {
+      "epoch": 0.5951969477079605,
+      "grad_norm": 0.8493059277534485,
+      "learning_rate": 8.699562180592481e-05,
+      "loss": 5.0947,
+      "num_input_tokens_seen": 475398144,
+      "step": 3627
+    },
+    {
+      "epoch": 0.5956892528756262,
+      "grad_norm": 0.6575183272361755,
+      "learning_rate": 8.695966577140451e-05,
+      "loss": 5.108,
+      "num_input_tokens_seen": 475791360,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5961815580432921,
+      "grad_norm": 0.7009572982788086,
+      "learning_rate": 8.692375428287271e-05,
+      "loss": 5.099,
+      "num_input_tokens_seen": 476184576,
+      "step": 3633
+    },
+    {
+      "epoch": 0.5966738632109579,
+      "grad_norm": 0.8006022572517395,
+      "learning_rate": 8.68878872484252e-05,
+      "loss": 5.0507,
+      "num_input_tokens_seen": 476577792,
+      "step": 3636
+    },
+    {
+      "epoch": 0.5971661683786237,
+      "grad_norm": 0.5653481483459473,
+      "learning_rate": 8.685206457642292e-05,
+      "loss": 5.0688,
+      "num_input_tokens_seen": 476971008,
+      "step": 3639
+    },
+    {
+      "epoch": 0.5976584735462895,
+      "grad_norm": 0.7987883687019348,
+      "learning_rate": 8.681628617549114e-05,
+      "loss": 5.0972,
+      "num_input_tokens_seen": 477364224,
+      "step": 3642
+    },
+    {
+      "epoch": 0.5981507787139553,
+      "grad_norm": 0.6354634165763855,
+      "learning_rate": 8.678055195451837e-05,
+      "loss": 5.0745,
+      "num_input_tokens_seen": 477757440,
+      "step": 3645
+    },
+    {
+      "epoch": 0.5986430838816211,
+      "grad_norm": 0.6245685815811157,
+      "learning_rate": 8.67448618226554e-05,
+      "loss": 5.07,
+      "num_input_tokens_seen": 478150656,
+      "step": 3648
+    },
+    {
+      "epoch": 0.599135389049287,
+      "grad_norm": 0.6478342413902283,
+      "learning_rate": 8.670921568931434e-05,
+      "loss": 5.1107,
+      "num_input_tokens_seen": 478543872,
+      "step": 3651
+    },
+    {
+      "epoch": 0.5996276942169527,
+      "grad_norm": 0.6710765957832336,
+      "learning_rate": 8.667361346416774e-05,
+      "loss": 5.0968,
+      "num_input_tokens_seen": 478937088,
+      "step": 3654
+    },
+    {
+      "epoch": 0.6001199993846186,
+      "grad_norm": 0.5611339807510376,
+      "learning_rate": 8.663805505714746e-05,
+      "loss": 5.0798,
+      "num_input_tokens_seen": 479330304,
+      "step": 3657
+    },
+    {
+      "epoch": 0.6006123045522843,
+      "grad_norm": 0.593101441860199,
+      "learning_rate": 8.660254037844386e-05,
+      "loss": 5.1296,
+      "num_input_tokens_seen": 479723520,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6011046097199502,
+      "grad_norm": 0.7235480546951294,
+      "learning_rate": 8.656706933850477e-05,
+      "loss": 5.0782,
+      "num_input_tokens_seen": 480116736,
+      "step": 3663
+    },
+    {
+      "epoch": 0.6015969148876159,
+      "grad_norm": 0.5439255237579346,
+      "learning_rate": 8.653164184803456e-05,
+      "loss": 5.0872,
+      "num_input_tokens_seen": 480509952,
+      "step": 3666
+    },
+    {
+      "epoch": 0.6020892200552818,
+      "grad_norm": 0.5817519426345825,
+      "learning_rate": 8.649625781799325e-05,
+      "loss": 5.1071,
+      "num_input_tokens_seen": 480903168,
+      "step": 3669
+    },
+    {
+      "epoch": 0.6025815252229476,
+      "grad_norm": 0.6808876395225525,
+      "learning_rate": 8.646091715959547e-05,
+      "loss": 5.0952,
+      "num_input_tokens_seen": 481296384,
+      "step": 3672
+    },
+    {
+      "epoch": 0.6030738303906134,
+      "grad_norm": 0.6932339668273926,
+      "learning_rate": 8.642561978430955e-05,
+      "loss": 5.0893,
+      "num_input_tokens_seen": 481689600,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6035661355582792,
+      "grad_norm": 0.6061970591545105,
+      "learning_rate": 8.63903656038567e-05,
+      "loss": 5.0866,
+      "num_input_tokens_seen": 482082816,
+      "step": 3678
+    },
+    {
+      "epoch": 0.604058440725945,
+      "grad_norm": 0.5691749453544617,
+      "learning_rate": 8.635515453020989e-05,
+      "loss": 5.0906,
+      "num_input_tokens_seen": 482476032,
+      "step": 3681
+    },
+    {
+      "epoch": 0.6045507458936108,
+      "grad_norm": 0.5950030088424683,
+      "learning_rate": 8.631998647559312e-05,
+      "loss": 5.0537,
+      "num_input_tokens_seen": 482869248,
+      "step": 3684
+    },
+    {
+      "epoch": 0.6050430510612766,
+      "grad_norm": 0.6372474431991577,
+      "learning_rate": 8.628486135248037e-05,
+      "loss": 5.0622,
+      "num_input_tokens_seen": 483262464,
+      "step": 3687
+    },
+    {
+      "epoch": 0.6055353562289424,
+      "grad_norm": 0.7356825470924377,
+      "learning_rate": 8.624977907359473e-05,
+      "loss": 5.1061,
+      "num_input_tokens_seen": 483655680,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6060276613966082,
+      "grad_norm": 0.604340136051178,
+      "learning_rate": 8.621473955190753e-05,
+      "loss": 5.0179,
+      "num_input_tokens_seen": 484048896,
+      "step": 3693
+    },
+    {
+      "epoch": 0.606519966564274,
+      "grad_norm": 0.665763258934021,
+      "learning_rate": 8.617974270063731e-05,
+      "loss": 5.0976,
+      "num_input_tokens_seen": 484442112,
+      "step": 3696
+    },
+    {
+      "epoch": 0.6070122717319398,
+      "grad_norm": 0.5729550719261169,
+      "learning_rate": 8.614478843324907e-05,
+      "loss": 5.1136,
+      "num_input_tokens_seen": 484835328,
+      "step": 3699
+    },
+    {
+      "epoch": 0.6075045768996057,
+      "grad_norm": 0.6027372479438782,
+      "learning_rate": 8.61098766634533e-05,
+      "loss": 5.0495,
+      "num_input_tokens_seen": 485228544,
+      "step": 3702
+    },
+    {
+      "epoch": 0.6079968820672714,
+      "grad_norm": 0.5599421858787537,
+      "learning_rate": 8.607500730520499e-05,
+      "loss": 5.1028,
+      "num_input_tokens_seen": 485621760,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6084891872349373,
+      "grad_norm": 0.7194706201553345,
+      "learning_rate": 8.604018027270296e-05,
+      "loss": 5.0837,
+      "num_input_tokens_seen": 486014976,
+      "step": 3708
+    },
+    {
+      "epoch": 0.608981492402603,
+      "grad_norm": 0.5925415754318237,
+      "learning_rate": 8.600539548038875e-05,
+      "loss": 5.0818,
+      "num_input_tokens_seen": 486408192,
+      "step": 3711
+    },
+    {
+      "epoch": 0.6094737975702689,
+      "grad_norm": 0.5674649477005005,
+      "learning_rate": 8.597065284294591e-05,
+      "loss": 5.1209,
+      "num_input_tokens_seen": 486801408,
+      "step": 3714
+    },
+    {
+      "epoch": 0.6099661027379347,
+      "grad_norm": 0.5196229815483093,
+      "learning_rate": 8.5935952275299e-05,
+      "loss": 5.0793,
+      "num_input_tokens_seen": 487194624,
+      "step": 3717
+    },
+    {
+      "epoch": 0.6104584079056005,
+      "grad_norm": 0.5827131271362305,
+      "learning_rate": 8.590129369261278e-05,
+      "loss": 5.0466,
+      "num_input_tokens_seen": 487587840,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6109507130732663,
+      "grad_norm": 0.6285468339920044,
+      "learning_rate": 8.586667701029127e-05,
+      "loss": 5.1242,
+      "num_input_tokens_seen": 487981056,
+      "step": 3723
+    },
+    {
+      "epoch": 0.6114430182409321,
+      "grad_norm": 0.5768407583236694,
+      "learning_rate": 8.583210214397702e-05,
+      "loss": 5.0845,
+      "num_input_tokens_seen": 488374272,
+      "step": 3726
+    },
+    {
+      "epoch": 0.6119353234085979,
+      "grad_norm": 0.6446322798728943,
+      "learning_rate": 8.57975690095501e-05,
+      "loss": 5.0994,
+      "num_input_tokens_seen": 488767488,
+      "step": 3729
+    },
+    {
+      "epoch": 0.6124276285762638,
+      "grad_norm": 0.5369075536727905,
+      "learning_rate": 8.57630775231273e-05,
+      "loss": 5.0974,
+      "num_input_tokens_seen": 489160704,
+      "step": 3732
+    },
+    {
+      "epoch": 0.6129199337439295,
+      "grad_norm": 0.6329900026321411,
+      "learning_rate": 8.572862760106127e-05,
+      "loss": 5.0761,
+      "num_input_tokens_seen": 489553920,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6134122389115954,
+      "grad_norm": 0.578373908996582,
+      "learning_rate": 8.569421915993972e-05,
+      "loss": 5.086,
+      "num_input_tokens_seen": 489947136,
+      "step": 3738
+    },
+    {
+      "epoch": 0.6139045440792611,
+      "grad_norm": 0.598923921585083,
+      "learning_rate": 8.565985211658447e-05,
+      "loss": 5.055,
+      "num_input_tokens_seen": 490340352,
+      "step": 3741
+    },
+    {
+      "epoch": 0.614396849246927,
+      "grad_norm": 0.6580232977867126,
+      "learning_rate": 8.562552638805071e-05,
+      "loss": 5.0599,
+      "num_input_tokens_seen": 490733568,
+      "step": 3744
+    },
+    {
+      "epoch": 0.6148891544145928,
+      "grad_norm": 0.5640982985496521,
+      "learning_rate": 8.559124189162605e-05,
+      "loss": 5.0757,
+      "num_input_tokens_seen": 491126784,
+      "step": 3747
+    },
+    {
+      "epoch": 0.6153814595822585,
+      "grad_norm": 0.7263691425323486,
+      "learning_rate": 8.555699854482974e-05,
+      "loss": 5.0805,
+      "num_input_tokens_seen": 491520000,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6158737647499244,
+      "grad_norm": 0.5789725184440613,
+      "learning_rate": 8.552279626541192e-05,
+      "loss": 5.0723,
+      "num_input_tokens_seen": 491913216,
+      "step": 3753
+    },
+    {
+      "epoch": 0.6163660699175901,
+      "grad_norm": 0.6067011952400208,
+      "learning_rate": 8.548863497135262e-05,
+      "loss": 5.0662,
+      "num_input_tokens_seen": 492306432,
+      "step": 3756
+    },
+    {
+      "epoch": 0.616858375085256,
+      "grad_norm": 0.6231025457382202,
+      "learning_rate": 8.545451458086107e-05,
+      "loss": 5.1009,
+      "num_input_tokens_seen": 492699648,
+      "step": 3759
+    },
+    {
+      "epoch": 0.6173506802529218,
+      "grad_norm": 0.7004355192184448,
+      "learning_rate": 8.542043501237481e-05,
+      "loss": 5.0582,
+      "num_input_tokens_seen": 493092864,
+      "step": 3762
+    },
+    {
+      "epoch": 0.6178429854205876,
+      "grad_norm": 0.6546366810798645,
+      "learning_rate": 8.53863961845589e-05,
+      "loss": 5.0931,
+      "num_input_tokens_seen": 493486080,
+      "step": 3765
+    },
+    {
+      "epoch": 0.6183352905882534,
+      "grad_norm": 0.6404768228530884,
+      "learning_rate": 8.535239801630506e-05,
+      "loss": 5.1405,
+      "num_input_tokens_seen": 493879296,
+      "step": 3768
+    },
+    {
+      "epoch": 0.6188275957559192,
+      "grad_norm": 0.6902008056640625,
+      "learning_rate": 8.531844042673096e-05,
+      "loss": 5.0707,
+      "num_input_tokens_seen": 494272512,
+      "step": 3771
+    },
+    {
+      "epoch": 0.619319900923585,
+      "grad_norm": 0.6566780209541321,
+      "learning_rate": 8.528452333517929e-05,
+      "loss": 5.1154,
+      "num_input_tokens_seen": 494665728,
+      "step": 3774
+    },
+    {
+      "epoch": 0.6198122060912508,
+      "grad_norm": 0.6058692932128906,
+      "learning_rate": 8.525064666121706e-05,
+      "loss": 5.0052,
+      "num_input_tokens_seen": 495058944,
+      "step": 3777
+    },
+    {
+      "epoch": 0.6203045112589166,
+      "grad_norm": 0.6108487844467163,
+      "learning_rate": 8.521681032463467e-05,
+      "loss": 5.0998,
+      "num_input_tokens_seen": 495452160,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6207968164265825,
+      "grad_norm": 0.6402431130409241,
+      "learning_rate": 8.518301424544526e-05,
+      "loss": 5.0671,
+      "num_input_tokens_seen": 495845376,
+      "step": 3783
+    },
+    {
+      "epoch": 0.6212891215942482,
+      "grad_norm": 0.6639608144760132,
+      "learning_rate": 8.514925834388382e-05,
+      "loss": 5.0585,
+      "num_input_tokens_seen": 496238592,
+      "step": 3786
+    },
+    {
+      "epoch": 0.6217814267619141,
+      "grad_norm": 0.6030707955360413,
+      "learning_rate": 8.511554254040647e-05,
+      "loss": 5.1078,
+      "num_input_tokens_seen": 496631808,
+      "step": 3789
+    },
+    {
+      "epoch": 0.6222737319295798,
+      "grad_norm": 0.6333598494529724,
+      "learning_rate": 8.508186675568954e-05,
+      "loss": 5.0646,
+      "num_input_tokens_seen": 497025024,
+      "step": 3792
+    },
+    {
+      "epoch": 0.6227660370972457,
+      "grad_norm": 0.6044871211051941,
+      "learning_rate": 8.504823091062899e-05,
+      "loss": 5.1171,
+      "num_input_tokens_seen": 497418240,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6232583422649115,
+      "grad_norm": 0.6012600064277649,
+      "learning_rate": 8.501463492633939e-05,
+      "loss": 5.045,
+      "num_input_tokens_seen": 497811456,
+      "step": 3798
+    },
+    {
+      "epoch": 0.6237506474325772,
+      "grad_norm": 0.6684429049491882,
+      "learning_rate": 8.49810787241534e-05,
+      "loss": 5.0726,
+      "num_input_tokens_seen": 498204672,
+      "step": 3801
+    },
+    {
+      "epoch": 0.6242429526002431,
+      "grad_norm": 0.5676344633102417,
+      "learning_rate": 8.494756222562075e-05,
+      "loss": 5.0832,
+      "num_input_tokens_seen": 498597888,
+      "step": 3804
+    },
+    {
+      "epoch": 0.6247352577679088,
+      "grad_norm": 0.606695294380188,
+      "learning_rate": 8.491408535250763e-05,
+      "loss": 5.0568,
+      "num_input_tokens_seen": 498991104,
+      "step": 3807
+    },
+    {
+      "epoch": 0.6252275629355747,
+      "grad_norm": 0.5882166028022766,
+      "learning_rate": 8.488064802679595e-05,
+      "loss": 5.0352,
+      "num_input_tokens_seen": 499384320,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6257198681032405,
+      "grad_norm": 0.6526467800140381,
+      "learning_rate": 8.484725017068234e-05,
+      "loss": 5.0533,
+      "num_input_tokens_seen": 499777536,
+      "step": 3813
+    },
+    {
+      "epoch": 0.6262121732709063,
+      "grad_norm": 0.593462347984314,
+      "learning_rate": 8.48138917065777e-05,
+      "loss": 5.0718,
+      "num_input_tokens_seen": 500170752,
+      "step": 3816
+    },
+    {
+      "epoch": 0.6267044784385721,
+      "grad_norm": 0.5729905366897583,
+      "learning_rate": 8.478057255710627e-05,
+      "loss": 5.0125,
+      "num_input_tokens_seen": 500563968,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6271967836062379,
+      "grad_norm": 0.6892577409744263,
+      "learning_rate": 8.474729264510482e-05,
+      "loss": 5.0931,
+      "num_input_tokens_seen": 500957184,
+      "step": 3822
+    },
+    {
+      "epoch": 0.6276890887739037,
+      "grad_norm": 0.5425217151641846,
+      "learning_rate": 8.471405189362207e-05,
+      "loss": 5.0371,
+      "num_input_tokens_seen": 501350400,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6281813939415696,
+      "grad_norm": 0.6370276808738708,
+      "learning_rate": 8.468085022591781e-05,
+      "loss": 5.0494,
+      "num_input_tokens_seen": 501743616,
+      "step": 3828
+    },
+    {
+      "epoch": 0.6286736991092353,
+      "grad_norm": 0.6339290142059326,
+      "learning_rate": 8.464768756546222e-05,
+      "loss": 5.0355,
+      "num_input_tokens_seen": 502136832,
+      "step": 3831
+    },
+    {
+      "epoch": 0.6291660042769012,
+      "grad_norm": 0.5908637046813965,
+      "learning_rate": 8.461456383593512e-05,
+      "loss": 5.069,
+      "num_input_tokens_seen": 502530048,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6296583094445669,
+      "grad_norm": 0.6436389088630676,
+      "learning_rate": 8.458147896122517e-05,
+      "loss": 5.0195,
+      "num_input_tokens_seen": 502923264,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6301506146122328,
+      "grad_norm": 0.5766847133636475,
+      "learning_rate": 8.454843286542926e-05,
+      "loss": 5.0878,
+      "num_input_tokens_seen": 503316480,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6306429197798986,
+      "grad_norm": 0.5753293037414551,
+      "learning_rate": 8.451542547285164e-05,
+      "loss": 5.1015,
+      "num_input_tokens_seen": 503709696,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6311352249475644,
+      "grad_norm": 0.6848548650741577,
+      "learning_rate": 8.448245670800332e-05,
+      "loss": 5.0775,
+      "num_input_tokens_seen": 504102912,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6316275301152302,
+      "grad_norm": 0.5692152380943298,
+      "learning_rate": 8.444952649560123e-05,
+      "loss": 5.0544,
+      "num_input_tokens_seen": 504496128,
+      "step": 3849
+    },
+    {
+      "epoch": 0.632119835282896,
+      "grad_norm": 0.7179746627807617,
+      "learning_rate": 8.441663476056757e-05,
+      "loss": 5.0092,
+      "num_input_tokens_seen": 504889344,
+      "step": 3852
+    },
+    {
+      "epoch": 0.6326121404505618,
+      "grad_norm": 0.7406941056251526,
+      "learning_rate": 8.438378142802908e-05,
+      "loss": 5.0623,
+      "num_input_tokens_seen": 505282560,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6331044456182277,
+      "grad_norm": 0.5824952721595764,
+      "learning_rate": 8.43509664233163e-05,
+      "loss": 5.1089,
+      "num_input_tokens_seen": 505675776,
+      "step": 3858
+    },
+    {
+      "epoch": 0.6335967507858934,
+      "grad_norm": 0.7226423621177673,
+      "learning_rate": 8.431818967196287e-05,
+      "loss": 5.1065,
+      "num_input_tokens_seen": 506068992,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6340890559535592,
+      "grad_norm": 0.6426088809967041,
+      "learning_rate": 8.42854510997049e-05,
+      "loss": 5.0762,
+      "num_input_tokens_seen": 506462208,
+      "step": 3864
+    },
+    {
+      "epoch": 0.634581361121225,
+      "grad_norm": 0.7668753266334534,
+      "learning_rate": 8.425275063248005e-05,
+      "loss": 5.0407,
+      "num_input_tokens_seen": 506855424,
+      "step": 3867
+    },
+    {
+      "epoch": 0.6350736662888908,
+      "grad_norm": 0.7967394590377808,
+      "learning_rate": 8.422008819642705e-05,
+      "loss": 5.0791,
+      "num_input_tokens_seen": 507248640,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6355659714565566,
+      "grad_norm": 0.6385906934738159,
+      "learning_rate": 8.418746371788493e-05,
+      "loss": 5.0588,
+      "num_input_tokens_seen": 507641856,
+      "step": 3873
+    },
+    {
+      "epoch": 0.6360582766242224,
+      "grad_norm": 0.6711165308952332,
+      "learning_rate": 8.415487712339226e-05,
+      "loss": 5.0421,
+      "num_input_tokens_seen": 508035072,
+      "step": 3876
+    },
+    {
+      "epoch": 0.6365505817918883,
+      "grad_norm": 0.6327312588691711,
+      "learning_rate": 8.412232833968649e-05,
+      "loss": 5.0499,
+      "num_input_tokens_seen": 508428288,
+      "step": 3879
+    },
+    {
+      "epoch": 0.637042886959554,
+      "grad_norm": 0.5863744616508484,
+      "learning_rate": 8.408981729370331e-05,
+      "loss": 5.0552,
+      "num_input_tokens_seen": 508821504,
+      "step": 3882
+    },
+    {
+      "epoch": 0.6375351921272199,
+      "grad_norm": 0.6245995163917542,
+      "learning_rate": 8.405734391257592e-05,
+      "loss": 5.0508,
+      "num_input_tokens_seen": 509214720,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6380274972948856,
+      "grad_norm": 0.7355886697769165,
+      "learning_rate": 8.40249081236343e-05,
+      "loss": 5.0546,
+      "num_input_tokens_seen": 509607936,
+      "step": 3888
+    },
+    {
+      "epoch": 0.6385198024625515,
+      "grad_norm": 0.5727904438972473,
+      "learning_rate": 8.399250985440458e-05,
+      "loss": 5.0456,
+      "num_input_tokens_seen": 510001152,
+      "step": 3891
+    },
+    {
+      "epoch": 0.6390121076302173,
+      "grad_norm": 0.6342762112617493,
+      "learning_rate": 8.396014903260839e-05,
+      "loss": 5.0593,
+      "num_input_tokens_seen": 510394368,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6395044127978831,
+      "grad_norm": 0.7297379970550537,
+      "learning_rate": 8.392782558616211e-05,
+      "loss": 5.0787,
+      "num_input_tokens_seen": 510787584,
+      "step": 3897
+    },
+    {
+      "epoch": 0.6399967179655489,
+      "grad_norm": 0.5754993557929993,
+      "learning_rate": 8.389553944317623e-05,
+      "loss": 5.0372,
+      "num_input_tokens_seen": 511180800,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6404890231332147,
+      "grad_norm": 0.6071540713310242,
+      "learning_rate": 8.386329053195467e-05,
+      "loss": 5.0784,
+      "num_input_tokens_seen": 511574016,
+      "step": 3903
+    },
+    {
+      "epoch": 0.6409813283008805,
+      "grad_norm": 0.5957722067832947,
+      "learning_rate": 8.383107878099417e-05,
+      "loss": 5.067,
+      "num_input_tokens_seen": 511967232,
+      "step": 3906
+    },
+    {
+      "epoch": 0.6414736334685464,
+      "grad_norm": 0.5961174964904785,
+      "learning_rate": 8.379890411898351e-05,
+      "loss": 5.0588,
+      "num_input_tokens_seen": 512360448,
+      "step": 3909
+    },
+    {
+      "epoch": 0.6419659386362121,
+      "grad_norm": 0.6179754734039307,
+      "learning_rate": 8.376676647480295e-05,
+      "loss": 5.0564,
+      "num_input_tokens_seen": 512753664,
+      "step": 3912
+    },
+    {
+      "epoch": 0.642458243803878,
+      "grad_norm": 0.5459384322166443,
+      "learning_rate": 8.373466577752348e-05,
+      "loss": 5.0854,
+      "num_input_tokens_seen": 513146880,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6429505489715437,
+      "grad_norm": 0.7513952255249023,
+      "learning_rate": 8.370260195640626e-05,
+      "loss": 5.089,
+      "num_input_tokens_seen": 513540096,
+      "step": 3918
+    },
+    {
+      "epoch": 0.6434428541392095,
+      "grad_norm": 0.6562470197677612,
+      "learning_rate": 8.367057494090192e-05,
+      "loss": 5.052,
+      "num_input_tokens_seen": 513933312,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6439351593068754,
+      "grad_norm": 0.6965239644050598,
+      "learning_rate": 8.363858466064986e-05,
+      "loss": 5.0638,
+      "num_input_tokens_seen": 514326528,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6444274644745411,
+      "grad_norm": 0.5959606766700745,
+      "learning_rate": 8.360663104547769e-05,
+      "loss": 5.0521,
+      "num_input_tokens_seen": 514719744,
+      "step": 3927
+    },
+    {
+      "epoch": 0.644919769642207,
+      "grad_norm": 0.54644376039505,
+      "learning_rate": 8.357471402540053e-05,
+      "loss": 5.0629,
+      "num_input_tokens_seen": 515112960,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6454120748098727,
+      "grad_norm": 0.5704362988471985,
+      "learning_rate": 8.354283353062033e-05,
+      "loss": 5.0272,
+      "num_input_tokens_seen": 515506176,
+      "step": 3933
+    },
+    {
+      "epoch": 0.6459043799775386,
+      "grad_norm": 0.522625744342804,
+      "learning_rate": 8.351098949152536e-05,
+      "loss": 5.0068,
+      "num_input_tokens_seen": 515899392,
+      "step": 3936
+    },
+    {
+      "epoch": 0.6463966851452044,
+      "grad_norm": 0.6017957329750061,
+      "learning_rate": 8.347918183868937e-05,
+      "loss": 5.0166,
+      "num_input_tokens_seen": 516292608,
+      "step": 3939
+    },
+    {
+      "epoch": 0.6468889903128702,
+      "grad_norm": 0.6562044024467468,
+      "learning_rate": 8.344741050287123e-05,
+      "loss": 5.0237,
+      "num_input_tokens_seen": 516685824,
+      "step": 3942
+    },
+    {
+      "epoch": 0.647381295480536,
+      "grad_norm": 0.6371638178825378,
+      "learning_rate": 8.341567541501397e-05,
+      "loss": 5.0629,
+      "num_input_tokens_seen": 517079040,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6478736006482018,
+      "grad_norm": 0.5510682463645935,
+      "learning_rate": 8.338397650624441e-05,
+      "loss": 5.0254,
+      "num_input_tokens_seen": 517472256,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6483659058158676,
+      "grad_norm": 0.6829474568367004,
+      "learning_rate": 8.335231370787243e-05,
+      "loss": 5.029,
+      "num_input_tokens_seen": 517865472,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6488582109835335,
+      "grad_norm": 0.5510388016700745,
+      "learning_rate": 8.332068695139037e-05,
+      "loss": 5.0445,
+      "num_input_tokens_seen": 518258688,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6493505161511992,
+      "grad_norm": 0.5481399893760681,
+      "learning_rate": 8.328909616847231e-05,
+      "loss": 5.0571,
+      "num_input_tokens_seen": 518651904,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6498428213188651,
+      "grad_norm": 0.6369343400001526,
+      "learning_rate": 8.325754129097364e-05,
+      "loss": 5.08,
+      "num_input_tokens_seen": 519045120,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6503351264865308,
+      "grad_norm": 0.5993247628211975,
+      "learning_rate": 8.322602225093026e-05,
+      "loss": 5.0518,
+      "num_input_tokens_seen": 519438336,
+      "step": 3963
+    },
+    {
+      "epoch": 0.6508274316541967,
+      "grad_norm": 0.64273601770401,
+      "learning_rate": 8.319453898055805e-05,
+      "loss": 5.0511,
+      "num_input_tokens_seen": 519831552,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6513197368218625,
+      "grad_norm": 0.6449026465415955,
+      "learning_rate": 8.316309141225228e-05,
+      "loss": 5.0171,
+      "num_input_tokens_seen": 520224768,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6518120419895282,
+      "grad_norm": 0.7018563151359558,
+      "learning_rate": 8.313167947858695e-05,
+      "loss": 5.0412,
+      "num_input_tokens_seen": 520617984,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6523043471571941,
+      "grad_norm": 0.5524522066116333,
+      "learning_rate": 8.310030311231415e-05,
+      "loss": 5.1299,
+      "num_input_tokens_seen": 521011200,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6527966523248598,
+      "grad_norm": 0.6293720006942749,
+      "learning_rate": 8.306896224636362e-05,
+      "loss": 5.0796,
+      "num_input_tokens_seen": 521404416,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6532889574925257,
+      "grad_norm": 0.6255269050598145,
+      "learning_rate": 8.303765681384188e-05,
+      "loss": 5.0136,
+      "num_input_tokens_seen": 521797632,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6537812626601914,
+      "grad_norm": 0.637909471988678,
+      "learning_rate": 8.300638674803195e-05,
+      "loss": 5.0853,
+      "num_input_tokens_seen": 522190848,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6542735678278573,
+      "grad_norm": 0.6187140345573425,
+      "learning_rate": 8.297515198239245e-05,
+      "loss": 5.0677,
+      "num_input_tokens_seen": 522584064,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6547658729955231,
+      "grad_norm": 0.5749291181564331,
+      "learning_rate": 8.294395245055722e-05,
+      "loss": 5.0805,
+      "num_input_tokens_seen": 522977280,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6552581781631889,
+      "grad_norm": 0.5552260875701904,
+      "learning_rate": 8.291278808633464e-05,
+      "loss": 5.0614,
+      "num_input_tokens_seen": 523370496,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6557504833308547,
+      "grad_norm": 0.5661478042602539,
+      "learning_rate": 8.288165882370701e-05,
+      "loss": 5.1098,
+      "num_input_tokens_seen": 523763712,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6562427884985205,
+      "grad_norm": 0.6904596090316772,
+      "learning_rate": 8.285056459683002e-05,
+      "loss": 5.0278,
+      "num_input_tokens_seen": 524156928,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6564068902210758,
+      "eval_accuracy": 0.22359713401726103,
+      "eval_loss": 5.324248313903809,
+      "eval_runtime": 113.075,
+      "eval_samples_per_second": 2.653,
+      "eval_steps_per_second": 1.327,
+      "num_input_tokens_seen": 524288000,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6567350936661863,
+      "grad_norm": 0.6449172496795654,
+      "learning_rate": 8.281950534003216e-05,
+      "loss": 5.0487,
+      "num_input_tokens_seen": 524550144,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6572273988338522,
+      "grad_norm": 0.6652653813362122,
+      "learning_rate": 8.278848098781413e-05,
+      "loss": 5.018,
+      "num_input_tokens_seen": 524943360,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6577197040015179,
+      "grad_norm": 0.6469755172729492,
+      "learning_rate": 8.275749147484824e-05,
+      "loss": 5.0477,
+      "num_input_tokens_seen": 525336576,
+      "step": 4008
+    },
+    {
+      "epoch": 0.6582120091691838,
+      "grad_norm": 0.6589633822441101,
+      "learning_rate": 8.272653673597785e-05,
+      "loss": 5.0356,
+      "num_input_tokens_seen": 525729792,
+      "step": 4011
+    },
+    {
+      "epoch": 0.6587043143368495,
+      "grad_norm": 0.7347235083580017,
+      "learning_rate": 8.269561670621681e-05,
+      "loss": 5.0088,
+      "num_input_tokens_seen": 526123008,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6591966195045154,
+      "grad_norm": 0.5741059184074402,
+      "learning_rate": 8.266473132074881e-05,
+      "loss": 5.0538,
+      "num_input_tokens_seen": 526516224,
+      "step": 4017
+    },
+    {
+      "epoch": 0.6596889246721812,
+      "grad_norm": 0.6932168006896973,
+      "learning_rate": 8.263388051492694e-05,
+      "loss": 5.0435,
+      "num_input_tokens_seen": 526909440,
+      "step": 4020
+    },
+    {
+      "epoch": 0.660181229839847,
+      "grad_norm": 0.6108718514442444,
+      "learning_rate": 8.260306422427303e-05,
+      "loss": 5.0209,
+      "num_input_tokens_seen": 527302656,
+      "step": 4023
+    },
+    {
+      "epoch": 0.6606735350075128,
+      "grad_norm": 0.7427350282669067,
+      "learning_rate": 8.257228238447704e-05,
+      "loss": 5.0025,
+      "num_input_tokens_seen": 527695872,
+      "step": 4026
+    },
+    {
+      "epoch": 0.6611658401751785,
+      "grad_norm": 0.6407040357589722,
+      "learning_rate": 8.254153493139666e-05,
+      "loss": 5.0381,
+      "num_input_tokens_seen": 528089088,
+      "step": 4029
+    },
+    {
+      "epoch": 0.6616581453428444,
+      "grad_norm": 0.6951069831848145,
+      "learning_rate": 8.251082180105658e-05,
+      "loss": 5.0644,
+      "num_input_tokens_seen": 528482304,
+      "step": 4032
+    },
+    {
+      "epoch": 0.6621504505105102,
+      "grad_norm": 0.64825439453125,
+      "learning_rate": 8.248014292964801e-05,
+      "loss": 5.0702,
+      "num_input_tokens_seen": 528875520,
+      "step": 4035
+    },
+    {
+      "epoch": 0.662642755678176,
+      "grad_norm": 0.631080687046051,
+      "learning_rate": 8.244949825352815e-05,
+      "loss": 5.0367,
+      "num_input_tokens_seen": 529268736,
+      "step": 4038
+    },
+    {
+      "epoch": 0.6631350608458418,
+      "grad_norm": 0.5960158109664917,
+      "learning_rate": 8.241888770921956e-05,
+      "loss": 5.0529,
+      "num_input_tokens_seen": 529661952,
+      "step": 4041
+    },
+    {
+      "epoch": 0.6636273660135076,
+      "grad_norm": 0.5926985144615173,
+      "learning_rate": 8.238831123340965e-05,
+      "loss": 5.0342,
+      "num_input_tokens_seen": 530055168,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6641196711811734,
+      "grad_norm": 0.6134880185127258,
+      "learning_rate": 8.235776876295013e-05,
+      "loss": 5.0454,
+      "num_input_tokens_seen": 530448384,
+      "step": 4047
+    },
+    {
+      "epoch": 0.6646119763488393,
+      "grad_norm": 0.6768664121627808,
+      "learning_rate": 8.232726023485646e-05,
+      "loss": 5.0676,
+      "num_input_tokens_seen": 530841600,
+      "step": 4050
+    },
+    {
+      "epoch": 0.665104281516505,
+      "grad_norm": 0.6263231635093689,
+      "learning_rate": 8.22967855863073e-05,
+      "loss": 5.03,
+      "num_input_tokens_seen": 531234816,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6655965866841709,
+      "grad_norm": 0.6137826442718506,
+      "learning_rate": 8.226634475464398e-05,
+      "loss": 5.0772,
+      "num_input_tokens_seen": 531628032,
+      "step": 4056
+    },
+    {
+      "epoch": 0.6660888918518366,
+      "grad_norm": 0.6005773544311523,
+      "learning_rate": 8.223593767736994e-05,
+      "loss": 5.0229,
+      "num_input_tokens_seen": 532021248,
+      "step": 4059
+    },
+    {
+      "epoch": 0.6665811970195025,
+      "grad_norm": 0.5843310356140137,
+      "learning_rate": 8.22055642921502e-05,
+      "loss": 5.0513,
+      "num_input_tokens_seen": 532414464,
+      "step": 4062
+    },
+    {
+      "epoch": 0.6670735021871683,
+      "grad_norm": 0.5929916501045227,
+      "learning_rate": 8.217522453681083e-05,
+      "loss": 5.0487,
+      "num_input_tokens_seen": 532807680,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6675658073548341,
+      "grad_norm": 0.5393611788749695,
+      "learning_rate": 8.214491834933838e-05,
+      "loss": 5.0281,
+      "num_input_tokens_seen": 533200896,
+      "step": 4068
+    },
+    {
+      "epoch": 0.6680581125224999,
+      "grad_norm": 0.6220096945762634,
+      "learning_rate": 8.21146456678794e-05,
+      "loss": 5.025,
+      "num_input_tokens_seen": 533594112,
+      "step": 4071
+    },
+    {
+      "epoch": 0.6685504176901657,
+      "grad_norm": 0.6115317940711975,
+      "learning_rate": 8.208440643073989e-05,
+      "loss": 5.0254,
+      "num_input_tokens_seen": 533987328,
+      "step": 4074
+    },
+    {
+      "epoch": 0.6690427228578315,
+      "grad_norm": 0.6211367845535278,
+      "learning_rate": 8.205420057638475e-05,
+      "loss": 5.0665,
+      "num_input_tokens_seen": 534380544,
+      "step": 4077
+    },
+    {
+      "epoch": 0.6695350280254974,
+      "grad_norm": 0.5687063932418823,
+      "learning_rate": 8.202402804343728e-05,
+      "loss": 5.038,
+      "num_input_tokens_seen": 534773760,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6700273331931631,
+      "grad_norm": 0.572433590888977,
+      "learning_rate": 8.199388877067867e-05,
+      "loss": 5.0709,
+      "num_input_tokens_seen": 535166976,
+      "step": 4083
+    },
+    {
+      "epoch": 0.670519638360829,
+      "grad_norm": 0.5648781061172485,
+      "learning_rate": 8.196378269704742e-05,
+      "loss": 5.0144,
+      "num_input_tokens_seen": 535560192,
+      "step": 4086
+    },
+    {
+      "epoch": 0.6710119435284947,
+      "grad_norm": 0.622725784778595,
+      "learning_rate": 8.193370976163886e-05,
+      "loss": 4.9888,
+      "num_input_tokens_seen": 535953408,
+      "step": 4089
+    },
+    {
+      "epoch": 0.6715042486961605,
+      "grad_norm": 0.6132870316505432,
+      "learning_rate": 8.190366990370464e-05,
+      "loss": 5.0087,
+      "num_input_tokens_seen": 536346624,
+      "step": 4092
+    },
+    {
+      "epoch": 0.6719965538638263,
+      "grad_norm": 0.5610283017158508,
+      "learning_rate": 8.187366306265222e-05,
+      "loss": 5.0339,
+      "num_input_tokens_seen": 536739840,
+      "step": 4095
+    },
+    {
+      "epoch": 0.6724888590314921,
+      "grad_norm": 0.5569362640380859,
+      "learning_rate": 8.184368917804431e-05,
+      "loss": 5.0178,
+      "num_input_tokens_seen": 537133056,
+      "step": 4098
+    },
+    {
+      "epoch": 0.672981164199158,
+      "grad_norm": 0.6829281449317932,
+      "learning_rate": 8.181374818959841e-05,
+      "loss": 5.0422,
+      "num_input_tokens_seen": 537526272,
+      "step": 4101
+    },
+    {
+      "epoch": 0.6734734693668237,
+      "grad_norm": 0.6499529480934143,
+      "learning_rate": 8.178384003718625e-05,
+      "loss": 5.0081,
+      "num_input_tokens_seen": 537919488,
+      "step": 4104
+    },
+    {
+      "epoch": 0.6739657745344896,
+      "grad_norm": 0.5391960740089417,
+      "learning_rate": 8.175396466083337e-05,
+      "loss": 5.026,
+      "num_input_tokens_seen": 538312704,
+      "step": 4107
+    },
+    {
+      "epoch": 0.6744580797021553,
+      "grad_norm": 0.5357514023780823,
+      "learning_rate": 8.17241220007185e-05,
+      "loss": 5.0033,
+      "num_input_tokens_seen": 538705920,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6749503848698212,
+      "grad_norm": 0.5889911651611328,
+      "learning_rate": 8.169431199717313e-05,
+      "loss": 4.9921,
+      "num_input_tokens_seen": 539099136,
+      "step": 4113
+    },
+    {
+      "epoch": 0.675442690037487,
+      "grad_norm": 0.5937342643737793,
+      "learning_rate": 8.1664534590681e-05,
+      "loss": 5.0344,
+      "num_input_tokens_seen": 539492352,
+      "step": 4116
+    },
+    {
+      "epoch": 0.6759349952051528,
+      "grad_norm": 0.6347099542617798,
+      "learning_rate": 8.163478972187763e-05,
+      "loss": 5.033,
+      "num_input_tokens_seen": 539885568,
+      "step": 4119
+    },
+    {
+      "epoch": 0.6764273003728186,
+      "grad_norm": 0.5977573990821838,
+      "learning_rate": 8.160507733154971e-05,
+      "loss": 5.0467,
+      "num_input_tokens_seen": 540278784,
+      "step": 4122
+    },
+    {
+      "epoch": 0.6769196055404844,
+      "grad_norm": 0.5776995420455933,
+      "learning_rate": 8.157539736063474e-05,
+      "loss": 5.0784,
+      "num_input_tokens_seen": 540672000,
+      "step": 4125
+    },
+    {
+      "epoch": 0.6774119107081502,
+      "grad_norm": 0.5840444564819336,
+      "learning_rate": 8.154574975022046e-05,
+      "loss": 5.0125,
+      "num_input_tokens_seen": 541065216,
+      "step": 4128
+    },
+    {
+      "epoch": 0.6779042158758161,
+      "grad_norm": 0.6594420671463013,
+      "learning_rate": 8.151613444154437e-05,
+      "loss": 5.0195,
+      "num_input_tokens_seen": 541458432,
+      "step": 4131
+    },
+    {
+      "epoch": 0.6783965210434818,
+      "grad_norm": 0.6229584217071533,
+      "learning_rate": 8.14865513759933e-05,
+      "loss": 5.0545,
+      "num_input_tokens_seen": 541851648,
+      "step": 4134
+    },
+    {
+      "epoch": 0.6788888262111477,
+      "grad_norm": 0.5304208993911743,
+      "learning_rate": 8.145700049510277e-05,
+      "loss": 5.0066,
+      "num_input_tokens_seen": 542244864,
+      "step": 4137
+    },
+    {
+      "epoch": 0.6793811313788134,
+      "grad_norm": 0.7174444794654846,
+      "learning_rate": 8.14274817405567e-05,
+      "loss": 5.0549,
+      "num_input_tokens_seen": 542638080,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6798734365464792,
+      "grad_norm": 0.5718116164207458,
+      "learning_rate": 8.13979950541868e-05,
+      "loss": 5.0239,
+      "num_input_tokens_seen": 543031296,
+      "step": 4143
+    },
+    {
+      "epoch": 0.6803657417141451,
+      "grad_norm": 0.6597188711166382,
+      "learning_rate": 8.136854037797212e-05,
+      "loss": 5.018,
+      "num_input_tokens_seen": 543424512,
+      "step": 4146
+    },
+    {
+      "epoch": 0.6808580468818108,
+      "grad_norm": 0.6705512404441833,
+      "learning_rate": 8.133911765403855e-05,
+      "loss": 5.0543,
+      "num_input_tokens_seen": 543817728,
+      "step": 4149
+    },
+    {
+      "epoch": 0.6813503520494767,
+      "grad_norm": 0.6903484463691711,
+      "learning_rate": 8.130972682465842e-05,
+      "loss": 5.0084,
+      "num_input_tokens_seen": 544210944,
+      "step": 4152
+    },
+    {
+      "epoch": 0.6818426572171424,
+      "grad_norm": 0.563361406326294,
+      "learning_rate": 8.128036783224992e-05,
+      "loss": 5.0479,
+      "num_input_tokens_seen": 544604160,
+      "step": 4155
+    },
+    {
+      "epoch": 0.6823349623848083,
+      "grad_norm": 0.5460705757141113,
+      "learning_rate": 8.125104061937669e-05,
+      "loss": 5.0404,
+      "num_input_tokens_seen": 544997376,
+      "step": 4158
+    },
+    {
+      "epoch": 0.6828272675524741,
+      "grad_norm": 0.5406615734100342,
+      "learning_rate": 8.122174512874733e-05,
+      "loss": 5.0741,
+      "num_input_tokens_seen": 545390592,
+      "step": 4161
+    },
+    {
+      "epoch": 0.6833195727201399,
+      "grad_norm": 0.6041008830070496,
+      "learning_rate": 8.119248130321494e-05,
+      "loss": 5.0259,
+      "num_input_tokens_seen": 545783808,
+      "step": 4164
+    },
+    {
+      "epoch": 0.6838118778878057,
+      "grad_norm": 0.6043838262557983,
+      "learning_rate": 8.116324908577667e-05,
+      "loss": 5.0465,
+      "num_input_tokens_seen": 546177024,
+      "step": 4167
+    },
+    {
+      "epoch": 0.6843041830554715,
+      "grad_norm": 0.6404688954353333,
+      "learning_rate": 8.113404841957315e-05,
+      "loss": 5.023,
+      "num_input_tokens_seen": 546570240,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6847964882231373,
+      "grad_norm": 0.5821146965026855,
+      "learning_rate": 8.110487924788816e-05,
+      "loss": 5.0608,
+      "num_input_tokens_seen": 546963456,
+      "step": 4173
+    },
+    {
+      "epoch": 0.6852887933908032,
+      "grad_norm": 0.704936146736145,
+      "learning_rate": 8.107574151414814e-05,
+      "loss": 5.091,
+      "num_input_tokens_seen": 547356672,
+      "step": 4176
+    },
+    {
+      "epoch": 0.6857810985584689,
+      "grad_norm": 0.5520325899124146,
+      "learning_rate": 8.104663516192164e-05,
+      "loss": 5.0377,
+      "num_input_tokens_seen": 547749888,
+      "step": 4179
+    },
+    {
+      "epoch": 0.6862734037261348,
+      "grad_norm": 0.5666207075119019,
+      "learning_rate": 8.101756013491894e-05,
+      "loss": 5.014,
+      "num_input_tokens_seen": 548143104,
+      "step": 4182
+    },
+    {
+      "epoch": 0.6867657088938005,
+      "grad_norm": 0.5827277302742004,
+      "learning_rate": 8.09885163769916e-05,
+      "loss": 5.008,
+      "num_input_tokens_seen": 548536320,
+      "step": 4185
+    },
+    {
+      "epoch": 0.6872580140614664,
+      "grad_norm": 0.5662521719932556,
+      "learning_rate": 8.095950383213192e-05,
+      "loss": 5.0114,
+      "num_input_tokens_seen": 548929536,
+      "step": 4188
+    },
+    {
+      "epoch": 0.6877503192291322,
+      "grad_norm": 0.6049064993858337,
+      "learning_rate": 8.093052244447264e-05,
+      "loss": 5.0039,
+      "num_input_tokens_seen": 549322752,
+      "step": 4191
+    },
+    {
+      "epoch": 0.688242624396798,
+      "grad_norm": 0.6402872204780579,
+      "learning_rate": 8.090157215828629e-05,
+      "loss": 5.0325,
+      "num_input_tokens_seen": 549715968,
+      "step": 4194
+    },
+    {
+      "epoch": 0.6887349295644638,
+      "grad_norm": 0.582022488117218,
+      "learning_rate": 8.08726529179849e-05,
+      "loss": 5.0176,
+      "num_input_tokens_seen": 550109184,
+      "step": 4197
+    },
+    {
+      "epoch": 0.6892272347321295,
+      "grad_norm": 0.577924370765686,
+      "learning_rate": 8.08437646681195e-05,
+      "loss": 5.043,
+      "num_input_tokens_seen": 550502400,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6897195398997954,
+      "grad_norm": 0.6571632623672485,
+      "learning_rate": 8.081490735337961e-05,
+      "loss": 4.9929,
+      "num_input_tokens_seen": 550895616,
+      "step": 4203
+    },
+    {
+      "epoch": 0.6902118450674611,
+      "grad_norm": 0.5845437049865723,
+      "learning_rate": 8.078608091859296e-05,
+      "loss": 5.0122,
+      "num_input_tokens_seen": 551288832,
+      "step": 4206
+    },
+    {
+      "epoch": 0.690704150235127,
+      "grad_norm": 0.5887883901596069,
+      "learning_rate": 8.075728530872482e-05,
+      "loss": 5.0236,
+      "num_input_tokens_seen": 551682048,
+      "step": 4209
+    },
+    {
+      "epoch": 0.6911964554027928,
+      "grad_norm": 0.5670571327209473,
+      "learning_rate": 8.072852046887776e-05,
+      "loss": 4.9989,
+      "num_input_tokens_seen": 552075264,
+      "step": 4212
+    },
+    {
+      "epoch": 0.6916887605704586,
+      "grad_norm": 0.6224859952926636,
+      "learning_rate": 8.069978634429111e-05,
+      "loss": 5.0138,
+      "num_input_tokens_seen": 552468480,
+      "step": 4215
+    },
+    {
+      "epoch": 0.6921810657381244,
+      "grad_norm": 0.610889732837677,
+      "learning_rate": 8.067108288034053e-05,
+      "loss": 5.0593,
+      "num_input_tokens_seen": 552861696,
+      "step": 4218
+    },
+    {
+      "epoch": 0.6926733709057902,
+      "grad_norm": 0.5440495610237122,
+      "learning_rate": 8.064241002253757e-05,
+      "loss": 5.0239,
+      "num_input_tokens_seen": 553254912,
+      "step": 4221
+    },
+    {
+      "epoch": 0.693165676073456,
+      "grad_norm": 0.5816980004310608,
+      "learning_rate": 8.061376771652931e-05,
+      "loss": 5.045,
+      "num_input_tokens_seen": 553648128,
+      "step": 4224
+    },
+    {
+      "epoch": 0.6936579812411219,
+      "grad_norm": 0.5322626233100891,
+      "learning_rate": 8.058515590809782e-05,
+      "loss": 5.0538,
+      "num_input_tokens_seen": 554041344,
+      "step": 4227
+    },
+    {
+      "epoch": 0.6941502864087876,
+      "grad_norm": 0.6580582857131958,
+      "learning_rate": 8.055657454315977e-05,
+      "loss": 5.0129,
+      "num_input_tokens_seen": 554434560,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6946425915764535,
+      "grad_norm": 0.5960021615028381,
+      "learning_rate": 8.052802356776606e-05,
+      "loss": 5.0385,
+      "num_input_tokens_seen": 554827776,
+      "step": 4233
+    },
+    {
+      "epoch": 0.6951348967441192,
+      "grad_norm": 0.6201943159103394,
+      "learning_rate": 8.049950292810128e-05,
+      "loss": 5.0126,
+      "num_input_tokens_seen": 555220992,
+      "step": 4236
+    },
+    {
+      "epoch": 0.6956272019117851,
+      "grad_norm": 0.6654148697853088,
+      "learning_rate": 8.047101257048339e-05,
+      "loss": 5.0154,
+      "num_input_tokens_seen": 555614208,
+      "step": 4239
+    },
+    {
+      "epoch": 0.6961195070794509,
+      "grad_norm": 0.5792354941368103,
+      "learning_rate": 8.044255244136322e-05,
+      "loss": 5.0095,
+      "num_input_tokens_seen": 556007424,
+      "step": 4242
+    },
+    {
+      "epoch": 0.6966118122471167,
+      "grad_norm": 0.5518591403961182,
+      "learning_rate": 8.041412248732407e-05,
+      "loss": 4.9994,
+      "num_input_tokens_seen": 556400640,
+      "step": 4245
+    },
+    {
+      "epoch": 0.6971041174147825,
+      "grad_norm": 0.6194866299629211,
+      "learning_rate": 8.038572265508136e-05,
+      "loss": 5.0041,
+      "num_input_tokens_seen": 556793856,
+      "step": 4248
+    },
+    {
+      "epoch": 0.6975964225824483,
+      "grad_norm": 0.6971485614776611,
+      "learning_rate": 8.035735289148207e-05,
+      "loss": 5.0282,
+      "num_input_tokens_seen": 557187072,
+      "step": 4251
+    },
+    {
+      "epoch": 0.6980887277501141,
+      "grad_norm": 0.5932923555374146,
+      "learning_rate": 8.032901314350443e-05,
+      "loss": 5.111,
+      "num_input_tokens_seen": 557580288,
+      "step": 4254
+    },
+    {
+      "epoch": 0.69858103291778,
+      "grad_norm": 0.6528842449188232,
+      "learning_rate": 8.030070335825747e-05,
+      "loss": 5.0172,
+      "num_input_tokens_seen": 557973504,
+      "step": 4257
+    },
+    {
+      "epoch": 0.6990733380854457,
+      "grad_norm": 0.658224880695343,
+      "learning_rate": 8.027242348298066e-05,
+      "loss": 5.036,
+      "num_input_tokens_seen": 558366720,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6995656432531115,
+      "grad_norm": 0.4988958537578583,
+      "learning_rate": 8.024417346504334e-05,
+      "loss": 4.9453,
+      "num_input_tokens_seen": 558759936,
+      "step": 4263
+    },
+    {
+      "epoch": 0.7000579484207773,
+      "grad_norm": 0.6576696038246155,
+      "learning_rate": 8.021595325194448e-05,
+      "loss": 5.0356,
+      "num_input_tokens_seen": 559153152,
+      "step": 4266
+    },
+    {
+      "epoch": 0.7005502535884431,
+      "grad_norm": 0.6395735144615173,
+      "learning_rate": 8.01877627913122e-05,
+      "loss": 5.041,
+      "num_input_tokens_seen": 559546368,
+      "step": 4269
+    },
+    {
+      "epoch": 0.701042558756109,
+      "grad_norm": 0.5595338940620422,
+      "learning_rate": 8.015960203090336e-05,
+      "loss": 4.9971,
+      "num_input_tokens_seen": 559939584,
+      "step": 4272
+    },
+    {
+      "epoch": 0.7015348639237747,
+      "grad_norm": 0.656607449054718,
+      "learning_rate": 8.013147091860318e-05,
+      "loss": 5.0286,
+      "num_input_tokens_seen": 560332800,
+      "step": 4275
+    },
+    {
+      "epoch": 0.7020271690914406,
+      "grad_norm": 0.6094218492507935,
+      "learning_rate": 8.010336940242475e-05,
+      "loss": 4.9691,
+      "num_input_tokens_seen": 560726016,
+      "step": 4278
+    },
+    {
+      "epoch": 0.7025194742591063,
+      "grad_norm": 0.632666289806366,
+      "learning_rate": 8.007529743050875e-05,
+      "loss": 5.0467,
+      "num_input_tokens_seen": 561119232,
+      "step": 4281
+    },
+    {
+      "epoch": 0.7030117794267722,
+      "grad_norm": 0.6173758506774902,
+      "learning_rate": 8.004725495112299e-05,
+      "loss": 5.021,
+      "num_input_tokens_seen": 561512448,
+      "step": 4284
+    },
+    {
+      "epoch": 0.703504084594438,
+      "grad_norm": 0.6932005286216736,
+      "learning_rate": 8.001924191266195e-05,
+      "loss": 5.0294,
+      "num_input_tokens_seen": 561905664,
+      "step": 4287
+    },
+    {
+      "epoch": 0.7039963897621038,
+      "grad_norm": 0.537021815776825,
+      "learning_rate": 7.999125826364651e-05,
+      "loss": 5.0358,
+      "num_input_tokens_seen": 562298880,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7044886949297696,
+      "grad_norm": 0.6217671632766724,
+      "learning_rate": 7.996330395272346e-05,
+      "loss": 5.0181,
+      "num_input_tokens_seen": 562692096,
+      "step": 4293
+    },
+    {
+      "epoch": 0.7049810000974354,
+      "grad_norm": 0.6611473560333252,
+      "learning_rate": 7.993537892866508e-05,
+      "loss": 5.0083,
+      "num_input_tokens_seen": 563085312,
+      "step": 4296
+    },
+    {
+      "epoch": 0.7054733052651012,
+      "grad_norm": 0.6191869378089905,
+      "learning_rate": 7.990748314036885e-05,
+      "loss": 5.0469,
+      "num_input_tokens_seen": 563478528,
+      "step": 4299
+    },
+    {
+      "epoch": 0.7059656104327671,
+      "grad_norm": 0.6508629322052002,
+      "learning_rate": 7.987961653685697e-05,
+      "loss": 5.0245,
+      "num_input_tokens_seen": 563871744,
+      "step": 4302
+    },
+    {
+      "epoch": 0.7064579156004328,
+      "grad_norm": 0.5989941954612732,
+      "learning_rate": 7.9851779067276e-05,
+      "loss": 5.0273,
+      "num_input_tokens_seen": 564264960,
+      "step": 4305
+    },
+    {
+      "epoch": 0.7069502207680987,
+      "grad_norm": 0.6267138123512268,
+      "learning_rate": 7.98239706808965e-05,
+      "loss": 4.9692,
+      "num_input_tokens_seen": 564658176,
+      "step": 4308
+    },
+    {
+      "epoch": 0.7074425259357644,
+      "grad_norm": 0.6717625260353088,
+      "learning_rate": 7.979619132711254e-05,
+      "loss": 4.9929,
+      "num_input_tokens_seen": 565051392,
+      "step": 4311
+    },
+    {
+      "epoch": 0.7079348311034303,
+      "grad_norm": 0.7013946771621704,
+      "learning_rate": 7.976844095544147e-05,
+      "loss": 4.9909,
+      "num_input_tokens_seen": 565444608,
+      "step": 4314
+    },
+    {
+      "epoch": 0.708427136271096,
+      "grad_norm": 0.589557945728302,
+      "learning_rate": 7.974071951552337e-05,
+      "loss": 5.0382,
+      "num_input_tokens_seen": 565837824,
+      "step": 4317
+    },
+    {
+      "epoch": 0.7089194414387618,
+      "grad_norm": 0.6243999600410461,
+      "learning_rate": 7.97130269571208e-05,
+      "loss": 5.0769,
+      "num_input_tokens_seen": 566231040,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7094117466064277,
+      "grad_norm": 0.6162127256393433,
+      "learning_rate": 7.968536323011831e-05,
+      "loss": 5.0046,
+      "num_input_tokens_seen": 566624256,
+      "step": 4323
+    },
+    {
+      "epoch": 0.7099040517740934,
+      "grad_norm": 0.598885178565979,
+      "learning_rate": 7.965772828452217e-05,
+      "loss": 5.0263,
+      "num_input_tokens_seen": 567017472,
+      "step": 4326
+    },
+    {
+      "epoch": 0.7103963569417593,
+      "grad_norm": 0.5507720708847046,
+      "learning_rate": 7.963012207045987e-05,
+      "loss": 5.0186,
+      "num_input_tokens_seen": 567410688,
+      "step": 4329
+    },
+    {
+      "epoch": 0.710888662109425,
+      "grad_norm": 0.5605682730674744,
+      "learning_rate": 7.960254453817985e-05,
+      "loss": 4.9507,
+      "num_input_tokens_seen": 567803904,
+      "step": 4332
+    },
+    {
+      "epoch": 0.7113809672770909,
+      "grad_norm": 0.6122700572013855,
+      "learning_rate": 7.957499563805107e-05,
+      "loss": 5.0315,
+      "num_input_tokens_seen": 568197120,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7118732724447567,
+      "grad_norm": 0.7516876459121704,
+      "learning_rate": 7.954747532056262e-05,
+      "loss": 4.9979,
+      "num_input_tokens_seen": 568590336,
+      "step": 4338
+    },
+    {
+      "epoch": 0.7123655776124225,
+      "grad_norm": 0.5879480242729187,
+      "learning_rate": 7.951998353632336e-05,
+      "loss": 5.018,
+      "num_input_tokens_seen": 568983552,
+      "step": 4341
+    },
+    {
+      "epoch": 0.7128578827800883,
+      "grad_norm": 0.7155163884162903,
+      "learning_rate": 7.949252023606159e-05,
+      "loss": 5.0276,
+      "num_input_tokens_seen": 569376768,
+      "step": 4344
+    },
+    {
+      "epoch": 0.7133501879477541,
+      "grad_norm": 0.5591400265693665,
+      "learning_rate": 7.946508537062463e-05,
+      "loss": 5.0113,
+      "num_input_tokens_seen": 569769984,
+      "step": 4347
+    },
+    {
+      "epoch": 0.7138424931154199,
+      "grad_norm": 0.5849490761756897,
+      "learning_rate": 7.943767889097847e-05,
+      "loss": 4.9933,
+      "num_input_tokens_seen": 570163200,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7143347982830858,
+      "grad_norm": 0.6266775727272034,
+      "learning_rate": 7.941030074820736e-05,
+      "loss": 5.0199,
+      "num_input_tokens_seen": 570556416,
+      "step": 4353
+    },
+    {
+      "epoch": 0.7148271034507515,
+      "grad_norm": 0.6590829491615295,
+      "learning_rate": 7.938295089351354e-05,
+      "loss": 5.0225,
+      "num_input_tokens_seen": 570949632,
+      "step": 4356
+    },
+    {
+      "epoch": 0.7153194086184174,
+      "grad_norm": 0.5544537901878357,
+      "learning_rate": 7.935562927821676e-05,
+      "loss": 5.0161,
+      "num_input_tokens_seen": 571342848,
+      "step": 4359
+    },
+    {
+      "epoch": 0.7158117137860831,
+      "grad_norm": 0.5898229479789734,
+      "learning_rate": 7.932833585375402e-05,
+      "loss": 5.0445,
+      "num_input_tokens_seen": 571736064,
+      "step": 4362
+    },
+    {
+      "epoch": 0.716304018953749,
+      "grad_norm": 0.6551963686943054,
+      "learning_rate": 7.930107057167912e-05,
+      "loss": 4.9816,
+      "num_input_tokens_seen": 572129280,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7167963241214148,
+      "grad_norm": 0.530617356300354,
+      "learning_rate": 7.927383338366234e-05,
+      "loss": 5.0351,
+      "num_input_tokens_seen": 572522496,
+      "step": 4368
+    },
+    {
+      "epoch": 0.7172886292890805,
+      "grad_norm": 0.6499817371368408,
+      "learning_rate": 7.924662424149012e-05,
+      "loss": 5.0164,
+      "num_input_tokens_seen": 572915712,
+      "step": 4371
+    },
+    {
+      "epoch": 0.7177809344567464,
+      "grad_norm": 0.6318878531455994,
+      "learning_rate": 7.921944309706458e-05,
+      "loss": 5.0125,
+      "num_input_tokens_seen": 573308928,
+      "step": 4374
+    },
+    {
+      "epoch": 0.7182732396244121,
+      "grad_norm": 0.5628673434257507,
+      "learning_rate": 7.919228990240331e-05,
+      "loss": 4.9925,
+      "num_input_tokens_seen": 573702144,
+      "step": 4377
+    },
+    {
+      "epoch": 0.718765544792078,
+      "grad_norm": 0.7107114791870117,
+      "learning_rate": 7.916516460963895e-05,
+      "loss": 4.9837,
+      "num_input_tokens_seen": 574095360,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7192578499597438,
+      "grad_norm": 0.6181208491325378,
+      "learning_rate": 7.913806717101879e-05,
+      "loss": 5.0152,
+      "num_input_tokens_seen": 574488576,
+      "step": 4383
+    },
+    {
+      "epoch": 0.7197501551274096,
+      "grad_norm": 0.6599327921867371,
+      "learning_rate": 7.911099753890446e-05,
+      "loss": 5.0476,
+      "num_input_tokens_seen": 574881792,
+      "step": 4386
+    },
+    {
+      "epoch": 0.7202424602950754,
+      "grad_norm": 0.5509748458862305,
+      "learning_rate": 7.90839556657716e-05,
+      "loss": 5.0521,
+      "num_input_tokens_seen": 575275008,
+      "step": 4389
+    },
+    {
+      "epoch": 0.7207347654627412,
+      "grad_norm": 0.5559970736503601,
+      "learning_rate": 7.905694150420948e-05,
+      "loss": 5.0064,
+      "num_input_tokens_seen": 575668224,
+      "step": 4392
+    },
+    {
+      "epoch": 0.721227070630407,
+      "grad_norm": 0.5684667825698853,
+      "learning_rate": 7.902995500692065e-05,
+      "loss": 5.0044,
+      "num_input_tokens_seen": 576061440,
+      "step": 4395
+    },
+    {
+      "epoch": 0.7217193757980729,
+      "grad_norm": 0.6226740479469299,
+      "learning_rate": 7.900299612672062e-05,
+      "loss": 5.0389,
+      "num_input_tokens_seen": 576454656,
+      "step": 4398
+    },
+    {
+      "epoch": 0.7220475792431834,
+      "eval_accuracy": 0.22640612278130598,
+      "eval_loss": 5.292022228240967,
+      "eval_runtime": 114.3223,
+      "eval_samples_per_second": 2.624,
+      "eval_steps_per_second": 1.312,
+      "num_input_tokens_seen": 576716800,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7222116809657386,
+      "grad_norm": 0.6325856447219849,
+      "learning_rate": 7.897606481653748e-05,
+      "loss": 5.0615,
+      "num_input_tokens_seen": 576847872,
+      "step": 4401
+    },
+    {
+      "epoch": 0.7227039861334045,
+      "grad_norm": 0.607279896736145,
+      "learning_rate": 7.894916102941156e-05,
+      "loss": 4.9912,
+      "num_input_tokens_seen": 577241088,
+      "step": 4404
+    },
+    {
+      "epoch": 0.7231962913010702,
+      "grad_norm": 0.5997862219810486,
+      "learning_rate": 7.892228471849507e-05,
+      "loss": 5.0194,
+      "num_input_tokens_seen": 577634304,
+      "step": 4407
+    },
+    {
+      "epoch": 0.7236885964687361,
+      "grad_norm": 0.6025493741035461,
+      "learning_rate": 7.889543583705186e-05,
+      "loss": 4.9893,
+      "num_input_tokens_seen": 578027520,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7241809016364019,
+      "grad_norm": 0.6023775339126587,
+      "learning_rate": 7.886861433845691e-05,
+      "loss": 4.9841,
+      "num_input_tokens_seen": 578420736,
+      "step": 4413
+    },
+    {
+      "epoch": 0.7246732068040677,
+      "grad_norm": 0.5165770649909973,
+      "learning_rate": 7.884182017619615e-05,
+      "loss": 4.9778,
+      "num_input_tokens_seen": 578813952,
+      "step": 4416
+    },
+    {
+      "epoch": 0.7251655119717335,
+      "grad_norm": 0.6931847333908081,
+      "learning_rate": 7.881505330386602e-05,
+      "loss": 4.9895,
+      "num_input_tokens_seen": 579207168,
+      "step": 4419
+    },
+    {
+      "epoch": 0.7256578171393993,
+      "grad_norm": 0.6051077842712402,
+      "learning_rate": 7.878831367517315e-05,
+      "loss": 4.9822,
+      "num_input_tokens_seen": 579600384,
+      "step": 4422
+    },
+    {
+      "epoch": 0.7261501223070651,
+      "grad_norm": 0.5695605874061584,
+      "learning_rate": 7.876160124393405e-05,
+      "loss": 4.9885,
+      "num_input_tokens_seen": 579993600,
+      "step": 4425
+    },
+    {
+      "epoch": 0.7266424274747308,
+      "grad_norm": 0.5569063425064087,
+      "learning_rate": 7.873491596407478e-05,
+      "loss": 4.9892,
+      "num_input_tokens_seen": 580386816,
+      "step": 4428
+    },
+    {
+      "epoch": 0.7271347326423967,
+      "grad_norm": 0.5842316746711731,
+      "learning_rate": 7.870825778963058e-05,
+      "loss": 4.9598,
+      "num_input_tokens_seen": 580780032,
+      "step": 4431
+    },
+    {
+      "epoch": 0.7276270378100625,
+      "grad_norm": 0.6461383104324341,
+      "learning_rate": 7.868162667474556e-05,
+      "loss": 4.989,
+      "num_input_tokens_seen": 581173248,
+      "step": 4434
+    },
+    {
+      "epoch": 0.7281193429777283,
+      "grad_norm": 0.6318100690841675,
+      "learning_rate": 7.865502257367235e-05,
+      "loss": 5.0184,
+      "num_input_tokens_seen": 581566464,
+      "step": 4437
+    },
+    {
+      "epoch": 0.7286116481453941,
+      "grad_norm": 0.6120467782020569,
+      "learning_rate": 7.862844544077183e-05,
+      "loss": 5.0424,
+      "num_input_tokens_seen": 581959680,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7291039533130599,
+      "grad_norm": 0.6271392107009888,
+      "learning_rate": 7.860189523051269e-05,
+      "loss": 5.0165,
+      "num_input_tokens_seen": 582352896,
+      "step": 4443
+    },
+    {
+      "epoch": 0.7295962584807257,
+      "grad_norm": 0.5989521741867065,
+      "learning_rate": 7.857537189747122e-05,
+      "loss": 5.0157,
+      "num_input_tokens_seen": 582746112,
+      "step": 4446
+    },
+    {
+      "epoch": 0.7300885636483916,
+      "grad_norm": 0.5683810114860535,
+      "learning_rate": 7.854887539633091e-05,
+      "loss": 4.9685,
+      "num_input_tokens_seen": 583139328,
+      "step": 4449
+    },
+    {
+      "epoch": 0.7305808688160573,
+      "grad_norm": 0.598704993724823,
+      "learning_rate": 7.852240568188216e-05,
+      "loss": 5.0173,
+      "num_input_tokens_seen": 583532544,
+      "step": 4452
+    },
+    {
+      "epoch": 0.7310731739837232,
+      "grad_norm": 0.5917366743087769,
+      "learning_rate": 7.849596270902193e-05,
+      "loss": 5.0061,
+      "num_input_tokens_seen": 583925760,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7315654791513889,
+      "grad_norm": 0.6544704437255859,
+      "learning_rate": 7.846954643275341e-05,
+      "loss": 4.9838,
+      "num_input_tokens_seen": 584318976,
+      "step": 4458
+    },
+    {
+      "epoch": 0.7320577843190548,
+      "grad_norm": 0.5870675444602966,
+      "learning_rate": 7.844315680818579e-05,
+      "loss": 4.988,
+      "num_input_tokens_seen": 584712192,
+      "step": 4461
+    },
+    {
+      "epoch": 0.7325500894867206,
+      "grad_norm": 0.7421653270721436,
+      "learning_rate": 7.841679379053378e-05,
+      "loss": 5.0049,
+      "num_input_tokens_seen": 585105408,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7330423946543864,
+      "grad_norm": 0.5522703528404236,
+      "learning_rate": 7.839045733511741e-05,
+      "loss": 5.0234,
+      "num_input_tokens_seen": 585498624,
+      "step": 4467
+    },
+    {
+      "epoch": 0.7335346998220522,
+      "grad_norm": 0.6532583832740784,
+      "learning_rate": 7.836414739736173e-05,
+      "loss": 4.9713,
+      "num_input_tokens_seen": 585891840,
+      "step": 4470
+    },
+    {
+      "epoch": 0.734027004989718,
+      "grad_norm": 0.6363163590431213,
+      "learning_rate": 7.833786393279637e-05,
+      "loss": 5.0337,
+      "num_input_tokens_seen": 586285056,
+      "step": 4473
+    },
+    {
+      "epoch": 0.7345193101573838,
+      "grad_norm": 0.7302047610282898,
+      "learning_rate": 7.831160689705535e-05,
+      "loss": 5.0239,
+      "num_input_tokens_seen": 586678272,
+      "step": 4476
+    },
+    {
+      "epoch": 0.7350116153250497,
+      "grad_norm": 0.5957253575325012,
+      "learning_rate": 7.828537624587667e-05,
+      "loss": 5.0409,
+      "num_input_tokens_seen": 587071488,
+      "step": 4479
+    },
+    {
+      "epoch": 0.7355039204927154,
+      "grad_norm": 0.675913393497467,
+      "learning_rate": 7.825917193510209e-05,
+      "loss": 5.004,
+      "num_input_tokens_seen": 587464704,
+      "step": 4482
+    },
+    {
+      "epoch": 0.7359962256603813,
+      "grad_norm": 0.6078398823738098,
+      "learning_rate": 7.823299392067672e-05,
+      "loss": 5.012,
+      "num_input_tokens_seen": 587857920,
+      "step": 4485
+    },
+    {
+      "epoch": 0.736488530828047,
+      "grad_norm": 0.7093663215637207,
+      "learning_rate": 7.82068421586488e-05,
+      "loss": 4.9781,
+      "num_input_tokens_seen": 588251136,
+      "step": 4488
+    },
+    {
+      "epoch": 0.7369808359957128,
+      "grad_norm": 0.5911455154418945,
+      "learning_rate": 7.81807166051693e-05,
+      "loss": 5.0096,
+      "num_input_tokens_seen": 588644352,
+      "step": 4491
+    },
+    {
+      "epoch": 0.7374731411633787,
+      "grad_norm": 0.5189294815063477,
+      "learning_rate": 7.815461721649169e-05,
+      "loss": 5.0144,
+      "num_input_tokens_seen": 589037568,
+      "step": 4494
+    },
+    {
+      "epoch": 0.7379654463310444,
+      "grad_norm": 0.5418742299079895,
+      "learning_rate": 7.812854394897162e-05,
+      "loss": 5.0046,
+      "num_input_tokens_seen": 589430784,
+      "step": 4497
+    },
+    {
+      "epoch": 0.7384577514987103,
+      "grad_norm": 0.6336432099342346,
+      "learning_rate": 7.810249675906653e-05,
+      "loss": 4.998,
+      "num_input_tokens_seen": 589824000,
+      "step": 4500
+    },
+    {
+      "epoch": 0.738950056666376,
+      "grad_norm": 0.5630114078521729,
+      "learning_rate": 7.807647560333547e-05,
+      "loss": 5.0358,
+      "num_input_tokens_seen": 590217216,
+      "step": 4503
+    },
+    {
+      "epoch": 0.7394423618340419,
+      "grad_norm": 0.7701685428619385,
+      "learning_rate": 7.80504804384387e-05,
+      "loss": 5.008,
+      "num_input_tokens_seen": 590610432,
+      "step": 4506
+    },
+    {
+      "epoch": 0.7399346670017077,
+      "grad_norm": 0.7231709361076355,
+      "learning_rate": 7.802451122113745e-05,
+      "loss": 5.036,
+      "num_input_tokens_seen": 591003648,
+      "step": 4509
+    },
+    {
+      "epoch": 0.7404269721693735,
+      "grad_norm": 0.5655918717384338,
+      "learning_rate": 7.799856790829355e-05,
+      "loss": 4.991,
+      "num_input_tokens_seen": 591396864,
+      "step": 4512
+    },
+    {
+      "epoch": 0.7409192773370393,
+      "grad_norm": 0.6875422596931458,
+      "learning_rate": 7.797265045686918e-05,
+      "loss": 5.0266,
+      "num_input_tokens_seen": 591790080,
+      "step": 4515
+    },
+    {
+      "epoch": 0.7414115825047051,
+      "grad_norm": 0.7806153297424316,
+      "learning_rate": 7.794675882392659e-05,
+      "loss": 4.9771,
+      "num_input_tokens_seen": 592183296,
+      "step": 4518
+    },
+    {
+      "epoch": 0.7419038876723709,
+      "grad_norm": 0.5837035775184631,
+      "learning_rate": 7.792089296662772e-05,
+      "loss": 5.0107,
+      "num_input_tokens_seen": 592576512,
+      "step": 4521
+    },
+    {
+      "epoch": 0.7423961928400368,
+      "grad_norm": 0.6467218399047852,
+      "learning_rate": 7.789505284223402e-05,
+      "loss": 5.0342,
+      "num_input_tokens_seen": 592969728,
+      "step": 4524
+    },
+    {
+      "epoch": 0.7428884980077025,
+      "grad_norm": 0.7257503867149353,
+      "learning_rate": 7.786923840810598e-05,
+      "loss": 5.0151,
+      "num_input_tokens_seen": 593362944,
+      "step": 4527
+    },
+    {
+      "epoch": 0.7433808031753684,
+      "grad_norm": 0.5904244184494019,
+      "learning_rate": 7.784344962170305e-05,
+      "loss": 4.9642,
+      "num_input_tokens_seen": 593756160,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7438731083430341,
+      "grad_norm": 0.6016355752944946,
+      "learning_rate": 7.781768644058319e-05,
+      "loss": 5.0225,
+      "num_input_tokens_seen": 594149376,
+      "step": 4533
+    },
+    {
+      "epoch": 0.7443654135107,
+      "grad_norm": 0.6705912947654724,
+      "learning_rate": 7.779194882240258e-05,
+      "loss": 5.0378,
+      "num_input_tokens_seen": 594542592,
+      "step": 4536
+    },
+    {
+      "epoch": 0.7448577186783657,
+      "grad_norm": 0.6506786942481995,
+      "learning_rate": 7.776623672491541e-05,
+      "loss": 5.0302,
+      "num_input_tokens_seen": 594935808,
+      "step": 4539
+    },
+    {
+      "epoch": 0.7453500238460316,
+      "grad_norm": 0.6254527568817139,
+      "learning_rate": 7.77405501059736e-05,
+      "loss": 4.9927,
+      "num_input_tokens_seen": 595329024,
+      "step": 4542
+    },
+    {
+      "epoch": 0.7458423290136974,
+      "grad_norm": 0.6810278296470642,
+      "learning_rate": 7.771488892352636e-05,
+      "loss": 4.9727,
+      "num_input_tokens_seen": 595722240,
+      "step": 4545
+    },
+    {
+      "epoch": 0.7463346341813631,
+      "grad_norm": 0.6676033139228821,
+      "learning_rate": 7.768925313562004e-05,
+      "loss": 4.952,
+      "num_input_tokens_seen": 596115456,
+      "step": 4548
+    },
+    {
+      "epoch": 0.746826939349029,
+      "grad_norm": 0.5896539688110352,
+      "learning_rate": 7.766364270039782e-05,
+      "loss": 4.972,
+      "num_input_tokens_seen": 596508672,
+      "step": 4551
+    },
+    {
+      "epoch": 0.7473192445166947,
+      "grad_norm": 0.5440996885299683,
+      "learning_rate": 7.763805757609938e-05,
+      "loss": 5.0342,
+      "num_input_tokens_seen": 596901888,
+      "step": 4554
+    },
+    {
+      "epoch": 0.7478115496843606,
+      "grad_norm": 0.6157152056694031,
+      "learning_rate": 7.761249772106066e-05,
+      "loss": 5.0039,
+      "num_input_tokens_seen": 597295104,
+      "step": 4557
+    },
+    {
+      "epoch": 0.7483038548520264,
+      "grad_norm": 0.5251792073249817,
+      "learning_rate": 7.758696309371352e-05,
+      "loss": 5.0235,
+      "num_input_tokens_seen": 597688320,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7487961600196922,
+      "grad_norm": 0.5628390908241272,
+      "learning_rate": 7.756145365258549e-05,
+      "loss": 4.9777,
+      "num_input_tokens_seen": 598081536,
+      "step": 4563
+    },
+    {
+      "epoch": 0.749288465187358,
+      "grad_norm": 0.5079121589660645,
+      "learning_rate": 7.753596935629956e-05,
+      "loss": 4.975,
+      "num_input_tokens_seen": 598474752,
+      "step": 4566
+    },
+    {
+      "epoch": 0.7497807703550238,
+      "grad_norm": 0.5366020798683167,
+      "learning_rate": 7.751051016357372e-05,
+      "loss": 5.0269,
+      "num_input_tokens_seen": 598867968,
+      "step": 4569
+    },
+    {
+      "epoch": 0.7502730755226896,
+      "grad_norm": 0.5857319235801697,
+      "learning_rate": 7.748507603322084e-05,
+      "loss": 4.9907,
+      "num_input_tokens_seen": 599261184,
+      "step": 4572
+    },
+    {
+      "epoch": 0.7507653806903555,
+      "grad_norm": 0.631243884563446,
+      "learning_rate": 7.745966692414832e-05,
+      "loss": 5.0049,
+      "num_input_tokens_seen": 599654400,
+      "step": 4575
+    },
+    {
+      "epoch": 0.7512576858580212,
+      "grad_norm": 0.554707407951355,
+      "learning_rate": 7.743428279535785e-05,
+      "loss": 4.9768,
+      "num_input_tokens_seen": 600047616,
+      "step": 4578
+    },
+    {
+      "epoch": 0.7517499910256871,
+      "grad_norm": 0.5789132118225098,
+      "learning_rate": 7.740892360594508e-05,
+      "loss": 5.0059,
+      "num_input_tokens_seen": 600440832,
+      "step": 4581
+    },
+    {
+      "epoch": 0.7522422961933528,
+      "grad_norm": 0.6106734275817871,
+      "learning_rate": 7.738358931509934e-05,
+      "loss": 5.0244,
+      "num_input_tokens_seen": 600834048,
+      "step": 4584
+    },
+    {
+      "epoch": 0.7527346013610187,
+      "grad_norm": 0.7959611415863037,
+      "learning_rate": 7.735827988210347e-05,
+      "loss": 4.9899,
+      "num_input_tokens_seen": 601227264,
+      "step": 4587
+    },
+    {
+      "epoch": 0.7532269065286845,
+      "grad_norm": 0.6123759150505066,
+      "learning_rate": 7.733299526633342e-05,
+      "loss": 5.0084,
+      "num_input_tokens_seen": 601620480,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7537192116963503,
+      "grad_norm": 0.8088337779045105,
+      "learning_rate": 7.730773542725799e-05,
+      "loss": 5.0053,
+      "num_input_tokens_seen": 602013696,
+      "step": 4593
+    },
+    {
+      "epoch": 0.7542115168640161,
+      "grad_norm": 0.6326596140861511,
+      "learning_rate": 7.72825003244387e-05,
+      "loss": 4.9837,
+      "num_input_tokens_seen": 602406912,
+      "step": 4596
+    },
+    {
+      "epoch": 0.7547038220316818,
+      "grad_norm": 0.8043537139892578,
+      "learning_rate": 7.72572899175293e-05,
+      "loss": 4.9436,
+      "num_input_tokens_seen": 602800128,
+      "step": 4599
+    },
+    {
+      "epoch": 0.7551961271993477,
+      "grad_norm": 0.70109623670578,
+      "learning_rate": 7.723210416627567e-05,
+      "loss": 5.0085,
+      "num_input_tokens_seen": 603193344,
+      "step": 4602
+    },
+    {
+      "epoch": 0.7556884323670136,
+      "grad_norm": 0.6619580984115601,
+      "learning_rate": 7.720694303051547e-05,
+      "loss": 5.0244,
+      "num_input_tokens_seen": 603586560,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7561807375346793,
+      "grad_norm": 0.6413691639900208,
+      "learning_rate": 7.718180647017793e-05,
+      "loss": 4.9864,
+      "num_input_tokens_seen": 603979776,
+      "step": 4608
+    },
+    {
+      "epoch": 0.7566730427023451,
+      "grad_norm": 0.6121329665184021,
+      "learning_rate": 7.715669444528345e-05,
+      "loss": 4.9885,
+      "num_input_tokens_seen": 604372992,
+      "step": 4611
+    },
+    {
+      "epoch": 0.7571653478700109,
+      "grad_norm": 0.5942865014076233,
+      "learning_rate": 7.713160691594354e-05,
+      "loss": 4.9964,
+      "num_input_tokens_seen": 604766208,
+      "step": 4614
+    },
+    {
+      "epoch": 0.7576576530376767,
+      "grad_norm": 0.7060537934303284,
+      "learning_rate": 7.710654384236036e-05,
+      "loss": 5.0532,
+      "num_input_tokens_seen": 605159424,
+      "step": 4617
+    },
+    {
+      "epoch": 0.7581499582053426,
+      "grad_norm": 0.6082600355148315,
+      "learning_rate": 7.708150518482654e-05,
+      "loss": 5.0,
+      "num_input_tokens_seen": 605552640,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7586422633730083,
+      "grad_norm": 0.5539378523826599,
+      "learning_rate": 7.705649090372494e-05,
+      "loss": 4.9649,
+      "num_input_tokens_seen": 605945856,
+      "step": 4623
+    },
+    {
+      "epoch": 0.7591345685406742,
+      "grad_norm": 0.6369094848632812,
+      "learning_rate": 7.703150095952836e-05,
+      "loss": 4.9854,
+      "num_input_tokens_seen": 606339072,
+      "step": 4626
+    },
+    {
+      "epoch": 0.7596268737083399,
+      "grad_norm": 0.6065540909767151,
+      "learning_rate": 7.700653531279927e-05,
+      "loss": 5.0692,
+      "num_input_tokens_seen": 606732288,
+      "step": 4629
+    },
+    {
+      "epoch": 0.7601191788760058,
+      "grad_norm": 0.6990041136741638,
+      "learning_rate": 7.698159392418949e-05,
+      "loss": 5.0129,
+      "num_input_tokens_seen": 607125504,
+      "step": 4632
+    },
+    {
+      "epoch": 0.7606114840436716,
+      "grad_norm": 0.5794108510017395,
+      "learning_rate": 7.695667675444007e-05,
+      "loss": 5.0128,
+      "num_input_tokens_seen": 607518720,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7611037892113374,
+      "grad_norm": 0.6325658559799194,
+      "learning_rate": 7.693178376438095e-05,
+      "loss": 5.0184,
+      "num_input_tokens_seen": 607911936,
+      "step": 4638
+    },
+    {
+      "epoch": 0.7615960943790032,
+      "grad_norm": 0.6222713589668274,
+      "learning_rate": 7.690691491493065e-05,
+      "loss": 5.0287,
+      "num_input_tokens_seen": 608305152,
+      "step": 4641
+    },
+    {
+      "epoch": 0.762088399546669,
+      "grad_norm": 0.7314369082450867,
+      "learning_rate": 7.688207016709613e-05,
+      "loss": 5.0145,
+      "num_input_tokens_seen": 608698368,
+      "step": 4644
+    },
+    {
+      "epoch": 0.7625807047143348,
+      "grad_norm": 0.6101865768432617,
+      "learning_rate": 7.685724948197242e-05,
+      "loss": 5.0326,
+      "num_input_tokens_seen": 609091584,
+      "step": 4647
+    },
+    {
+      "epoch": 0.7630730098820006,
+      "grad_norm": 0.6560894846916199,
+      "learning_rate": 7.683245282074242e-05,
+      "loss": 4.9619,
+      "num_input_tokens_seen": 609484800,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7635653150496664,
+      "grad_norm": 0.5975686311721802,
+      "learning_rate": 7.68076801446767e-05,
+      "loss": 5.0259,
+      "num_input_tokens_seen": 609878016,
+      "step": 4653
+    },
+    {
+      "epoch": 0.7640576202173323,
+      "grad_norm": 0.5379384160041809,
+      "learning_rate": 7.678293141513307e-05,
+      "loss": 4.9856,
+      "num_input_tokens_seen": 610271232,
+      "step": 4656
+    },
+    {
+      "epoch": 0.764549925384998,
+      "grad_norm": 0.5923475623130798,
+      "learning_rate": 7.675820659355657e-05,
+      "loss": 5.0146,
+      "num_input_tokens_seen": 610664448,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7650422305526638,
+      "grad_norm": 0.5225929617881775,
+      "learning_rate": 7.673350564147898e-05,
+      "loss": 5.037,
+      "num_input_tokens_seen": 611057664,
+      "step": 4662
+    },
+    {
+      "epoch": 0.7655345357203296,
+      "grad_norm": 0.5816430449485779,
+      "learning_rate": 7.670882852051876e-05,
+      "loss": 4.9901,
+      "num_input_tokens_seen": 611450880,
+      "step": 4665
+    },
+    {
+      "epoch": 0.7660268408879954,
+      "grad_norm": 0.5520696043968201,
+      "learning_rate": 7.668417519238068e-05,
+      "loss": 4.9269,
+      "num_input_tokens_seen": 611844096,
+      "step": 4668
+    },
+    {
+      "epoch": 0.7665191460556613,
+      "grad_norm": 0.5480571985244751,
+      "learning_rate": 7.66595456188556e-05,
+      "loss": 4.9748,
+      "num_input_tokens_seen": 612237312,
+      "step": 4671
+    },
+    {
+      "epoch": 0.767011451223327,
+      "grad_norm": 0.609230101108551,
+      "learning_rate": 7.663493976182024e-05,
+      "loss": 4.9698,
+      "num_input_tokens_seen": 612630528,
+      "step": 4674
+    },
+    {
+      "epoch": 0.7675037563909929,
+      "grad_norm": 0.7071127891540527,
+      "learning_rate": 7.661035758323693e-05,
+      "loss": 5.0464,
+      "num_input_tokens_seen": 613023744,
+      "step": 4677
+    },
+    {
+      "epoch": 0.7679960615586586,
+      "grad_norm": 0.7234042286872864,
+      "learning_rate": 7.658579904515338e-05,
+      "loss": 4.9778,
+      "num_input_tokens_seen": 613416960,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7684883667263245,
+      "grad_norm": 0.652027428150177,
+      "learning_rate": 7.656126410970232e-05,
+      "loss": 4.971,
+      "num_input_tokens_seen": 613810176,
+      "step": 4683
+    },
+    {
+      "epoch": 0.7689806718939903,
+      "grad_norm": 0.7037466168403625,
+      "learning_rate": 7.653675273910141e-05,
+      "loss": 5.0228,
+      "num_input_tokens_seen": 614203392,
+      "step": 4686
+    },
+    {
+      "epoch": 0.7694729770616561,
+      "grad_norm": 0.5814207196235657,
+      "learning_rate": 7.651226489565297e-05,
+      "loss": 4.9948,
+      "num_input_tokens_seen": 614596608,
+      "step": 4689
+    },
+    {
+      "epoch": 0.7699652822293219,
+      "grad_norm": 0.7240739464759827,
+      "learning_rate": 7.64878005417436e-05,
+      "loss": 5.0276,
+      "num_input_tokens_seen": 614989824,
+      "step": 4692
+    },
+    {
+      "epoch": 0.7704575873969877,
+      "grad_norm": 0.7340640425682068,
+      "learning_rate": 7.646335963984411e-05,
+      "loss": 4.984,
+      "num_input_tokens_seen": 615383040,
+      "step": 4695
+    },
+    {
+      "epoch": 0.7709498925646535,
+      "grad_norm": 0.6027811765670776,
+      "learning_rate": 7.643894215250911e-05,
+      "loss": 4.9804,
+      "num_input_tokens_seen": 615776256,
+      "step": 4698
+    },
+    {
+      "epoch": 0.7714421977323194,
+      "grad_norm": 0.5704403519630432,
+      "learning_rate": 7.641454804237697e-05,
+      "loss": 4.9516,
+      "num_input_tokens_seen": 616169472,
+      "step": 4701
+    },
+    {
+      "epoch": 0.7719345028999851,
+      "grad_norm": 0.5998306274414062,
+      "learning_rate": 7.639017727216942e-05,
+      "loss": 4.991,
+      "num_input_tokens_seen": 616562688,
+      "step": 4704
+    },
+    {
+      "epoch": 0.772426808067651,
+      "grad_norm": 0.6468780040740967,
+      "learning_rate": 7.636582980469131e-05,
+      "loss": 5.0081,
+      "num_input_tokens_seen": 616955904,
+      "step": 4707
+    },
+    {
+      "epoch": 0.7729191132353167,
+      "grad_norm": 0.5996934771537781,
+      "learning_rate": 7.634150560283054e-05,
+      "loss": 4.9508,
+      "num_input_tokens_seen": 617349120,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7734114184029826,
+      "grad_norm": 0.7018175721168518,
+      "learning_rate": 7.631720462955758e-05,
+      "loss": 5.0056,
+      "num_input_tokens_seen": 617742336,
+      "step": 4713
+    },
+    {
+      "epoch": 0.7739037235706484,
+      "grad_norm": 0.6839625239372253,
+      "learning_rate": 7.629292684792543e-05,
+      "loss": 4.9534,
+      "num_input_tokens_seen": 618135552,
+      "step": 4716
+    },
+    {
+      "epoch": 0.7743960287383141,
+      "grad_norm": 0.652540922164917,
+      "learning_rate": 7.62686722210693e-05,
+      "loss": 4.965,
+      "num_input_tokens_seen": 618528768,
+      "step": 4719
+    },
+    {
+      "epoch": 0.77488833390598,
+      "grad_norm": 0.7679561972618103,
+      "learning_rate": 7.624444071220638e-05,
+      "loss": 4.9414,
+      "num_input_tokens_seen": 618921984,
+      "step": 4722
+    },
+    {
+      "epoch": 0.7753806390736457,
+      "grad_norm": 0.7038884162902832,
+      "learning_rate": 7.62202322846356e-05,
+      "loss": 4.9849,
+      "num_input_tokens_seen": 619315200,
+      "step": 4725
+    },
+    {
+      "epoch": 0.7758729442413116,
+      "grad_norm": 0.6602532267570496,
+      "learning_rate": 7.619604690173746e-05,
+      "loss": 5.0283,
+      "num_input_tokens_seen": 619708416,
+      "step": 4728
+    },
+    {
+      "epoch": 0.7763652494089774,
+      "grad_norm": 0.6303278803825378,
+      "learning_rate": 7.617188452697369e-05,
+      "loss": 4.9665,
+      "num_input_tokens_seen": 620101632,
+      "step": 4731
+    },
+    {
+      "epoch": 0.7768575545766432,
+      "grad_norm": 0.669486403465271,
+      "learning_rate": 7.614774512388709e-05,
+      "loss": 4.9801,
+      "num_input_tokens_seen": 620494848,
+      "step": 4734
+    },
+    {
+      "epoch": 0.777349859744309,
+      "grad_norm": 0.6238739490509033,
+      "learning_rate": 7.61236286561013e-05,
+      "loss": 5.0117,
+      "num_input_tokens_seen": 620888064,
+      "step": 4737
+    },
+    {
+      "epoch": 0.7778421649119748,
+      "grad_norm": 0.5390612483024597,
+      "learning_rate": 7.609953508732051e-05,
+      "loss": 4.966,
+      "num_input_tokens_seen": 621281280,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7783344700796406,
+      "grad_norm": 0.6080785393714905,
+      "learning_rate": 7.607546438132936e-05,
+      "loss": 4.9709,
+      "num_input_tokens_seen": 621674496,
+      "step": 4743
+    },
+    {
+      "epoch": 0.7788267752473064,
+      "grad_norm": 0.5530922412872314,
+      "learning_rate": 7.605141650199254e-05,
+      "loss": 4.9836,
+      "num_input_tokens_seen": 622067712,
+      "step": 4746
+    },
+    {
+      "epoch": 0.7793190804149722,
+      "grad_norm": 0.5536843538284302,
+      "learning_rate": 7.602739141325468e-05,
+      "loss": 4.9132,
+      "num_input_tokens_seen": 622460928,
+      "step": 4749
+    },
+    {
+      "epoch": 0.7798113855826381,
+      "grad_norm": 0.6630847454071045,
+      "learning_rate": 7.600338907914008e-05,
+      "loss": 4.9765,
+      "num_input_tokens_seen": 622854144,
+      "step": 4752
+    },
+    {
+      "epoch": 0.7803036907503038,
+      "grad_norm": 0.5388240814208984,
+      "learning_rate": 7.597940946375254e-05,
+      "loss": 4.9774,
+      "num_input_tokens_seen": 623247360,
+      "step": 4755
+    },
+    {
+      "epoch": 0.7807959959179697,
+      "grad_norm": 0.5779323577880859,
+      "learning_rate": 7.595545253127499e-05,
+      "loss": 4.9994,
+      "num_input_tokens_seen": 623640576,
+      "step": 4758
+    },
+    {
+      "epoch": 0.7812883010856354,
+      "grad_norm": 0.622727632522583,
+      "learning_rate": 7.593151824596948e-05,
+      "loss": 5.0106,
+      "num_input_tokens_seen": 624033792,
+      "step": 4761
+    },
+    {
+      "epoch": 0.7817806062533013,
+      "grad_norm": 0.6555529236793518,
+      "learning_rate": 7.590760657217675e-05,
+      "loss": 4.9756,
+      "num_input_tokens_seen": 624427008,
+      "step": 4764
+    },
+    {
+      "epoch": 0.7822729114209671,
+      "grad_norm": 0.6330806612968445,
+      "learning_rate": 7.588371747431617e-05,
+      "loss": 5.0433,
+      "num_input_tokens_seen": 624820224,
+      "step": 4767
+    },
+    {
+      "epoch": 0.7827652165886329,
+      "grad_norm": 0.5727117657661438,
+      "learning_rate": 7.585985091688538e-05,
+      "loss": 4.9545,
+      "num_input_tokens_seen": 625213440,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7832575217562987,
+      "grad_norm": 0.5896506309509277,
+      "learning_rate": 7.583600686446016e-05,
+      "loss": 4.9836,
+      "num_input_tokens_seen": 625606656,
+      "step": 4773
+    },
+    {
+      "epoch": 0.7837498269239644,
+      "grad_norm": 0.5335837602615356,
+      "learning_rate": 7.581218528169419e-05,
+      "loss": 4.9992,
+      "num_input_tokens_seen": 625999872,
+      "step": 4776
+    },
+    {
+      "epoch": 0.7842421320916303,
+      "grad_norm": 0.6446117758750916,
+      "learning_rate": 7.578838613331881e-05,
+      "loss": 4.9502,
+      "num_input_tokens_seen": 626393088,
+      "step": 4779
+    },
+    {
+      "epoch": 0.7847344372592961,
+      "grad_norm": 0.5382349491119385,
+      "learning_rate": 7.576460938414284e-05,
+      "loss": 4.9647,
+      "num_input_tokens_seen": 626786304,
+      "step": 4782
+    },
+    {
+      "epoch": 0.7852267424269619,
+      "grad_norm": 0.7169778943061829,
+      "learning_rate": 7.574085499905227e-05,
+      "loss": 4.9854,
+      "num_input_tokens_seen": 627179520,
+      "step": 4785
+    },
+    {
+      "epoch": 0.7857190475946277,
+      "grad_norm": 0.5940950512886047,
+      "learning_rate": 7.571712294301016e-05,
+      "loss": 5.0078,
+      "num_input_tokens_seen": 627572736,
+      "step": 4788
+    },
+    {
+      "epoch": 0.7862113527622935,
+      "grad_norm": 0.6756063103675842,
+      "learning_rate": 7.56934131810564e-05,
+      "loss": 4.9913,
+      "num_input_tokens_seen": 627965952,
+      "step": 4791
+    },
+    {
+      "epoch": 0.7867036579299593,
+      "grad_norm": 0.5659769177436829,
+      "learning_rate": 7.566972567830736e-05,
+      "loss": 5.0014,
+      "num_input_tokens_seen": 628359168,
+      "step": 4794
+    },
+    {
+      "epoch": 0.7871959630976252,
+      "grad_norm": 0.595694363117218,
+      "learning_rate": 7.564606039995587e-05,
+      "loss": 4.9739,
+      "num_input_tokens_seen": 628752384,
+      "step": 4797
+    },
+    {
+      "epoch": 0.7876882682652909,
+      "grad_norm": 0.6841439604759216,
+      "learning_rate": 7.562241731127086e-05,
+      "loss": 4.9732,
+      "num_input_tokens_seen": 629145600,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7876882682652909,
+      "eval_accuracy": 0.22760788145253216,
+      "eval_loss": 5.267364978790283,
+      "eval_runtime": 115.6402,
+      "eval_samples_per_second": 2.594,
+      "eval_steps_per_second": 1.297,
+      "num_input_tokens_seen": 629145600,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7881805734329568,
+      "grad_norm": 0.5944079160690308,
+      "learning_rate": 7.559879637759725e-05,
+      "loss": 4.956,
+      "num_input_tokens_seen": 629538816,
+      "step": 4803
+    },
+    {
+      "epoch": 0.7886728786006225,
+      "grad_norm": 0.6290807723999023,
+      "learning_rate": 7.557519756435562e-05,
+      "loss": 4.9765,
+      "num_input_tokens_seen": 629932032,
+      "step": 4806
+    },
+    {
+      "epoch": 0.7891651837682884,
+      "grad_norm": 0.6066911220550537,
+      "learning_rate": 7.555162083704213e-05,
+      "loss": 5.0129,
+      "num_input_tokens_seen": 630325248,
+      "step": 4809
+    },
+    {
+      "epoch": 0.7896574889359542,
+      "grad_norm": 0.5937833786010742,
+      "learning_rate": 7.552806616122819e-05,
+      "loss": 4.9783,
+      "num_input_tokens_seen": 630718464,
+      "step": 4812
+    },
+    {
+      "epoch": 0.79014979410362,
+      "grad_norm": 0.6029826402664185,
+      "learning_rate": 7.550453350256031e-05,
+      "loss": 4.9904,
+      "num_input_tokens_seen": 631111680,
+      "step": 4815
+    },
+    {
+      "epoch": 0.7906420992712858,
+      "grad_norm": 0.6355852484703064,
+      "learning_rate": 7.548102282675988e-05,
+      "loss": 4.9274,
+      "num_input_tokens_seen": 631504896,
+      "step": 4818
+    },
+    {
+      "epoch": 0.7911344044389516,
+      "grad_norm": 0.6588253378868103,
+      "learning_rate": 7.545753409962299e-05,
+      "loss": 4.9691,
+      "num_input_tokens_seen": 631898112,
+      "step": 4821
+    },
+    {
+      "epoch": 0.7916267096066174,
+      "grad_norm": 0.5973790884017944,
+      "learning_rate": 7.543406728702016e-05,
+      "loss": 4.9812,
+      "num_input_tokens_seen": 632291328,
+      "step": 4824
+    },
+    {
+      "epoch": 0.7921190147742833,
+      "grad_norm": 0.6492071151733398,
+      "learning_rate": 7.541062235489617e-05,
+      "loss": 4.9996,
+      "num_input_tokens_seen": 632684544,
+      "step": 4827
+    },
+    {
+      "epoch": 0.792611319941949,
+      "grad_norm": 0.5668359994888306,
+      "learning_rate": 7.538719926926982e-05,
+      "loss": 4.9706,
+      "num_input_tokens_seen": 633077760,
+      "step": 4830
+    },
+    {
+      "epoch": 0.7931036251096149,
+      "grad_norm": 0.6291391849517822,
+      "learning_rate": 7.536379799623378e-05,
+      "loss": 5.001,
+      "num_input_tokens_seen": 633470976,
+      "step": 4833
+    },
+    {
+      "epoch": 0.7935959302772806,
+      "grad_norm": 0.5828115940093994,
+      "learning_rate": 7.534041850195436e-05,
+      "loss": 4.9702,
+      "num_input_tokens_seen": 633864192,
+      "step": 4836
+    },
+    {
+      "epoch": 0.7940882354449464,
+      "grad_norm": 0.6520923376083374,
+      "learning_rate": 7.531706075267127e-05,
+      "loss": 4.9241,
+      "num_input_tokens_seen": 634257408,
+      "step": 4839
+    },
+    {
+      "epoch": 0.7945805406126123,
+      "grad_norm": 0.5931693911552429,
+      "learning_rate": 7.529372471469743e-05,
+      "loss": 4.9394,
+      "num_input_tokens_seen": 634650624,
+      "step": 4842
+    },
+    {
+      "epoch": 0.795072845780278,
+      "grad_norm": 0.684350311756134,
+      "learning_rate": 7.527041035441882e-05,
+      "loss": 4.9617,
+      "num_input_tokens_seen": 635043840,
+      "step": 4845
+    },
+    {
+      "epoch": 0.7955651509479439,
+      "grad_norm": 0.5431036949157715,
+      "learning_rate": 7.52471176382942e-05,
+      "loss": 4.9856,
+      "num_input_tokens_seen": 635437056,
+      "step": 4848
+    },
+    {
+      "epoch": 0.7960574561156096,
+      "grad_norm": 0.5861454606056213,
+      "learning_rate": 7.522384653285496e-05,
+      "loss": 4.972,
+      "num_input_tokens_seen": 635830272,
+      "step": 4851
+    },
+    {
+      "epoch": 0.7965497612832755,
+      "grad_norm": 0.6265558004379272,
+      "learning_rate": 7.520059700470488e-05,
+      "loss": 4.9853,
+      "num_input_tokens_seen": 636223488,
+      "step": 4854
+    },
+    {
+      "epoch": 0.7970420664509412,
+      "grad_norm": 0.6842302083969116,
+      "learning_rate": 7.517736902051998e-05,
+      "loss": 4.981,
+      "num_input_tokens_seen": 636616704,
+      "step": 4857
+    },
+    {
+      "epoch": 0.7975343716186071,
+      "grad_norm": 0.5554177761077881,
+      "learning_rate": 7.515416254704823e-05,
+      "loss": 4.9695,
+      "num_input_tokens_seen": 637009920,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7980266767862729,
+      "grad_norm": 0.6592026352882385,
+      "learning_rate": 7.513097755110948e-05,
+      "loss": 4.9942,
+      "num_input_tokens_seen": 637403136,
+      "step": 4863
+    },
+    {
+      "epoch": 0.7985189819539387,
+      "grad_norm": 0.5816788077354431,
+      "learning_rate": 7.510781399959514e-05,
+      "loss": 5.0034,
+      "num_input_tokens_seen": 637796352,
+      "step": 4866
+    },
+    {
+      "epoch": 0.7990112871216045,
+      "grad_norm": 0.6009346842765808,
+      "learning_rate": 7.508467185946803e-05,
+      "loss": 4.9395,
+      "num_input_tokens_seen": 638189568,
+      "step": 4869
+    },
+    {
+      "epoch": 0.7995035922892703,
+      "grad_norm": 0.6724573373794556,
+      "learning_rate": 7.506155109776222e-05,
+      "loss": 5.0087,
+      "num_input_tokens_seen": 638582784,
+      "step": 4872
+    },
+    {
+      "epoch": 0.7999958974569361,
+      "grad_norm": 0.6114869117736816,
+      "learning_rate": 7.503845168158274e-05,
+      "loss": 4.9479,
+      "num_input_tokens_seen": 638976000,
+      "step": 4875
+    },
+    {
+      "epoch": 0.800488202624602,
+      "grad_norm": 0.7219544649124146,
+      "learning_rate": 7.501537357810552e-05,
+      "loss": 4.9821,
+      "num_input_tokens_seen": 639369216,
+      "step": 4878
+    },
+    {
+      "epoch": 0.8009805077922677,
+      "grad_norm": 0.6012791395187378,
+      "learning_rate": 7.499231675457698e-05,
+      "loss": 4.9551,
+      "num_input_tokens_seen": 639762432,
+      "step": 4881
+    },
+    {
+      "epoch": 0.8014728129599336,
+      "grad_norm": 0.6682071685791016,
+      "learning_rate": 7.496928117831408e-05,
+      "loss": 4.9587,
+      "num_input_tokens_seen": 640155648,
+      "step": 4884
+    },
+    {
+      "epoch": 0.8019651181275993,
+      "grad_norm": 0.6674060821533203,
+      "learning_rate": 7.494626681670401e-05,
+      "loss": 4.9892,
+      "num_input_tokens_seen": 640548864,
+      "step": 4887
+    },
+    {
+      "epoch": 0.8024574232952651,
+      "grad_norm": 0.6695443391799927,
+      "learning_rate": 7.492327363720392e-05,
+      "loss": 4.9558,
+      "num_input_tokens_seen": 640942080,
+      "step": 4890
+    },
+    {
+      "epoch": 0.802949728462931,
+      "grad_norm": 0.6352146863937378,
+      "learning_rate": 7.490030160734087e-05,
+      "loss": 5.0049,
+      "num_input_tokens_seen": 641335296,
+      "step": 4893
+    },
+    {
+      "epoch": 0.8034420336305967,
+      "grad_norm": 0.6130344271659851,
+      "learning_rate": 7.487735069471156e-05,
+      "loss": 5.0011,
+      "num_input_tokens_seen": 641728512,
+      "step": 4896
+    },
+    {
+      "epoch": 0.8039343387982626,
+      "grad_norm": 0.7044432163238525,
+      "learning_rate": 7.485442086698213e-05,
+      "loss": 4.9738,
+      "num_input_tokens_seen": 642121728,
+      "step": 4899
+    },
+    {
+      "epoch": 0.8044266439659283,
+      "grad_norm": 0.5631691217422485,
+      "learning_rate": 7.483151209188803e-05,
+      "loss": 4.9646,
+      "num_input_tokens_seen": 642514944,
+      "step": 4902
+    },
+    {
+      "epoch": 0.8049189491335942,
+      "grad_norm": 0.6033437848091125,
+      "learning_rate": 7.480862433723372e-05,
+      "loss": 4.9488,
+      "num_input_tokens_seen": 642908160,
+      "step": 4905
+    },
+    {
+      "epoch": 0.80541125430126,
+      "grad_norm": 0.577086329460144,
+      "learning_rate": 7.478575757089261e-05,
+      "loss": 4.9401,
+      "num_input_tokens_seen": 643301376,
+      "step": 4908
+    },
+    {
+      "epoch": 0.8059035594689258,
+      "grad_norm": 0.6318846940994263,
+      "learning_rate": 7.47629117608068e-05,
+      "loss": 5.0187,
+      "num_input_tokens_seen": 643694592,
+      "step": 4911
+    },
+    {
+      "epoch": 0.8063958646365916,
+      "grad_norm": 0.635599672794342,
+      "learning_rate": 7.474008687498688e-05,
+      "loss": 4.9588,
+      "num_input_tokens_seen": 644087808,
+      "step": 4914
+    },
+    {
+      "epoch": 0.8068881698042574,
+      "grad_norm": 0.5748466849327087,
+      "learning_rate": 7.471728288151176e-05,
+      "loss": 4.9887,
+      "num_input_tokens_seen": 644481024,
+      "step": 4917
+    },
+    {
+      "epoch": 0.8073804749719232,
+      "grad_norm": 0.5932797789573669,
+      "learning_rate": 7.469449974852852e-05,
+      "loss": 4.974,
+      "num_input_tokens_seen": 644874240,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8078727801395891,
+      "grad_norm": 0.6433725953102112,
+      "learning_rate": 7.467173744425214e-05,
+      "loss": 4.9916,
+      "num_input_tokens_seen": 645267456,
+      "step": 4923
+    },
+    {
+      "epoch": 0.8083650853072548,
+      "grad_norm": 0.6424766778945923,
+      "learning_rate": 7.46489959369654e-05,
+      "loss": 5.0115,
+      "num_input_tokens_seen": 645660672,
+      "step": 4926
+    },
+    {
+      "epoch": 0.8088573904749207,
+      "grad_norm": 0.6171865463256836,
+      "learning_rate": 7.462627519501864e-05,
+      "loss": 5.0431,
+      "num_input_tokens_seen": 646053888,
+      "step": 4929
+    },
+    {
+      "epoch": 0.8093496956425864,
+      "grad_norm": 0.7216885089874268,
+      "learning_rate": 7.460357518682957e-05,
+      "loss": 4.988,
+      "num_input_tokens_seen": 646447104,
+      "step": 4932
+    },
+    {
+      "epoch": 0.8098420008102523,
+      "grad_norm": 0.7182767987251282,
+      "learning_rate": 7.458089588088317e-05,
+      "loss": 4.967,
+      "num_input_tokens_seen": 646840320,
+      "step": 4935
+    },
+    {
+      "epoch": 0.8103343059779181,
+      "grad_norm": 0.6302428841590881,
+      "learning_rate": 7.455823724573136e-05,
+      "loss": 4.9635,
+      "num_input_tokens_seen": 647233536,
+      "step": 4938
+    },
+    {
+      "epoch": 0.8108266111455839,
+      "grad_norm": 0.6370172500610352,
+      "learning_rate": 7.453559924999299e-05,
+      "loss": 4.9557,
+      "num_input_tokens_seen": 647626752,
+      "step": 4941
+    },
+    {
+      "epoch": 0.8113189163132497,
+      "grad_norm": 0.6621046662330627,
+      "learning_rate": 7.451298186235347e-05,
+      "loss": 4.9624,
+      "num_input_tokens_seen": 648019968,
+      "step": 4944
+    },
+    {
+      "epoch": 0.8118112214809154,
+      "grad_norm": 0.6274940967559814,
+      "learning_rate": 7.449038505156474e-05,
+      "loss": 4.9531,
+      "num_input_tokens_seen": 648413184,
+      "step": 4947
+    },
+    {
+      "epoch": 0.8123035266485813,
+      "grad_norm": 0.6614890098571777,
+      "learning_rate": 7.446780878644506e-05,
+      "loss": 4.9893,
+      "num_input_tokens_seen": 648806400,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8127958318162471,
+      "grad_norm": 0.7011924386024475,
+      "learning_rate": 7.444525303587875e-05,
+      "loss": 4.9168,
+      "num_input_tokens_seen": 649199616,
+      "step": 4953
+    },
+    {
+      "epoch": 0.8132881369839129,
+      "grad_norm": 0.5855690836906433,
+      "learning_rate": 7.442271776881608e-05,
+      "loss": 4.9858,
+      "num_input_tokens_seen": 649592832,
+      "step": 4956
+    },
+    {
+      "epoch": 0.8137804421515787,
+      "grad_norm": 0.7736572027206421,
+      "learning_rate": 7.440020295427308e-05,
+      "loss": 4.9729,
+      "num_input_tokens_seen": 649986048,
+      "step": 4959
+    },
+    {
+      "epoch": 0.8142727473192445,
+      "grad_norm": 0.7245569229125977,
+      "learning_rate": 7.437770856133132e-05,
+      "loss": 4.9601,
+      "num_input_tokens_seen": 650379264,
+      "step": 4962
+    },
+    {
+      "epoch": 0.8147650524869103,
+      "grad_norm": 0.5991591811180115,
+      "learning_rate": 7.435523455913788e-05,
+      "loss": 4.9826,
+      "num_input_tokens_seen": 650772480,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8152573576545761,
+      "grad_norm": 0.680020272731781,
+      "learning_rate": 7.433278091690488e-05,
+      "loss": 4.9621,
+      "num_input_tokens_seen": 651165696,
+      "step": 4968
+    },
+    {
+      "epoch": 0.8157496628222419,
+      "grad_norm": 0.5399065613746643,
+      "learning_rate": 7.431034760390964e-05,
+      "loss": 4.9357,
+      "num_input_tokens_seen": 651558912,
+      "step": 4971
+    },
+    {
+      "epoch": 0.8162419679899078,
+      "grad_norm": 0.7673788666725159,
+      "learning_rate": 7.428793458949426e-05,
+      "loss": 4.98,
+      "num_input_tokens_seen": 651952128,
+      "step": 4974
+    },
+    {
+      "epoch": 0.8167342731575735,
+      "grad_norm": 0.5334210395812988,
+      "learning_rate": 7.426554184306555e-05,
+      "loss": 4.9979,
+      "num_input_tokens_seen": 652345344,
+      "step": 4977
+    },
+    {
+      "epoch": 0.8172265783252394,
+      "grad_norm": 0.7658263444900513,
+      "learning_rate": 7.424316933409486e-05,
+      "loss": 4.994,
+      "num_input_tokens_seen": 652738560,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8177188834929051,
+      "grad_norm": 0.6794727444648743,
+      "learning_rate": 7.42208170321178e-05,
+      "loss": 4.9735,
+      "num_input_tokens_seen": 653131776,
+      "step": 4983
+    },
+    {
+      "epoch": 0.818211188660571,
+      "grad_norm": 0.6948509812355042,
+      "learning_rate": 7.419848490673427e-05,
+      "loss": 4.9359,
+      "num_input_tokens_seen": 653524992,
+      "step": 4986
+    },
+    {
+      "epoch": 0.8187034938282368,
+      "grad_norm": 0.7119221687316895,
+      "learning_rate": 7.417617292760803e-05,
+      "loss": 4.9919,
+      "num_input_tokens_seen": 653918208,
+      "step": 4989
+    },
+    {
+      "epoch": 0.8191957989959026,
+      "grad_norm": 0.6246024966239929,
+      "learning_rate": 7.415388106446673e-05,
+      "loss": 4.9472,
+      "num_input_tokens_seen": 654311424,
+      "step": 4992
+    },
+    {
+      "epoch": 0.8196881041635684,
+      "grad_norm": 0.5838922262191772,
+      "learning_rate": 7.413160928710163e-05,
+      "loss": 4.9537,
+      "num_input_tokens_seen": 654704640,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8201804093312342,
+      "grad_norm": 0.6566455960273743,
+      "learning_rate": 7.410935756536754e-05,
+      "loss": 4.9545,
+      "num_input_tokens_seen": 655097856,
+      "step": 4998
+    },
+    {
+      "epoch": 0.8206727144989,
+      "grad_norm": 0.6249347925186157,
+      "learning_rate": 7.408712586918248e-05,
+      "loss": 4.9567,
+      "num_input_tokens_seen": 655491072,
+      "step": 5001
+    },
+    {
+      "epoch": 0.8211650196665659,
+      "grad_norm": 0.604857325553894,
+      "learning_rate": 7.406491416852765e-05,
+      "loss": 4.955,
+      "num_input_tokens_seen": 655884288,
+      "step": 5004
+    },
+    {
+      "epoch": 0.8216573248342316,
+      "grad_norm": 0.6290884017944336,
+      "learning_rate": 7.404272243344716e-05,
+      "loss": 4.9201,
+      "num_input_tokens_seen": 656277504,
+      "step": 5007
+    },
+    {
+      "epoch": 0.8221496300018974,
+      "grad_norm": 0.5492717623710632,
+      "learning_rate": 7.4020550634048e-05,
+      "loss": 5.0007,
+      "num_input_tokens_seen": 656670720,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8226419351695632,
+      "grad_norm": 0.5919643044471741,
+      "learning_rate": 7.39983987404997e-05,
+      "loss": 4.979,
+      "num_input_tokens_seen": 657063936,
+      "step": 5013
+    },
+    {
+      "epoch": 0.823134240337229,
+      "grad_norm": 0.6006259322166443,
+      "learning_rate": 7.39762667230343e-05,
+      "loss": 4.9535,
+      "num_input_tokens_seen": 657457152,
+      "step": 5016
+    },
+    {
+      "epoch": 0.8236265455048949,
+      "grad_norm": 0.5834858417510986,
+      "learning_rate": 7.395415455194607e-05,
+      "loss": 4.9279,
+      "num_input_tokens_seen": 657850368,
+      "step": 5019
+    },
+    {
+      "epoch": 0.8241188506725606,
+      "grad_norm": 0.5802402496337891,
+      "learning_rate": 7.393206219759145e-05,
+      "loss": 4.9848,
+      "num_input_tokens_seen": 658243584,
+      "step": 5022
+    },
+    {
+      "epoch": 0.8246111558402265,
+      "grad_norm": 0.6450897455215454,
+      "learning_rate": 7.390998963038879e-05,
+      "loss": 5.0166,
+      "num_input_tokens_seen": 658636800,
+      "step": 5025
+    },
+    {
+      "epoch": 0.8251034610078922,
+      "grad_norm": 0.5459268689155579,
+      "learning_rate": 7.388793682081826e-05,
+      "loss": 4.9584,
+      "num_input_tokens_seen": 659030016,
+      "step": 5028
+    },
+    {
+      "epoch": 0.8255957661755581,
+      "grad_norm": 0.6466140747070312,
+      "learning_rate": 7.386590373942163e-05,
+      "loss": 4.9798,
+      "num_input_tokens_seen": 659423232,
+      "step": 5031
+    },
+    {
+      "epoch": 0.8260880713432239,
+      "grad_norm": 0.6367524266242981,
+      "learning_rate": 7.384389035680211e-05,
+      "loss": 4.9722,
+      "num_input_tokens_seen": 659816448,
+      "step": 5034
+    },
+    {
+      "epoch": 0.8265803765108897,
+      "grad_norm": 0.5187392234802246,
+      "learning_rate": 7.382189664362424e-05,
+      "loss": 4.9654,
+      "num_input_tokens_seen": 660209664,
+      "step": 5037
+    },
+    {
+      "epoch": 0.8270726816785555,
+      "grad_norm": 0.5242530107498169,
+      "learning_rate": 7.379992257061364e-05,
+      "loss": 4.9528,
+      "num_input_tokens_seen": 660602880,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8275649868462213,
+      "grad_norm": 0.7018054723739624,
+      "learning_rate": 7.377796810855694e-05,
+      "loss": 4.9836,
+      "num_input_tokens_seen": 660996096,
+      "step": 5043
+    },
+    {
+      "epoch": 0.8280572920138871,
+      "grad_norm": 0.6890904903411865,
+      "learning_rate": 7.37560332283015e-05,
+      "loss": 5.0295,
+      "num_input_tokens_seen": 661389312,
+      "step": 5046
+    },
+    {
+      "epoch": 0.828549597181553,
+      "grad_norm": 0.5935149192810059,
+      "learning_rate": 7.373411790075538e-05,
+      "loss": 5.0017,
+      "num_input_tokens_seen": 661782528,
+      "step": 5049
+    },
+    {
+      "epoch": 0.8290419023492187,
+      "grad_norm": 0.6082046031951904,
+      "learning_rate": 7.371222209688707e-05,
+      "loss": 4.9301,
+      "num_input_tokens_seen": 662175744,
+      "step": 5052
+    },
+    {
+      "epoch": 0.8295342075168846,
+      "grad_norm": 0.6416882872581482,
+      "learning_rate": 7.36903457877254e-05,
+      "loss": 4.9764,
+      "num_input_tokens_seen": 662568960,
+      "step": 5055
+    },
+    {
+      "epoch": 0.8300265126845503,
+      "grad_norm": 0.6466403603553772,
+      "learning_rate": 7.366848894435935e-05,
+      "loss": 4.9977,
+      "num_input_tokens_seen": 662962176,
+      "step": 5058
+    },
+    {
+      "epoch": 0.8305188178522162,
+      "grad_norm": 0.5742791891098022,
+      "learning_rate": 7.364665153793785e-05,
+      "loss": 4.932,
+      "num_input_tokens_seen": 663355392,
+      "step": 5061
+    },
+    {
+      "epoch": 0.831011123019882,
+      "grad_norm": 0.5598244667053223,
+      "learning_rate": 7.362483353966971e-05,
+      "loss": 4.9783,
+      "num_input_tokens_seen": 663748608,
+      "step": 5064
+    },
+    {
+      "epoch": 0.8315034281875477,
+      "grad_norm": 0.6468334794044495,
+      "learning_rate": 7.360303492082339e-05,
+      "loss": 4.9809,
+      "num_input_tokens_seen": 664141824,
+      "step": 5067
+    },
+    {
+      "epoch": 0.8319957333552136,
+      "grad_norm": 0.7428439855575562,
+      "learning_rate": 7.358125565272689e-05,
+      "loss": 4.9878,
+      "num_input_tokens_seen": 664535040,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8324880385228793,
+      "grad_norm": 0.656428873538971,
+      "learning_rate": 7.355949570676748e-05,
+      "loss": 4.9558,
+      "num_input_tokens_seen": 664928256,
+      "step": 5073
+    },
+    {
+      "epoch": 0.8329803436905452,
+      "grad_norm": 0.6178997755050659,
+      "learning_rate": 7.353775505439173e-05,
+      "loss": 5.0094,
+      "num_input_tokens_seen": 665321472,
+      "step": 5076
+    },
+    {
+      "epoch": 0.8334726488582109,
+      "grad_norm": 0.6288952827453613,
+      "learning_rate": 7.351603366710516e-05,
+      "loss": 4.9676,
+      "num_input_tokens_seen": 665714688,
+      "step": 5079
+    },
+    {
+      "epoch": 0.8339649540258768,
+      "grad_norm": 0.5963147878646851,
+      "learning_rate": 7.349433151647226e-05,
+      "loss": 4.936,
+      "num_input_tokens_seen": 666107904,
+      "step": 5082
+    },
+    {
+      "epoch": 0.8344572591935426,
+      "grad_norm": 0.6190651655197144,
+      "learning_rate": 7.347264857411617e-05,
+      "loss": 4.9584,
+      "num_input_tokens_seen": 666501120,
+      "step": 5085
+    },
+    {
+      "epoch": 0.8349495643612084,
+      "grad_norm": 0.6707886457443237,
+      "learning_rate": 7.345098481171862e-05,
+      "loss": 4.9514,
+      "num_input_tokens_seen": 666894336,
+      "step": 5088
+    },
+    {
+      "epoch": 0.8354418695288742,
+      "grad_norm": 0.6272398829460144,
+      "learning_rate": 7.342934020101978e-05,
+      "loss": 4.9309,
+      "num_input_tokens_seen": 667287552,
+      "step": 5091
+    },
+    {
+      "epoch": 0.83593417469654,
+      "grad_norm": 0.6572047472000122,
+      "learning_rate": 7.340771471381805e-05,
+      "loss": 4.9722,
+      "num_input_tokens_seen": 667680768,
+      "step": 5094
+    },
+    {
+      "epoch": 0.8364264798642058,
+      "grad_norm": 0.644632875919342,
+      "learning_rate": 7.338610832196996e-05,
+      "loss": 4.982,
+      "num_input_tokens_seen": 668073984,
+      "step": 5097
+    },
+    {
+      "epoch": 0.8369187850318717,
+      "grad_norm": 0.6848341822624207,
+      "learning_rate": 7.336452099738994e-05,
+      "loss": 4.9322,
+      "num_input_tokens_seen": 668467200,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8374110901995374,
+      "grad_norm": 0.576561450958252,
+      "learning_rate": 7.334295271205029e-05,
+      "loss": 4.9732,
+      "num_input_tokens_seen": 668860416,
+      "step": 5103
+    },
+    {
+      "epoch": 0.8379033953672033,
+      "grad_norm": 0.6817082166671753,
+      "learning_rate": 7.332140343798091e-05,
+      "loss": 4.9491,
+      "num_input_tokens_seen": 669253632,
+      "step": 5106
+    },
+    {
+      "epoch": 0.838395700534869,
+      "grad_norm": 0.716584324836731,
+      "learning_rate": 7.329987314726919e-05,
+      "loss": 4.9788,
+      "num_input_tokens_seen": 669646848,
+      "step": 5109
+    },
+    {
+      "epoch": 0.8388880057025349,
+      "grad_norm": 0.5894129872322083,
+      "learning_rate": 7.327836181205987e-05,
+      "loss": 4.9469,
+      "num_input_tokens_seen": 670040064,
+      "step": 5112
+    },
+    {
+      "epoch": 0.8393803108702007,
+      "grad_norm": 0.5741015672683716,
+      "learning_rate": 7.325686940455489e-05,
+      "loss": 5.0015,
+      "num_input_tokens_seen": 670433280,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8398726160378664,
+      "grad_norm": 0.601706326007843,
+      "learning_rate": 7.323539589701322e-05,
+      "loss": 4.9758,
+      "num_input_tokens_seen": 670826496,
+      "step": 5118
+    },
+    {
+      "epoch": 0.8403649212055323,
+      "grad_norm": 0.6591434478759766,
+      "learning_rate": 7.321394126175073e-05,
+      "loss": 4.915,
+      "num_input_tokens_seen": 671219712,
+      "step": 5121
+    },
+    {
+      "epoch": 0.840857226373198,
+      "grad_norm": 0.6301112174987793,
+      "learning_rate": 7.319250547113997e-05,
+      "loss": 4.9975,
+      "num_input_tokens_seen": 671612928,
+      "step": 5124
+    },
+    {
+      "epoch": 0.8413495315408639,
+      "grad_norm": 0.6425953507423401,
+      "learning_rate": 7.317108849761018e-05,
+      "loss": 4.9452,
+      "num_input_tokens_seen": 672006144,
+      "step": 5127
+    },
+    {
+      "epoch": 0.8418418367085297,
+      "grad_norm": 0.6861438155174255,
+      "learning_rate": 7.314969031364696e-05,
+      "loss": 4.9237,
+      "num_input_tokens_seen": 672399360,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8423341418761955,
+      "grad_norm": 0.6459963917732239,
+      "learning_rate": 7.312831089179222e-05,
+      "loss": 4.9467,
+      "num_input_tokens_seen": 672792576,
+      "step": 5133
+    },
+    {
+      "epoch": 0.8428264470438613,
+      "grad_norm": 0.5769563913345337,
+      "learning_rate": 7.3106950204644e-05,
+      "loss": 4.9517,
+      "num_input_tokens_seen": 673185792,
+      "step": 5136
+    },
+    {
+      "epoch": 0.8433187522115271,
+      "grad_norm": 0.6215202212333679,
+      "learning_rate": 7.30856082248564e-05,
+      "loss": 4.9513,
+      "num_input_tokens_seen": 673579008,
+      "step": 5139
+    },
+    {
+      "epoch": 0.8438110573791929,
+      "grad_norm": 0.5655343532562256,
+      "learning_rate": 7.306428492513929e-05,
+      "loss": 4.9578,
+      "num_input_tokens_seen": 673972224,
+      "step": 5142
+    },
+    {
+      "epoch": 0.8443033625468588,
+      "grad_norm": 0.6174660325050354,
+      "learning_rate": 7.304298027825828e-05,
+      "loss": 4.9143,
+      "num_input_tokens_seen": 674365440,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8447956677145245,
+      "grad_norm": 0.5804635286331177,
+      "learning_rate": 7.302169425703454e-05,
+      "loss": 4.9371,
+      "num_input_tokens_seen": 674758656,
+      "step": 5148
+    },
+    {
+      "epoch": 0.8452879728821904,
+      "grad_norm": 0.6029345989227295,
+      "learning_rate": 7.300042683434464e-05,
+      "loss": 4.9934,
+      "num_input_tokens_seen": 675151872,
+      "step": 5151
+    },
+    {
+      "epoch": 0.8457802780498561,
+      "grad_norm": 0.5814540386199951,
+      "learning_rate": 7.297917798312041e-05,
+      "loss": 4.9335,
+      "num_input_tokens_seen": 675545088,
+      "step": 5154
+    },
+    {
+      "epoch": 0.846272583217522,
+      "grad_norm": 0.6849837303161621,
+      "learning_rate": 7.29579476763488e-05,
+      "loss": 4.9467,
+      "num_input_tokens_seen": 675938304,
+      "step": 5157
+    },
+    {
+      "epoch": 0.8467648883851878,
+      "grad_norm": 0.5766465663909912,
+      "learning_rate": 7.293673588707179e-05,
+      "loss": 4.9537,
+      "num_input_tokens_seen": 676331520,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8472571935528536,
+      "grad_norm": 0.6713130474090576,
+      "learning_rate": 7.291554258838608e-05,
+      "loss": 4.9514,
+      "num_input_tokens_seen": 676724736,
+      "step": 5163
+    },
+    {
+      "epoch": 0.8477494987205194,
+      "grad_norm": 0.5405411720275879,
+      "learning_rate": 7.289436775344317e-05,
+      "loss": 4.97,
+      "num_input_tokens_seen": 677117952,
+      "step": 5166
+    },
+    {
+      "epoch": 0.8482418038881852,
+      "grad_norm": 0.6186428666114807,
+      "learning_rate": 7.287321135544904e-05,
+      "loss": 4.9795,
+      "num_input_tokens_seen": 677511168,
+      "step": 5169
+    },
+    {
+      "epoch": 0.848734109055851,
+      "grad_norm": 0.6150509715080261,
+      "learning_rate": 7.285207336766414e-05,
+      "loss": 5.0054,
+      "num_input_tokens_seen": 677904384,
+      "step": 5172
+    },
+    {
+      "epoch": 0.8492264142235169,
+      "grad_norm": 0.5928124785423279,
+      "learning_rate": 7.283095376340306e-05,
+      "loss": 4.9536,
+      "num_input_tokens_seen": 678297600,
+      "step": 5175
+    },
+    {
+      "epoch": 0.8497187193911826,
+      "grad_norm": 0.6393954753875732,
+      "learning_rate": 7.280985251603467e-05,
+      "loss": 4.976,
+      "num_input_tokens_seen": 678690816,
+      "step": 5178
+    },
+    {
+      "epoch": 0.8502110245588484,
+      "grad_norm": 0.583487331867218,
+      "learning_rate": 7.278876959898168e-05,
+      "loss": 4.98,
+      "num_input_tokens_seen": 679084032,
+      "step": 5181
+    },
+    {
+      "epoch": 0.8507033297265142,
+      "grad_norm": 0.6160304546356201,
+      "learning_rate": 7.276770498572075e-05,
+      "loss": 4.9436,
+      "num_input_tokens_seen": 679477248,
+      "step": 5184
+    },
+    {
+      "epoch": 0.85119563489418,
+      "grad_norm": 0.6791704893112183,
+      "learning_rate": 7.274665864978213e-05,
+      "loss": 4.9429,
+      "num_input_tokens_seen": 679870464,
+      "step": 5187
+    },
+    {
+      "epoch": 0.8516879400618458,
+      "grad_norm": 0.5873450636863708,
+      "learning_rate": 7.272563056474974e-05,
+      "loss": 4.9702,
+      "num_input_tokens_seen": 680263680,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8521802452295116,
+      "grad_norm": 0.5966980457305908,
+      "learning_rate": 7.270462070426084e-05,
+      "loss": 4.9582,
+      "num_input_tokens_seen": 680656896,
+      "step": 5193
+    },
+    {
+      "epoch": 0.8526725503971775,
+      "grad_norm": 0.5830942392349243,
+      "learning_rate": 7.268362904200598e-05,
+      "loss": 4.9787,
+      "num_input_tokens_seen": 681050112,
+      "step": 5196
+    },
+    {
+      "epoch": 0.8531648555648432,
+      "grad_norm": 0.5953894853591919,
+      "learning_rate": 7.266265555172888e-05,
+      "loss": 4.9375,
+      "num_input_tokens_seen": 681443328,
+      "step": 5199
+    },
+    {
+      "epoch": 0.8533289572873985,
+      "eval_accuracy": 0.2291695163654128,
+      "eval_loss": 5.2417755126953125,
+      "eval_runtime": 124.158,
+      "eval_samples_per_second": 2.416,
+      "eval_steps_per_second": 1.208,
+      "num_input_tokens_seen": 681574400,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8536571607325091,
+      "grad_norm": 0.6928351521492004,
+      "learning_rate": 7.264170020722628e-05,
+      "loss": 4.9173,
+      "num_input_tokens_seen": 681836544,
+      "step": 5202
+    },
+    {
+      "epoch": 0.8541494659001748,
+      "grad_norm": 0.5389966368675232,
+      "learning_rate": 7.262076298234773e-05,
+      "loss": 4.9654,
+      "num_input_tokens_seen": 682229760,
+      "step": 5205
+    },
+    {
+      "epoch": 0.8546417710678407,
+      "grad_norm": 0.6330162882804871,
+      "learning_rate": 7.259984385099556e-05,
+      "loss": 4.9327,
+      "num_input_tokens_seen": 682622976,
+      "step": 5208
+    },
+    {
+      "epoch": 0.8551340762355065,
+      "grad_norm": 0.5867341160774231,
+      "learning_rate": 7.257894278712468e-05,
+      "loss": 4.934,
+      "num_input_tokens_seen": 683016192,
+      "step": 5211
+    },
+    {
+      "epoch": 0.8556263814031723,
+      "grad_norm": 0.5803040862083435,
+      "learning_rate": 7.255805976474242e-05,
+      "loss": 4.942,
+      "num_input_tokens_seen": 683409408,
+      "step": 5214
+    },
+    {
+      "epoch": 0.8561186865708381,
+      "grad_norm": 0.6194251179695129,
+      "learning_rate": 7.253719475790852e-05,
+      "loss": 4.9473,
+      "num_input_tokens_seen": 683802624,
+      "step": 5217
+    },
+    {
+      "epoch": 0.8566109917385039,
+      "grad_norm": 0.638539731502533,
+      "learning_rate": 7.25163477407348e-05,
+      "loss": 4.9615,
+      "num_input_tokens_seen": 684195840,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8571032969061697,
+      "grad_norm": 0.6786993741989136,
+      "learning_rate": 7.24955186873852e-05,
+      "loss": 4.9214,
+      "num_input_tokens_seen": 684589056,
+      "step": 5223
+    },
+    {
+      "epoch": 0.8575956020738356,
+      "grad_norm": 0.5903679728507996,
+      "learning_rate": 7.247470757207554e-05,
+      "loss": 4.9546,
+      "num_input_tokens_seen": 684982272,
+      "step": 5226
+    },
+    {
+      "epoch": 0.8580879072415013,
+      "grad_norm": 0.6937805414199829,
+      "learning_rate": 7.245391436907346e-05,
+      "loss": 4.9727,
+      "num_input_tokens_seen": 685375488,
+      "step": 5229
+    },
+    {
+      "epoch": 0.8585802124091672,
+      "grad_norm": 0.6725658774375916,
+      "learning_rate": 7.243313905269826e-05,
+      "loss": 4.9675,
+      "num_input_tokens_seen": 685768704,
+      "step": 5232
+    },
+    {
+      "epoch": 0.8590725175768329,
+      "grad_norm": 0.6803759336471558,
+      "learning_rate": 7.241238159732069e-05,
+      "loss": 4.9197,
+      "num_input_tokens_seen": 686161920,
+      "step": 5235
+    },
+    {
+      "epoch": 0.8595648227444987,
+      "grad_norm": 0.7206405401229858,
+      "learning_rate": 7.239164197736292e-05,
+      "loss": 4.9093,
+      "num_input_tokens_seen": 686555136,
+      "step": 5238
+    },
+    {
+      "epoch": 0.8600571279121646,
+      "grad_norm": 0.6930558681488037,
+      "learning_rate": 7.237092016729838e-05,
+      "loss": 4.9747,
+      "num_input_tokens_seen": 686948352,
+      "step": 5241
+    },
+    {
+      "epoch": 0.8605494330798303,
+      "grad_norm": 0.6751710176467896,
+      "learning_rate": 7.235021614165161e-05,
+      "loss": 4.9482,
+      "num_input_tokens_seen": 687341568,
+      "step": 5244
+    },
+    {
+      "epoch": 0.8610417382474962,
+      "grad_norm": 0.5579334497451782,
+      "learning_rate": 7.232952987499815e-05,
+      "loss": 4.9393,
+      "num_input_tokens_seen": 687734784,
+      "step": 5247
+    },
+    {
+      "epoch": 0.8615340434151619,
+      "grad_norm": 0.5883084535598755,
+      "learning_rate": 7.230886134196436e-05,
+      "loss": 4.8905,
+      "num_input_tokens_seen": 688128000,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8620263485828278,
+      "grad_norm": 0.6635749936103821,
+      "learning_rate": 7.228821051722736e-05,
+      "loss": 4.9485,
+      "num_input_tokens_seen": 688521216,
+      "step": 5253
+    },
+    {
+      "epoch": 0.8625186537504936,
+      "grad_norm": 0.6582626104354858,
+      "learning_rate": 7.226757737551486e-05,
+      "loss": 4.9598,
+      "num_input_tokens_seen": 688914432,
+      "step": 5256
+    },
+    {
+      "epoch": 0.8630109589181594,
+      "grad_norm": 0.7006992101669312,
+      "learning_rate": 7.224696189160501e-05,
+      "loss": 4.9825,
+      "num_input_tokens_seen": 689307648,
+      "step": 5259
+    },
+    {
+      "epoch": 0.8635032640858252,
+      "grad_norm": 0.654933512210846,
+      "learning_rate": 7.222636404032635e-05,
+      "loss": 4.9345,
+      "num_input_tokens_seen": 689700864,
+      "step": 5262
+    },
+    {
+      "epoch": 0.863995569253491,
+      "grad_norm": 0.6875039935112,
+      "learning_rate": 7.220578379655756e-05,
+      "loss": 5.0074,
+      "num_input_tokens_seen": 690094080,
+      "step": 5265
+    },
+    {
+      "epoch": 0.8644878744211568,
+      "grad_norm": 0.7432628870010376,
+      "learning_rate": 7.218522113522744e-05,
+      "loss": 4.9506,
+      "num_input_tokens_seen": 690487296,
+      "step": 5268
+    },
+    {
+      "epoch": 0.8649801795888227,
+      "grad_norm": 0.6537792086601257,
+      "learning_rate": 7.216467603131472e-05,
+      "loss": 4.9174,
+      "num_input_tokens_seen": 690880512,
+      "step": 5271
+    },
+    {
+      "epoch": 0.8654724847564884,
+      "grad_norm": 0.6503370404243469,
+      "learning_rate": 7.214414845984798e-05,
+      "loss": 4.9177,
+      "num_input_tokens_seen": 691273728,
+      "step": 5274
+    },
+    {
+      "epoch": 0.8659647899241543,
+      "grad_norm": 0.6077583432197571,
+      "learning_rate": 7.212363839590548e-05,
+      "loss": 4.9793,
+      "num_input_tokens_seen": 691666944,
+      "step": 5277
+    },
+    {
+      "epoch": 0.86645709509182,
+      "grad_norm": 0.5995897054672241,
+      "learning_rate": 7.210314581461502e-05,
+      "loss": 4.9067,
+      "num_input_tokens_seen": 692060160,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8669494002594859,
+      "grad_norm": 0.5458451509475708,
+      "learning_rate": 7.208267069115388e-05,
+      "loss": 4.9198,
+      "num_input_tokens_seen": 692453376,
+      "step": 5283
+    },
+    {
+      "epoch": 0.8674417054271517,
+      "grad_norm": 0.6200904250144958,
+      "learning_rate": 7.206221300074863e-05,
+      "loss": 4.9502,
+      "num_input_tokens_seen": 692846592,
+      "step": 5286
+    },
+    {
+      "epoch": 0.8679340105948175,
+      "grad_norm": 0.5652337670326233,
+      "learning_rate": 7.204177271867505e-05,
+      "loss": 4.9348,
+      "num_input_tokens_seen": 693239808,
+      "step": 5289
+    },
+    {
+      "epoch": 0.8684263157624833,
+      "grad_norm": 0.5828121304512024,
+      "learning_rate": 7.202134982025796e-05,
+      "loss": 5.0021,
+      "num_input_tokens_seen": 693633024,
+      "step": 5292
+    },
+    {
+      "epoch": 0.868918620930149,
+      "grad_norm": 0.6214671730995178,
+      "learning_rate": 7.200094428087114e-05,
+      "loss": 4.9013,
+      "num_input_tokens_seen": 694026240,
+      "step": 5295
+    },
+    {
+      "epoch": 0.8694109260978149,
+      "grad_norm": 0.6146253347396851,
+      "learning_rate": 7.198055607593714e-05,
+      "loss": 4.8906,
+      "num_input_tokens_seen": 694419456,
+      "step": 5298
+    },
+    {
+      "epoch": 0.8699032312654806,
+      "grad_norm": 0.6827002167701721,
+      "learning_rate": 7.196018518092727e-05,
+      "loss": 4.9346,
+      "num_input_tokens_seen": 694812672,
+      "step": 5301
+    },
+    {
+      "epoch": 0.8703955364331465,
+      "grad_norm": 0.6271741986274719,
+      "learning_rate": 7.193983157136133e-05,
+      "loss": 4.9341,
+      "num_input_tokens_seen": 695205888,
+      "step": 5304
+    },
+    {
+      "epoch": 0.8708878416008123,
+      "grad_norm": 0.6707040667533875,
+      "learning_rate": 7.191949522280763e-05,
+      "loss": 4.9357,
+      "num_input_tokens_seen": 695599104,
+      "step": 5307
+    },
+    {
+      "epoch": 0.8713801467684781,
+      "grad_norm": 0.5915789604187012,
+      "learning_rate": 7.189917611088272e-05,
+      "loss": 4.9366,
+      "num_input_tokens_seen": 695992320,
+      "step": 5310
+    },
+    {
+      "epoch": 0.8718724519361439,
+      "grad_norm": 0.6474863290786743,
+      "learning_rate": 7.187887421125144e-05,
+      "loss": 4.9588,
+      "num_input_tokens_seen": 696385536,
+      "step": 5313
+    },
+    {
+      "epoch": 0.8723647571038097,
+      "grad_norm": 0.5836812853813171,
+      "learning_rate": 7.185858949962659e-05,
+      "loss": 4.9235,
+      "num_input_tokens_seen": 696778752,
+      "step": 5316
+    },
+    {
+      "epoch": 0.8728570622714755,
+      "grad_norm": 0.6480825543403625,
+      "learning_rate": 7.183832195176905e-05,
+      "loss": 4.9712,
+      "num_input_tokens_seen": 697171968,
+      "step": 5319
+    },
+    {
+      "epoch": 0.8733493674391414,
+      "grad_norm": 0.5924190878868103,
+      "learning_rate": 7.181807154348743e-05,
+      "loss": 4.9862,
+      "num_input_tokens_seen": 697565184,
+      "step": 5322
+    },
+    {
+      "epoch": 0.8738416726068071,
+      "grad_norm": 0.5856485962867737,
+      "learning_rate": 7.179783825063807e-05,
+      "loss": 4.907,
+      "num_input_tokens_seen": 697958400,
+      "step": 5325
+    },
+    {
+      "epoch": 0.874333977774473,
+      "grad_norm": 0.581123948097229,
+      "learning_rate": 7.177762204912492e-05,
+      "loss": 4.9002,
+      "num_input_tokens_seen": 698351616,
+      "step": 5328
+    },
+    {
+      "epoch": 0.8748262829421387,
+      "grad_norm": 0.5788645148277283,
+      "learning_rate": 7.17574229148994e-05,
+      "loss": 4.93,
+      "num_input_tokens_seen": 698744832,
+      "step": 5331
+    },
+    {
+      "epoch": 0.8753185881098046,
+      "grad_norm": 0.5507704615592957,
+      "learning_rate": 7.173724082396026e-05,
+      "loss": 4.9437,
+      "num_input_tokens_seen": 699138048,
+      "step": 5334
+    },
+    {
+      "epoch": 0.8758108932774704,
+      "grad_norm": 0.5941615104675293,
+      "learning_rate": 7.171707575235344e-05,
+      "loss": 4.9439,
+      "num_input_tokens_seen": 699531264,
+      "step": 5337
+    },
+    {
+      "epoch": 0.8763031984451362,
+      "grad_norm": 0.5601658821105957,
+      "learning_rate": 7.169692767617206e-05,
+      "loss": 4.9691,
+      "num_input_tokens_seen": 699924480,
+      "step": 5340
+    },
+    {
+      "epoch": 0.876795503612802,
+      "grad_norm": 0.6255508065223694,
+      "learning_rate": 7.167679657155616e-05,
+      "loss": 4.9546,
+      "num_input_tokens_seen": 700317696,
+      "step": 5343
+    },
+    {
+      "epoch": 0.8772878087804677,
+      "grad_norm": 0.5593776702880859,
+      "learning_rate": 7.165668241469273e-05,
+      "loss": 4.9272,
+      "num_input_tokens_seen": 700710912,
+      "step": 5346
+    },
+    {
+      "epoch": 0.8777801139481336,
+      "grad_norm": 0.5787507891654968,
+      "learning_rate": 7.163658518181542e-05,
+      "loss": 4.937,
+      "num_input_tokens_seen": 701104128,
+      "step": 5349
+    },
+    {
+      "epoch": 0.8782724191157995,
+      "grad_norm": 0.589417576789856,
+      "learning_rate": 7.161650484920457e-05,
+      "loss": 4.9483,
+      "num_input_tokens_seen": 701497344,
+      "step": 5352
+    },
+    {
+      "epoch": 0.8787647242834652,
+      "grad_norm": 0.5617730021476746,
+      "learning_rate": 7.159644139318704e-05,
+      "loss": 4.9215,
+      "num_input_tokens_seen": 701890560,
+      "step": 5355
+    },
+    {
+      "epoch": 0.879257029451131,
+      "grad_norm": 0.6318708658218384,
+      "learning_rate": 7.157639479013606e-05,
+      "loss": 4.9201,
+      "num_input_tokens_seen": 702283776,
+      "step": 5358
+    },
+    {
+      "epoch": 0.8797493346187968,
+      "grad_norm": 0.6035711765289307,
+      "learning_rate": 7.155636501647111e-05,
+      "loss": 4.9469,
+      "num_input_tokens_seen": 702676992,
+      "step": 5361
+    },
+    {
+      "epoch": 0.8802416397864626,
+      "grad_norm": 0.5588794946670532,
+      "learning_rate": 7.153635204865795e-05,
+      "loss": 4.9162,
+      "num_input_tokens_seen": 703070208,
+      "step": 5364
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.6277647614479065,
+      "learning_rate": 7.151635586320828e-05,
+      "loss": 4.917,
+      "num_input_tokens_seen": 703463424,
+      "step": 5367
+    },
+    {
+      "epoch": 0.8812262501217942,
+      "grad_norm": 0.5754868984222412,
+      "learning_rate": 7.149637643667977e-05,
+      "loss": 4.9523,
+      "num_input_tokens_seen": 703856640,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8817185552894601,
+      "grad_norm": 0.6343851089477539,
+      "learning_rate": 7.147641374567592e-05,
+      "loss": 4.971,
+      "num_input_tokens_seen": 704249856,
+      "step": 5373
+    },
+    {
+      "epoch": 0.8822108604571258,
+      "grad_norm": 0.5441321134567261,
+      "learning_rate": 7.145646776684588e-05,
+      "loss": 4.969,
+      "num_input_tokens_seen": 704643072,
+      "step": 5376
+    },
+    {
+      "epoch": 0.8827031656247917,
+      "grad_norm": 0.6083698272705078,
+      "learning_rate": 7.143653847688449e-05,
+      "loss": 5.0037,
+      "num_input_tokens_seen": 705036288,
+      "step": 5379
+    },
+    {
+      "epoch": 0.8831954707924575,
+      "grad_norm": 0.5505862832069397,
+      "learning_rate": 7.141662585253196e-05,
+      "loss": 4.9558,
+      "num_input_tokens_seen": 705429504,
+      "step": 5382
+    },
+    {
+      "epoch": 0.8836877759601233,
+      "grad_norm": 0.5923398733139038,
+      "learning_rate": 7.13967298705739e-05,
+      "loss": 4.9021,
+      "num_input_tokens_seen": 705822720,
+      "step": 5385
+    },
+    {
+      "epoch": 0.8841800811277891,
+      "grad_norm": 0.5669833421707153,
+      "learning_rate": 7.137685050784117e-05,
+      "loss": 4.9494,
+      "num_input_tokens_seen": 706215936,
+      "step": 5388
+    },
+    {
+      "epoch": 0.8846723862954549,
+      "grad_norm": 0.7324468493461609,
+      "learning_rate": 7.135698774120979e-05,
+      "loss": 4.9386,
+      "num_input_tokens_seen": 706609152,
+      "step": 5391
+    },
+    {
+      "epoch": 0.8851646914631207,
+      "grad_norm": 0.6149271130561829,
+      "learning_rate": 7.133714154760073e-05,
+      "loss": 4.963,
+      "num_input_tokens_seen": 707002368,
+      "step": 5394
+    },
+    {
+      "epoch": 0.8856569966307866,
+      "grad_norm": 0.6040337681770325,
+      "learning_rate": 7.131731190397995e-05,
+      "loss": 4.9305,
+      "num_input_tokens_seen": 707395584,
+      "step": 5397
+    },
+    {
+      "epoch": 0.8861493017984523,
+      "grad_norm": 0.5804337859153748,
+      "learning_rate": 7.129749878735813e-05,
+      "loss": 4.8961,
+      "num_input_tokens_seen": 707788800,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8866416069661182,
+      "grad_norm": 0.6081647276878357,
+      "learning_rate": 7.127770217479066e-05,
+      "loss": 4.9379,
+      "num_input_tokens_seen": 708182016,
+      "step": 5403
+    },
+    {
+      "epoch": 0.8871339121337839,
+      "grad_norm": 0.5394389033317566,
+      "learning_rate": 7.125792204337751e-05,
+      "loss": 4.9666,
+      "num_input_tokens_seen": 708575232,
+      "step": 5406
+    },
+    {
+      "epoch": 0.8876262173014497,
+      "grad_norm": 0.5974504947662354,
+      "learning_rate": 7.123815837026311e-05,
+      "loss": 4.9122,
+      "num_input_tokens_seen": 708968448,
+      "step": 5409
+    },
+    {
+      "epoch": 0.8881185224691155,
+      "grad_norm": 0.5746287107467651,
+      "learning_rate": 7.121841113263623e-05,
+      "loss": 4.9027,
+      "num_input_tokens_seen": 709361664,
+      "step": 5412
+    },
+    {
+      "epoch": 0.8886108276367813,
+      "grad_norm": 0.6007826328277588,
+      "learning_rate": 7.119868030772991e-05,
+      "loss": 4.9593,
+      "num_input_tokens_seen": 709754880,
+      "step": 5415
+    },
+    {
+      "epoch": 0.8891031328044472,
+      "grad_norm": 0.5774713158607483,
+      "learning_rate": 7.117896587282125e-05,
+      "loss": 4.9635,
+      "num_input_tokens_seen": 710148096,
+      "step": 5418
+    },
+    {
+      "epoch": 0.8895954379721129,
+      "grad_norm": 0.636423647403717,
+      "learning_rate": 7.115926780523142e-05,
+      "loss": 4.9186,
+      "num_input_tokens_seen": 710541312,
+      "step": 5421
+    },
+    {
+      "epoch": 0.8900877431397788,
+      "grad_norm": 0.6212565898895264,
+      "learning_rate": 7.11395860823255e-05,
+      "loss": 4.9337,
+      "num_input_tokens_seen": 710934528,
+      "step": 5424
+    },
+    {
+      "epoch": 0.8905800483074445,
+      "grad_norm": 0.6649476885795593,
+      "learning_rate": 7.111992068151236e-05,
+      "loss": 4.9409,
+      "num_input_tokens_seen": 711327744,
+      "step": 5427
+    },
+    {
+      "epoch": 0.8910723534751104,
+      "grad_norm": 0.6479114890098572,
+      "learning_rate": 7.110027158024453e-05,
+      "loss": 4.9311,
+      "num_input_tokens_seen": 711720960,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8915646586427762,
+      "grad_norm": 0.6290377974510193,
+      "learning_rate": 7.108063875601819e-05,
+      "loss": 4.9444,
+      "num_input_tokens_seen": 712114176,
+      "step": 5433
+    },
+    {
+      "epoch": 0.892056963810442,
+      "grad_norm": 0.6628143787384033,
+      "learning_rate": 7.106102218637291e-05,
+      "loss": 4.9459,
+      "num_input_tokens_seen": 712507392,
+      "step": 5436
+    },
+    {
+      "epoch": 0.8925492689781078,
+      "grad_norm": 0.6291846632957458,
+      "learning_rate": 7.104142184889171e-05,
+      "loss": 4.8912,
+      "num_input_tokens_seen": 712900608,
+      "step": 5439
+    },
+    {
+      "epoch": 0.8930415741457736,
+      "grad_norm": 0.7024665474891663,
+      "learning_rate": 7.10218377212008e-05,
+      "loss": 4.9319,
+      "num_input_tokens_seen": 713293824,
+      "step": 5442
+    },
+    {
+      "epoch": 0.8935338793134394,
+      "grad_norm": 0.6190261840820312,
+      "learning_rate": 7.100226978096957e-05,
+      "loss": 4.9276,
+      "num_input_tokens_seen": 713687040,
+      "step": 5445
+    },
+    {
+      "epoch": 0.8940261844811053,
+      "grad_norm": 0.6652445197105408,
+      "learning_rate": 7.098271800591048e-05,
+      "loss": 4.9636,
+      "num_input_tokens_seen": 714080256,
+      "step": 5448
+    },
+    {
+      "epoch": 0.894518489648771,
+      "grad_norm": 0.5882824659347534,
+      "learning_rate": 7.096318237377886e-05,
+      "loss": 4.9664,
+      "num_input_tokens_seen": 714473472,
+      "step": 5451
+    },
+    {
+      "epoch": 0.8950107948164369,
+      "grad_norm": 0.5919190645217896,
+      "learning_rate": 7.094366286237293e-05,
+      "loss": 4.9181,
+      "num_input_tokens_seen": 714866688,
+      "step": 5454
+    },
+    {
+      "epoch": 0.8955030999841026,
+      "grad_norm": 0.6125955581665039,
+      "learning_rate": 7.092415944953361e-05,
+      "loss": 4.9809,
+      "num_input_tokens_seen": 715259904,
+      "step": 5457
+    },
+    {
+      "epoch": 0.8959954051517685,
+      "grad_norm": 0.5722345113754272,
+      "learning_rate": 7.090467211314446e-05,
+      "loss": 4.9237,
+      "num_input_tokens_seen": 715653120,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8964877103194343,
+      "grad_norm": 0.5687103271484375,
+      "learning_rate": 7.088520083113153e-05,
+      "loss": 4.9173,
+      "num_input_tokens_seen": 716046336,
+      "step": 5463
+    },
+    {
+      "epoch": 0.8969800154871,
+      "grad_norm": 0.662970244884491,
+      "learning_rate": 7.08657455814633e-05,
+      "loss": 4.9096,
+      "num_input_tokens_seen": 716439552,
+      "step": 5466
+    },
+    {
+      "epoch": 0.8974723206547659,
+      "grad_norm": 0.601774275302887,
+      "learning_rate": 7.08463063421505e-05,
+      "loss": 4.9254,
+      "num_input_tokens_seen": 716832768,
+      "step": 5469
+    },
+    {
+      "epoch": 0.8979646258224316,
+      "grad_norm": 0.6241858005523682,
+      "learning_rate": 7.082688309124617e-05,
+      "loss": 4.9082,
+      "num_input_tokens_seen": 717225984,
+      "step": 5472
+    },
+    {
+      "epoch": 0.8984569309900975,
+      "grad_norm": 0.5964360237121582,
+      "learning_rate": 7.080747580684533e-05,
+      "loss": 4.9056,
+      "num_input_tokens_seen": 717619200,
+      "step": 5475
+    },
+    {
+      "epoch": 0.8989492361577633,
+      "grad_norm": 0.6412652730941772,
+      "learning_rate": 7.078808446708505e-05,
+      "loss": 4.8908,
+      "num_input_tokens_seen": 718012416,
+      "step": 5478
+    },
+    {
+      "epoch": 0.8994415413254291,
+      "grad_norm": 0.5989963412284851,
+      "learning_rate": 7.076870905014429e-05,
+      "loss": 4.9292,
+      "num_input_tokens_seen": 718405632,
+      "step": 5481
+    },
+    {
+      "epoch": 0.8999338464930949,
+      "grad_norm": 0.7762419581413269,
+      "learning_rate": 7.074934953424378e-05,
+      "loss": 4.9499,
+      "num_input_tokens_seen": 718798848,
+      "step": 5484
+    },
+    {
+      "epoch": 0.9004261516607607,
+      "grad_norm": 0.5652441382408142,
+      "learning_rate": 7.073000589764593e-05,
+      "loss": 4.898,
+      "num_input_tokens_seen": 719192064,
+      "step": 5487
+    },
+    {
+      "epoch": 0.9009184568284265,
+      "grad_norm": 0.6855098009109497,
+      "learning_rate": 7.071067811865475e-05,
+      "loss": 4.9768,
+      "num_input_tokens_seen": 719585280,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9014107619960924,
+      "grad_norm": 0.7517238855361938,
+      "learning_rate": 7.069136617561571e-05,
+      "loss": 4.9106,
+      "num_input_tokens_seen": 719978496,
+      "step": 5493
+    },
+    {
+      "epoch": 0.9019030671637581,
+      "grad_norm": 0.5679114460945129,
+      "learning_rate": 7.067207004691567e-05,
+      "loss": 4.9333,
+      "num_input_tokens_seen": 720371712,
+      "step": 5496
+    },
+    {
+      "epoch": 0.902395372331424,
+      "grad_norm": 0.6112279295921326,
+      "learning_rate": 7.065278971098276e-05,
+      "loss": 4.9465,
+      "num_input_tokens_seen": 720764928,
+      "step": 5499
+    },
+    {
+      "epoch": 0.9028876774990897,
+      "grad_norm": 0.5691413879394531,
+      "learning_rate": 7.063352514628629e-05,
+      "loss": 4.8991,
+      "num_input_tokens_seen": 721158144,
+      "step": 5502
+    },
+    {
+      "epoch": 0.9033799826667556,
+      "grad_norm": 0.7350879907608032,
+      "learning_rate": 7.06142763313366e-05,
+      "loss": 4.9199,
+      "num_input_tokens_seen": 721551360,
+      "step": 5505
+    },
+    {
+      "epoch": 0.9038722878344214,
+      "grad_norm": 0.5895538926124573,
+      "learning_rate": 7.059504324468505e-05,
+      "loss": 4.9424,
+      "num_input_tokens_seen": 721944576,
+      "step": 5508
+    },
+    {
+      "epoch": 0.9043645930020872,
+      "grad_norm": 0.6411991119384766,
+      "learning_rate": 7.057582586492387e-05,
+      "loss": 4.9006,
+      "num_input_tokens_seen": 722337792,
+      "step": 5511
+    },
+    {
+      "epoch": 0.904856898169753,
+      "grad_norm": 0.5808833837509155,
+      "learning_rate": 7.055662417068605e-05,
+      "loss": 4.9046,
+      "num_input_tokens_seen": 722731008,
+      "step": 5514
+    },
+    {
+      "epoch": 0.9053492033374188,
+      "grad_norm": 0.5693103671073914,
+      "learning_rate": 7.05374381406452e-05,
+      "loss": 4.8972,
+      "num_input_tokens_seen": 723124224,
+      "step": 5517
+    },
+    {
+      "epoch": 0.9058415085050846,
+      "grad_norm": 0.5854239463806152,
+      "learning_rate": 7.051826775351563e-05,
+      "loss": 4.9348,
+      "num_input_tokens_seen": 723517440,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9063338136727503,
+      "grad_norm": 0.7103624939918518,
+      "learning_rate": 7.049911298805197e-05,
+      "loss": 4.9114,
+      "num_input_tokens_seen": 723910656,
+      "step": 5523
+    },
+    {
+      "epoch": 0.9068261188404162,
+      "grad_norm": 0.591009259223938,
+      "learning_rate": 7.047997382304934e-05,
+      "loss": 5.0049,
+      "num_input_tokens_seen": 724303872,
+      "step": 5526
+    },
+    {
+      "epoch": 0.907318424008082,
+      "grad_norm": 0.6358263492584229,
+      "learning_rate": 7.046085023734305e-05,
+      "loss": 4.9576,
+      "num_input_tokens_seen": 724697088,
+      "step": 5529
+    },
+    {
+      "epoch": 0.9078107291757478,
+      "grad_norm": 0.671718180179596,
+      "learning_rate": 7.044174220980871e-05,
+      "loss": 4.8979,
+      "num_input_tokens_seen": 725090304,
+      "step": 5532
+    },
+    {
+      "epoch": 0.9083030343434136,
+      "grad_norm": 0.5911664962768555,
+      "learning_rate": 7.042264971936185e-05,
+      "loss": 4.9507,
+      "num_input_tokens_seen": 725483520,
+      "step": 5535
+    },
+    {
+      "epoch": 0.9087953395110794,
+      "grad_norm": 0.6500189304351807,
+      "learning_rate": 7.040357274495808e-05,
+      "loss": 4.8758,
+      "num_input_tokens_seen": 725876736,
+      "step": 5538
+    },
+    {
+      "epoch": 0.9092876446787452,
+      "grad_norm": 0.5432878732681274,
+      "learning_rate": 7.038451126559289e-05,
+      "loss": 4.9449,
+      "num_input_tokens_seen": 726269952,
+      "step": 5541
+    },
+    {
+      "epoch": 0.9097799498464111,
+      "grad_norm": 0.5804717540740967,
+      "learning_rate": 7.036546526030153e-05,
+      "loss": 4.9024,
+      "num_input_tokens_seen": 726663168,
+      "step": 5544
+    },
+    {
+      "epoch": 0.9102722550140768,
+      "grad_norm": 0.5860951542854309,
+      "learning_rate": 7.034643470815894e-05,
+      "loss": 4.9393,
+      "num_input_tokens_seen": 727056384,
+      "step": 5547
+    },
+    {
+      "epoch": 0.9107645601817427,
+      "grad_norm": 0.6187010407447815,
+      "learning_rate": 7.032741958827968e-05,
+      "loss": 4.8962,
+      "num_input_tokens_seen": 727449600,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9112568653494084,
+      "grad_norm": 0.5530468225479126,
+      "learning_rate": 7.030841987981778e-05,
+      "loss": 4.9029,
+      "num_input_tokens_seen": 727842816,
+      "step": 5553
+    },
+    {
+      "epoch": 0.9117491705170743,
+      "grad_norm": 0.6535037755966187,
+      "learning_rate": 7.02894355619667e-05,
+      "loss": 4.9293,
+      "num_input_tokens_seen": 728236032,
+      "step": 5556
+    },
+    {
+      "epoch": 0.9122414756847401,
+      "grad_norm": 0.5901381969451904,
+      "learning_rate": 7.027046661395916e-05,
+      "loss": 4.8875,
+      "num_input_tokens_seen": 728629248,
+      "step": 5559
+    },
+    {
+      "epoch": 0.9127337808524059,
+      "grad_norm": 0.6690804958343506,
+      "learning_rate": 7.025151301506713e-05,
+      "loss": 4.9228,
+      "num_input_tokens_seen": 729022464,
+      "step": 5562
+    },
+    {
+      "epoch": 0.9132260860200717,
+      "grad_norm": 0.6213147044181824,
+      "learning_rate": 7.02325747446017e-05,
+      "loss": 4.9315,
+      "num_input_tokens_seen": 729415680,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9137183911877375,
+      "grad_norm": 0.6875079870223999,
+      "learning_rate": 7.021365178191292e-05,
+      "loss": 4.9614,
+      "num_input_tokens_seen": 729808896,
+      "step": 5568
+    },
+    {
+      "epoch": 0.9142106963554033,
+      "grad_norm": 0.5743534564971924,
+      "learning_rate": 7.019474410638983e-05,
+      "loss": 4.9304,
+      "num_input_tokens_seen": 730202112,
+      "step": 5571
+    },
+    {
+      "epoch": 0.9147030015230692,
+      "grad_norm": 0.6975659728050232,
+      "learning_rate": 7.017585169746028e-05,
+      "loss": 4.9321,
+      "num_input_tokens_seen": 730595328,
+      "step": 5574
+    },
+    {
+      "epoch": 0.9151953066907349,
+      "grad_norm": 0.6249175071716309,
+      "learning_rate": 7.015697453459085e-05,
+      "loss": 4.8852,
+      "num_input_tokens_seen": 730988544,
+      "step": 5577
+    },
+    {
+      "epoch": 0.9156876118584008,
+      "grad_norm": 0.6176156401634216,
+      "learning_rate": 7.013811259728677e-05,
+      "loss": 4.9137,
+      "num_input_tokens_seen": 731381760,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9161799170260665,
+      "grad_norm": 0.5664033889770508,
+      "learning_rate": 7.011926586509181e-05,
+      "loss": 4.9301,
+      "num_input_tokens_seen": 731774976,
+      "step": 5583
+    },
+    {
+      "epoch": 0.9166722221937323,
+      "grad_norm": 0.5968320369720459,
+      "learning_rate": 7.010043431758822e-05,
+      "loss": 4.9191,
+      "num_input_tokens_seen": 732168192,
+      "step": 5586
+    },
+    {
+      "epoch": 0.9171645273613982,
+      "grad_norm": 0.5811692476272583,
+      "learning_rate": 7.008161793439657e-05,
+      "loss": 4.9041,
+      "num_input_tokens_seen": 732561408,
+      "step": 5589
+    },
+    {
+      "epoch": 0.9176568325290639,
+      "grad_norm": 0.631001889705658,
+      "learning_rate": 7.006281669517578e-05,
+      "loss": 4.9357,
+      "num_input_tokens_seen": 732954624,
+      "step": 5592
+    },
+    {
+      "epoch": 0.9181491376967298,
+      "grad_norm": 0.5815140604972839,
+      "learning_rate": 7.004403057962285e-05,
+      "loss": 4.9433,
+      "num_input_tokens_seen": 733347840,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9186414428643955,
+      "grad_norm": 0.5923864841461182,
+      "learning_rate": 7.002525956747294e-05,
+      "loss": 4.9322,
+      "num_input_tokens_seen": 733741056,
+      "step": 5598
+    },
+    {
+      "epoch": 0.9189696463095061,
+      "eval_accuracy": 0.23118384627910762,
+      "eval_loss": 5.216609477996826,
+      "eval_runtime": 125.2559,
+      "eval_samples_per_second": 2.395,
+      "eval_steps_per_second": 1.198,
+      "num_input_tokens_seen": 734003200,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9191337480320614,
+      "grad_norm": 0.6140096187591553,
+      "learning_rate": 7.000650363849917e-05,
+      "loss": 4.9178,
+      "num_input_tokens_seen": 734134272,
+      "step": 5601
+    },
+    {
+      "epoch": 0.9196260531997272,
+      "grad_norm": 0.6851264238357544,
+      "learning_rate": 6.998776277251258e-05,
+      "loss": 4.8845,
+      "num_input_tokens_seen": 734527488,
+      "step": 5604
+    },
+    {
+      "epoch": 0.920118358367393,
+      "grad_norm": 0.6549767851829529,
+      "learning_rate": 6.996903694936202e-05,
+      "loss": 4.8942,
+      "num_input_tokens_seen": 734920704,
+      "step": 5607
+    },
+    {
+      "epoch": 0.9206106635350588,
+      "grad_norm": 0.6244533061981201,
+      "learning_rate": 6.995032614893404e-05,
+      "loss": 4.9043,
+      "num_input_tokens_seen": 735313920,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9211029687027246,
+      "grad_norm": 0.7099424600601196,
+      "learning_rate": 6.993163035115284e-05,
+      "loss": 4.8973,
+      "num_input_tokens_seen": 735707136,
+      "step": 5613
+    },
+    {
+      "epoch": 0.9215952738703904,
+      "grad_norm": 0.6875070929527283,
+      "learning_rate": 6.991294953598019e-05,
+      "loss": 4.9278,
+      "num_input_tokens_seen": 736100352,
+      "step": 5616
+    },
+    {
+      "epoch": 0.9220875790380562,
+      "grad_norm": 0.5664244890213013,
+      "learning_rate": 6.989428368341524e-05,
+      "loss": 4.9095,
+      "num_input_tokens_seen": 736493568,
+      "step": 5619
+    },
+    {
+      "epoch": 0.922579884205722,
+      "grad_norm": 0.7427223920822144,
+      "learning_rate": 6.987563277349452e-05,
+      "loss": 4.9275,
+      "num_input_tokens_seen": 736886784,
+      "step": 5622
+    },
+    {
+      "epoch": 0.9230721893733879,
+      "grad_norm": 0.668783962726593,
+      "learning_rate": 6.985699678629191e-05,
+      "loss": 4.9108,
+      "num_input_tokens_seen": 737280000,
+      "step": 5625
+    },
+    {
+      "epoch": 0.9235644945410536,
+      "grad_norm": 0.6262427568435669,
+      "learning_rate": 6.983837570191838e-05,
+      "loss": 4.9542,
+      "num_input_tokens_seen": 737673216,
+      "step": 5628
+    },
+    {
+      "epoch": 0.9240567997087195,
+      "grad_norm": 0.6482858061790466,
+      "learning_rate": 6.981976950052198e-05,
+      "loss": 4.9115,
+      "num_input_tokens_seen": 738066432,
+      "step": 5631
+    },
+    {
+      "epoch": 0.9245491048763852,
+      "grad_norm": 0.7812525629997253,
+      "learning_rate": 6.980117816228785e-05,
+      "loss": 4.9108,
+      "num_input_tokens_seen": 738459648,
+      "step": 5634
+    },
+    {
+      "epoch": 0.925041410044051,
+      "grad_norm": 0.5804930329322815,
+      "learning_rate": 6.978260166743796e-05,
+      "loss": 4.9111,
+      "num_input_tokens_seen": 738852864,
+      "step": 5637
+    },
+    {
+      "epoch": 0.9255337152117169,
+      "grad_norm": 0.563565194606781,
+      "learning_rate": 6.976403999623119e-05,
+      "loss": 4.9069,
+      "num_input_tokens_seen": 739246080,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9260260203793826,
+      "grad_norm": 0.555899441242218,
+      "learning_rate": 6.974549312896306e-05,
+      "loss": 4.8704,
+      "num_input_tokens_seen": 739639296,
+      "step": 5643
+    },
+    {
+      "epoch": 0.9265183255470485,
+      "grad_norm": 0.5783566832542419,
+      "learning_rate": 6.972696104596579e-05,
+      "loss": 4.9685,
+      "num_input_tokens_seen": 740032512,
+      "step": 5646
+    },
+    {
+      "epoch": 0.9270106307147142,
+      "grad_norm": 0.5310668349266052,
+      "learning_rate": 6.97084437276082e-05,
+      "loss": 4.9193,
+      "num_input_tokens_seen": 740425728,
+      "step": 5649
+    },
+    {
+      "epoch": 0.9275029358823801,
+      "grad_norm": 0.5854000449180603,
+      "learning_rate": 6.96899411542955e-05,
+      "loss": 4.9021,
+      "num_input_tokens_seen": 740818944,
+      "step": 5652
+    },
+    {
+      "epoch": 0.9279952410500459,
+      "grad_norm": 0.6978147029876709,
+      "learning_rate": 6.967145330646938e-05,
+      "loss": 4.8868,
+      "num_input_tokens_seen": 741212160,
+      "step": 5655
+    },
+    {
+      "epoch": 0.9284875462177117,
+      "grad_norm": 0.5259647965431213,
+      "learning_rate": 6.965298016460775e-05,
+      "loss": 4.9378,
+      "num_input_tokens_seen": 741605376,
+      "step": 5658
+    },
+    {
+      "epoch": 0.9289798513853775,
+      "grad_norm": 0.8059858679771423,
+      "learning_rate": 6.963452170922476e-05,
+      "loss": 4.9439,
+      "num_input_tokens_seen": 741998592,
+      "step": 5661
+    },
+    {
+      "epoch": 0.9294721565530433,
+      "grad_norm": 0.6470625996589661,
+      "learning_rate": 6.961607792087073e-05,
+      "loss": 4.9094,
+      "num_input_tokens_seen": 742391808,
+      "step": 5664
+    },
+    {
+      "epoch": 0.9299644617207091,
+      "grad_norm": 0.8249083757400513,
+      "learning_rate": 6.959764878013196e-05,
+      "loss": 4.9275,
+      "num_input_tokens_seen": 742785024,
+      "step": 5667
+    },
+    {
+      "epoch": 0.930456766888375,
+      "grad_norm": 0.5849491953849792,
+      "learning_rate": 6.957923426763075e-05,
+      "loss": 4.9589,
+      "num_input_tokens_seen": 743178240,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9309490720560407,
+      "grad_norm": 0.7007787227630615,
+      "learning_rate": 6.956083436402524e-05,
+      "loss": 4.9004,
+      "num_input_tokens_seen": 743571456,
+      "step": 5673
+    },
+    {
+      "epoch": 0.9314413772237066,
+      "grad_norm": 0.6450519561767578,
+      "learning_rate": 6.954244905000938e-05,
+      "loss": 4.949,
+      "num_input_tokens_seen": 743964672,
+      "step": 5676
+    },
+    {
+      "epoch": 0.9319336823913723,
+      "grad_norm": 0.689784049987793,
+      "learning_rate": 6.95240783063128e-05,
+      "loss": 4.9121,
+      "num_input_tokens_seen": 744357888,
+      "step": 5679
+    },
+    {
+      "epoch": 0.9324259875590382,
+      "grad_norm": 0.5879255533218384,
+      "learning_rate": 6.950572211370075e-05,
+      "loss": 4.9665,
+      "num_input_tokens_seen": 744751104,
+      "step": 5682
+    },
+    {
+      "epoch": 0.932918292726704,
+      "grad_norm": 0.6298468708992004,
+      "learning_rate": 6.948738045297404e-05,
+      "loss": 4.9033,
+      "num_input_tokens_seen": 745144320,
+      "step": 5685
+    },
+    {
+      "epoch": 0.9334105978943698,
+      "grad_norm": 0.541263997554779,
+      "learning_rate": 6.946905330496889e-05,
+      "loss": 4.9259,
+      "num_input_tokens_seen": 745537536,
+      "step": 5688
+    },
+    {
+      "epoch": 0.9339029030620356,
+      "grad_norm": 0.6083950996398926,
+      "learning_rate": 6.945074065055687e-05,
+      "loss": 4.8963,
+      "num_input_tokens_seen": 745930752,
+      "step": 5691
+    },
+    {
+      "epoch": 0.9343952082297013,
+      "grad_norm": 0.60537189245224,
+      "learning_rate": 6.943244247064488e-05,
+      "loss": 4.9197,
+      "num_input_tokens_seen": 746323968,
+      "step": 5694
+    },
+    {
+      "epoch": 0.9348875133973672,
+      "grad_norm": 0.6046431064605713,
+      "learning_rate": 6.941415874617496e-05,
+      "loss": 4.9622,
+      "num_input_tokens_seen": 746717184,
+      "step": 5697
+    },
+    {
+      "epoch": 0.935379818565033,
+      "grad_norm": 0.5610204339027405,
+      "learning_rate": 6.939588945812431e-05,
+      "loss": 4.9442,
+      "num_input_tokens_seen": 747110400,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9358721237326988,
+      "grad_norm": 0.5891165733337402,
+      "learning_rate": 6.937763458750514e-05,
+      "loss": 4.9079,
+      "num_input_tokens_seen": 747503616,
+      "step": 5703
+    },
+    {
+      "epoch": 0.9363644289003646,
+      "grad_norm": 0.5680641531944275,
+      "learning_rate": 6.93593941153646e-05,
+      "loss": 4.9396,
+      "num_input_tokens_seen": 747896832,
+      "step": 5706
+    },
+    {
+      "epoch": 0.9368567340680304,
+      "grad_norm": 0.6169180870056152,
+      "learning_rate": 6.934116802278468e-05,
+      "loss": 4.9221,
+      "num_input_tokens_seen": 748290048,
+      "step": 5709
+    },
+    {
+      "epoch": 0.9373490392356962,
+      "grad_norm": 0.5749015212059021,
+      "learning_rate": 6.932295629088219e-05,
+      "loss": 4.9258,
+      "num_input_tokens_seen": 748683264,
+      "step": 5712
+    },
+    {
+      "epoch": 0.9378413444033621,
+      "grad_norm": 0.5365791916847229,
+      "learning_rate": 6.930475890080862e-05,
+      "loss": 4.8815,
+      "num_input_tokens_seen": 749076480,
+      "step": 5715
+    },
+    {
+      "epoch": 0.9383336495710278,
+      "grad_norm": 0.5742934942245483,
+      "learning_rate": 6.928657583375008e-05,
+      "loss": 4.8828,
+      "num_input_tokens_seen": 749469696,
+      "step": 5718
+    },
+    {
+      "epoch": 0.9388259547386937,
+      "grad_norm": 0.6056268215179443,
+      "learning_rate": 6.92684070709272e-05,
+      "loss": 4.9585,
+      "num_input_tokens_seen": 749862912,
+      "step": 5721
+    },
+    {
+      "epoch": 0.9393182599063594,
+      "grad_norm": 0.5517481565475464,
+      "learning_rate": 6.925025259359513e-05,
+      "loss": 4.8756,
+      "num_input_tokens_seen": 750256128,
+      "step": 5724
+    },
+    {
+      "epoch": 0.9398105650740253,
+      "grad_norm": 0.5459040999412537,
+      "learning_rate": 6.923211238304328e-05,
+      "loss": 4.9259,
+      "num_input_tokens_seen": 750649344,
+      "step": 5727
+    },
+    {
+      "epoch": 0.940302870241691,
+      "grad_norm": 0.640325129032135,
+      "learning_rate": 6.92139864205954e-05,
+      "loss": 4.898,
+      "num_input_tokens_seen": 751042560,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9407951754093569,
+      "grad_norm": 0.7371880412101746,
+      "learning_rate": 6.919587468760951e-05,
+      "loss": 4.885,
+      "num_input_tokens_seen": 751435776,
+      "step": 5733
+    },
+    {
+      "epoch": 0.9412874805770227,
+      "grad_norm": 0.6491408348083496,
+      "learning_rate": 6.917777716547768e-05,
+      "loss": 4.9081,
+      "num_input_tokens_seen": 751828992,
+      "step": 5736
+    },
+    {
+      "epoch": 0.9417797857446885,
+      "grad_norm": 0.5965387225151062,
+      "learning_rate": 6.915969383562604e-05,
+      "loss": 4.9093,
+      "num_input_tokens_seen": 752222208,
+      "step": 5739
+    },
+    {
+      "epoch": 0.9422720909123543,
+      "grad_norm": 0.6211044788360596,
+      "learning_rate": 6.914162467951475e-05,
+      "loss": 4.9024,
+      "num_input_tokens_seen": 752615424,
+      "step": 5742
+    },
+    {
+      "epoch": 0.94276439608002,
+      "grad_norm": 0.6093422174453735,
+      "learning_rate": 6.912356967863777e-05,
+      "loss": 4.9659,
+      "num_input_tokens_seen": 753008640,
+      "step": 5745
+    },
+    {
+      "epoch": 0.9432567012476859,
+      "grad_norm": 0.7026922106742859,
+      "learning_rate": 6.910552881452296e-05,
+      "loss": 4.8806,
+      "num_input_tokens_seen": 753401856,
+      "step": 5748
+    },
+    {
+      "epoch": 0.9437490064153518,
+      "grad_norm": 0.5915043950080872,
+      "learning_rate": 6.908750206873184e-05,
+      "loss": 4.8699,
+      "num_input_tokens_seen": 753795072,
+      "step": 5751
+    },
+    {
+      "epoch": 0.9442413115830175,
+      "grad_norm": 0.7234703898429871,
+      "learning_rate": 6.90694894228596e-05,
+      "loss": 4.8763,
+      "num_input_tokens_seen": 754188288,
+      "step": 5754
+    },
+    {
+      "epoch": 0.9447336167506833,
+      "grad_norm": 0.5900036096572876,
+      "learning_rate": 6.905149085853502e-05,
+      "loss": 4.895,
+      "num_input_tokens_seen": 754581504,
+      "step": 5757
+    },
+    {
+      "epoch": 0.9452259219183491,
+      "grad_norm": 0.7614732384681702,
+      "learning_rate": 6.903350635742038e-05,
+      "loss": 4.9233,
+      "num_input_tokens_seen": 754974720,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9457182270860149,
+      "grad_norm": 0.6385030746459961,
+      "learning_rate": 6.901553590121132e-05,
+      "loss": 4.8984,
+      "num_input_tokens_seen": 755367936,
+      "step": 5763
+    },
+    {
+      "epoch": 0.9462105322536808,
+      "grad_norm": 0.6103296279907227,
+      "learning_rate": 6.899757947163688e-05,
+      "loss": 4.9036,
+      "num_input_tokens_seen": 755761152,
+      "step": 5766
+    },
+    {
+      "epoch": 0.9467028374213465,
+      "grad_norm": 0.5251742005348206,
+      "learning_rate": 6.897963705045933e-05,
+      "loss": 4.9414,
+      "num_input_tokens_seen": 756154368,
+      "step": 5769
+    },
+    {
+      "epoch": 0.9471951425890124,
+      "grad_norm": 0.6542143821716309,
+      "learning_rate": 6.896170861947415e-05,
+      "loss": 4.9107,
+      "num_input_tokens_seen": 756547584,
+      "step": 5772
+    },
+    {
+      "epoch": 0.9476874477566781,
+      "grad_norm": 0.6727720499038696,
+      "learning_rate": 6.894379416050985e-05,
+      "loss": 4.8905,
+      "num_input_tokens_seen": 756940800,
+      "step": 5775
+    },
+    {
+      "epoch": 0.948179752924344,
+      "grad_norm": 0.5717095732688904,
+      "learning_rate": 6.892589365542804e-05,
+      "loss": 4.9338,
+      "num_input_tokens_seen": 757334016,
+      "step": 5778
+    },
+    {
+      "epoch": 0.9486720580920098,
+      "grad_norm": 0.6129010915756226,
+      "learning_rate": 6.890800708612326e-05,
+      "loss": 4.8975,
+      "num_input_tokens_seen": 757727232,
+      "step": 5781
+    },
+    {
+      "epoch": 0.9491643632596756,
+      "grad_norm": 0.6094048619270325,
+      "learning_rate": 6.889013443452292e-05,
+      "loss": 4.9282,
+      "num_input_tokens_seen": 758120448,
+      "step": 5784
+    },
+    {
+      "epoch": 0.9496566684273414,
+      "grad_norm": 0.5988612771034241,
+      "learning_rate": 6.887227568258717e-05,
+      "loss": 4.9152,
+      "num_input_tokens_seen": 758513664,
+      "step": 5787
+    },
+    {
+      "epoch": 0.9501489735950072,
+      "grad_norm": 0.6123887300491333,
+      "learning_rate": 6.885443081230899e-05,
+      "loss": 4.9025,
+      "num_input_tokens_seen": 758906880,
+      "step": 5790
+    },
+    {
+      "epoch": 0.950641278762673,
+      "grad_norm": 0.6150357127189636,
+      "learning_rate": 6.883659980571393e-05,
+      "loss": 4.944,
+      "num_input_tokens_seen": 759300096,
+      "step": 5793
+    },
+    {
+      "epoch": 0.9511335839303389,
+      "grad_norm": 0.558800995349884,
+      "learning_rate": 6.881878264486008e-05,
+      "loss": 4.9404,
+      "num_input_tokens_seen": 759693312,
+      "step": 5796
+    },
+    {
+      "epoch": 0.9516258890980046,
+      "grad_norm": 0.5992735624313354,
+      "learning_rate": 6.880097931183812e-05,
+      "loss": 4.9464,
+      "num_input_tokens_seen": 760086528,
+      "step": 5799
+    },
+    {
+      "epoch": 0.9521181942656705,
+      "grad_norm": 0.597300112247467,
+      "learning_rate": 6.878318978877102e-05,
+      "loss": 4.9496,
+      "num_input_tokens_seen": 760479744,
+      "step": 5802
+    },
+    {
+      "epoch": 0.9526104994333362,
+      "grad_norm": 0.5881853103637695,
+      "learning_rate": 6.876541405781422e-05,
+      "loss": 4.8984,
+      "num_input_tokens_seen": 760872960,
+      "step": 5805
+    },
+    {
+      "epoch": 0.953102804601002,
+      "grad_norm": 0.5540621280670166,
+      "learning_rate": 6.874765210115533e-05,
+      "loss": 4.8732,
+      "num_input_tokens_seen": 761266176,
+      "step": 5808
+    },
+    {
+      "epoch": 0.9535951097686679,
+      "grad_norm": 0.5574669241905212,
+      "learning_rate": 6.872990390101416e-05,
+      "loss": 4.9345,
+      "num_input_tokens_seen": 761659392,
+      "step": 5811
+    },
+    {
+      "epoch": 0.9540874149363336,
+      "grad_norm": 0.6479876637458801,
+      "learning_rate": 6.871216943964268e-05,
+      "loss": 4.9443,
+      "num_input_tokens_seen": 762052608,
+      "step": 5814
+    },
+    {
+      "epoch": 0.9545797201039995,
+      "grad_norm": 0.586850106716156,
+      "learning_rate": 6.869444869932488e-05,
+      "loss": 4.9271,
+      "num_input_tokens_seen": 762445824,
+      "step": 5817
+    },
+    {
+      "epoch": 0.9550720252716652,
+      "grad_norm": 0.736300528049469,
+      "learning_rate": 6.867674166237667e-05,
+      "loss": 4.973,
+      "num_input_tokens_seen": 762839040,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9555643304393311,
+      "grad_norm": 0.5725162029266357,
+      "learning_rate": 6.865904831114593e-05,
+      "loss": 4.8662,
+      "num_input_tokens_seen": 763232256,
+      "step": 5823
+    },
+    {
+      "epoch": 0.9560566356069969,
+      "grad_norm": 0.6230468153953552,
+      "learning_rate": 6.86413686280123e-05,
+      "loss": 4.8975,
+      "num_input_tokens_seen": 763625472,
+      "step": 5826
+    },
+    {
+      "epoch": 0.9565489407746627,
+      "grad_norm": 0.5963659882545471,
+      "learning_rate": 6.862370259538721e-05,
+      "loss": 4.9168,
+      "num_input_tokens_seen": 764018688,
+      "step": 5829
+    },
+    {
+      "epoch": 0.9570412459423285,
+      "grad_norm": 0.6411994695663452,
+      "learning_rate": 6.86060501957137e-05,
+      "loss": 4.9349,
+      "num_input_tokens_seen": 764411904,
+      "step": 5832
+    },
+    {
+      "epoch": 0.9575335511099943,
+      "grad_norm": 0.6455360651016235,
+      "learning_rate": 6.858841141146649e-05,
+      "loss": 4.9462,
+      "num_input_tokens_seen": 764805120,
+      "step": 5835
+    },
+    {
+      "epoch": 0.9580258562776601,
+      "grad_norm": 0.7855854034423828,
+      "learning_rate": 6.857078622515172e-05,
+      "loss": 4.93,
+      "num_input_tokens_seen": 765198336,
+      "step": 5838
+    },
+    {
+      "epoch": 0.9585181614453259,
+      "grad_norm": 0.6942725777626038,
+      "learning_rate": 6.855317461930706e-05,
+      "loss": 4.9313,
+      "num_input_tokens_seen": 765591552,
+      "step": 5841
+    },
+    {
+      "epoch": 0.9590104666129917,
+      "grad_norm": 0.645770788192749,
+      "learning_rate": 6.853557657650157e-05,
+      "loss": 4.9185,
+      "num_input_tokens_seen": 765984768,
+      "step": 5844
+    },
+    {
+      "epoch": 0.9595027717806576,
+      "grad_norm": 0.6724271178245544,
+      "learning_rate": 6.851799207933553e-05,
+      "loss": 4.9469,
+      "num_input_tokens_seen": 766377984,
+      "step": 5847
+    },
+    {
+      "epoch": 0.9599950769483233,
+      "grad_norm": 0.6254433393478394,
+      "learning_rate": 6.850042111044057e-05,
+      "loss": 4.9137,
+      "num_input_tokens_seen": 766771200,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9604873821159892,
+      "grad_norm": 0.5859279036521912,
+      "learning_rate": 6.848286365247937e-05,
+      "loss": 4.8999,
+      "num_input_tokens_seen": 767164416,
+      "step": 5853
+    },
+    {
+      "epoch": 0.9609796872836549,
+      "grad_norm": 0.6462947130203247,
+      "learning_rate": 6.846531968814576e-05,
+      "loss": 4.9484,
+      "num_input_tokens_seen": 767557632,
+      "step": 5856
+    },
+    {
+      "epoch": 0.9614719924513208,
+      "grad_norm": 0.5213596820831299,
+      "learning_rate": 6.844778920016459e-05,
+      "loss": 4.9116,
+      "num_input_tokens_seen": 767950848,
+      "step": 5859
+    },
+    {
+      "epoch": 0.9619642976189866,
+      "grad_norm": 0.6776494979858398,
+      "learning_rate": 6.843027217129164e-05,
+      "loss": 4.9351,
+      "num_input_tokens_seen": 768344064,
+      "step": 5862
+    },
+    {
+      "epoch": 0.9624566027866523,
+      "grad_norm": 0.5639019012451172,
+      "learning_rate": 6.841276858431358e-05,
+      "loss": 4.8733,
+      "num_input_tokens_seen": 768737280,
+      "step": 5865
+    },
+    {
+      "epoch": 0.9629489079543182,
+      "grad_norm": 0.6252104640007019,
+      "learning_rate": 6.839527842204787e-05,
+      "loss": 4.8784,
+      "num_input_tokens_seen": 769130496,
+      "step": 5868
+    },
+    {
+      "epoch": 0.9634412131219839,
+      "grad_norm": 0.6042758822441101,
+      "learning_rate": 6.837780166734271e-05,
+      "loss": 4.9082,
+      "num_input_tokens_seen": 769523712,
+      "step": 5871
+    },
+    {
+      "epoch": 0.9639335182896498,
+      "grad_norm": 0.6575847268104553,
+      "learning_rate": 6.836033830307697e-05,
+      "loss": 4.8967,
+      "num_input_tokens_seen": 769916928,
+      "step": 5874
+    },
+    {
+      "epoch": 0.9644258234573156,
+      "grad_norm": 0.728636622428894,
+      "learning_rate": 6.834288831216011e-05,
+      "loss": 4.9266,
+      "num_input_tokens_seen": 770310144,
+      "step": 5877
+    },
+    {
+      "epoch": 0.9649181286249814,
+      "grad_norm": 0.5777960419654846,
+      "learning_rate": 6.832545167753211e-05,
+      "loss": 4.9058,
+      "num_input_tokens_seen": 770703360,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9654104337926472,
+      "grad_norm": 0.7681834697723389,
+      "learning_rate": 6.830802838216338e-05,
+      "loss": 4.9101,
+      "num_input_tokens_seen": 771096576,
+      "step": 5883
+    },
+    {
+      "epoch": 0.965902738960313,
+      "grad_norm": 0.5604771971702576,
+      "learning_rate": 6.829061840905477e-05,
+      "loss": 4.9314,
+      "num_input_tokens_seen": 771489792,
+      "step": 5886
+    },
+    {
+      "epoch": 0.9663950441279788,
+      "grad_norm": 0.5763574838638306,
+      "learning_rate": 6.82732217412374e-05,
+      "loss": 4.8776,
+      "num_input_tokens_seen": 771883008,
+      "step": 5889
+    },
+    {
+      "epoch": 0.9668873492956447,
+      "grad_norm": 0.580155611038208,
+      "learning_rate": 6.825583836177263e-05,
+      "loss": 4.9113,
+      "num_input_tokens_seen": 772276224,
+      "step": 5892
+    },
+    {
+      "epoch": 0.9673796544633104,
+      "grad_norm": 0.6300919651985168,
+      "learning_rate": 6.823846825375201e-05,
+      "loss": 4.8782,
+      "num_input_tokens_seen": 772669440,
+      "step": 5895
+    },
+    {
+      "epoch": 0.9678719596309763,
+      "grad_norm": 0.5544500350952148,
+      "learning_rate": 6.822111140029719e-05,
+      "loss": 4.8604,
+      "num_input_tokens_seen": 773062656,
+      "step": 5898
+    },
+    {
+      "epoch": 0.968364264798642,
+      "grad_norm": 0.6302488446235657,
+      "learning_rate": 6.820376778455987e-05,
+      "loss": 4.9071,
+      "num_input_tokens_seen": 773455872,
+      "step": 5901
+    },
+    {
+      "epoch": 0.9688565699663079,
+      "grad_norm": 0.6206812858581543,
+      "learning_rate": 6.81864373897217e-05,
+      "loss": 4.9005,
+      "num_input_tokens_seen": 773849088,
+      "step": 5904
+    },
+    {
+      "epoch": 0.9693488751339737,
+      "grad_norm": 0.638460636138916,
+      "learning_rate": 6.816912019899426e-05,
+      "loss": 4.9282,
+      "num_input_tokens_seen": 774242304,
+      "step": 5907
+    },
+    {
+      "epoch": 0.9698411803016395,
+      "grad_norm": 0.6319560408592224,
+      "learning_rate": 6.815181619561888e-05,
+      "loss": 4.8894,
+      "num_input_tokens_seen": 774635520,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9703334854693053,
+      "grad_norm": 0.5597271919250488,
+      "learning_rate": 6.813452536286677e-05,
+      "loss": 4.9043,
+      "num_input_tokens_seen": 775028736,
+      "step": 5913
+    },
+    {
+      "epoch": 0.970825790636971,
+      "grad_norm": 0.8066889643669128,
+      "learning_rate": 6.811724768403874e-05,
+      "loss": 4.9336,
+      "num_input_tokens_seen": 775421952,
+      "step": 5916
+    },
+    {
+      "epoch": 0.9713180958046369,
+      "grad_norm": 0.6452884078025818,
+      "learning_rate": 6.809998314246527e-05,
+      "loss": 4.9215,
+      "num_input_tokens_seen": 775815168,
+      "step": 5919
+    },
+    {
+      "epoch": 0.9718104009723028,
+      "grad_norm": 0.5996139645576477,
+      "learning_rate": 6.80827317215064e-05,
+      "loss": 4.9279,
+      "num_input_tokens_seen": 776208384,
+      "step": 5922
+    },
+    {
+      "epoch": 0.9723027061399685,
+      "grad_norm": 0.7165923118591309,
+      "learning_rate": 6.806549340455164e-05,
+      "loss": 4.9125,
+      "num_input_tokens_seen": 776601600,
+      "step": 5925
+    },
+    {
+      "epoch": 0.9727950113076343,
+      "grad_norm": 0.6175607442855835,
+      "learning_rate": 6.804826817501996e-05,
+      "loss": 4.8999,
+      "num_input_tokens_seen": 776994816,
+      "step": 5928
+    },
+    {
+      "epoch": 0.9732873164753001,
+      "grad_norm": 0.6167675256729126,
+      "learning_rate": 6.803105601635961e-05,
+      "loss": 4.9054,
+      "num_input_tokens_seen": 777388032,
+      "step": 5931
+    },
+    {
+      "epoch": 0.9737796216429659,
+      "grad_norm": 0.6145119667053223,
+      "learning_rate": 6.801385691204829e-05,
+      "loss": 4.878,
+      "num_input_tokens_seen": 777781248,
+      "step": 5934
+    },
+    {
+      "epoch": 0.9742719268106318,
+      "grad_norm": 0.5707881450653076,
+      "learning_rate": 6.799667084559273e-05,
+      "loss": 4.8818,
+      "num_input_tokens_seen": 778174464,
+      "step": 5937
+    },
+    {
+      "epoch": 0.9747642319782975,
+      "grad_norm": 0.5859269499778748,
+      "learning_rate": 6.797949780052896e-05,
+      "loss": 4.8894,
+      "num_input_tokens_seen": 778567680,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9752565371459634,
+      "grad_norm": 0.560761570930481,
+      "learning_rate": 6.796233776042202e-05,
+      "loss": 4.9154,
+      "num_input_tokens_seen": 778960896,
+      "step": 5943
+    },
+    {
+      "epoch": 0.9757488423136291,
+      "grad_norm": 0.5537312626838684,
+      "learning_rate": 6.794519070886606e-05,
+      "loss": 4.8822,
+      "num_input_tokens_seen": 779354112,
+      "step": 5946
+    },
+    {
+      "epoch": 0.976241147481295,
+      "grad_norm": 0.6321630477905273,
+      "learning_rate": 6.792805662948407e-05,
+      "loss": 4.9161,
+      "num_input_tokens_seen": 779747328,
+      "step": 5949
+    },
+    {
+      "epoch": 0.9767334526489607,
+      "grad_norm": 0.6216323375701904,
+      "learning_rate": 6.791093550592807e-05,
+      "loss": 4.912,
+      "num_input_tokens_seen": 780140544,
+      "step": 5952
+    },
+    {
+      "epoch": 0.9772257578166266,
+      "grad_norm": 0.5987245440483093,
+      "learning_rate": 6.789382732187882e-05,
+      "loss": 4.9255,
+      "num_input_tokens_seen": 780533760,
+      "step": 5955
+    },
+    {
+      "epoch": 0.9777180629842924,
+      "grad_norm": 0.6332117319107056,
+      "learning_rate": 6.787673206104584e-05,
+      "loss": 4.8713,
+      "num_input_tokens_seen": 780926976,
+      "step": 5958
+    },
+    {
+      "epoch": 0.9782103681519582,
+      "grad_norm": 0.5283944010734558,
+      "learning_rate": 6.785964970716741e-05,
+      "loss": 4.9124,
+      "num_input_tokens_seen": 781320192,
+      "step": 5961
+    },
+    {
+      "epoch": 0.978702673319624,
+      "grad_norm": 0.5525709390640259,
+      "learning_rate": 6.784258024401038e-05,
+      "loss": 4.9321,
+      "num_input_tokens_seen": 781713408,
+      "step": 5964
+    },
+    {
+      "epoch": 0.9791949784872898,
+      "grad_norm": 0.5929332971572876,
+      "learning_rate": 6.782552365537023e-05,
+      "loss": 4.9264,
+      "num_input_tokens_seen": 782106624,
+      "step": 5967
+    },
+    {
+      "epoch": 0.9796872836549556,
+      "grad_norm": 0.6105230450630188,
+      "learning_rate": 6.780847992507089e-05,
+      "loss": 4.9408,
+      "num_input_tokens_seen": 782499840,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9801795888226215,
+      "grad_norm": 0.6432685256004333,
+      "learning_rate": 6.779144903696476e-05,
+      "loss": 4.8833,
+      "num_input_tokens_seen": 782893056,
+      "step": 5973
+    },
+    {
+      "epoch": 0.9806718939902872,
+      "grad_norm": 0.6585260033607483,
+      "learning_rate": 6.77744309749326e-05,
+      "loss": 4.915,
+      "num_input_tokens_seen": 783286272,
+      "step": 5976
+    },
+    {
+      "epoch": 0.981164199157953,
+      "grad_norm": 0.6275220513343811,
+      "learning_rate": 6.775742572288348e-05,
+      "loss": 4.9577,
+      "num_input_tokens_seen": 783679488,
+      "step": 5979
+    },
+    {
+      "epoch": 0.9816565043256188,
+      "grad_norm": 0.6095502972602844,
+      "learning_rate": 6.774043326475473e-05,
+      "loss": 4.8891,
+      "num_input_tokens_seen": 784072704,
+      "step": 5982
+    },
+    {
+      "epoch": 0.9821488094932846,
+      "grad_norm": 0.632768988609314,
+      "learning_rate": 6.772345358451186e-05,
+      "loss": 4.9223,
+      "num_input_tokens_seen": 784465920,
+      "step": 5985
+    },
+    {
+      "epoch": 0.9826411146609505,
+      "grad_norm": 0.6620305180549622,
+      "learning_rate": 6.770648666614851e-05,
+      "loss": 4.8904,
+      "num_input_tokens_seen": 784859136,
+      "step": 5988
+    },
+    {
+      "epoch": 0.9831334198286162,
+      "grad_norm": 0.6766555309295654,
+      "learning_rate": 6.768953249368636e-05,
+      "loss": 4.8817,
+      "num_input_tokens_seen": 785252352,
+      "step": 5991
+    },
+    {
+      "epoch": 0.9836257249962821,
+      "grad_norm": 0.5768389105796814,
+      "learning_rate": 6.767259105117506e-05,
+      "loss": 4.8491,
+      "num_input_tokens_seen": 785645568,
+      "step": 5994
+    },
+    {
+      "epoch": 0.9841180301639478,
+      "grad_norm": 0.6032988429069519,
+      "learning_rate": 6.765566232269226e-05,
+      "loss": 4.9208,
+      "num_input_tokens_seen": 786038784,
+      "step": 5997
+    },
+    {
+      "epoch": 0.9846103353316137,
+      "grad_norm": 0.6204380393028259,
+      "learning_rate": 6.763874629234341e-05,
+      "loss": 4.8818,
+      "num_input_tokens_seen": 786432000,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9846103353316137,
+      "eval_accuracy": 0.2314671877544374,
+      "eval_loss": 5.198113918304443,
+      "eval_runtime": 127.6077,
+      "eval_samples_per_second": 2.351,
+      "eval_steps_per_second": 1.175,
+      "num_input_tokens_seen": 786432000,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9851026404992795,
+      "grad_norm": 0.5865297317504883,
+      "learning_rate": 6.762184294426182e-05,
+      "loss": 4.8824,
+      "num_input_tokens_seen": 786825216,
+      "step": 6003
+    },
+    {
+      "epoch": 0.9855949456669453,
+      "grad_norm": 0.6484345197677612,
+      "learning_rate": 6.760495226260847e-05,
+      "loss": 4.8941,
+      "num_input_tokens_seen": 787218432,
+      "step": 6006
+    },
+    {
+      "epoch": 0.9860872508346111,
+      "grad_norm": 0.59868985414505,
+      "learning_rate": 6.75880742315721e-05,
+      "loss": 4.9025,
+      "num_input_tokens_seen": 787611648,
+      "step": 6009
+    },
+    {
+      "epoch": 0.9865795560022769,
+      "grad_norm": 0.5721721053123474,
+      "learning_rate": 6.757120883536902e-05,
+      "loss": 4.8878,
+      "num_input_tokens_seen": 788004864,
+      "step": 6012
+    },
+    {
+      "epoch": 0.9870718611699427,
+      "grad_norm": 0.647430956363678,
+      "learning_rate": 6.755435605824312e-05,
+      "loss": 4.9164,
+      "num_input_tokens_seen": 788398080,
+      "step": 6015
+    },
+    {
+      "epoch": 0.9875641663376086,
+      "grad_norm": 0.6465507745742798,
+      "learning_rate": 6.753751588446576e-05,
+      "loss": 4.9158,
+      "num_input_tokens_seen": 788791296,
+      "step": 6018
+    },
+    {
+      "epoch": 0.9880564715052743,
+      "grad_norm": 0.6825433969497681,
+      "learning_rate": 6.752068829833576e-05,
+      "loss": 4.9292,
+      "num_input_tokens_seen": 789184512,
+      "step": 6021
+    },
+    {
+      "epoch": 0.9885487766729402,
+      "grad_norm": 0.5541806221008301,
+      "learning_rate": 6.750387328417927e-05,
+      "loss": 4.9168,
+      "num_input_tokens_seen": 789577728,
+      "step": 6024
+    },
+    {
+      "epoch": 0.9890410818406059,
+      "grad_norm": 0.6401199698448181,
+      "learning_rate": 6.748707082634982e-05,
+      "loss": 4.8989,
+      "num_input_tokens_seen": 789970944,
+      "step": 6027
+    },
+    {
+      "epoch": 0.9895333870082718,
+      "grad_norm": 0.5784100294113159,
+      "learning_rate": 6.747028090922809e-05,
+      "loss": 4.8809,
+      "num_input_tokens_seen": 790364160,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9900256921759376,
+      "grad_norm": 0.6119440197944641,
+      "learning_rate": 6.745350351722202e-05,
+      "loss": 4.8772,
+      "num_input_tokens_seen": 790757376,
+      "step": 6033
+    },
+    {
+      "epoch": 0.9905179973436034,
+      "grad_norm": 0.5223164558410645,
+      "learning_rate": 6.743673863476671e-05,
+      "loss": 4.9168,
+      "num_input_tokens_seen": 791150592,
+      "step": 6036
+    },
+    {
+      "epoch": 0.9910103025112692,
+      "grad_norm": 0.6039283275604248,
+      "learning_rate": 6.74199862463242e-05,
+      "loss": 4.8858,
+      "num_input_tokens_seen": 791543808,
+      "step": 6039
+    },
+    {
+      "epoch": 0.9915026076789349,
+      "grad_norm": 0.5842203497886658,
+      "learning_rate": 6.740324633638366e-05,
+      "loss": 4.9642,
+      "num_input_tokens_seen": 791937024,
+      "step": 6042
+    },
+    {
+      "epoch": 0.9919949128466008,
+      "grad_norm": 0.5527199506759644,
+      "learning_rate": 6.738651888946112e-05,
+      "loss": 4.8472,
+      "num_input_tokens_seen": 792330240,
+      "step": 6045
+    },
+    {
+      "epoch": 0.9924872180142666,
+      "grad_norm": 0.7194064259529114,
+      "learning_rate": 6.736980389009957e-05,
+      "loss": 4.9052,
+      "num_input_tokens_seen": 792723456,
+      "step": 6048
+    },
+    {
+      "epoch": 0.9929795231819324,
+      "grad_norm": 0.5239654779434204,
+      "learning_rate": 6.735310132286876e-05,
+      "loss": 4.8818,
+      "num_input_tokens_seen": 793116672,
+      "step": 6051
+    },
+    {
+      "epoch": 0.9934718283495982,
+      "grad_norm": 0.5916581153869629,
+      "learning_rate": 6.733641117236525e-05,
+      "loss": 4.9061,
+      "num_input_tokens_seen": 793509888,
+      "step": 6054
+    },
+    {
+      "epoch": 0.993964133517264,
+      "grad_norm": 0.5269445776939392,
+      "learning_rate": 6.731973342321227e-05,
+      "loss": 4.8932,
+      "num_input_tokens_seen": 793903104,
+      "step": 6057
+    },
+    {
+      "epoch": 0.9944564386849298,
+      "grad_norm": 0.5389668941497803,
+      "learning_rate": 6.73030680600597e-05,
+      "loss": 4.8693,
+      "num_input_tokens_seen": 794296320,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9949487438525956,
+      "grad_norm": 0.5677322745323181,
+      "learning_rate": 6.728641506758407e-05,
+      "loss": 4.9268,
+      "num_input_tokens_seen": 794689536,
+      "step": 6063
+    },
+    {
+      "epoch": 0.9954410490202614,
+      "grad_norm": 0.5324602723121643,
+      "learning_rate": 6.726977443048832e-05,
+      "loss": 4.9348,
+      "num_input_tokens_seen": 795082752,
+      "step": 6066
+    },
+    {
+      "epoch": 0.9959333541879273,
+      "grad_norm": 0.5456458330154419,
+      "learning_rate": 6.725314613350202e-05,
+      "loss": 4.8845,
+      "num_input_tokens_seen": 795475968,
+      "step": 6069
+    },
+    {
+      "epoch": 0.996425659355593,
+      "grad_norm": 0.5815337896347046,
+      "learning_rate": 6.723653016138096e-05,
+      "loss": 4.8419,
+      "num_input_tokens_seen": 795869184,
+      "step": 6072
+    },
+    {
+      "epoch": 0.9969179645232589,
+      "grad_norm": 0.6590341329574585,
+      "learning_rate": 6.721992649890743e-05,
+      "loss": 4.9176,
+      "num_input_tokens_seen": 796262400,
+      "step": 6075
+    },
+    {
+      "epoch": 0.9974102696909246,
+      "grad_norm": 0.5114787817001343,
+      "learning_rate": 6.720333513088994e-05,
+      "loss": 4.9158,
+      "num_input_tokens_seen": 796655616,
+      "step": 6078
+    },
+    {
+      "epoch": 0.9979025748585905,
+      "grad_norm": 0.6558582186698914,
+      "learning_rate": 6.71867560421633e-05,
+      "loss": 4.8964,
+      "num_input_tokens_seen": 797048832,
+      "step": 6081
+    },
+    {
+      "epoch": 0.9983948800262563,
+      "grad_norm": 0.5851151347160339,
+      "learning_rate": 6.717018921758838e-05,
+      "loss": 4.9178,
+      "num_input_tokens_seen": 797442048,
+      "step": 6084
+    },
+    {
+      "epoch": 0.9988871851939221,
+      "grad_norm": 0.6128942370414734,
+      "learning_rate": 6.715363464205227e-05,
+      "loss": 4.9023,
+      "num_input_tokens_seen": 797835264,
+      "step": 6087
+    },
+    {
+      "epoch": 0.9993794903615879,
+      "grad_norm": 0.6216326355934143,
+      "learning_rate": 6.713709230046812e-05,
+      "loss": 4.8119,
+      "num_input_tokens_seen": 798228480,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9998717955292536,
+      "grad_norm": 0.656932532787323,
+      "learning_rate": 6.712056217777502e-05,
+      "loss": 4.9169,
+      "num_input_tokens_seen": 798621696,
+      "step": 6093
+    },
+    {
+      "epoch": 0.9998717955292536,
+      "num_input_tokens_seen": 798621696,
+      "step": 6093,
+      "total_flos": 4.855554488590664e+17,
+      "train_loss": 5.304058988399935,
+      "train_runtime": 127910.3478,
+      "train_samples_per_second": 3.049,
+      "train_steps_per_second": 0.048
+    }
+  ],
+  "logging_steps": 3,
+  "max_steps": 6093,
+  "num_input_tokens_seen": 798621696,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 4.855554488590664e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}